{
    "9805/astro-ph9805133_arXiv.txt": {
        "abstract": "The global development of magnetohydrodynamic turbulence in an accretion disk is studied within a simplified disk model that omits vertical stratification. Starting with a weak vertical seed field, a saturated state is obtained after a few tens of orbits in which the energy in the predominantly toroidal magnetic field is still subthermal. The efficiency of angular momentum transport, parameterized by the Shakura-Sunyaev $\\alpha$ parameter, is of the order of $10^{-1}$. The dominant contribution to $\\alpha$ comes from magnetic stresses, which are enhanced by the presence of weak net vertical fields. The power spectra of the magnetic fields are flat or decline only slowly towards the largest scales accessible in the calculation, suggesting that the viscosity arising from MHD turbulence may not be a locally determined quantity. I discuss how these results compare with observationally inferred values of $\\alpha$, and possible implications for models of jet formation. ",
        "introduction": "The Balbus-Hawley instability is the most generally applicable mechanism known to initiate turbulence and outward angular momentum transport in accretion disks (Balbus \\& Hawley 1991). This is a linear, local instability that exists for rotating flows threaded by a weak magnetic field with ${\\rm d} \\Omega^2 / {\\rm d} r < 0$, conditions satisfied in disks (for earlier discussions see Velikhov 1959, Chandrasekhar 1961). A vigorous growth rate is obtained for a wide variety of initial magnetic field configurations (Balbus \\& Hawley 1992; Ogilvie \\& Pringle 1996; Terquem \\& Papaloizou 1996), implying that the instability is inescapable for ionized disks where the field is well-coupled to the gas.  Extensive numerical simulations have explored the nonlinear development of the instability within the local, shearing box approximation (for a review, see e.g. Gammie 1998). Such simulations have convincingly established that the nonlinear development of the Balbus-Hawley instability leads to sustained turbulence and significant angular momentum transport, typically finding a Shakura-Sunyaev (1973) $\\alpha \\approx 10^{-2}$ (Hawley, Gammie \\& Balbus 1995, 1996; Stone et al 1996; Brandenburg et al. 1995). There is some evidence for cyclic behavior that might have important implications for disk variability (Brandenburg et al. 1996). Equally important has been the final elimination of convection (Stone \\& Balbus 1996), and the near-elimination of nonlinear hydrodynamic turbulence (Balbus, Hawley \\& Stone 1996), as plausible rival mechanisms for angular momentum transport in accretion disks. Progress has also been made in trying to understand how the rich phenomenology of accretion disk variability can arise within a dynamo driven disk model (Armitage, Livio \\& Pringle 1996; Gammie \\& Menou 1998), although much more remains to be done in this area.  There are many further questions that one may hope simulations will address, and not all of them are amenable to a local treatment. Most obviously, is the angular momentum transport in a disk locally determined? What is the structure of the spatial and time variability of the disk fields, and are they suitable for launching a magnetically driven disk wind or jet? Unsurprisingly, the global calculations needed to investigate these issues are extremely demanding, both as a consequence of the larger computational domain and, especially, because of the need to simulate regions of low density where the high Alfv\\'en speed severely limits the timestep of explicit numerical codes.  In this paper, results are presented from a vertically unstratified global simulation of accretion disk turbulence. Such a calculation is evidently missing essential physics. There is no buoyancy, no possibility of Parker instability (Parker 1979), and no magnetically dominated disk corona -- all features that are expected to arise in a full disk model and which may be crucial for the disk dynamo problem (Tout \\& Pringle 1992). However the lesser computational demands permit a preliminary investigation of some of the important questions raised by previous, local, simulations. ",
        "conclusions": "In this paper, we have reported on a global simulation of an unstratified magnetized accretion disk. As expected from analytic considerations (Curry \\& Pudritz 1994, 1995, 1996) and local simulations, MHD instabilities generate sustained turbulence that leads to outward transport of angular momentum. The generated fields possess considerable power in azimuthal modes of low $m$, which correspond to physical scales considerably in excess of $H$, the disk semi-thickness, and display a ragged spiral structure. The current simulation does not admit the development of the Parker instability, which might depress the power on large scales, but with this caveat the results suggest that a viscosity originating from MHD turbulence may not be a locally determined quantity. The dominant magnetic field component is toroidal, and the interaction of this fluctuating internal field with a magnetosphere is likely to be an important complication to the already complex picture of star-disk interaction in magnetic systems (Miller \\& Stone 1997; Torkelsson 1998).  The efficiency of angular momentum transport seen in the calculation, parameterized by the Shakura-Sunyaev $\\alpha$ prescription, is $\\alpha \\approx 10^{-1}$. This is larger than the value obtained from local calculations (Gammie 1998) with zero net vertical magnetic fields, though we have noted that the influence of a vertical field on the current simulation is likely to have boosted the value of $\\alpha$ significantly. More realistic simulations with demonstrated numerical convergence are evidently required. However there is no obvious discrepancy with the values of $\\alpha$ inferred for dwarf novae, where modeling of disk outbursts suggests $\\alpha = 0.1 - 0.3$ (Cannizzo 1993), and for Active Galactic Nuclei, where the admittedly poorer observations are consistent with an $\\alpha$ of $10^{-2}$ (Siemiginowska \\& Czerny 1989). Conversely it is hard to see why a viscosity derived from MHD turbulence should be two or three orders of magnitude {\\em lower} in the ionized inner regions of protostellar disks, as required to match the timescales of FU Orionis outbursts within disk instability models (Bell \\& Lin 1994). This may call into question the {\\em self-regulated} aspect of the thermal disk instability picture for FU Orionis events. The generation of magnetic fields of large scale may additionally be important for models of jet formation (eg. Blandford \\& Payne 1982; Ouyed, Pudritz \\& Stone 1997; Matsumoto \\& Shibata 1997; Konigl 1997), which if generated via a disk dynamo would be expected to be most efficient in relatively thick disks or advection dominated flows (Narayan \\& Yi 1995). Observations of which systems produce jets appear to be broadly consistent with a model in which $(H/R)$ is a controlling parameter, though many other possibilities are also viable (Livio 1997)."
    },
    "9805/astro-ph9805243_arXiv.txt": {
        "abstract": "In order to clarify the role of the magnetic field in generating abundance inhomogeneities in the atmospheres of Ap stars, we present new abundance Doppler images and an approximate magnetic field geometry for the Ap star $\\iota$~Cas. ",
        "introduction": "The inhomogeneous distributions of surface chemical abundance observed in the Ap and Bp stars are thought to result from a complex interplay between gravitational and radiative diffusion,  mass loss, turbulence and circulation processes in their atmospheres (Michaud \\& Proffitt 1993). By determining the magnetic field geometries and surface chemical abundance distributions of a representative sample of these objects, important new contraints can be placed upon the manner and degree to which these processes interact with the magnetic field.  $\\iota$~Cas (HD 15089) is classified as A5p in the {\\em Henry Draper Catalogue}. Its projected rotational velocity is moderately high ($\\sim$ 50~\\kms), making it an ideal candidate for Doppler imaging. Borra \\& Landstreet (1980) obtained four Balmer-line magnetometer measurements of the longitudinal magnetic field of this star, with $1\\sigma$ uncertainties around 150~G. All but one are consistent with zero field, indicating that the magnetic field is quite weak. The photometric period (which we assume to be the rotational period) has been previously reported by a number of authors. ",
        "conclusions": ""
    },
    "9805/astro-ph9805075_arXiv.txt": {
        "abstract": "Using Hipparcos parallaxes, we show that the metallic A-F giants found by Hauck (1986) on the basis of their high $\\Delta m_2$ index in Geneva photometry are on average more evolved than their non-metallic counterparts. Their mass distribution, rate of binaries and $v\\sin i$ are shown to be incompatible with those of Am stars, so that they cannot be descendants of the latter. They might be former normal stars going through a short metal-rich phase at the very end of their life on the Main Sequence. ",
        "introduction": "The metallic A-F giants, found by Hauck (1986) on the basis of their high $\\Delta m_2$ index, have an abundance pattern similar to that of Am stars, except for Ca and Sc which have a more or less solar abundance (Berthet 1990, 1991). Their chemical anomalies closely resemble those of the $\\delta Del$ stars, which seem to be evolved Am stars (Kurtz 1976). Therefore, it appeared natural to consider the metallic A-F giants as evolved Am stars too. Indeed, the theory of radiative diffusion foresees that calcium, which had sinked to large atmospheric depths in the beginning of the star's life would be finally dredged up by the increasingly deeper outer convective zone as the star would reach the giant phase. This scenario was advocated by Berthet (1992).  Hereafter, we reconsider this question from a different standpoint and examine the evolutionary state of the metallic A-F giants in the light of the Hipparcos results. ",
        "conclusions": "There is a set of rather compelling arguments against the idea that metallic A-F giants would be evolved Am stars. On the other hand, the precise Hipparcos parallaxes allow for the first time to pinpoint these metallic giants in the HR diagram and to show that they all have $\\log g \\leq 3.8$. This seems to add some weight to the alternative idea that these giants have nothing to do with Am stars but may be former ``normal'' A stars going through a short phase where, for some as yet unclear reason, radiative diffusion is allowed to enhance the metallic abundance in their atmosphere. \\begin{figure}[hbt] \\psfig{figure=p25f1.ps,height=7.8cm} \\caption{HR diagram of all giants considered by Hauck (1986). Full dots: $\\Delta m_2 > 0.013$, open dots: $\\Delta m_2\\leq 0.013$. ZAMS, TAMS and evolutionary tracks for $Z=0.020$ are from Schaller et al. (1992).} \\label{fp} \\end{figure}"
    },
    "9805/hep-ph9805360_arXiv.txt": {
        "abstract": "The time evolution of the correlation functions of an ensemble of anharmonic N-component  oscillators with O(N) symmetry is described by a flow equation, exact up to corrections of order $1/N^2$. We find effective irreversibility. Nevertheless, analytical and numerical investigation reveals that the system does not reach thermal equilibrium for large times, even when $N\\rightarrow \\infty$. Depending on the initial distribution, the dynamics is asymptotically stable or it exhibits growing modes which break the conditions for the validity of the $1/N$ expansion for large time. We investigate both classical and quantum systems, the latter being the limit of an $O(N)$ symmetric scalar quantum field theory in zero spatial dimensions. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805239_arXiv.txt": {
        "abstract": "We show that upcoming CMB satellite experiments and large redshift surveys can be used together to yield 5\\% determinations of $H_0$ and $\\om$, an order of magnitude improvement over CMB data alone. CMB anisotropies provide the sound horizon at recombination as a standard ruler.  For reasonable baryon fractions, this scale is imprinted on the galaxy power spectrum as a series of spectral features. Measuring these features in redshift space determines the Hubble constant, which in turn yields $\\om$ once combined with CMB data. Since the oscillations in both power spectra are frozen in at recombination, this test is insensitive to low-redshift cosmology. ",
        "introduction": "In the usual cosmological paradigm, the cosmic microwave background (CMB) contains a vast amount of information about cosmological parameters (\\cite{Hu97}\\ 1997).  With upcoming experiments, most notably the two satellite missions MAP\\footnote{http://map.gsfc.nasa.gov} and Planck\\footnote{http://astro.estec.esa.nl/SA-general/Projects/Planck}, detailed measurements of the angular power spectra of its anisotropy and polarization may accurately determine many cosmological parameters (\\cite{Jun96}\\ 1996; \\cite{Bon97}\\ 1997; \\cite{Zal97}\\ 1997). However, certain changes in the cosmological parameters can conspire to leave the CMB power spectra unchanged, resulting in degenerate directions in the parameter space (\\cite{Bon94}\\ 1994, 1997; \\cite{Zal97}\\ 1997; \\cite{Hue98}\\ 1998).  For example, since the Hubble constant $H_0$ and the matter density $\\om$ can be varied while keeping the angular diameter distance and the matter-radiation ratio fixed, their values remain uncertain but highly correlated. Such degeneracies must be broken with cosmological information from other sources.  Upcoming redshift surveys for the study of large-scale structure hold the potential for resolving this issue.  In particular, the 2dF survey\\footnote{http://meteor.anu.edu.au/$\\sim$colless/2dF} and the Sloan Digital Sky Survey (SDSS)\\footnote{http://www.astro.princeton.edu/BBOOK} should measure the galaxy power spectrum on large enough scales to allow detailed comparisons to the mass power spectra predicted by cosmological theories.  In this {\\it Letter} and a companion paper (Eisenstein, Hu, \\& Tegmark 1998, hereafter \\cite{EHT}), we explore the potential of combining redshift surveys and CMB anisotropy data for the purpose of parameter estimation. Here, we focus on the dramatic improvement possible in the measurement of $H_0$ and $\\om$.  Neither data set yields tight limits by itself, yet together they could yield errors better than 5\\% on $H_0$ and 10\\% on $\\om$.  The key to this improvement is the presence of features in the matter power spectrum on scales exceeding $60 h^{-1}\\mpc$.  With a non-negligible baryon fraction, the acoustic oscillations that exist before recombination are imprinted not only on CMB anisotropies but also on the linear power spectrum (\\cite{Hol89} 1989, \\cite{Hu96} 1996, \\cite{Eis98a}\\ 1998a). CMB anisotropies accurately calibrate their characteristic length scale; measurement of this standard ruler in the redshift survey power spectrum yields $H_0$.  With this added information, the CMB returns a significantly more precise measure of $\\om$. ",
        "conclusions": "\\label{sec:Discussion}  Detection of acoustic oscillations in the matter power spectrum would be a triumph for cosmology, as it would confirm the standard thermal history and the gravitational instability paradigm. Moreover, because the matter power spectrum displays these oscillations in a different manner than does the CMB, we would gain new leverage on cosmological parameters. In particular, we have shown in this {\\it Letter} that the combination of power spectrum measurements from a galaxy redshift survey with anisotropy measurements from CMB satellite experiments could yield a precision measurement of $H_0$ and $\\om$.  The potential measurement of $H_0$ and $\\om$ depends critically on the ability of the redshift survey to detect the baryonic features in the linear power spectrum.  The best possible error bars are a strong function of the baryon fraction but are surprisingly good even if the fraction is $\\sim\\!10$\\%, roughly the minimum implied by cluster observations.  For such cases, the fractional limits achievable with the SDSS are 5\\% for $H_0$ and 10\\% for $\\om$ if only the first acoustic peak in $P(k)$ is detected.  Detecting the smaller-scale peaks could allow an additional factor of 3 refinement; the exact limits would depend upon the scale at which non-linear effects smooth out the power spectrum.  The results depend only mildly on the details of the CMB experiment:  we find only slight gains as our presumed CMB data set improves from MAP without polarization to Planck with polarization.  While we have quoted numbers for SDSS, it is possible that the 2dF survey will be able to make significant progress on the detection of features in the power spectrum on very large scales. Unfortunately, the hints of excess power on $100\\hmpc$ scales are not likely to be due to baryons (\\cite{Eis98c} 1998).  We have treated the galaxy power spectrum assuming linear bias on large scales.  There is some theoretical motivation for this (\\cite{Sch98}\\ 1998); moreover, if bias tends towards unity as structure grows (\\cite{Fry96} 1996; \\cite{Teg98a}\\ 1998), then scale dependences in the bias at the time of formation will be suppressed. Most importantly, this method of measuring $H_0$ and $\\om$ depends upon extracting an oscillatory feature from the power spectrum. While one cannot prove that scale-dependent bias should be monotonic on the largest scales, this seems more likely than an oscillation! Finally, the assumption of linearity can be tested by constructing the power spectrum with different types of galaxies (e.g., \\cite{Pea97}\\ 1997); future redshift surveys will allow this to be done on very large scales with good statistics.  The method proposed here yields $H_0$ independent of local distance measurements and $\\om$ without the complications inherent in dynamical methods.  In that, it is free of many confusing astrophysical problems.  On the other hand, it does depend upon restricting oneself to a class of models with observable acoustic oscillations in both CMB anisotropies and the galaxy power spectrum. This assumption will be definitively tested from the data itself.  If the method described in this {\\it Letter} can yield tight constraints on $H_0$ and $\\om$, it will then be very important to compare these to other measurements of these quantities.  In the coming decade, there will be a number of paths toward a precision measure of $H_0$, such as the local distance ladder (e.g., \\cite{Fre98}\\ 1998), gravitational lensing (e.g., \\cite{Bla96}\\ 1996), and the S-Z effect (e.g., \\cite{Coo98}\\ 1998). Similarly, good estimates of $\\om$ may be possible from velocity fields (e.g., \\cite{Dek97}\\ 1997), cluster evolution (\\cite{Car97a}\\ 1997a; \\cite{Bah97}\\ 1997), and $M/L$ measurements (e.g., \\cite{Car97b}\\ 1997b). If the results from these diverse sets of measurements are found to agree, we will have a secure foundation upon which to base our cosmology.   Acknowledgements: Numerical power spectra were generated with CMBFAST (\\cite{Sel96b} 1996). We thank Martin White for useful discussions. D.J.E.\\ is supported by a Frank and Peggy Taplin Membership; D.J.E.\\ and W.H.\\ by NSF-9513835; W.H.\\ by the Keck Foundation and a Sloan Fellowship; M.T.\\ by NASA through grant NAG5-6034 and Hubble Fellowship HF-01084.01-96A from STScI, operated by AURA, Inc. under NASA contract NAS4-26555."
    },
    "9805/astro-ph9805149_arXiv.txt": {
        "abstract": "A new method of $T_{\\rm eff}$ determination for CP stars is proposed. The method is based on the fact that the slope of the energy distribution in the Balmer continuum near the Balmer jump is identical for \"normal\" main sequence stars and for CP~stars with the same $T_{\\rm eff}$. It is shown that the $T_{\\rm eff}$ of CP stars derived by this method are in good agreement with those derived by other methods. ",
        "introduction": "A review of the various methods of effective temperature determination for chemically peculiar (CP) stars shows once more the difficulty of deriving the $T_{\\rm eff}$ of these stars. If one uses methods taking into consideration the blanketing effect, the temperature obtained is close to the effective one. In the infrared flux method (IRFM) first proposed by Blackwell \\& Shallis~(1977), a monochromatic flux is measured in the infrared region to minimize the blanketing effect. The method proposed by Stepien \\& Dominiczak~(1989) takes into account the blanketing effect as well. The photometric methods may be useful, since it is possible to apply a correction to the color (or model) temperature and to give relatively good estimates of $T_{\\rm eff}$ (Hauck \\& North~1993). Another way is to use an observational parameter which is not affected by peculiarities and can be applicable both to the \"normal\" main sequence stars and to the CP stars. The Balmer continuum slope near the Balmer jump ($\\it\\Phi_{\\rm u}$) can be a useful tool for determination of $T_{\\rm eff}$ for CP stars (Sokolov 1995). The determination of the effective temperatures of CP stars using the $\\it\\Phi_{\\rm u}$ is discussed. ",
        "conclusions": "The calibration curve described above was applied to 50 CP2 stars from the catalog of stellar spectrophotometry (Adelman et al. 1989) and to 18 CP2 stars from the Pulkovo spectrophotometric catalog of bright stars (Alekseeva et al. 1996). To test the validity of the proposed method of determination of $T_{\\rm eff}$ for CP2 stars, the temperatures derived from $\\it\\Phi$$_{\\rm u}$ were compared with those derived from the IRFM, from the method of Stepien~\\&~Dominiczak~(1989) and from the (B2-G) color index of Geneva photometry.  In the literature we found five papers concerning CP2 stars for which the effective temperature is derived using the IRFM. The values of $T_{\\rm eff}$ derived from $\\it\\Phi$$_{\\rm u}$ are compared with those obtained by IRFM for 13 common stars (see Fig.~1a). One can see that the agreement appears to be very good. So, the mean effective temperature difference is $\\Delta T_{\\rm eff}~=~ T_{\\rm eff}$($\\it\\Phi$$_{\\rm u}$)~-~$T_{\\rm eff}$(IRFM)~ =~41$\\pm$127~K, with a linear correlation coefficient {\\it r}~=~0.972, and $\\alpha$~=~0.904 for the slope of the regression line of $T_{\\rm eff}$($\\it\\Phi$$_{\\rm u}$) versus $T_{\\rm eff}$(IRFM). \\begin{figure} [t] \\centering{ \\vspace{-1.1cm} \\hspace*{-1.1cm} \\vbox{\\psfig{figure=sokolovf.ps,angle=270,height=6.0cm}}\\par } {\\small Fig. 1: Comparison of $T_{\\rm eff}$ derived from $\\it\\Phi_{\\rm u}$ with those derived by infrared flux method - (a), by the method proposed by Stepien \\& Dominiczak - (b) and from (B2-G) color index of Geneva photometry - (c).} \\end{figure}  Stepien~\\&~Dominiczak (1989) proposed a new method to determine the effective temperatures of CP2 stars. In order to have a wider sample of stars for the comparison, this method was applied to the model temperatures ($T_{\\rm M}$) obtained by Adelman~(1985). The value of $T_{\\rm M}$ was calculated as the average of $T$(PC) and $T$(BJ) for all stars of Table~2 of Adelman~(1985). After that, this mean value of $T_{\\rm M}$ was corrected for the blanketing effect by using Eq.~12 from the paper of Stepien~\\&~Dominiczak. The resulting temperatures ($T_{\\rm eff}$(S\\&D)) are compared with $T_{\\rm eff}$ derived from $\\it\\Phi$$_{\\rm u}$. Figure~1b gives a plot of $T_{\\rm eff}$($\\it\\Phi$$_{\\rm u}$) versus $T_{\\rm eff}$(S\\&D) for 47 common stars. Basically, there are no systematic differences between the two sets of data, as confirmed by the following results: $\\Delta$$T_{\\rm eff}$~=~38$\\pm$69~K, {\\it r}~=~0.952, $\\alpha$~=~0.965.  In order to estimate photometrically the effective temperatures of CP2 stars the (${\\it B}{\\rm 2-}{\\it G}$) color index was used. Figure~1c gives a plot of the $T_{\\rm eff}$ derived from $\\it\\Phi_{\\rm u}$ versus $T_{\\rm eff}$(${\\it B}{\\rm 2-}{\\it G}$) for the 59 common stars. One can see that there is no systematic difference between the two sets of data, though the scatter of the points on Figure~1c is rather high (up to 1000~K), especially for the stars with $T_{\\rm eff}>9500$~K. For the stars in our sample we have $\\Delta T_{\\rm eff}$~=~102$\\pm$76~K, {\\it r}~=~0.938, and $\\alpha$~=~0.975.  Generally, there is no significant systematic difference between the temperatures derived from $\\it\\Phi$$_{\\rm u}$ and those derived from fluxes by other methods. The temperature calibration derived for B, A and F main sequence stars is applicable to CP2 stars as well. The temperature derived from $\\it\\Phi$$_{\\rm u}$ for CP2 stars may be identified with their effective one, because the influence of the stars's peculiarity on the Balmer continuum slope near the Balmer jump is negligible. In our study only CP2~stars were used, but this method can be extended to other types of CP~stars, for which the blanketing effect is less pronounced: for them, the temperature derived from $\\it\\Phi_{\\rm u}$ should then be even closer to the effective one."
    },
    "9805/astro-ph9805194_arXiv.txt": {
        "abstract": "I present a specific worked example of evolution through inflation to the initial conditions for an isocurvature CDM model for structure formation. The model invokes three scalar fields, one that drives power law inflation, one that survives to become the present-day CDM, and one that gives the CDM field a mass that slowly decreases during inflation and so ``tilts'' the primeval mass fluctuation spectrum of the CDM. The functional forms for the potentials and the parameter values that lead to an observationally acceptable model for structure formation do not seem to be out of line with current ideas about the physics of the very early universe. I argue in an accompanying paper that the model offers an acceptable fit to main observational constraints. ",
        "introduction": "This paper with its companion (Peebles 1998; hereafter Paper II) is the latest in a series of studies of isocurvature models for structure formation. (Earlier papers may be traced back from Peebles~1997). There are two motivations for this work. First, it is important to know that there is an observationally viable and theoretically not unreasonable alternative to the commonly discussed adiabatic cold dark matter (ACDM) model for structure formation. As long as there are alternatives it demonstrates that we do not have an established standard model for the early universe, that there is more to be discovered than tighter constraints on parameters and functional forms for a potential. Second, the isocurvature models seem to be better adapted to galaxy formation at high redshift, a condition I find attractive. This paper is meant to demonstrate that initial conditions for the isocurvature case, which I shall call ICDM, can be given an explicit and not unreasonable basis in a physical model for inflation. In Paper~II I argue that the model offers an acceptable fit to the observational constraints that can be applied without the use of numerical simulations.  There is considerable activity in the study of the rich variety of ideas for a specific model for inflation motivated by current ideas in particle physics (as reviewed by Randall 1997). I make use of functional forms for potentials that commonly appear in these discussions, but the motivation is to obtain a specific model that could have come out of inflation to compare to the rich suite of observational evidence. The two cultures --- based in particle physics and astronomical phenomenology --- will meet if it becomes possible to establish that the narrow range of models from fully acceptable physics overlaps the narrow range of models that are observationally acceptable.  The dynamical actors in the structure formation model to be discussed here and in Paper~II are the same as in the family of adiabatic CDM (ACDM) models --- baryons, radiation (the CBR) with initially homogeneous entropy per baryon, cold dark matter, and three families of neutrinos --- and in about the same amounts. In the adiabatic version, a scale-invariant departure from a homogeneous primeval mass distribution has power spectrum $P\\propto k$. In the ICDM model a scale-invariant spectrum of the distribution of the CDM would be $P\\propto k^{-3}$, while the net mass density is homogeneous. Early discussions of inflation accepted the proposition that adiabatic and isocurvature scenarios are equally well motivated (eg. Steinhardt \\& Turner 1983, Linde 1985; Seckel \\&\\ Turner 1985), but when Efstathiou \\&\\ Bond (1986) demonstrated that a scale-invariant isocurvature model violates the bound on the CBR anisotropy attention naturally turned to the scale-invariant adiabatic case. The COBE detection of the CBR anisotropy (Smoot et al. 1992) showed that if the universe were Einstein-de~Sitter the ACDM spectrum with a reasonable bias would have  to be ``tilted'' from scale-invariance, an arguably natural adjustment of the inflation picture (eg. Crittenden {\\it et al.} 1993). Tilt is not needed if the mean mass density (excluding a term in the stress-energy tensor that acts like a cosmological constant) is well below the Einstein-de~Sitter value, but the precedent has been set: consider tilting the ICDM spectrum to fit the observations. Models for inflation that tilt the spectrum in the wanted direction have been discussed by Kofman \\&\\ Pogosyan (1988), Salopek, Bond, \\&\\ Bardeen (1989), and Linde \\&\\ Mukhanov (1997), and a model fitted to the observational constraints has been presented in Peebles~(1997). Here and in Paper~II I present a more detailed discussion along the lines of this last paper.  For the purpose of displaying a specific worked example I adopt definite values of the parameters in the cosmology and the structure formation model and derive from them the parameters in the inflation model, the latter including the needed initial conditions early in inflation. The parameters will have to be reconsidered with each improvement of the observations, of course. The hope is that such adjustments driven by advances in the observations and physics may back us into the corner of model and parameter space that is a reasonable approximation to reality.  The elements of the inflation scenario are presented in \\S 2. Section 3 starts from adopted values of the parameters in the models for cosmology and structure formation and presents derived values for the parameters in the inflation model. Except where otherwise indicated units are chosen so $\\hbar = 1=c$, and I follow the notation in Peebles~(1993), \\S 17. ",
        "conclusions": "In the adiabatic CDM family of models (that I have termed ACDM) the cold dark matter particles can have been produced out of entropy at any time after inflation ended, a broadly general situation. The isocurvature CDM (ICDM) model is more specific --- the CDM is a remnant of a field present during inflation --- and hence arguably less likely in the absence of evidence for or against the nature of the CDM. This consideration may become more compelling when we have better experimental constraints on the dark matter.  Assessments of the relative degree of ``fine tuning'' of the adiabatic and isocurvature inflation models also seem to be of doubtful significance at this stage of our understanding, so I note only four aspects of the ICDM model. First, there is precedent in the literature for the CDM potential  $V(\\psi ,\\phi )$ in equation~(\\ref{eq:V}). The same functional form, applied in a different context, appears in the Hybrid Inflation model (Copeland et al. 1994; Linde 1994), and similar forms are not uncommon (eg. Kofman \\& Linde 1987; Hodges et al. 1990; Randall, Solja\\u ci\\'c \\& Guth 1996). Second, the conditions on an acceptable set of dimensionless parameters for the ICDM model are $\\gamma\\sim\\beta\\ll\\epsilon\\sim 0.1$. These do not seem unduly severe or artificial. Third, the model requires that the initial value of the field $\\psi$ be large enough that $\\psi$ approaches the solution in  equation~(\\ref{eq:psiearly}) well before field fluctuations are frozen on scales we can observe, and that the initial value of $\\phi$ be small enough that $\\psi$ can drive it to zero mean value before the field is squeezed on scales of interest. I do not know how to judge whether these conditions are likely outcomes of the physical situation prior to inflation. Finally, an acceptable set of values of the characteristic masses that appear in the model is \\beqa M \\sim 10^{19}\\hbox{ GeV},\\qquad T_r \\sim  10^{13}\\hbox{ GeV},\\nonumber\\\\ \\mu \\sim  10^{9}\\hbox{ GeV},\\qquad m \\sim  1 \\hbox{ GeV}. \\eeqa The broad range of values is impressive, but so is the range of measured characteristic masses in particle physics and those that appear in physics-motivated models for inflation (Randall 1997).  The greatly tightened constraints from observational programs in progress are going to make it much harder to invent acceptable structure formation models such as ACDM and ICDM from largely theoretical and aesthetic considerations. Some already known member of the CDM family --- in which I would include ACDM and ICDM --- may survive the precision tests in progress and by its success compel acceptance as a truly valid approximation to reality. If this does not happen, perhaps we will be lucky enough to see in the results from observations and particle physics guidance to the formulation of more promising models.  I conclude that the ICDM model has as valid a pedigree from accepted ideas about the very early universe as might be expected for a model that is motivated by astronomical phenomenology rather than physics. The observational situation is discussed in Paper~II."
    },
    "9805/astro-ph9805061_arXiv.txt": {
        "abstract": "Abundances of Mg, Ca, Sc, Cr, Fe, and Ni are derived for A stars of five nearby open clusters of various ages using high resolution spectroscopy. We point out a correlation between the abundance of Ca and that of Sc, suggesting that the abundance anomalies of these elements arise from the same physical process. Pronounced Am patterns are rather found in the oldest cluster stars whereas younger targets show weaker Am anomalies and atypical patterns for some of them. ",
        "introduction": "The abundance anomalies of the Am stars (and, as a rule, of most of the Chemically Peculiar stars) are nowadays broadly considered to be produced by microscopic diffusion. Indeed, the observed anomalies are consistent with the computed diffusion velocities for many elements. Nevertheless, the detailed understanding of the Am phenomenon requires more thorough studies since the stratification process is time-dependent and affected by the other physical processes at play in the stellar medium. Calcium and scandium are elements of special interest since their deficiencies are usually used to detect Am stars (Conti 1970; Preston 1974). The evolution of the abundances of Ca and Sc in the superficial layers of an A star has been computed by Alecian (1996) assuming no helium convection zone, as suggested by the diffusion model for Am stars. His results show that the behaviour of both elements are strongly dependent on the strength of the large scale motions introduced in the computations, namely, a mass-loss and the extension of the superficial mixing zone beneath the convective zone. In some cases, phases of overabundance occur when the star is around 10 million years old. So, the classification criterion of Am stars based on the calcium or scandium deficiency is questionable for young A stars.  Abundance determinations in young main-sequence A stars can constrain such computations. Open cluster stars are the best candidates since their age is known with much greater accuracy than for field stars. Few abundance studies in open clusters have been done up to now (see Burkhart \\& Coupry 1997 and references therein). The use of electronic detectors (Reticon and CCDs) and the subsequent improvement in the quality of the spectra made possible the study of the lithium abundance and, by the way, renewed the interest for open clusters: observations of A-type stars were carried out by Boesgaard (1987) in Coma, and by Burkhart \\& Coupry (1989; 1997) in the Hyades and Pleiades.  This paper summarizes the work undertaken in collaboration with G. Alecian and C. Burkhart concerning the abundance of calcium and scandium in open cluster A stars. More details are provided in Hui-Bon-Hoa et al. (1997) and Hui-Bon-Hoa \\& Alecian (1998). ",
        "conclusions": "\\begin{figure}[ht] \\psfig{figure=huifig2.eps,width=10cm} \\caption{Abundance patterns for the stars of: \\textbf{a} $\\alpha$ Per; \\textbf{b} the Pleiades; \\textbf{c} Coma; \\textbf{d} the Hyades; \\textbf{e} Praesepe; \\textbf{f} field stars. The metallicity of the Hyades is indicated by a dash-dotted line.} \\end{figure} As usual, the script [X] for any quantity X means $\\mathrm{log\\,(X)_{*}\\,-\\,log\\,(X)_{\\sun}}$. In Fig.~1, we can see a loose correlation between [Ca/H] and [Sc/H] for our cluster stars. This suggests that the anomalies of these two elements come from the same physical process. Besides, the more pronounced deficiencies (left part of the graph) are found in the oldest stars (members of the Hyades and Praesepe). Younger targets (in $\\alpha$ Per, Pleiades) show weak underabundances or marginal overabundances (right part of the graph).  The abundance patterns for our sample stars are shown in Fig.~2. In ordinates are the logarithmic differences between the abundance value in the star of concern and the solar one for Mg, Ca, Sc, Cr, Fe, and Ni. The corresponding points are linked for each star. An arrow means upper limit and error bars ended by arrows denote very uncertain values. We can see that the Am pattern is well-marked in the oldest cluster stars (Hyades and Praesepe) of our sample as well as for the field Am stars. In the youngest clusters ($\\alpha$ Per, Pleiades), the Am pattern is almost absent and several objects show atypical patterns with marginal overabundances of Ca and/or Sc. This would suggest that stars of these clusters are in transient phases of the stratification process.  The marginal overabundances of Ca and Sc in $\\alpha$ Per and Pleiades are not strong enough to confirm the phase of overabundance predicted by Alecian (1996). Either the youngest clusters of our sample are already too old and their stars have passed the phase of overabundance or this phase does not exist. In this last case, the extension of the mixing zone should be less than one pressure scale height and the mass-loss rate about $\\mathrm{10^{-14}M\\sun/yr}$.  These conclusions need to be confirmed by observations of more clusters and younger targets. Also, elements that could help a better understanding of the Am phenomenon deserve to be studied (rare earths for instance)."
    },
    "9805/gr-qc9805026_arXiv.txt": {
        "abstract": "We perform simulations of relativistic binary stars in post-Newtonian gravity to investigate their dynamical stability prior to merger against gravitational collapse in a tidal field. In general, our equations are only strictly accurate to first post-Newtonian order, but they recover full general relativity for spherical, static stars. We study both corotational and irrotational binary configurations of identical stars in circular orbits.  We adopt a soft, adiabatic equation of state with $\\Gamma = 1.4$, for which the onset of instability occurs at a sufficiently small value of the compaction $M/R$ that a post-Newtonian approximation is quite accurate. For such a soft equation of state there is no innermost stable circular orbit, so that we can study arbitrarily close binaries. This choice still allows us to study all the qualitative features exhibited by any adiabatic equation of state regarding stability against gravitational collapse.  We demonstrate that, independent of the internal stellar velocity profile, the tidal field from a binary companion stabilizes a star against gravitational collapse. ",
        "introduction": "Binary neutron stars are known to exist and for some of these systems in our own galaxy (including PSR B1913+16 and B1534+12), general relativistic effects in the binary orbit have been measured to high precision~\\cite{tamt93,tw89}. Interest in binary neutron stars has been stimulated by the prospect of future observations of extragalactic systems by gravitational wave interferometers like LIGO\\cite{LIGO}, VIRGO\\cite{VIRGO}, TAMA\\cite{TAMA} and GEO\\cite{GEO}. Binary neutron stars are among the most promising sources of gravitational waves for these detectors, and therefore it is important to predict theoretically the gravitational waveform emitted during the inspiral and the final coalescence of the two stars. Interest in these systems also arises on a more fundamental level, since the two-body problem is one of the outstanding unsolved problems in classical general relativity.  Considerable effort has gone into understanding binary neutron stars. Most of this work has been performed within the framework of Newtonian and post-Newtonian gravity (see, e.g.,~\\cite{BCSSTc} for a review and list of references). General relativistic treatments are currently only in their infancy. Recently, Wilson, Mathews and Marronetti~\\cite{wilson} (hereafter WMM) reported results obtained with a relativistic hydrodynamics code. Their code assumed several simplifying physical and mathematical approximations.  Their results suggest that the central densities of the stars increase as the stars approach each other and that massive neutron stars, stable in isolation, individually collapse to black holes prior to merger. WMM therefore find that in general relativity, the presence of a companion star and its tidal field tend to destabilize the stars in a binary system.  This conclusion is contrary to what is expected from Newtonian~\\cite{lrs93}, post-Newtonian~\\cite{lai,lrs97,wiseman}, perturbative~\\cite{brady} and matched asymptotic expansion~\\cite{EEF,KIP} treatments of the problem. Constructing self-consistent, fully relativistic initial data for two neutron stars in a circular, quasi-equilibrium orbit does not show any evidence of this ``crushing effect'' either~\\cite{BCSSTa}. Moreover, applying energy turning-point methods to sequences of these initial data suggests that inspiraling neutron star binaries are {\\em secularly} stable all the way down to the innermost stable circular orbit~\\cite{BCSSTb}. To summarize, most researchers currently believe that the maximum allowed rest mass of neutron stars in close binaries is {\\em larger} than in isolation, and that their central density is {\\em smaller} than in isolation. If there exists any destabilizing, relativistic effect at high post-Newtonian order, then this effect is much smaller than the dominating stabilizing effect of the tidal field.  However, to date, the only fully {\\em dynamical} treatment of the problem in general relativity -- that of WMM -- reports a star-crushing effect. In this paper, we perform a new, fully dynamical simulation for binary stars in post-Newtonian gravity.  We use a formalism in which (1) all first post-Newtonian terms are taken into account, and (2) sufficient nonlinearity is retained, so that spherical, static stars satisfy the fully general relativistic equations exactly. As explained in section II below, this formalism is very suitable for studying binary neutron stars. We study relativistic effects in binary stars with $M/R \\ll 1$, where $M$ and $R$ are typical values of the stellar mass and radius, so that a post-Newtonian treatment is completely adequate.  By performing a fully dynamical calculation, we can relax various constraints assumed in previous treatments. For example, Wiseman assumed the stars to remain spherically symmetric~\\cite{wiseman}, Baumgarte {\\em et al.}~\\cite{BCSSTc} assumed the binary stars to be corotating, and Thorne~\\cite{KIP} assumed the stars' orbital separation to be much larger than the stars' radius. Here, we relax all these assumptions and study tidally deformed stars, both corotational and irrotational, at arbitrarily small separations. We still find that the presence of the tidal field of a companion star tends to stabilize neutron stars against catastrophic collapse.  To establish the stability of binary stars against collapse, we construct quasi-equilibrium initial data for identical binary neutron stars in a close, circular orbit. The idea is to show whether stars in a binary formed from the inspiral of objects which are stable in isolation remain stable at close separation.  Our models have rest masses near the maximum allowed rest mass for spherical stars in isolation and thus provide the best candidates for collapse if the tidal field is destabilizing (stars with rest masses well below the maximum allowed value are unambiguously stable).  In order to demonstrate that these stars are dynamically stable, we need to locate the onset of instability in the binary, and compare it with the onset of instability for isolated stars. Since the shift is fairly small, a very careful treatment with high numerical accuracy is necessary. We detail our method of locating the onset of instability in section IV.  The paper is organized as follows: In section II, we present the post-Newtonian formalism adopted in this paper.  We calibrate our code in section III by locating the analytically known onset of radial instability of relativistic spherical stars against gravitational collapse~\\cite{Chandra}. In section IV, we study the dynamical stability against gravitational collapse of close binary stars, and briefly summarize our results in section V. ",
        "conclusions": "We perform post-Newtonian, dynamical simulations of close binaries in circular orbit. In particular, we study the stability of the individual stars against gravitational collapse in both corotational and irrotational systems containing stars of equal mass.  We have chosen a soft, adiabatic equation of state with $\\Gamma = 1.4$, for which there is no innermost stable circular orbit, so that the binary {\\em orbit} is stable even when the stars are in contact, and for which the onset of instability for a spherical star in isolation occurs at a very small value of the compaction $M/R$. We can therefore study the individual stars' stability properties in near contact binaries, for which the tidal effects are strongest, and in a regime in which a post-Newtonian approximation is very accurate.  We do not find any crushing effect as reported by WMM~\\cite{wilson}. In contrast, the maximum density in both corotational and irrotational binaries is smaller than that of spherical stars in isolation.  We find that stars in binaries can support more mass than in isolation. Moreover, all stars that are stable against radial perturbations in isolation, will also be dynamically stable when put into a binary.  All these results are in complete agreement with, for example, the findings of Baumgarte {\\em et~al.}~\\cite{BCSSTa,BCSSTb}, Flanagan~\\cite{EEF}, and Thorne~\\cite{KIP}.  For the most part, their discussions rigorously address {\\em secular} stability only. Several different arguments can be invoked to suggest that {\\em secularly} stable binaries are also {\\em dynamically} stable, but this is strictly proven only in Newtonian theory (see also~\\cite{lrs97}).  Our dynamical calculations reported in this paper are the first to directly confirm dynamical stability, at least within our post-Newtonian approximation.  We compare, in a near-equilibrium approximation, corotational and irrotational binary models.  As expected, stars in corotational binaries can support slightly more mass than in irrotational binaries, but apart from these small differences we do not find any qualitative difference in their radial stability properties.  A more rigorous treatment will require the construction of post-Newtonian, irrotational equilibrium binary models for initial data.  Since our computations have been performed in nondimensional units, our results apply not only to neutron star binaries, but also to binaries of white dwarfs and supermassive stars. In fact, the equations of state of massive white dwarfs (ideal degenerate, extremely relativistic electrons) and supermassive stars (radiation $\\gg$ thermal pressure) are closely approximated by the value $\\Gamma = 1.4$ that we have adopted.  These binaries may be important low frequency gravitational wave sources for future space-based gravitational waved detectors, like LISA."
    },
    "9805/astro-ph9805127_arXiv.txt": {
        "abstract": "The $I$-band brightness $M_I$ of clump stars is a possible distance indicator for stellar populations. Investigations have shown that $M_I$ is almost insensitive to the $(V-I)$ colour within the clump. Based on this, it was assumed that $M_I$ was insensitive to age and composition of the stellar population and therefore an ideal standard candle, which could be calibrated with local clump stars, whose absolute brightness is known from {\\em Hipparcos} parallaxes. This resulted in a distance to the LMC about 15\\% shorter than usually determined.  In the present paper we show that with a population synthesis approach we can reproduce the constancy of $M_I$ with colour for the local {\\sl Hipparcos} clump sample. Nevertheless, $M_I$ is not a constant among different populations, but depends on metallicity. As a result, the determined distance modulus to the LMC of $18.28\\pm0.18$ mag is in better agreement with standard values. This resolves, at least partially, the controversial result obtained by the assumption of a universal value for $M_I$.  Particularly remarkable is our prediction that stars slightly heavier than the maximum mass for developing degenerate He cores, \\Mhef, should define a secondary, clumpy structure, about 0.3~mag below the bluest extremity of the red clump. Both features are well separated in the \\mi\\ vs.\\ \\vi\\ diagram of metal-rich stellar populations. Indeed, this secondary clump can be clearly identified in the {\\em Hipparcos} database of stars with reliable $I$ photometry and parallax errors smaller than 10\\%.  Since the stars in this feature should represent a narrow range of masses, their mass determination, e.g.\\ by the use of binary systems, can provide information about the efficiency of convective overshooting from stellar cores.  Our investigation demonstrates that the RGB clump cannot be used as a distance indicator without proper knowledge and modelling of the population under investigation. In addition, there remain unsolved problems in the models, such as correct bolometric corrections and colour transformations. \\vspace{2.0cm}  \\keywords stars: evolution -- Hertzsprung-Russell (HR) diagram -- \\comment{stars:horizontal branch --} \\comment{stars: luminosity function, mass function --} solar neighbourhood -- Galaxy: stellar content -- galaxies: distances and redshifts -- Magellanic Clouds ",
        "introduction": "\\label{sec_intro}  The red giant clump is an easily recognizable feature in many colour-magnitude diagrams (CMD). It consists of stars of rather low mass, which are currently undergoing their central helium burning. Physically, it is identical to the horizontal branch in globular clusters made up by less massive and more metal-poor stars.  Paczy\\'nski \\& Stanek (1998) recently determined the mean absolute brightness ($M_I$) of local clump stars making use of {\\em Hipparcos} parallaxes. They selected the stars from the {\\em Hipparcos} catalogue (ESA 1997) with parallax measured to better than 10\\%. Then they determined the mean \\mi\\ magnitude of the clump stars, $\\mimax=-0.185\\pm0.016$, by fitting a gaussian-like curve to the magnitude distribution of 657 stars inside the box defined by $0.8<(\\vi)<1.25$, $1.1>\\mi>-1.4$. Selecting stars in two colour subintervals, $0.8<(\\vi)<1.0$ and $1.0<(\\vi)<1.25$, they found values of \\mimax\\ formally indistinguishable from the mean. The same independence of the apparent $I$-band brightness \\imax\\ in the \\vi\\ colour was also found for the clump stars in Baade's Window, over the entire $0.8<(\\vi)<1.4$ interval. Based on these results, Paczy\\'nski \\& Stanek (1998) assumed \\mimax\\ to be independent of the properties of the observed stellar populations, at least for stars with $0.8<(\\vi)<1.25$. Under this assumption, the galactocentric distance was determined by comparing the apparent $I$ magnitude of Baade's Window clump stars, \\imax, with the reference value obtained from the {\\em Hipparcos} sample.  The same process was repeated for stars in M31 (Stanek \\& Garnavich 1998), the Magellanic Clouds (Udalski et al.\\ 1998), and the LMC alone (Stanek, Zaritski \\& Harris 1998). In all these cases, the mean \\imax\\ value was found to be nearly independent of the \\vi\\ colour sampled.  For both the bulge and M31, the distances obtained were essentially in agreement with those obtained from other methods. In the case of the Magellanic Clouds, however, distances turned out to be significantly shorter than those derived by Cepheid stars and commonly accepted: Udalski et al.\\ (1998) and Stanek, Zaritsky \\& Harris (1998) find that the relatively well-settled distance modulus of the LMC of about $\\dmo=18.5$~mag could be overestimated by 0.45~mag, or a factor of about 15\\% in distance, with respect to the real one. Since clump stars provide a one-step distance, this lead them to claim that {\\em other} distance indicators (such as Cepheids) were erroneous and should be re-investigated.  Since Magellanic Cloud stars have mean metallicities well below those of local stars and of stars that define the clump in the bulge and in M31 over the $0.8<(\\vi)<1.25$ interval, the suspicion that metallicity effects may be causing the discrepant results, is obvious. According to the above authors, theoretical models show weak dependence of \\mbol\\ and \\mi\\ (for clump stars) on either age or chemical composition. That is a surprising statement, since models in the literature (e.g.\\ Sweigart \\& Gross 1976; Seidel, Demarque \\& Weinberg 1987; Bertelli et al.\\ 1994, and references therein) show that clump stars of different masses and metallicities may have luminosities differing of up to 0.5~mag (see \\refsec{sec_stars}).  The main support for using \\mimax\\ as a standard candle comes, instead, from the observed independence of \\mimax\\ on \\vi\\ (and hence supposedly metallicity).  Cole (1998) accordingly proposed a revision of the clump distance to the LMC, based on the mean age and metallicity differences between the LMC and local stars. Considering these differences, and making use of the (theoretical) dependence of clump magnitude on both parameters, he shows that the LMC red clump should be about 0.32~mag brighter than the local disk one, and obtains a distance modulus of $18.36\\pm0.17$~mag to the LMC. Beaulieu \\& Sackett (1998) obtained a good model for the LMC clump by adopting a distance modulus of $18.3$ from isochrone fitting.  However, this kind of first-order explanation seems to be in contrast with the observation that \\mimax\\ is almost constant at different \\vi\\ colours in different stellar systems. Since the clump colour is usually considered as indicative of metallicity, the \\mimax\\ constancy with colour strongly suggests that it is, in reality, inpedendent of metallicity. The question arises if theoretical models can explain this constancy in a composite stellar population, but simultaneously predict brighter clumps at lower metallicities, as required to obtain the usual LMC distance. That is one of the questions we are going to address in the following.  In this work, we intend to examine the fundamentals of the red clump method, with the aid of evolutionary models and isochrone calculations which should represent, as far as possible, the standard theoretical predictions for the behaviour of clump stars in stellar populations of different ages and metallicities. In \\refsec{sec_stars} we briefly describe the theoretical stellar models and isochrones we used, and the general predictions for the \\mi\\ magnitude and colours of the clump stars; in \\refsec{sec_clump} we show that the models predict a fine structure of the red clump, which is indeed observed in the \\mi\\ vs.\\ \\vi\\ diagram from {\\em Hipparcos} data. In the light of these results, in \\refsec{sec_lf} we show how the mean \\mi\\ for a composite stellar population can be, at the same time, almost independent of the \\vi\\ colour sampled, and dependent on the mean metallicity of the population observed (being in general brighter for lower metallicities); this behaviour, together with our present knowledge on the star formation history of the LMC, helps to put the distance modulus of this galaxy closer to the more traditional value of about 18.5~mag (\\refsec{sec_lmc}).  Finally, in \\refsec{sec_comments} we comment on the accuracy in distance determinations that can be obtained from the red clump method. ",
        "conclusions": ""
    },
    "9805/astro-ph9805311_arXiv.txt": {
        "abstract": "We present optical (6500-9200\\AA) spectroscopy of eight cool dwarfs detected in a 231 square degree ``Mini-survey'' of the Deep NEar Infrared Survey (DENIS) data. We are able to confirm that the spectral types derived from the Mini-survey infrared spectroscopy are meaningful. We provide a spectral sequence which extends beyond the M-dwarf range and into the proposed ``L'' class of dwarfs. The dominant spectral features in the optical for these L-type dwarfs are resonance lines of Cs\\,I and molecular band heads of CrH and FeH. The other dominant feature in these L-type spectra is a broad 600\\,\\AA\\ absorption dip centered on 7700\\,\\AA, which we identify with extremely strong (equivalent width $\\sim$ several hundred \\AA) absorption associated with the 7664,7698\\,\\AA\\ resonance doublet of K\\,I. We find that model atmospheres which include the effects of molecular condensation without dust opacity (to simulate rapid gravitational settling of dust grains) produce significantly better agreement with observed optical spectra for L-type dwarfs, than models including dust opacity. This suggests gravitational settling of dust grains  plays an important role in L-dwarf photospheres. The extreme strength of the K\\,I resonance doublet, and disappearance of TiO and VO, and the consequent dominance of CrH and FeH in L-dwarf spectra offer considerable prospects as sensitive effective temperature diagnostics, even at low spectral resolution. ",
        "introduction": "The history of the study of very-low mass (VLM) stars and brown dwarfs has shown again and again that when new technologies are implemented, new objects with previously unseen properties are discovered. Examples include the use of wide-field photographic surveys and digital scanning machines to discover the first VLM stars (Luyten 1979; Reid \\& Gilmore 1981; Probst \\& Liebert 1983; Bessell 1991; Irwin \\etal\\ 1991); the use of infrared spectroscopy to discover the importance of H$_2$O absorption in VLM stars (Berriman \\& Reid 1987); and, the use of infrared imaging, adaptive optics and coronography to discover the proto-typical cold brown dwarf Gl\\,229B, which further confirmed the importance of CH$_4$ in cold brown dwarfs (Nakajima \\etal\\ 1995; Oppenheimer \\etal\\ 1995). The next major breakthrough will be the identification of significant numbers of brown dwarfs by the coming generation of infrared all-sky surveys -- in particular DENIS (Epchtein 1997) and 2MASS (Skrutskie \\etal\\ 1997). \\\\  DENIS will be a complete near infrared survey of the southern sky (Epchtein \\etal\\ 1994; Epchtein 1997) in the I, J and K$^\\prime$ bands, to approximate 3-$\\sigma$ limits of I=18, J=16, and K=13.5. The products of this survey will be databases of calibrated images, extended sources, and small objects. The survey started in January 1996 and is expected to be completed within five years.  We have carried out a ``Mini-survey'' with infrared spectroscopic follow up on the very low-mass (VLM) star and brown dwarf candidates contained in $\\approx$1\\% of the DENIS survey data (Delfosse \\etal\\ 1998, 1997). The image data from the high latitude part ($|b_{II}|>20$-30\\degr) of 47 survey strips, were processed and used to identify a sample of objects for which infrared H- and K-band spectroscopy was carried out in order to estimate luminosities/temperatures. In this paper we present optical spectroscopy for a sample of cool dwarfs identified in a 231 square degree ``Mini-survey'' of the data from the DEep Near-Infrared Survey (DENIS), and discuss the significant features these spectra reveal. ",
        "conclusions": "We have shown that optical spectroscopy confirms the luminosity classifications derived from the infrared spectra of brown dwarf candidates by Delfosse \\etal\\ (1998). In particular we confirm the detection of at least two objects which are as late as the latest known field M-dwarfs. The coolest objects studied in this work show prominent features of Cs\\,I, K\\,I, CrH and FeH, and a sequence of decreasing strength in TiO and VO features with decreasing temperature. The coolest object in our sample also shows decreased strength in CrH. The ``on-off'' behaviour of CrH, which is driven by the equilibria of molecular condensation as effective temperature decreases, implies it may be useful as a temperature diagnostic which can be readily modelled. Lines of the K\\,I doublet become {\\em extremely} strong ($\\sim$300\\,\\AA\\ equivalent width) in the coolest DENIS objects (T$_eff \\simlt 1800$K), and also offer the possibility of being powerful temperature diagnostics. And lastly, the improved match of models assuming rapid gravitational settling of dust grains, over models with no settling of dust grains, would seem to indicate that settling effects play a role in 2000-1000\\,K photospheres.\\\\  It is clear that an entirely new sequence of spectral sub-classes for the ``L''-type dwarfs will be required. Assignment of specific sub-classes must await both improvements in atmospheric models to detail the effects of dust condensation and settling on radiative transfer and the equation of state, and an increase in the sample of L-dwarfs with optical spectra. The latter will be achieved by the 2MASS and DENIS surveys within the next two to three years, promising exciting developments in our understanding of the behaviour and properties of brown dwarfs."
    },
    "9805/hep-ph9805449_arXiv.txt": {
        "abstract": "A simple hadronic axion model is proposed in the framework of gauge-mediated supersymmetry breaking. Dynamics of Peccei-Quinn symmetry breaking is governed by supersymmetry breaking effects and the Peccei-Quinn breaking scale $f_{PQ}$ is inversely proportional to the gravitino mass. The gravitino mass range which corresponds to the axion window $f_{PQ} \\simeq 10^{9}$ GeV -- $10^{13}$ GeV lies in the region predicted by gauge-mediated supersymmetry breaking models. The model is also shown to be cosmologically viable. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805276_arXiv.txt": {
        "abstract": " ",
        "introduction": "Ascertaining the core collapse supernova mechanism is a long-standing problem in astrophysics. The current paradigm begins with the collapse of a massive star's iron core and the generation of an outwardly propagating shock wave that results from core rebound. Because of nuclear dissociation and neutrino losses, the shock stagnates. This sets the stage for a shock reheating mechanism whereby neutrino energy deposition via electron neutrino and antineutrino absorption on nucleons behind the shock reenergizes it (Bethe \\& Wilson 1985; Wilson 1985).  The shock reheating phase is essential to the supernova's success, but it is precisely this phase that is difficult to simulate realistically. During shock reheating, core electron neutrinos and antineutrinos are radiated from their respective neutrinospheres, and a small fraction of this radiated energy is absorbed in the exterior shocked mantle. The shock reheating depends sensitively on the electron neutrino and antineutrino luminosities, spectra (best characterized by the {\\small RMS} energies), and angular distributions in the region behind the shock (e.g., see Burrows \\& Goshy 1993, Janka \\& M\\\"{u}ller 1996, Mezzacappa et al. 1998b). These, in turn, depend on the neutrino transport in the semitransparent region encompassing the neutrinospheres, necessitating a neutrino transport treatment that is able to transit accurately and seamlessly between neutrino-thick and neutrino-thin regions.  Various neutrino transport approximations have been implemented in simulating core collapse supernovae. The most sophisticated approximation, which naturally has been used in detailed one-dimensional simulations, is multigroup flux-limited diffusion ({\\small MGFLD}; e.g., Bowers \\& Wilson 1982, Bruenn 1985, Myra et al. 1987). {\\small MGFLD} closes the neutrino radiation hydrodynamics hierarchy of equations at the level of the first moment (the neutrino flux) by imposing a relationship between the flux and the gradient of the neutrino energy density (the zeroth moment).  For example,  \\begin{equation} F_{\\nu}=-\\frac{c\\Lambda}{3}\\frac{\\partial U_{\\nu}}{\\partial r}+..., \\label{eq:mgfld} \\end{equation}  \\begin{equation} \\Lambda = \\frac{1}{1/\\lambda + |\\partial U_{\\nu}/\\partial r|/3U_{\\nu}}, \\label{eq:lambda} \\end{equation}  \\noindent where $\\lambda$ is the neutrino mean free path, and $U_{\\nu}$ and $F_{\\nu}$ are the neutrino energy density and flux (Bruenn 1985). [Other forms for the flux-limiter $\\Lambda$ can be found in Bowers \\& Wilson (1982), Levermore \\& Pomraning (1981), and Myra et al (1987).] Whereas the limits $\\lambda \\rightarrow 0$ and $\\lambda \\rightarrow \\infty$  produce the correct diffusion and free streaming fluxes, it is in the critical intermediate regime where the {\\small MGFLD} approximation is of unknown accuracy. Unfortunately, the quantities central to the postshock neutrino heating mentioned earlier are determined in this regime, and given the sensitivity of the neutrino heating to these quantities it becomes necessary to consider more accurate transport schemes.  Moreover, in detailed one-dimensional simulations that have implemented elaborate {\\small MGFLD} neutrino transport (e.g., see Bruenn 1993, Wilson \\& Mayle 1993, and Swesty \\& Lattimer 1994), explosions were not obtained unless the neutrino heating was boosted by additional phenomena, such as convection. This leaves us with at least two possibilities to consider:  (1) Failures to produce explosions in the absence of additional phenomena, such as convection, have resulted from inexact neutrino transport.  (2) Additional phenomena may be essential in obtaining explosions.  Option (1) requires further comment.  All investigators  agree convection will occur during the shock reheating, explosion initiation phase in core collapse supernovae (Herant, Benz, \\& Colgate 1992; Miller et al. 1993; Herant et al. 1994; Burrows et al. 1995, Janka \\& M\\\"{u}ller 1996, Mezzacappa et al. 1998b).  Therefore, strictly speaking, all investigators agree the flow will not be spherically symmetric. However, although convection will certainly occur, it may play no significant role in {\\em initiating} the explosion.  It is with this distinction in mind that option (1) need be considered.  For example, significant neutrino-driven convection was seen in recent multidimensional simulations employing one-dimensional {\\small MGFLD} neutrino transport (Mezzacappa et al. 1998b); however, the angle-averaged shock radius, among other quantities, did not differ significantly from its one-dimensional counterpart, and no explosion was obtained.  Ultimately, any successful model of core collapse supernovae will have to reproduce observables that clearly do not originate from spherically symmetric explosions, the most obvious of which is neutron star kicks.  At this point, whether or not these kicks are generated during the initiation of the explosion or shortly thereafter is an open question.  Note in this regard that simulations that have invoked multidimensional effects such as convection to explain such kicks have had difficulty generating, for example, adequate kick velocities. Mechanisms invoking convection, or aspherical neutrino emission resulting from convection, have not been able to produce kicks in excess of about 300 km/s, which therefore cannot account for the fastest pulsars --- for example, PSR 2224+65, which has a velocity around 800 km/s --- (Janka \\& M\\\"{u}ller 1994; Burrows \\& Hayes 1995, 1996). Moreover, definitive predictions of neutron star kick velocities from aspherical supernovae will require three-dimensional simulations. (The aforementioned simulations were carried out in two dimensions.) Recent simulations of neutrino-driven convection in two and three dimensions demonstrate that, as expected, longer-wavelength modes break up in three dimensions, rendering the angle-averaged flow qualitatively much closer to spherically symmetric (Knerr et al. 1998; see also M\\\"{u}ller 1993 and Janka and M\\\"{u}ller 1996). In light of this, it is difficult to see how the already low values for neutron star kick velocities obtained by invoking convection and/or convection-induced anisotropic neutrino emission during the explosion itself could be enhanced when these same simulations are carried out in three dimensions.  In this paper, we compare three-flavor multigroup Boltzmann neutrino transport ({\\small MGBT}) and (Bruenn's) {\\small MGFLD} in spherically symmetric, hydrostatic, thermally frozen, postbounce profiles, with an eye toward quantities central to the postbounce neutrino heating mechanism for reviving the stalled shock. In particular, for both transport methods, we compute and compare the neutrino luminosities, {\\small RMS} energies, mean inverse flux factors, and net heating rates as functions of radius, time, and precollapse model. We then discuss the ramifications our results have for the supernova mechanism. This work is a continuation and extension of core collapse simulations (Mezzacappa and Matzner 1989, Mezzacappa and Bruenn 1993a,b,c), in which exact Boltzmann neutrino transport and multigroup flux-limited diffusion were compared. ",
        "conclusions": "Comparing three-flavor {\\small MGBT} and three-flavor {\\small MGFLD} in postbounce supernova environments, we find that {\\small MGBT} leads to a significant increase/decrease in the {\\it net} heating/cooling rate, particularly above/below the gain radius. The {\\small MGBT} net heating  rate can be as much as $\\sim$2 times the {\\small MGFLD} net heating rate above the gain radius, with net cooling rates that are typically $\\sim$0.8 times the {\\small MGFLD} rate below the gain radius.  These differences stem primarily from differences in the neutrino luminosities and mean inverse flux factors;  the heating rate is linearly proportional to both these quantities, and differences in both add to produce a significant difference in the net heating rate.  In Figure 10,  we plot the sum of the electron neutrino and antineutrino luminosities computed in the  {\\small MGFLD} S15s7b dynamic simulation for several different postbounce times.  It is important to note that the total luminosity changes by $\\sim$5--15\\% between 100 km and 200 km on time scales $\\sim$30 ms.  Moreover, the light crossing time between 100 km and 200 km is $\\sim$1/3 ms.  Therefore, the neutrino source in our simulations changes on time scales that are two orders of magnitude greater than the time scales on which stationary state is established in this region. This suggests our stationary state results closely reflect what will occur in dynamic simulations.  We also observe that the differences in the net heating rate are greatest at earlier postbounce times for a given progenitor mass, and greater at any given postbounce time for greater progenitor mass. This is illustrated in Table 1. The  increase in net heating with increased progenitor mass is advantageous because of the slower fall-off in the preshock accretion ram pressure.  These results have at least two important ramifications for the supernova mechanism:  (1) With the dramatic increase in net heating above the gain radius, which is seen in all of our postbounce slices, it may be possible to obtain explosions in one dimension without multidimensional effects such as convection; this will be aided by the decrease in net cooling below the gain radius. It should be noted that our postbounce slices come from full radiation hydrodynamics simulations implementing {\\small MGFLD} that marginally failed to produce explosions (Bruenn 1993). The marginality of Bruenn's simulations is an important motivating factor in comparing our {\\small MGBT} results  solely with his {\\small MGFLD} results. All else being equal, increases in net heating  of the magnitude documented here would most likely lead to explosions. However, simulations coupling {\\small MGBT} and the core hydrodynamics must be carried out in order to compute any feedbacks. It remains to be seen whether the {\\small MGBT} heating rate will remain sufficiently high to generate an explosion.  Also, the effects of using general relativistic gravity, hydrodynamics, and neutrino transport must be explored, especially if explosions are obtained in the Newtonian limit. For example, the redshifted neutrino energies, the smaller gain region, and the greater infall velocities in the gain region will most likely conspire to make explosions more difficult to obtain (Mezzacappa et al. 1998b, DeNisco, Bruenn, \\& Mezzacappa 1998, Bruenn, DeNisco, \\& Mezzacappa 1998).  (2) With the dramatic increase in net heating occurring near the base of the gain region, we anticipate that {\\small MGBT} coupled to two-dimensional hydrodynamics will yield more vigorous neutrino-driven convection than was seen in Mezzacappa et al. (1998b), where two-dimensional hydrodynamics was coupled to one-dimensional {\\small MGFLD}. The combination of increased net heating and more vigorous neutrino-driven convection would be more favorable for shock revival.  In closing, our results are promising, and their ramifications for core collapse supernovae and, in particular, for the postbounce neutrino-heating, shock-revival mechanism await one- and two-dimensional dynamical simulations with {\\small MGBT} coupled to the core hydrodynamics. One-dimensional simulations are currently underway, and we plan to report on them soon."
    },
    "9805/astro-ph9805040_arXiv.txt": {
        "abstract": "Each time diffusion of elements is invoked in explaining abundance anomalies in a star, this supposes implicitly that a stratification process is in progress somewhere in that star. This means also, that the element abundances can still be evolving according to the star's age and fundamental parameters. Moreover, it has been shown that the superficial abundances may have complex temporal behavior. This should be detectable through new observations. In some cases, it may be already apparent in available data.  The building up of the elements' stratification is a very difficult process to study. This is due to the existence of strong non-linearities in the time-dependent equations, which must be solved numerically. We will discuss some works that tackle this problem for Ap and Am stars, and we will present some results concerning mostly Am stars. Future desirable improvements in these studies will be considered. ",
        "introduction": "\\subsection{The observational facts} Here, we shall consider mainly the three well-known groups of Chemically Peculiar stars on the main-sequence of the HR diagram: FmAm, HgMn (non-magnetic Ap or weak magnetic field), and magnetic Ap stars. They show clearly abundance anomalies of metals with respect to solar abundances. These anomalies extend from factors of about 2 to 10 (in Am stars) and up to some 10$^{6}$ (in Ap stars).  Many abundance determinations have been done for a large number of Ap and Bp stars, covering a large sample of effective temperatures, for a large number of heavy elements. One of the most striking aspects of the observational results is the rather wide scatter of abundances determined in these stars for a given element (see for instance Takada-Hidai 1990). This scatter is partly due to errors in the abundance determinations and to inaccuracies in the effective temperatures determinations, but not only. Despite this scatter, a clear correlation with respect to the effective temperature is well established for some elements like Mn (see for instance Smith and Dworetsky, 1993).  \\subsection{The diffusion model} It is generally accepted that diffusion processes of elements provide the best explanation for CP stars anomalies. Of course, microscopic diffusion interacts with other processes in stars, such as large scale motions (convection, turbulence, wind) and it cannot be considered alone: the anomalies are the result of a complex stratification process involving all these motions.  The diffusion model was first proposed by G. Michaud (1970), it is based on two basic findings: on the one hand, the abundance anomalies are related to atmospheric parameters like effective temperature and gravity, on another hand, there is a clear correlation between the superficial abundance anomalies and the radiative accelerations (one of the main terms involved in the diffusion velocity) in stellar external layers. The radiative accelerations are different according to the elements and they depend on atmospheric parameters. The anomalies which are found on main-sequence stars are not found among evolved stars, and this suggests that they are confined mostly to external layers. Moreover, in the HR diagram, the CP stars are located where the stars' outer convection zone is supposed to be the smallest. All of these arguments (with some others which are not discussed here) give a consistent outline for the diffusion model: when the stars evolve, the external layers are mixed with deeper ones and the abundance stratifications are lost.  In the diffusion model, the common property of the CP's is the weakness of the superficial helium abundance (due to helium gravitational settling). This helium underabundance is supposed to lead to the decrease (Am stars) or the disappearance (HgMn and magnetic Ap stars) of the superficial convection zone. Then, diffusion can occur in layers where stratification time scales are much shorter than in normal stars. ",
        "conclusions": ""
    },
    "9805/astro-ph9805089_arXiv.txt": {
        "abstract": "Smoothed Particle Hydrodynamics is a multidimensional Lagrangian method of numerical hydrodynamics that has been used to tackle a wide variety of problems in astrophysics. Here we develop the basic equations of the SPH scheme, and we discuss some of its numerical properties and limitations. As an illustration of typical astrophysical applications, we discuss recent calculations of stellar interactions, including collisions between main sequence stars and the coalescence of compact binaries. ",
        "introduction": "Smoothed Particle Hydrodynamics (SPH) is a Lagrangian method that was introduced specifically to simulate self-gravitating fluids moving freely in three dimensions. The key idea of SPH is to calculate pressure gradient forces by kernel estimation, directly from the particle positions, rather than by finite differencing on a grid, as in older particle methods such as PIC. SPH was first introduced by Lucy (1977) and Gingold \\& Monaghan (1977), who used it to study dynamical fission instabilities in rapidly rotating stars. Since then, a wide variety of astrophysical fluid dynamics problems have been tackled using SPH (see Monaghan 1992 for an overview). In addition to the stellar interaction problems described in \\S2, these have included planet and star formation (Nelson et al.\\ 1998; Burkert et al.\\ 1997), supernova explosions (Herant \\& Benz 1992; Garcia-Senz et al.\\ 1998), large-scale cosmological structure formation (Katz et al.\\ 1996; Shapiro et al.\\ 1996), and galaxy formation (Katz 1992; Steinmetz 1996).  \\subsection{SPH from a Variational Principle}  A straightforward derivation of the basic SPH equations can be obtained from a Lagrangian formulation of hydrodynamics (Gingold \\& Monaghan 1982). Consider for simplicity the adiabatic evolution of an ideal fluid with equation of state \\begin{equation} p=A\\rho^\\gamma, \\label{eos} \\end{equation} where $p$ is the pressure, $\\rho$ is the density, $\\gamma$ is the adiabatic exponent, and $A$ (assumed here to be constant in space and time) is related to the specific entropy ($s \\propto \\ln A$). The Euler equations of motion, \\begin{equation} {{\\d}\\vec{v}\\over \\d t}= {\\partial\\vec{v}\\over\\partial t}+(\\vec{v}\\cdot\\nabla)\\vec{v} =-{1\\over\\rho}\\nabla p, \\end{equation} can be derived from a variational principle with the Lagrangian \\begin{equation} L=\\int\\left\\{{1\\over2}v^2 - u[\\rho(\\vec{r})]\\right\\}\\,\\rho\\,{\\d}^3x. \\end{equation} Here $u[\\rho]=p/[(\\gamma-1)\\rho]=A\\rho^{\\gamma-1}/(\\gamma-1)$ is the specific internal energy of the fluid.  The basic idea in SPH is to use the discrete representation \\begin{equation} L_{SPH}=\\sum_{i=1}^N\\, m_i\\left[{1\\over2}v_i^2 -u(\\rho_i)\\right] \\label{lsph} \\end{equation} for the Lagrangian, where the sum is over a large but discrete number of small fluid elements, or ``particles,'' covering the volume of the fluid. Here $m_i$ is the mass and $\\vec{v}_i$ is the velocity of the particle with position $\\vec{r}_i$.  For expression~(\\ref{lsph}) to become the Lagrangian of a system with a finite number $N$ of degrees of freedom, we need a prescription to compute the density $\\rho_i$ at the position of any given particle $i$, as a function of the masses and positions of neighboring particles.  In SPH, the density at any position is typically calculated as the local average \\begin{equation} \\rho(\\vec{r})=\\sum_j m_j W(\\vec{r}-\\vec{r}_j;\\, h),\\label{rho} \\end{equation} where $W(\\vec{x};\\,h)$ is an interpolation, or smoothing, kernel of width $\\sim h$.  Necessary constraints on the kernel $W(\\vec{x};\\,h)$ are that (i) it integrates to unity (consequently the integral of eq. (\\ref{rho}) over all space automatically gives the total mass of the system), and (ii) it approaches the Dirac delta function $\\delta(\\vec{x})$ in the limit $h\\rightarrow 0$.  Equation (\\ref{rho}) gives, in particular, the density in the vicinity of particle $i$ as $\\rho_i=\\rho(\\vec{r}_i)$, and we can now obtain the equations of motion for all the particles. Deriving the Euler-Lagrange equations from $L_{SPH}$ we get \\begin{equation} {{\\d}\\vec{v}_i\\over \\d t}=-\\sum_j m_j \\,\\left({p_i\\over\\rho_i^2} +{p_j\\over\\rho_j^2}\\right)\\, \\nabla_i W_{ij}, \\label{simple} \\end{equation} where $W_{ij}= W(\\vec{r}_i-\\vec{r}_j;\\, h)$ and we have assumed that the form of $W$ is such that $W_{ij}=W_{ji}$. The expression on the right-hand side of eq.~(\\ref{simple}) is a sum over neighboring particles (within a distance $\\sim h$ of $\\vec{r}_i$) representing a discrete approximation to the pressure gradient force $[-(1/\\rho)\\nabla p]_i$ acting on particle $i$.  The following energy and momentum conservation laws are satisfied {\\em exactly\\/} by the simple SPH equations of motion given above \\begin{equation} {\\d\\over \\d t}\\left(\\sum_{i=1}^N m_i \\vec{v}_i \\right) =0, \\end{equation} and \\begin{equation} {\\d\\over \\d t}\\left(\\sum_{i=1}^N m_i\\, [{1\\over2}v_i^2 +u_i]\\right) =0, \\end{equation} where $u_i=p_i/[(\\gamma-1)\\rho_i]$. Note that energy and momentum conservation in this simple version of SPH is independent of the number of particles $N$.  Typically, a full implementation of SPH for astrophysical problems will add to eq.~(\\ref{simple}) a treatment of self-gravity (e.g., using one of the many grid-based or tree-based algorithms developed for N-body simulations) and an artificial viscosity term to allow for entropy production in shocks. In addition, we have assumed here that the smoothing length $h$ is constant in time and the same for all particles. In practice, individual and time-varying smoothing lengths $h_i(t)$ are almost always used, so that the local spatial resolution can be adapted to the (time-varying) density of SPH particles (see Nelson \\& Papaloizou 1994 for a rigorous derivation of the equations of motion from a variational principle in this case). Other derivations of the SPH equations, based on the application of smoothing operators to the fluid equations (and without the use of a variational principle), are also possible (see, e.g., Hernquist \\& Katz 1989).  \\subsection{Basic SPH Equations}  In this section, we summarize the basic equations for various forms of the SPH scheme currently in use, incorporating gravity, artificial viscosity, and individual smoothing lengths.  \\subsubsection{Density and Pressure}  The SPH estimate of the fluid density at $\\vec{r}_i$ is calculated as $\\rho_i=\\sum_j m_j W_{ij}$ [cf.\\ eq.~(\\ref{rho})]. Many recent implementations of SPH use a form for $W_{ij}$ proposed by Hernquist \\& Katz (1989), \\begin{equation} W_{ij}={1\\over2}\\left[W(|\\vec{r}_i-\\vec{r}_j|;\\,h_i)+W(|\\vec{r}_i- \\vec{r}_j|;\\,h_j)\\right]. \\end{equation} This choice guarantees symmetric weights $W_{ij}=W_{ji}$ even between particles $i$ and $j$ with different smoothing lengths. For the interpolation kernel $W(r;\\,h)$, the cubic spline \\begin{equation} W(r;\\,h)={1\\over\\pi h^3} \\cases{1-{3\\over2}\\left({r\\over h}\\right)^2 +{3\\over4}\\left({r\\over h}\\right)^3,        & $0\\le{r\\over h}<1$,\\cr {1\\over4}\\left[2-\\left({r\\over h}\\right)\\right]^3,& $1\\le{r\\over h}<2$,\\cr 0,                                          & ${r\\over h}\\ge2$,\\cr} \\label{WML} \\end{equation} (Monaghan \\& Lattanzio 1985) is a common choice. Eq.~(\\ref{WML}) is sometimes called a ``second-order accurate'' kernel. Indeed, when the true density $\\rho(\\vec{r})$ of the fluid is represented by an appropriate distribution of particle positions, masses, and smoothing lengths, one can show that $\\rho_i=\\rho(\\vec{r}_i)+O(h_i^2)$ (see, e.g., Monaghan 1985).  Depending on which thermodynamic evolution equation is integrated [see eqs.~(\\ref{udot}) and~(\\ref{adot}) below], particle $i$ also carries either the parameter $u_i$, the internal energy per unit mass in the fluid at $\\vec{r}_i$, or $A_i$, the entropic variable, a function of the specific entropy in the fluid at $\\vec{r}_i$. Although arbitrary equations of state can be implemented in SPH, here, for simplicity, we consider only polytropic equations of state. The pressure $p_i$ at $\\vec{r}_i$ is therefore related to the density by \\begin{equation} p_i=(\\gamma-1)\\,\\rho_i\\, u_i, \\end{equation} or \\begin{equation} p_i=A_i\\,\\rho_i^\\gamma. \\end{equation} The speed of sound in the fluid at $\\vec{r}_i$ is $c_i=(\\gamma p_i/\\rho_i)^{1/2}$.  \\subsubsection{Dynamical Equations and Gravity}  Particle positions are updated either by \\begin{equation} {{\\d}\\vec{r}_i \\over\\d t}= \\vec{v}_i, \\label{rdot} \\end{equation} or the more general XSPH method \\begin{equation} {{\\d}\\vec{r}_i\\over \\d t} = \\vec{v}_i+\\epsilon \\sum_j m_j{\\vec{v}_j-\\vec{v}_i\\over \\rho_{ij}}W_{ij} \\label{XSPH} \\end{equation} where $\\rho_{ij}=(\\rho_i+\\rho_j)/2$ and $\\epsilon$ is a constant parameter in the range $0 < \\epsilon < 1$ (Monaghan 1989).  Eq.~(\\ref{XSPH}), in contrast to eq.~(\\ref{rdot}), changes particle positions at a rate closer to the local smoothed velocity.  The XSPH method was originally proposed as a way to minimize spurious interparticle penetration across the interface of two colliding fluid streams.  Generalizing equation (\\ref{simple}) to account for gravitational forces and artificial viscosity (hereafter AV), the velocity of particle $i$ is updated according to \\begin{equation} {{\\d} \\vec{v}_i\\over \\d t} = \\vec{a}^{(Grav)}_i+\\vec{a}^{(SPH)}_i \\end{equation} where $\\vec{a}^{(Grav)}_i$ is the gravitational acceleration and \\begin{equation} \\vec{a}^{(SPH)}_i=-\\sum_j m_j \\left[\\left({p_i\\over\\rho_i^2}+ {p_j\\over\\rho_j^2}\\right)+\\Pi_{ij}\\right]{\\bf \\nabla}_i W_{ij}. \\label{fsph} \\end{equation} Various forms for the AV term $\\Pi_{ij}$ are discussed below. The AV ensures that correct jump conditions are satisfied across (smoothed) shock fronts, while the rest of equation~(\\ref{fsph}) represents one of many possible SPH-estimators for the acceleration due to the local pressure gradient (see, e.g., Monaghan 1985).  To provide reasonable accuracy, an SPH code must solve the equations of motion of a large number of particles (typically $N>>1000$). This rules out a direct summation method for calculating the gravitational field of the system, unless special purpose hardware such as the GRAPE is used (Steinmetz 1996; Klessen 1997).  In most implementations of SPH, particle-mesh algorithms (Evrard 1988; Rasio \\& Shapiro 1992; Couchman et al.\\ 1995) or tree-based algorithms (Hernquist \\& Katz 1989; Dave et al.\\ 1997) are used to calculate the gravitational accelerations $\\vec{a}^{(Grav)}_i$. Tree-based algorithms perform better for problems involving large dynamic ranges in density, such as star formation and large-scale cosmological simulations. For stellar interaction problems like those described in \\S 2, density contrasts rarely exceed a factor $\\sim10^2-10^3$ and in those cases grid-based algorithms and direct solvers are generally faster. Tree-based and grid-based algorithms are also used to calculate lists of nearest neighbors for each particle exactly as in gravitational $N$-body simulations.  \\subsubsection{Artificial Viscosity}  For the AV, a symmetrized version of the form proposed by Monaghan (1989) is often adopted, \\begin{equation} \\Pi_{ij}={-\\alpha\\mu_{ij}c_{ij}+\\beta\\mu_{ij}^2\\over\\rho_{ij}}, \\label{pi} \\end{equation} where $\\alpha$ and $\\beta$ are constant parameters, $c_{ij}=(c_i+c_j)/2$, and \\begin{equation} \\mu_{ij}=\\cases{ {\\left(\\vec{v}_i-\\vec{v}_j\\right)\\cdot(\\vec{r}_i-\\vec{r}_j)\\over h_{ij}\\left(|\\vec{r}_i -\\vec{r}_j|^2/h_{ij}^2+\\eta^2\\right)}& if $(\\vec{v}_i- \\vec{v}_j)\\cdot(\\vec{r}_i-\\vec{r}_j)<0$\\cr 0& if $(\\vec{v}_i-\\vec{v}_j)\\cdot(\\vec{r}_i- \\vec{r}_j)\\ge0$\\cr} \\label{mu} \\end{equation} with $h_{ij}=(h_i+h_j)/2$. This form represents a combination of a bulk viscosity (linear in $\\mu_{ij}$) and a von~Neumann-Richtmyer viscosity (quadratic in $\\mu_{ij}$).  The von~Neumann-Richtmyer viscosity was initially introduced to suppress particle interpenetration in the presence of strong shocks.  Eq.~(\\ref{pi}) provides a good treatment of shocks when $\\alpha\\approx1$, $\\beta\\approx2$ and $\\eta^2\\sim10^{-2}$ (Monaghan 1989; Hernquist \\& Katz 1989).  A well known problem with the classical AV of eq.~(\\ref{pi}) is that it can generate large amounts of spurious shear viscosity.  For this reason, Hernquist \\& Katz (1989) introduced another form for the AV: \\begin{equation} \\Pi_{ij}=\\cases{ {q_i\\over\\rho_{i}^2}+{q_j\\over\\rho_{j}^2}& if $( \\vec{v}_i-\\vec{v}_j)\\cdot(\\vec{r}_i-\\vec{r}_j)<0$\\cr 0& if $(\\vec{v}_i-\\vec{v}_j)\\cdot(\\vec{r}_i- \\vec{r}_j)\\ge0$\\cr}, \\label{pi2} \\end{equation} where \\begin{equation} q_i=\\cases{ \\alpha \\rho_i c_i h_i |{\\bf \\nabla}\\cdot \\vec{v}|_i+ \\beta \\rho_i h_i^2 |{\\bf \\nabla}\\cdot \\vec{v}|_i^2 & if $\\left({\\bf \\nabla}\\cdot \\vec{v}\\right)_i<0$\\cr 0& if $\\left({\\bf \\nabla}\\cdot \\vec{v}\\right)_i\\ge0$\\cr} \\label{q} \\end{equation} and \\begin{equation} ({\\bf \\nabla}\\cdot \\vec{v})_i={1 \\over \\rho_i}\\sum_j m_j (\\vec{v}_j-\\vec{v}_i)\\cdot{\\bf \\nabla}_i W_{ij}. \\label{divv} \\end{equation} Although this form provides a slightly less accurate description of shocks than equation (\\ref{pi}), it does exhibit less shear viscosity.  More recently, Balsara (1995) has proposed the AV \\begin{equation} \\Pi_{ij}= \\left({p_i\\over\\rho_i^2}+{p_j\\over\\rho_j^2}\\right) \\left(-\\alpha \\mu_{ij} + \\beta \\mu_{ij}^2\\right), \\label{piDB} \\end{equation} where \\begin{equation} \\mu_{ij}=\\cases{ {(\\vec{v}_i-\\vec{v}_j)\\cdot(\\vec{r}_i-\\vec{r}_j)\\over h_{ij}\\left(|\\vec{r}_i -\\vec{r}_j|^2/h_{ij}^2+\\eta^2\\right)}{f_i+f_j \\over 2 c_{ij}}& if $(\\vec{v}_i-\\vec{v}_j)\\cdot(\\vec{r}_i-\\vec{r}_j)<0$\\cr 0& if $(\\vec{v}_i-\\vec{v}_j)\\cdot(\\vec{r}_i-\\vec{r}_j)\\ge0$\\cr}. \\label{muDB} \\end{equation} Here $f_i$ is the form function for particle $i$ defined by \\begin{equation} f_i={|{\\bf \\nabla}\\cdot \\vec{v}|_i \\over |{\\bf \\nabla}\\cdot \\vec{v}|_i +|{\\bf \\nabla}\\times \\vec{v}|_i + \\eta' c_i/h_i }, \\label{fi} \\end{equation} where the factor $\\eta'\\sim 10^{-4}-10^{-5}$ prevents numerical divergences, $({\\bf \\nabla}\\cdot \\vec{v})_i$ is given by equation (\\ref{divv}), and \\begin{equation} ({\\bf \\nabla}\\times \\vec{v})_i={1 \\over \\rho_i}\\sum_j m_j (\\vec{v}_i-\\vec{v}_j)\\times{\\bf \\nabla}_i W_{ij}. \\label{curlv} \\end{equation} The form function $f_i$ acts as a switch, approaching unity in regions of strong compression ($|{\\bf \\nabla}\\cdot \\vec{v}|_i >>|{\\bf \\nabla}\\times \\vec{v}|_i$) and vanishing in regions of large vorticity ($|{\\bf \\nabla}\\times \\vec{v}|_i >>|{\\bf \\nabla}\\cdot \\vec{v}|_i$).  Consequently, this AV has the advantage that it is suppressed in shear layers. Note that since $(p_i/\\rho_i^2+p_j/\\rho_j^2)\\approx 2c_{ij}^2/(\\gamma\\rho_{ij})$, equation~(\\ref{piDB}) behaves like equation (\\ref{pi}) when $|{\\bf \\nabla}\\cdot \\vec{v}|_i >> |{\\bf \\nabla}\\times \\vec{v}|_i$, provided one rescales the $\\alpha$ and $\\beta$ in equation (\\ref{piDB}) to be a factor of $\\gamma/2$ times the $\\alpha$ and $\\beta$ in equation (\\ref{pi}).  \\subsubsection{Thermodynamics}  To complete the description of the fluid, either $u_i$ or $A_i$ is evolved according to a discretized version of the first law of thermodynamics.  Although various forms of these evolution equations exist, the most commonly used are \\begin{equation} {{\\d} u_i\\over \\d t}= {1\\over 2}\\sum_j m_j \\left({p_i\\over\\rho_i^2}+{p_j\\over\\rho_j^2}+ \\Pi_{ij}\\right)\\,(\\vec{v}_i-\\vec{v}_j)\\cdot{\\bf \\nabla}_i W_{ij}, \\label{udot} \\end{equation} and \\begin{equation} {{\\d} A_i\\over \\d t}={\\gamma-1\\over 2\\rho_i^{\\gamma-1}}\\, \\sum_jm_j\\,\\Pi_{ij}\\,\\,(\\vec{v}_i-\\vec{v}_j)\\cdot{\\bf \\nabla}_i W_{ij}. \\label{adot} \\end{equation} We call equation~(\\ref{udot}) the ``energy equation,'' while equation (\\ref{adot}) is the ``entropy equation.''  Which equation one should integrate depends upon the problem being treated.  Each has its own advantages and disadvantages. Note that the derivation of equations~(\\ref{udot}) and (\\ref{adot}) neglects terms proportional to the time derivative of $h_i$.  Therefore if we integrate the energy equation, even in the absence of AV, the total entropy of the system will not be strictly conserved if the particle smoothing lengths are allowed to vary in time; if the entropy equation is used to evolve the system, the total entropy would then be strictly conserved when $\\Pi_{ij}=0$, but not the total energy (Rasio 1991; Hernquist 1993). For more accurate treatments involving time-dependent smoothing lengths, see Nelson \\& Papaloizou (1994) and Serna et al. (1996). The energy equation has the advantage that other thermodynamic processes such as heating and cooling (Katz et al.\\ 1996) and nuclear burning Garcia-Senz et al.\\ 1998) can be incorporated more easily.   \\subsubsection{Integration in Time}  The results of SPH simulations involving only hydrodynamic forces and gravity do not depend strongly on the actual time-stepping routine used, as long as the routine remains stable and accurate.  A simple second-order explicit leap-frog scheme is often employed. Implicit schemes must be used when other processes such as heating and cooling are coupled to the dynamics (Katz et al.\\ 1996). A low order scheme is appropriate for SPH because pressure gradient forces are subject to numerical noise.  For stability, the timestep must satisfy a modified Courant condition, with $h_i$ replacing the usual grid separation. For accuracy, the timestep must be a small enough fraction of the dynamical time.  Among the many possible choices for determining the timestep, the prescription proposed by Monaghan (1989) is recommended. This sets \\begin{equation} \\Delta t=C_N\\,{\\rm Min}(\\Delta t_1,\\Delta t_2), \\label{dt} \\end{equation} where the constant dimensionless Courant number $C_N$ typically satisfies $0.1\\lo C_N \\lo 0.8$, and where \\begin{eqnarray} \\Delta t_1 &=&{\\rm Min}_i\\,(h_i/\\dot v_i)^{1/2}, \\label{dt1} \\\\ \\Delta t_2 &=&{\\rm Min}_i\\left( {h_i \\over c_i+k\\left(\\alpha c_i+\\beta {\\rm Max_j}|\\mu_{ij}|\\right)} \\right),  \\label{good.dt} \\end{eqnarray} with $k$ being a constant of order unity. If the Hernquist \\& Katz AV [eq.~(\\ref{pi2})] is used, the quantity Max$_j|\\mu_{ij}|$ in equation (\\ref{good.dt}) can be replaced by $h_i|{\\bf \\nabla}\\cdot \\vec{v}|_i$ if $({\\bf \\nabla}\\cdot \\vec{v})_i<0$, and by $0$ otherwise.  By accounting for AV-induced diffusion, the $\\alpha$ and $\\beta$ terms in the denominator of equation (\\ref{good.dt}) allow for a more efficient use of computational resources than simply using a smaller value of $C_N$.  \\subsubsection{Smoothing Lengths and Accuracy}  The size of the smoothing lengths is often chosen such that particles roughly maintain some predetermined number of neighbors $N_N$.  Typical values of $N_N$ range from about 20 to 100.  If a particle interacts with too few neighbors, then the forces on it are sporadic, a poor approximation to the forces on a true fluid element. In general, one finds that, for given physical conditions, the noise level in a calculation always decreases when $N_N$ is increased.  At the other extreme, large neighbor numbers degrade the resolution by requiring unreasonably large smoothing lengths. However,  higher accuracy is obtained in SPH calculations only when {\\em both\\/} the number of particles $N$ {\\em and\\/} the number of neighbors $N_N$ are increased, with $N$ increasing faster than $N_N$ so that the smoothing lengths $h_i$ decrease. Otherwise (e.g., if $N$ is increased while maintaining $N_N$ constant) the SPH method is {\\em inconsistent\\/}, i.e., it converges to an unphysical limit. This can be shown easily by deriving the dispersion relation for sound waves propagating in simple SPH systems (Rasio 1991). The choice of $N_N$ for a given calculation is therefore dictated by a compromise between an acceptable level of numerical noise and the desired spatial resolution (which is $\\approx h\\propto 1/N_N^{1/d}$ in $d$ dimensions) and level of accuracy.   \\subsection{Results of Recent Test Calculations}  The authors and their collaborators have performed a series of systematic tests to evaluate the effects of spurious transport in SPH calculations. These tests are presented in detail in Lombardi et al.\\ (1999), while here we summarize the main results. Our tests include (i) particle diffusion measurements, (ii) shock-tube tests, (iii) numerical viscosity measurements, and (iv) measurements of the spurious transport of angular momentum due to AV in differentially rotating, self-gravitating configurations. The results are useful for quantifying the accuracy of the SPH scheme, especially for problems where shear flows or shocks are present, as well as for problems where true mixing is relevant.  Other recent tests of SPH include those by Hernquist \\& Katz (1989) and by Steinmetz \\& M\\\"uller (1993).  \\subsubsection{Particle Diffusion}  Many of our tests focus on spurious diffusion, the motion of SPH particles introduced as an artifact of the numerical scheme.  Often applications require a careful tracing of the particle positions, and in these cases it is essential that spurious diffusion be small.  For example, SPH simulations can be used to establish the degree of fluid mixing during stellar collisions, which is of primary importance in determining the subsequent stellar evolution of the merger remnants (see \\S 2.1).  It must be stressed that the amount of mixing determined by SPH calculations is always an upper limit. In particular, low-resolution calculations tend to be noisy, and this noise can lead to spurious diffusion of particles, independent of any real physical mixing of fluid elements.  We have analyzed spurious diffusion by using SPH particles in a box with periodic boundary conditions to model a stationary fluid of infinite extent. For various noise levels (particle velocity dispersions) and neighbor numbers $N_N$, we measure the rate of diffusion, quantified by the diffusion coefficient \\begin{equation} D\\equiv\\left\\langle{{\\d}\\Delta r^2\\over \\d t}\\right\\rangle. \\end{equation} Here the brackets $\\langle\\rangle$ denote a time average, and $\\Delta r=(\\Delta x^2+\\Delta y^2+\\Delta z^2)^{1/2}$ is the total distance traveled by a particle due to spurious diffusion. Although strong shocks and AV in SPH calculations can lead to additional particle mixing (Monaghan 1989), particle diffusion is the dominant contribution to spurious mixing in weakly shocked fluids.  Once expressed in terms of the number density of SPH particles and the sound speed, these diffusion coefficients can therefore be used to estimate spurious deviations in particle positions in a wide variety of applications, including self-gravitating systems. For each particle in some large-scale simulation, this spurious deviation is estimated simply by numerically integrating \\begin{equation} \\Delta r^2\\approx \\int{ D \\d t}. \\label{integral} \\end{equation} The coefficient $D$ in the integrand of equation (\\ref{integral}) depends on the particle's velocity deviation from the local flow, the local number density $n$ of particles, and the local sound speed $c_s$, so that these quantities need to be monitored for each particle during the simulation.  Such a scheme was successfully used to estimate spurious mixing in the context of stellar collisions (Lombardi et al. 1996), where typically (with $N=3\\times 10^4$ and $N_N\\approx 64$) the diffusion coefficient was very roughly $D\\sim 0.05 c_s n^{-1/3}$.  For sufficiently low noise levels, the diffusion coefficient essentially vanishes, as the particles simply oscillate around equilibrium lattice sites.  We say that such a system has ``crystallized.'' For a neighbor number $N_N\\approx 64$, a system of SPH particles will crystallize if the root mean square velocity dispersion is less than about 3--4\\% of the sound speed. We find that crystallized cubic lattices are unstable against perturbations, while lattice types with large packing fractions, such as hexagonal close-packed, are stable.  For this reason it may sometimes be better to construct initial data by placing particles in an hexagonal close-packed lattice, rather than in a cubic lattice as is often done.  The diffusion coefficients have been measured using equal-mass particles.  Sometimes, however, SPH simulations use particles of unequal mass so that less dense regions can still be highly resolved. To test the effects of unequal mass particles in a self-gravitating system, we constructed an equilibrium $n=1.5$ polytrope (a polytrope is an idealized model for a spherical star, characterized by a relation of the form $P=\\rho^\\gamma$ between pressure $P$ and density $\\rho$; the polytropic index $n$ is defined by $\\gamma=1+1/n$), using particle masses which increased with radius in the initial configuration. Allowing the system to evolve, we observed that the heaviest particles gradually migrated towards the center of the star, exchanging places with less massive particles.  For a polytrope modeled with $N\\approx 1.4\\times 10^4$ particles and a neighbor number $N_N\\approx 64$, the distribution of particle masses is reversed within roughly 80 dynamical timescales. This is caused by the interactions among neighboring particles via the smoothing kernel. These interactions allow energy exchange, and equipartition of energy then requires the heavier particles to sink into the gravitational potential well.  Spurious mixing is therefore a more complicated process in simulations which use unequal mass particles: each particle has a preferred direction to migrate, and in a dynamical application this direction can be continually changing.  For simulations in which fluid mixing is important, equal-mass particles are an appropriate choice.  \\subsubsection{Shock Tube Tests}  The diffusion tests just described are all done in the absence of shocks and without AV.  To test the AV schemes described in \\S1.2, we turn to a periodic version of the 1-D Riemann shock-tube problem.  Initially, fluid slabs with constant (and alternating) density $\\rho$ and pressure $p$ are separated by an infinite number of planar, parallel, and equally spaced interfaces.  We treat this inherently 1-D problem with both a 1-D and a 3-D SPH code. The 1-D code is naturally more accurate, and provides a benchmark against which we can compare the results of our 3-D code. In both cases, periodic boundary conditions allow us to model the infinite number of slabs.  Using various values of $\\alpha$ and $\\beta$, we performed a number of such shock tube calculations with our 3-D code, at both Mach numbers ${\\cal M}\\approx1.6$ and ${\\cal M}\\approx 13.2$. We then compared the time variation of the internal energy and entropy of the system against that of the 1-D simulation. Furthermore, since any motion perpendicular to the bulk fluid flow is spurious, we were also able to examine spurious mixing in these simulations. We find that all three forms of AV can handle shocks well.  For example, with $N=10^4$ and $N_N\\approx 64$, there is better than 2\\% agreement with the 1-D code's internal energy vs.~time curve when ${\\cal M}\\approx 1.6$, and agreement at about the 3\\% level when ${\\cal M}\\approx 13.2$. We also find that both equations (\\ref{pi}) and (\\ref{piDB}), as compared to equation (\\ref{pi2}), allow less spurious mixing and do somewhat better at reproducing the 1-D code's results.  Such simulations are a useful and realistic way to calibrate spurious transport, since the test problem, which includes shocks and significant fluid motion, has many of the same properties as real astrophysical problems. In fact, the recoil shocks in stellar collisions do tend to be nearly planar, so that even the 1-D geometry of the shock fronts is realistic.  The periodic boundary conditions play the role of gravity in the sense that they prevent the gas from expanding to infinity.  \\subsubsection{Shear Flows}  To test the various AV forms in the presence of a shear flow, we impose the so-called slipping boundary conditions on a periodic box, as is commonly done in molecular dynamics (see, e.g., Naitoh \\& Ono 1976). The resulting ``stationary Couette flow'' has a velocity field close to $(v_x,v_y,v_z)=(v_0y/L,0,0)$ and allows us to measure the numerical viscosity of the particles.  As in the shock tube tests, we also examine spurious mixing in the direction perpendicular to the fluid flow. These shear tests therefore allow us to further investigate the accuracy of our SPH code as a function of the AV parameters and scheme.  We find that both the Hernquist \\& Katz AV [eq.~(\\ref{pi2})] and the Balsara AV [eq.~(\\ref{piDB})] exhibit less viscosity than the classical AV [eq.~(\\ref{pi})].  However, the classical AV does allow significantly less spurious mixing than the other forms. For all three forms of the AV, increasing $\\alpha$ and $\\beta$ tends to damp out the noise and consequently decrease spurious mixing, but it also increases the spurious shear viscosity.  Rotation plays an important role in many hydrodynamic processes.  For instance, a collision between stars can yield a rapidly and differentially rotating merger remnant.  Even in the absence of shocks, AV tends to damp away differential rotation due to the relative velocity of neighboring particles at slightly different radii, and an initially differentially rotating system will tend towards rigid rotation on the viscous dissipation timescale. In systems best modeled with a perfect fluid, ideally with a viscous timescale $\\tau=\\infty$, any such angular momentum transport introduced by the SPH scheme is spurious.  As a concrete example, we consider an axisymmetric equilibrium configuration differentially rotating with an angular velocity profile $\\Omega(\\varpi) \\propto \\varpi^{-\\lambda}$, where $\\varpi$ is the distance from the rotation axis and $\\lambda$ is a constant of order unity. We then analytically estimate the viscous dissipation timescale for each of the three AVs discussed in \\S 1.2. These analytic estimates are found to closely match numerically measured values of the timescale.  Both the Hernquist \\& Katz AV [eq.~(\\ref{pi2})] and the Balsara AV [eq.~(\\ref{piDB})] yield longer viscous timescales than the classical AV [eq.~(\\ref{pi})], and hence are better at maintaining the angular velocity profile.  The Balsara AV clearly does best in this regard, with a viscous timescale roughly $N_N^{1/2}$ times larger than for the classical AV.  When choosing values of AV parameters, one must weigh the relative importance of shocks, shear, and fluid mixing. For this reason, it is an application-dependent, somewhat subjective matter to specify ``optimal values'' of $\\alpha$ and $\\beta$.  We do, however, roughly delineate the boundaries of the region in parameter space that gives acceptable results in Lombardi et al.\\ (1999).  Our results concerning the various AV forms can be summarized as follows (see Lombardi et al.\\ 1999 for more details). We find that the AVs defined by equations (\\ref{pi}) and (\\ref{piDB}) do equally well both in their handling of shocks and in their controlling of spurious mixing, and do slightly better than equation (\\ref{pi2}).  Furthermore, both equations (\\ref{pi2}) and (\\ref{piDB}) do introduce less numerical viscosity than equation (\\ref{pi}). Since equation (\\ref{piDB}), Balsara's form of AV, does indeed significantly decrease the amount of shear viscosity without sacrificing accuracy in the treatment of shocks, we conclude that it is an appropriate choice for a broad range of problems.  This is consistent with the successful use of Balsara's AV reported by Navarro \\& Steinmetz (1997) in their models of rotating galaxies. ",
        "conclusions": ""
    },
    "9805/astro-ph9805106_arXiv.txt": {
        "abstract": "It is shown how laboratory experiments performed with high intensity femtosecond lasers can probe the physics of black holes in the near-horizon regime. The acceleration generated by the high intensity laser ranging from $10^{13}$g to more than $10^{18}$g is identified with the gravitational acceleration at stretched horizons. In the black-hole's asymptotic region, the stretched-horizon-reflected light shows a measurable universal phase acceleration of $c^4/4GM$.  \\noindent PACS numbers: 04.70.Bw, 97.60.Lf, 52.40.Nk, 52.50.Jm ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805112_arXiv.txt": {
        "abstract": "We present a deep X-ray observation of the young Galactic supernova remnant Cas\\,A, acquired with the ROSAT High Resolution Imager\\@.  This high dynamic range (232 ks) image reveals low-surface-brightness X-ray structure, which appears qualitatively similar to corresponding radio features.  We consider the correlation between the X-ray and radio morphologies and its physical implications.   After correcting for the inhomogeneous absorption across the remnant, we performed a point by point (4\\arcsec\\ resolution) surface brightness comparison between the X-ray and radio images.  We find a strong (r = 0.75) log-log correlation, implying an overall relationship of $\\log(\\Sigma_{_{\\rm X-ray}}) \\propto (2.21\\pm0.05) \\times \\log(\\Sigma_{_{\\rm radio}})$\\@.  This is consistent with proportionate partition (and possibly equipartition) between the local magnetic field and the hot gas --- implying that Cas\\,A's plasma is fully turbulent and continuously amplifying the magnetic field. ",
        "introduction": "A comparison of the X-ray and radio emission of young supernova remnants (SNRs) provides a powerful tool for investigating the physical relationships among the thermal plasma, cosmic ray electrons and the magnetic field (e.g., \\markcite{1982ApJ...257..145D}{Dickel et al.\\ 1982}; \\markcite{1984ApJ...287..295M}{Matsui et al.\\ 1984}; \\markcite{1990ApJ...350..266A}{Arendt et al. 1990}; \\markcite{1998ApJ...Dyer} {Dyer \\& Reynolds 1998})\\@.  Radio emission is governed primarily by the density of relativistic electrons and the magnetic field strength, while the intensity in soft X-rays is dictated by the gas density.   Shell SNRs provide excellent laboratories for investigating the interaction among these physical processes, with Cas\\,A being the natural launching point for such investigations, because of its high surface brightness in both the radio and X-ray wavelength regimes.  Previous work has shown that the soft X-ray morphology of Cas\\,A, on angular scales $\\gtrsim$30\\arcsec\\ (or 0.5 pc at a distance of 3.4 kpc), is dominated by absorption effects (\\markcite{1996ApJ...466..309K}{Keohane, Rudnick, \\& Anderson 1996}, hereafter \\markcite{1996ApJ...466..309K}{KRA})\\@.  When this absorption is taken into account, a higher degree of intrinsic correlation between the X-ray and radio images is found.  Similarly, \\markcite{1994PASJ...46L.151H}{Holt et al.\\ (1994)} found a correlation between the radio and the hard X-ray ($E$$>$$4.5$\\,keV) surface brightness, which is not affected by absorption.  These correlations demonstrate that a simple relationship may exist among the underlying physical parameters.  In this Letter we investigate the relationship between the radio and X-ray emission on smaller angular scales.  We use a newly acquired ROSAT High Resolution Imager (HRI) image of Cas\\,A, with a 40 times longer exposure than used in \\markcite{1996ApJ...466..309K}{KRA}\\@. To account for the inhomogeneous column density towards Cas\\,A, we ``deabsorb'' the image using H{\\sc~i} and OH absorption data (\\S\\ref{deabsorption.sec})\\@.  We perform a point by point (4\\arcsec\\ resolution) surface brightness comparison between the X-ray and radio images and find a statistically significant correlation (\\S\\ref{radio_comparison.sec})\\@.  We discuss the physical implications of this correlation (\\S\\ref{discussion}) and its implications for future observational and theoretical studies of Cas\\,A and other young SNRs (\\S\\ref{conclusion})\\@.   \\pagebreak[3] ",
        "conclusions": "\\label{conclusion} We have performed a comparison of Cas\\,A's radio and X-ray emission to a limiting resolution of 4\\arcsec\\ (0.07 pc)\\@.  The strong correlation between radio and X-ray surface brightness can be explained by the scenario that Cas\\,A has ``on average'' a spatially constant relativistic electron density and proportionate partition on small scales between its thermal and magnetic energy densities, as would be expected from a fully turbulent MHD system.  These results may have important implications for theoretical work. The complex plasma physics of SNRs often must be simplified in order for computer simulations to run in a cost-effective manner. Unfortunately, most often one assumption is spherical or cylindrical symmetry, which is incompatible with turbulent flow.   It may be possible, instead, to use the proportionate partition relation suggested by our result, along with some other characteristics of fully turbulent systems, to produce a more realistic first-order description of young, core-collapse, supernova remnants."
    },
    "9805/astro-ph9805324_arXiv.txt": {
        "abstract": "We present axisymmetric dynamical models of the edge-on S0 galaxy NGC~4342.  This small low-luminosity galaxy harbors, in addition to its outer disk, a bright nuclear stellar disk.  A combination of observations from the ground and with the Hubble Space Telescope (HST) has shown that NGC~4342 rotates rapidly and has a strong central increase in velocity dispersion.  We construct simple two-integral Jeans models as well as fully general, three-integral models.  The latter are built using a modified version of Schwarzschild's orbit-superposition technique developed by Rix \\etal and Cretton \\etal These models allow us to reproduce the full line-of-sight velocity distributions, or `velocity profiles' (VPs), which we parameterize by a Gauss-Hermite series. The modeling takes seeing convolution and pixel binning into account.  The two-integral Jeans models suggest a black hole (BH) mass between $3$ and $6\\times 10^8 \\Msun$, depending on the data set used to constrain the model, but they fail to fit the details of the observed kinematics. The three-integral models can fit all ground-based and HST data simultaneously, but only when a central BH is included.  Models without BH are ruled out to a confidence level better than $99.73$ per cent. We determine a BH mass of $3.0^{+1.7}_{-1.0} \\times 10^8 \\Msun$, where the errors are the formal 68.3 per cent confidence levels. This corresponds to 2.6 per cent of the total mass of the bulge, making NGC~4342 one of the galaxies with the highest BH mass to bulge mass ratio currently known.  The models that best fit the data do not have a two-integral phase-space distribution function. They have rather complex dynamical structures: the velocity anisotropies are strong functions of radius reflecting the multi-component structure of this galaxy.  When no central BH is included the best fit model tries to fit the high central velocity dispersion by placing stars on radial orbits. The high rotation velocities measured, however, restrict the amount of radial anisotropy such that the central velocity dispersion measured with the HST can only be fit when a massive BH is included in the models. ",
        "introduction": "Several lines of evidence suggest that active galactic nuclei (AGNs) are powered by accretion onto a super-massive black hole (BH) (Lynden-Bell 1969; Rees 1984). The much higher volume-number density of AGNs observed at redshift $z \\approx 2$ than at $z=0$, suggests that many quiescent (or `normal') galaxies today must have gone through an active phase in the past, and therefore harbor a massive BH as well. Such a BH will significantly influence the dynamics of the galaxy inside a radius of influence, $r_{\\rm BH} = G M_{\\rm BH}/\\sigma^2$, where $\\sigma$ is a characteristic velocity dispersion of the stars in the center. In particular, hydrostatic equilibrium requires that the rms velocities of the stars surrounding a massive BH follow an $r^{-1/2}$ power-law (Bahcall \\& Wolf 1976; Young 1980).  Since the late 70s, combined imaging and spectroscopy of the central regions of galaxies has suggested that massive BHs should be present in a number of early-type galaxies (see Kormendy \\& Richstone 1995 for a review). Conclusive dynamical evidence for the presence of a central BH requires that a model with a BH can fit all observations (photometric and kinematic), and that no model without a BH can provide an equally good fit. Such conclusive evidence can only be inferred from observations that probe well inside the radius where the BH dominates the dynamics. Up to a few years ago, most claimed BH detections were based on observations with spatial resolutions of similar size as the radii of influence of the inferred BH masses (Rix 1993). This, together with the limited amount of freedom in the models used to interpret the data, has hampered an unambiguous proof for the presence of these BHs (i.e., the observed kinematics could not be confronted with all possible dynamical configurations without a BH). Often spherical models were used even when the observed flattening was significant. If the models were axisymmetric, the distribution function (hereafter DF) was often assumed to depend only on the two classical integrals of motion, energy and vertical angular momentum; $f = f(E,L_z)$. This implies that the velocity dispersions in the radial and vertical directions are equal (i.e., $\\sigma_R = \\sigma_z$). It is well-known that strong radial anisotropy in the center of a galaxy results in a high central velocity dispersion, mimicking the presence of a massive BH (cf. Binney \\& Mamon 1982). Conclusive evidence for a BH therefore requires that one can rule out radial anisotropy as the cause of the high velocity dispersions measured, and models must thus be sufficiently general.  Recently two major breakthroughs have initiated a new era in the search for massive BHs in normal galaxies. First of all, we can now obtain kinematics at much higher spatial resolution (down to FWHM $\\sim 0.1''$), using specially-designed spectrographs, such as the Subarcsecond Imaging Spectrograph (SIS) on the Canada-France Hawaii Telescope, or the Faint Object Spectrograph (FOS) and STIS aboard the HST. This allows us to probe the gravitational potential much closer to the center, where the BH dominates the dynamics. Not only has this improved the evidence for massive BHs in several old BH-candidate galaxies (M31, Ford \\etal 1998; M32, van der Marel \\etal 1997, 1998; M87, Harms \\etal 1994, Macchetto \\etal 1997; NGC~3115, Kormendy \\etal 1996a; NGC~4594, Kormendy \\etal 1996b), but it has also provided new cases (M84, Bower \\etal 1998; NGC~3377, Kormendy \\etal 1998; NGC~3379, Gebhardt \\etal 1998; NGC~4261, Ferrarese, Ford \\& Jaffe 1996; NGC~4486B, Kormendy \\etal 1997; NGC~6251, Ferrarese, Ford \\& Jaffe 1998; and NGC~7052, van der Marel \\& van den Bosch 1998).  Secondly, the revolutionary increase in computer power has made it possible to investigate a large number of fully general, three-integral models based on the orbit-superposition method (Schwarzschild 1979). In the past decade, this method has been used to build a variety of spherical, axisymmetric and triaxial models (e.g., Schwarzschild 1982; Pfenniger 1984; Richstone \\& Tremaine 1984, 1988; Zhao 1996).  Levison \\& Richstone (1985), Richstone \\& Tremaine (1985), and Pfenniger (1984) showed how to include rotation velocities and velocity dispersions as kinematic constraints.  More recently, Rix \\etal (1997) and Cretton \\etal (1998) extended this modeling technique even further by fitting to the {\\it entire} velocity profiles (see also Richstone 1997).  Van der Marel \\etal (1997, 1998) used this to build fully general, axisymmetric models of M32, and showed convincingly that M32 harbors a massive BH.  Recent review papers on this rapidly evolving field include Ford \\etal (1998), Ho (1998), Richstone (1998), and van der Marel (1998).  In many galaxies where the presence of a BH has been suggested, a nuclear disk, seen close to edge-on, is present. These disks are either in gaseous form (M84, M87, NGC~4261, NGC~4594, NGC~6251, NGC~7052), or made up of stars (NGC~3115). It is easier to detect BHs in edge-on systems with disks, where one can use both the measured rotation velocities and the velocity dispersions to determine the central mass density. It is therefore not surprising that BHs have predominantly been found in galaxies with nuclear disks. Furthermore, nuclear disks allow a good determination of the central mass density of their host galaxies.  Gaseous disks have the advantage that their kinematics can be easily measured from emission lines. Since gas in a steady-state disk can only move on non-intersecting orbits, the measured rotation velocities of a settled gas disk, in the equatorial plane of an axisymmetric potential, correspond to the circular velocities, $V_c(R) = \\sqrt{R {\\rm d}\\Phi/{\\rm d}R}$.  The rotation curve of a nuclear gas disk therefore provides a direct measure of the central potential gradient, and thus of the central mass density. However, often the gas disks are not in a steady state; many show a distorted morphology (e.g. M87, see Ford \\etal 1994), and non-gravitational motion, such as outflow, inflow or turbulence can be present and complicate the dynamical analysis (e.g., NGC~4261, Jaffe \\etal 1996; NGC~7052, van den Bosch \\& van der Marel 1995). Nuclear {\\it stellar} disks do not suffer from this, but have the disadvantage that their kinematics are much harder to measure.  First of all, the kinematics have to be determined from absorption lines rather than emission lines, and secondly, the line-of-sight velocity distributions, or velocity profiles (VPs), measured are `contaminated' by light from the bulge component.  However, van den Bosch \\& de Zeeuw (1996) showed that with sufficient spatial and spectral resolution one can resolve the VPs in a broad bulge-component and a narrow disk-component. From these VPs the rotation curve of the nuclear disk can be derived, providing an accurate measure for the central mass density. Therefore, galaxies with an embedded nuclear disk (either gaseous or stellar) observed close to edge-on are ideal systems to investigate the presence of massive BHs.  In this paper we discuss the case of NGC~4342; a small, low-luminosity ($M_B = -17.47$) S0 galaxy in the Virgo cluster. The galaxy is listed as IC~3256 in both the Second and Third Reference Catalogues of Bright Galaxies, since in the past it has occasionally been confused with NGC~4341 and NGC~4343 (see Zwicky \\& Herzog 1966).  At a projected distance of $\\sim 30''$ SE of NGC~4342, a small galaxy is visible. It is uncertain whether this is a real companion of NGC~4342 or whether it is merely close in projection. HST images of NGC~4342 revealed both an outer disk, as well as a very bright nuclear stellar disk inside $\\sim 1''$ (van den Bosch \\etal 1994; Scorza \\& van den Bosch 1998). It is a normal galaxy, with no detected ISM (Roberts \\etal 1991), and with small color-gradients (van den Bosch, Jaffe \\& van der Marel 1998, hereafter BJM98).  For its size and luminosity, it does however reveal a remarkably large central velocity dispersion and a very steep rotation-curve (see BJM98). Unfortunately, the spectral resolution of the available kinematic data is insufficient to actually resolve the VPs in disk and bulge components. In order to determine the central mass density in NGC~4342, we thus have to construct dynamical models of the entire system: bulge and disk components.  Here we present simple two-integral Jeans models as well as fully general three-integral models, and we provide evidence for the presence of a central massive dark object (MDO) of $\\sim 3\\times 10^8 \\Msun$. Throughout this paper we assume the MDO to be a BH, but we discuss alternatives in Section~7.2  In Section~2 we briefly discuss the data used to constrain the models and in Section~3 we describe our mass model. In Section~4 we show the results of some simple two-integral modeling, and we discuss its shortcomings. Section~5 describes the general outline of the three-integral modeling technique. In Section~6 we discuss shortcomings of the velocity-profile parameterization used when applied to dynamically cold systems, and present a modified approach. The results of the three--integral modeling are discussed in Section~7. Finally, in Section~8, we sum up and present our conclusions. Throughout this paper we adopt a distance of 15 Mpc for NGC~4342, consistent with the distance of the Virgo cluster (Jacoby, Ciardullo \\& Ford 1990). ",
        "conclusions": "Spectra obtained with the WHT and HST/FOS of the edge-on S0 galaxy NGC~4342 have revealed a very steep central rotation curve and a strong central increase in velocity dispersion. These data suggest a large central mass concentration. In this paper we presented detailed dynamical models of NGC~4342 used to investigate whether its nucleus harbors a massive BH.  We model the luminous density distribution of NGC~4342 with multiple Gaussian components. After projection and PSF convolution this model provides an excellent fit to the HST $I$-band surface brightness distribution. The parameters of this model were derived with the MGE method.  Simple isotropic Jeans models suggest that NGC~4342 harbors a massive BH of a few times $10^8 \\Msun$. The actual mass of the BH depends on the data-set fitted: the WHT data suggest $M_{\\rm BH} \\approx 3 \\times 10^8 \\Msun$; the HST/FOS data suggest a somewhat larger BH mass of $\\sim 6 \\times 10^8 \\Msun$. This discrepancy already suggests that the assumptions underlying the Jeans models, i.e., $f=f(E,L_z)$ and therefore $\\sigma_R = \\sigma_z$, are incorrect. This is also evident from the fact that the Jeans models cannot accurately fit the major-axis rms velocities measured with the WHT. These rms velocities are independent of the freedom in the anisotropy $\\sigma_{\\phi}/\\sigma_{R}$ allowed in the Jeans modeling. We find that for a mass-to-light ratio $\\Upsilon_I = 6.2 \\Msun/\\Lsun$ and an isotropic velocity distribution ($\\sigma_R = \\sigma_z = \\sigma_{\\phi}$) the Jeans model with $M_{\\rm BH} = 3 \\times 10^8 \\Msun$ provides the best fit to the observed WHT velocity dispersions along the major axis.  However, the rotation velocities are not very well fitted and we find a correlation between $V_{\\rm obs}/V_{\\rm mod}$ and the local ellipticity of the projected surface brightness, such that the model underpredicts the rotation velocities in the highly flattened regions (dominated by the disk light) and overpredicts them in the less flattened region (dominated by the bulge light). This suggests that the different components in NGC~4342 have different velocity anisotropies.  We thus constructed three-integral axisymmetric models of NGC~4342 in order to examine the mass of a possible BH and the dynamical structure of the different components. The modeling technique is an extension of Schwarzschild's orbit-superposition technique, and is based on finding the ensemble of orbits that best fits the observations. These models make no assumption about the dynamical structure and are fully general. This technique, developed by Rix \\etal (1997) and Cretton \\etal (1998), has previously been used to prove the existence of a massive BH of $(3.4 \\pm 0.7) \\times 10^6 \\Msun$ in the compact elliptical M32 (van der Marel \\etal 1998). We have constructed a range of dynamical ($M_{\\rm BH},\\Upsilon_I$)-models of NGC~4342 to determine a central BH mass of $3.0^{+1.7}_{-1.0} \\times 10^8 \\Msun$ and an $I$-band mass-to-light ratio of $6.3^{+0.5}_{-0.4} \\Msun/\\Lsun$. The high spatial resolution of the HST/FOS data allow us to rule out models without a BH to a confidence level better than $99.73$ per cent.  With a similar confidence we can rule out models with a BH more massive than $7 \\times 10^8 \\Msun$. This upper limit on the BH mass is mainly due to the VP shape parameters $h_3$ and $h_4$.  With the current data we can not rule out alternatives to a massive BH, such as a cluster of brown dwarfs or stellar remnants.  Nevertheless, the QSO paradigm, together with the fact that the presence of massive BHs in galactic nuclei has unambiguously been demonstrated in a few galaxies were the inferred central densities are high enough to rule out dark clusters as alternatives (see discussion in Maoz 1997), make the interpretation of the inferred MDO in NGC~4342 in terms of a massive BH the most likely.  We computed the intrinsic mean velocities and velocity dispersions of the three-integral models. The dynamical structures of the best fitting models vary strongly with radius, reflecting the multi-component structure of NGC~4342. Between $2''$ and $12''$ in the equatorial plane the best fitting models change from azimuthally anisotropic to radially anisotropic, while $\\sigma_z / \\sigma_R \\approx 0.9$. This explains the correlation between the projected ellipticity and the failure of the isotropic Jeans models to fit the observed rotation velocities along the major axis. The bulge in the best fitting model without BH is radially anisotropic. However, we have shown that even without the constraints of the measured HST/FOS rotation velocities, models without BH cannot fit the central HST/FOS velocity dispersion. The rotation velocities measured from the ground already constrain the amount of central radial anisotropy such that models without a BH cannot fit the high central velocity dispersion measured with the HST/FOS.  The BH mass thus derived contributes a fraction of $2.6^{+1.5}_{-0.9}$ per cent to the total mass of the bulge ($1.2 \\times 10^{10} \\Msun$). With this BH mass, NGC~4342 has one of the highest ratios of BH mass over bulge mass. Currently, the BH in our own galaxy has, with $0.02$ per cent, the lowest BH mass to bulge mass ratio known: the scatter in the $M_{\\rm BH}$ vs. $M_{\\rm bulge}$ relation seems to be as large as two orders of magnitude.  Extremely high spatial resolution is required in order to investigate if other galaxies have even lower values of $M_{\\rm BH}/M_{\\rm bulge}$. In conclusion, current data are consistent with a relation between bulge mass and BH mass, but the scatter is very large, and it is likely that the current $M_{\\rm BH}/M_{\\rm bulge}$ ratios found are biased towards an upper limit. Although the newly installed Space Telescope Imaging Spectrograph (STIS) is likely to detect many more BH cases in the coming years, detection of BHs with masses of the order of a $0.02$ per cent of the bulge mass or less in galaxies in Virgo or beyond, will probably have to await a next generation space telescope."
    },
    "9805/astro-ph9805054_arXiv.txt": {
        "abstract": "Since the identification of these stars by Morgan et al. in 1943, various definitions have been proposed for the stars of the Lambda Bootis group. We present here the various definitions which have been given to these objects in order to induce a general discussion on this topic. ",
        "introduction": "\\label{intr} The criteria to detect this class of peculiar A-type stars rely upon the choices made by various authors in the last 50 years. Therefore several definitions of lambda Boo stars are found in the literature.  Both photometric and spectroscopic criteria have been used. Some definitions so far proposed, concern only stars of spectral type near A0 while others include A and F stars; no restrictions appear about the luminosity class and therefore the evolutionary stage of the lambda Boo stars.  The common character of these stars, according to the various definitions, is the weakness of the metallic lines; however requirements such as high $v\\sin i$ and deficiency of specific elements are introduced by some, but not all authors; the same remark concerns their kinematic properties.  We shall present below the criteria used by various authors in order to understand the differences between the different lists of such stars published up to now, and to open a discussion for the future. ",
        "conclusions": ""
    },
    "9805/astro-ph9805262_arXiv.txt": {
        "abstract": "We have performed a detailed \\vvmax test for a sample of the Canada-France Redshift Survey (CFRS) for the purpose of examining whether the comoving number density of field galaxies changes significantly at redshifts of $z<1$.  Taking into account the luminosity evolution of galaxies which depends on their morphological type through different history of star formation, we obtain \\avmax$\\approx 0.5$ in the range of $0.3<z<0.8$, where reliable redshifts were secured by spectroscopy of either absorption or emission lines for the CFRS sample. This indicates that a picture of mild evolution of field galaxies without significant mergers is consistent with the CFRS data. Early-type galaxies, selected by their $(V-I)_{\\rm AB}$ color, become unnaturally deficient in number at $z>0.8$ due to the selection bias, thereby causing a fictitious decrease of \\avmax. We therefore conclude that a reasonable choice of upper bound of redshift $z\\sim 0.8$ in the \\vvmax test saves the picture of passive evolution for field ellipticals in the CFRS sample, which was rejected by Kauffman, Charlot, \\& White (1996) without confining the redshift range. However, about 10\\% of the CFRS sample consists of galaxies having colors much bluer than predicted for irregular galaxies, and their \\avmax is significantly larger than 0.5.  We discuss this population of extremely blue galaxies in terms of starburst that has just turned on at their observed redshifts. ",
        "introduction": "The standard scenario of formation of elliptical galaxies is an initial starburst in dissipative gas collapse at very high redshift, followed by passive luminosity evolution to the present (Larson 1974; Tinsley \\& Gunn 1976).  A galaxy evolution model of stellar population synthesis based on such a scenario well reproduces the present spectral energy distribution (SED) of elliptical galaxies and naturally explains their color--magnitude relation owing to the galactic wind which stops the starburst (Arimoto \\& Yoshii 1987; Matteucci \\& Tornamb\\'{e} 1987).  An obvious consequence of the standard scenario is that the comoving number density of elliptical galaxies does not change at redshifts of $z<1$. However, Kauffmann, Charlot, and White (1996, hereafter KCW) recently performed a \\vvmax test (Schmidt 1968) for a sample of ellipticals in the Canada-France Redshift Survey (CFRS, Lilly et al. 1995a), and reported a striking result that their number density should significantly decrease towards $z\\sim 1$. Such a number evolution of field ellipticals obviously contradicts with the standard scenario and might be explained alternatively by mergers of smaller stellar systems and/or gaseous disks (Toomre \\& Toomre 1972; Kauffmann, White, \\& Guiderdoni 1993; Baugh, Cole, \\& Frenk 1996).  In KCW's analysis, elliptical galaxies are selected from the CFRS sample if their observed $(V-I)_{\\rm AB}$ colors are redder than a color boundary which separates ellipticals from other morphological types in the $(V-I)_{\\rm AB}-z$ diagram.  They placed the boundary using a Bruzual--Charlot model of population synthesis with 0.1 Gyr single starburst and 50 \\% solar metallicity.  This boundary certainly gives a reasonable fraction of ellipticals which agrees with {\\it local} galaxy surveys.   However, the choice of burst duration and metallicity is only {\\it ad hoc} without any physical basis, therefore rendering some doubts as to whether the KCW's color boundary works successfully up to $z\\sim 1$.  On the other hand, local galaxies are known to have different colors from earlier to later types along the Hubble sequence, and this color difference is attributed to type-dependent variation of star formation history in galaxies (Tinsley 1980; Kennicutt, Tamblyn, \\& Congdon 1994).  Therefore, the evolution models which reproduce the present colors would give a more natural way to distinguish various types in the $(V-I)_{\\rm AB}-z$ diagram. In this Letter, using the type-dependent evolution models of galaxies developed by Arimoto \\& Yoshii (1987, hereafter AY) and Arimoto, Yoshii, \\& Takahara (1992, hereafter AYT), we perform a \\vvmax test for the whole CFRS sample consisting of various types as a mixture. ",
        "conclusions": "In addition to the selection bias in the CFRS mentioned in the previous section, our use of AY model for E/S0 is also responsible for the conclusion presented here which is in sharp contrast with KCW's conclusion. The behavior of KCW's evolution model at $z<0.6$ is similar to the AY model, except for $\\sim$ 0.5 mag blueward shift probably due to their choice of a constant value of lower metallicity for all stellar populations.  It should be noted that an average of absolute magnitude for the CFRS galaxies of E/S0 type is $M_{I_{AB}} = -21.9$, after corrected to $z=0$, which obviously corresponds to the AY $10^{11-12}M_\\odot$ (baryon) models with the luminosity-weighted average of stellar metallicities equal to $130-180\\%$ of the solar (see Table 3 of AY).  On the other hand, KCW's use of much lower metallicity of $30-70\\%$ of the solar is equivalent to using the AY $10^{9-10}M_\\odot$ (baryon) model with $M_{I_{AB}}\\sim -17$ which corresponds to much smaller E/S0 galaxies not observed in the CFRS.  This suggests that KCW's evolution model is inappropriate to place a color boundary between E/S0 and other types for the CFRS sample.  In fact, KCW have also noticed the selection bias against early-type galaxies at redshifts close to unity in the CFRS sample.  Since the $(V-I)_{AB}$ colors are available for the CFRS galaxies with no redshift identifications, KCW evaluated a maximum redshift for which each unidentified galaxy would still lie above the KCW's curve in Fig. 1 (the dotted line) and be classified as early type.  They performed the \\vvmax test including the unidentified galaxies with these maximum redshifts and obtained \\avmax = 0.451 which is still smaller than 0.5. It should be noted that KCW's estimte of the maximum redshifts is heavily based on the steep rise of their model at $z \\gtilde 0.6$ in Fig. 1. In contrast, our models of E/S0 and Sab, which are considered to be more appropriate for analyzing the CFRS sample, do not show such a steep rise probably because of the longer duration of starbursts (0.7Gyr compared to 0.1 Gyr in KCW) and hence it is difficult or even impossible to define the maximum redshift for the unidentified, red galaxies.  Therefore, we consider that the estimate of the maximum redshifts is highly uncertain, and the incompleteness should be avoided by performing the \\vvmax test only in the range in which spectroscopic redshifts are secured with confidence.  Let us discuss more about the nature of EBGs, which are distributed widely in the redshift range of $0<z<1$ (Fig. \\ref{fig:color-z}).  The fact that their \\avmax value is larger than 0.5 suggests that the EBGs, soon after observed at $z$, must disappear or change their color to be classified as redder galaxies.  A likely mechanism for such sudden fading or reddening is that the EBGs are being observed during intensive starburst as inferred from their extremely blue colors.  In either cases the evolution of the EBGs may not be predicted by the usual Sdm models including ours, in which  the star formation rate is nearly constant to make the present colors as blue as local Sdm galaxies.  However, if the star formation is suddenly stopped on the way, the colors turn considerably redder than predicted by the Sdm model. It is therefore tempting to examine whether such starburst models for the EBGs lead to \\avmax$\\sim 0.5$ for the whole CFRS galaxies.  In this letter, we have performed a detailed \\vvmax test for the CFRS sample taking into account the AY and AYT models of galaxy evolution. Our hypothesis that all field galaxies evolve mildly according to type-dependent star formation histories without significant mergers is consistent with the CFRS data in $0.3<z<0.8$ excluding EBGs, especially in the $\\Lambda$-dominated universe. Consequently, the rejetion of passive evolution of ellipticals claimed by KCW is not valid. The CFRS data do not allow us to discriminate between the passive evolution and the number evolution of ellipticals claimed by KCW.  If we parametrize the evolution of comoving number density of E/S0 galaxies as $\\propto (1+z)^\\gamma$, then the analysis in the limited redshift range of $0.3<z<0.8$ gives $\\gamma_{\\rm E/S0}=0.6\\pm 1.7$, $-0.1\\pm 1.7$, and $-0.8\\pm 1.7 $ for the EdS, open, and $\\Lambda$-dominated universe models, respectively. The errors are larger than KCW's errors, because the redshift range is limited in our analysis.  For all the CFRS galaxies excluding the EBGs, the analysis within the same redshift range gives $\\gamma_{\\rm all}=1.8\\pm 0.7$, $1.1\\pm 0.7$, and $0.5\\pm 0.7$ in the same order as above. These $\\gamma$ values, though not so significant as claimed by KCW, would impose an important constraint on the number evolution of field galaxies in the framework of the hierarchical clustering of galaxies in the universe.  This work has been supported in part by the Grant-in-Aid for Scientific Research (3730) and Center-of-Excellence (COE) research (07CE2002) of the Ministry of Education, Science, and Culture of Japan. TT has been supported by JSPS Research Fellowships for Young Scientists.  \\begin{table} \\scriptsize \\begin{center} \\caption{The \\vvmax Statistics} \\begin{tabular}{cccccccc} \\hline  \\hline Cosmology & Galaxy & \\multicolumn{6}{c}{\\avmax \\tablenotemark{\\it a}} \\\\ \\cline{3-8} $(h, \\Omega_M, \\Omega_\\Lambda)$ &  Evolution &  E/S0  &  Sab  &  Sbc  &  Scd  &  Sdm  &  All Types  \\\\ \\hline (0.5, 1.0, 0.0) & on & .519 $\\pm$     .038 & .532 $\\pm$     .034 & .540 $\\pm$     .034 & .550 $\\pm$     .034 & .569 $\\pm$     .033 & .542 $\\pm$     .016   \\\\  (0.6, 0.2, 0.0)  & on & .494 $\\pm$     .038 & .508 $\\pm$     .034 & .530 $\\pm$     .033 & .524 $\\pm$     .033 & .575 $\\pm$     .034 & .526 $\\pm$     .016   \\\\  (0.7, 0.2, 0.8) & on & .478 $\\pm$     .039 & .496 $\\pm$     .036 & .518 $\\pm$     .034 & .513 $\\pm$     .033 & .547 $\\pm$     .033 & .511 $\\pm$     .016   \\\\  (0.7, 0.2, 0.8) & off & .471 $\\pm$     .038 & .501 $\\pm$     .038 & .537 $\\pm$     .033 & .536 $\\pm$     .031 & .539 $\\pm$     .037 & .517 $\\pm$     .016  \\\\ \\hline \\hline \\end{tabular} \\end{center} \\tablenotetext{a}{ Galaxy types used here are based on the color-aided classification (\\S 2) and are slightly different from the local surveys. The EBGs are not included in the statistics (see text for detail).}   \\end{table}"
    },
    "9805/astro-ph9805218_arXiv.txt": {
        "abstract": "We investigate the dependence of the local galaxy luminosity function on morphology using 5,404 galaxies from the recently enlarged Second Southern Sky Redshift Survey (SSRS2). Over the range $-22 \\le M_{B} \\le -14$ ($H_0 = 100\\,\\,\\rm km\\, s^{-1}\\,Mpc^{-1}$), the luminosity function of early-type galaxies is well fitted by a Schechter function with parameters $M_*=-19.37^{+0.10}_{-0.11}$, $\\alpha=-1.00^{+0.09}_{-0.09}$ and $\\phi_*=4.4\\pm 0.8 \\times 10^{-3} \\, \\rm{Mpc}^{-3}$.  The spiral luminosity function is very similar and is well fitted by the parameters $M_*=-19.43^{+0.08}_{-0.08}$, $\\alpha=-1.11^{+0.07}_{-0.06}$ and $\\phi_*=8.0 \\pm 1.4 \\times 10^{-3} \\, \\rm{Mpc}^{-3}$ over the same range in absolute magnitude.  The flat faint end of the early-type luminosity function is consistent with earlier measurements from the CfA Redshift Survey (Marzke \\etal 1994a) but is significantly steeper than the slope measured in the Stromlo-APM survey (Loveday \\etal 1992).  Combined with the increased normalization of the overall LF measured from intermediate redshift surveys, the flat faint-end slope of the E/S0 LF produces no-evolution models which reproduce the deep HST counts of early-type galaxies remarkably well.  However, the observed normalization of the SSRS2 LF is consistent with the low value measured in other local redshift surveys. The cause of this low-redshift anomaly remains unknown.  The luminosity function of irregular and peculiar galaxies in the SSRS2 is very steep: $M_* =-19.78^{+0.40}_{-0.50}$, $\\alpha=-1.81^{+0.24}_{-0.24}$ and $\\phi_*=0.2 \\pm 0.08 \\times 10^{-3} \\, \\rm{Mpc}^{-3}$. The steep slope at the faint end is consistent with the LFs measured for Sm-Im galaxies in the CfA survey, UV-selected galaxies (Treyer \\etal 1997), star-forming field galaxies (Bromley \\etal 1997) and the bluest galaxies in the SSRS2 (Marzke \\& da Costa 1997). As shown by Driver, Windhorst \\& Griffiths (1995$b$), the steep LF reduces the observed excess of faint irregulars over no-evolution predictions but cannot explain it entirely. ",
        "introduction": "The luminosity function (LF) of low-redshift galaxies is a primary benchmark for models of galaxy formation. The detailed shape of the LF reflects a host of physical processes ranging from the collapse of dark-matter halos to the complex cycle of gas cooling, star formation, and feedback into the interstellar medium.  The dependence of the LF on other observables such as morphology, color and emission-line strength provide further constraints on the input physics. Along with dynamical measures of galaxy masses, the LF for various types along the Hubble sequence anchors the theory of galaxy formation at low redshift.  Deep counts and redshift surveys of faint galaxies provide a wealth of constraints on galaxy evolution. Although recent surveys have been designed to cover a wide range in redshift and apparent magnitude (\\eg Lilly \\etal 1995 and Ellis \\etal 1996), limits on available telescope time force a compromise between the redshift baseline and the sampling at each redshift.  Surveys targeting intermediate-redshift galaxies are inevitably quite sparse at redshifts less than $z \\approx 0.1$. The Autofib survey (Ellis \\etal 1996) is perhaps the best compromise; the solid angle of that survey decreases with limiting magnitude in order to produce a catalog with similar numbers of galaxies in each broad redshift bin. However, in the regime $z \\le 0.1$, wide-angle surveys at relatively bright apparent magnitudes remain the best approach for measuring the local LF and its variations from one part of the local universe to another.  Very local samples ($z \\le 0.05$) allow the most detailed investigations of individual galaxy properties at the present-epoch; nearby galaxies are conveniently large and bright. The cost of this acuity is the unavoidable fact that smaller volumes are less likely to be fair samplings of the universe. Because fluctuations in the density field decrease with scale, samples covering larger volumes yield better estimates of the galaxy density.  Furthermore, the properties of individual galaxies correlate with local density on relatively small scales (Dressler 1980, Postman \\& Geller 1984, Maia \\& da Costa 1990, Hashimoto \\etal 1997). Correlations between galaxy properties and density on large scales are only vaguely understood (Park \\etal 1994, Willmer, da Costa \\& Pellegrini 1997), but it seems clear that fair estimates of the local LF and its dependence on galaxy properties requires averaging over a large number of large-scale structures.  Limits to our knowledge of the present-epoch universe are fundamental: in order to resolve the evolutionary history of the galaxy population from $\\sim 1\\,\\rm Gyr$ ago up to the present day, we are consigned by light-travel time alone to a region $\\sim 600 \\hmpc$ across.  Recent HST surveys of faint field galaxies have opened new windows on galaxy morphology at high redshift (Driver \\etal 1995$a,b$, Glazebrook \\etal 1995, Abraham \\etal 1996). Because different types of galaxies have quite different spectral energy distributions, apparent magnitude-limits impose redshift-dependent filters upon the observed distribution of galaxy types.  Predictions of the morphological composition of intermediate and high-redshift samples require (at the very least) a detailed understanding of the distribution of morphologies nearby.  In this paper, we measure the luminosity function and its dependence on galaxy morphology using the 5404 galaxies in the combined SSRS2 samples in the northern and southern Galactic caps. We summarize the data in $\\S2$ and review our computational techniques in $\\S3$. Section $\\S4$ summarizes the results, and in $\\S5$ we discuss the implications for deep HST counts.  We conclude in $\\S6$. ",
        "conclusions": "We have used a new sample of 5,404 galaxies with rough but complete morphological classifications to determine the galaxy luminosity function for different morphologies over the range $-22 \\le M_{B} \\le -14$.  We conclude that the faint-end slope for both E/S0 and spiral galaxies is essentially flat over this range.  The LF of irregular and peculiar galaxies is very steep ($M_*-19.78^{+0.40}_{-0.50}+5\\log{h}$, $\\alpha=1.81^{+0.24}_{-0.24}$, $\\phi_*=0.2 \\pm 0.08 \\times  10^{-3} h^3\\, \\rm{Mpc}^{-3}$). The faint-end slope of the Irr/Pec LF is consistent with earlier measurements from the CfA Redshift Survey, however, an excess of bright irregulars relative to the CfA LF leads to a brighter value of $M_*$ for this class in the SSRS2. This pattern appears in each morphological class and may be evidence that bright galaxies are generally underrepresented in the CfA Survey.  The flat faint-end slope of the E/S0 LF supports earlier claims that the Stromlo-APM LF under-represents faint early-type galaxies.  As a result, the no-evolution predictions of faint E/S0 counts based on the SSRS2 exceed the predictions based on the Stromlo-APM and, assuming that the high normalization obtained for the intermediate-redshift LFs is representative, the SSRS2 predictions are consistent with the observed counts of Es and S0s to $I=25$. As with other surveys to similar depths, however, the normalization obtained directly from the SSRS2 is low, and the explanation of this low-redshift anomaly remains elusive. Until the biases in the present-epoch luminosity function are better understood, the degree of evolution inferred from deep counts will remain uncertain."
    },
    "9805/astro-ph9805168_arXiv.txt": {
        "abstract": "Large-angle fluctuations in the cosmic X-ray background are investigated by a new formalism with a simple model of the X-ray sources. Our method is formulated from the Boltzmann equation and a simple extension of the work by Lahav et~al. to be applicable to a hyperbolic (open) universe. The low multipole fluctuations due to the source clustering are analyzed in various cosmological models in both numerical and analytic way. The fluctuations strongly depend on the X-ray sources evolution model, as pointed out previously. It turns out that the nearby ($z \\lesssim 0.1$) sources are the dominant contributors to the large-angle fluctuations. If these nearby sources are removed in an observed X-ray map, the dipole (low multipole) moment of the fluctuation drastically decreases. In this case the Compton-Getting effect of an observer's motion can be a dominant contribution to the dipole fluctuation. This feature of fluctuation, relating to the matter power spectrum, is discussed.  \\noindent {\\bf keywords} ~cosmology: theory --- cosmic X-ray background --- large-scale structure ",
        "introduction": "One of the most important problems in cosmology is to explain the formation of the cosmic structure. The discovery of anisotropies in the cosmic microwave background (CMB) supports the hypothesis that the structure of the Universe formed via gravitational instability from small density fluctuations (e.g.,~White, Scott, \\& Silk 1994). It is also argued that the cosmological models with cold dark matter (CDM) are favorable models with which to explain the large-scale structure of the galaxy distribution and the formation of the cosmic objects at the high-$z$ Universe (e.g.,~Dodelson, Gates, \\& Turner 1996; Peacock \\& Dodds 1994).  The origin of the cosmic X-ray background (CXB) radiation is also explored (e.g.,~Fabian \\& Barcons 1992;Hayashida 1989). The deep surveys of the 0.5-2 keV~X-ray range were carried out with ROSAT, and most of the background sources were resolved into AGN and other X-ray luminous galaxies (e.g., Hasinger 1997). From the isotropy of the observation, it was concluded that a good deal of the unresolved fraction of the CXB has come from redshifts $z>1$ (e.g., Carrera, Fabian \\& Barcons 1997). The conclusion may be supported in the higher X-ray energies band with ASCA observations (Georgantopoulos et al. 1997). The finite solution of the source problem will be obtained in the next generation of X-ray instruments (e.g., Charles \\& Seward 1995).  The CXB fluctuations are an important probe of the large-scale structure of the universe (~Barcons \\& Fabian 1988; Carrera et~al. 1997; Barcons, Fabian \\& Carrera 1997; Lahav, Piran \\& Treyer 1997). For example, the relation between the angular correlation function of the intensity fluctuations and the two-point spatial correlation function of galaxies $\\xi(r)$ was studied. (De~Zotti et~al. 1990; Martin-Mirones et~al. 1991; Persic et~al. 1989; Shafer 1983). The cosmological models with a cosmological constant can be tested by the cross-correlation between the X-ray and the microwave background(Boughn, Crittenden \\& Turok 1997).  Recently, Lahav et~al.(1997) have investigated the large-angle fluctuations in the CXB. They have calculated the expected values of the fluctuations in a statistical way. Treyer et~al.(1998) have compared the predicted CXB fluctuations with the HEAO1-A2 measurements. In the prediction of CXB fluctuations there are many uncertainties. The evolution of the X-ray sources are not completely understood. A simple (power-law) source evolution model is assumed in their paper (Lahav et al. 1997; Treyer et al.1998). Furthermore only a simple flat cosmological model is assumed. The difference in the cosmological model might yield the large difference in amplitude of the fluctuation. For example, if the X-ray sources at the high-$z$ universe are the dominant contributors to the angular fluctuations, the cosmological parameters, e.g., the curvature of the universe and the cosmological constant $\\Lambda$, are significant factors. It is therefore worth examining how the fluctuations depend on the cosmological parameters. We also develop a useful formalism which is applicable to a hyperbolic (open) universe in order to extend the work by Lahav et al. to various cosmological models, which is described in Section 2. Our formalism is based on the Boltzmann approach, which is familiar in analysis of the cosmic microwave anisotropies (e.g.,~Hu \\& Sugiyama 1995a;1995b). A simple model for the source distribution is introduced. In Section 3 we solve the Boltzmann equation and obtain the expression for the root mean square of multipole moments of the fluctuations. In Section 4 the fluctuations are analyzed in various cosmological models. Section 5 is devoted to summary and discussions.    Throughout this paper we will use the units $c=\\hbar=k_B=1$; however, we occasionally use the Planck constant $h_{\\rm P}(=2\\pi\\hbar)$ to make clear the meaning of equations. ",
        "conclusions": "In this paper, we have investigated the large-angle fluctuations in the CXB due to the X-ray source clustering. We have developed the formalism to describe the CXB fluctuations using Boltzmann equation under a simple model of the X-ray sources. Our formalism is a simple extension of that by Lahav et~al. (1997) to be applicable to an universe with hyperbolic geometry. The dependence of the fluctuations on the model parameters has been examined in various cosmological models. The fluctuation does not strongly depend on the cosmological parameters. It is quite sensitive to the parameters for the X-ray sources (Lahav et~al. 1997). The fluctuation is determined by the ratio of anisotropic X-ray flux to the isotropic one, i.e., $\\sqrt{C_l}\\sim \\Delta I_\\nu/ I^{(0)}_\\nu$. $\\Delta I_\\nu$ is essentially determined by the nearby sources at low redshift for the low multipole moment, while $I^{(0)}$ is done by the high-$z$ sources. The large redshift evolution of the X-ray emissivity (e.g., $p=3$, and $z_{\\rm max}$ is large) makes the flux from the far sources large and, as a result, the amplitude of fluctuation becomes small.  We have also pointed out the importance of the source distribution parameter $z_{\\rm min}$. The low multipole anisotropies are sensitive to it. This parameter is closely related to the reconstruction of the X-ray map by removing nearby bright sources. The dipole moment with $z_{\\rm min}=0.1$ is smaller by order of magnitude compared with the case $z_{\\rm min}=0$. If the sources sufficiently close to the our Galaxy $z~\\sim 0.1$ are removed, it may be expected that the effect of the peculiar motion of the observer (Compton-Getting (C-G) effect) is dominant in the CXB dipole.  Let us roughly compare the effects which contribute to the dipole anisotropies except for the source clustering effect. The observer's peculiar motion relative to the CMB was measured by using the COBE four-year data (Lineweaver et al. 1996), which gave the peculiar velocity $V_{\\rm obs}=368.9 \\pm 2.5 ~{\\rm km/s}$ in the direction ($l=264^{\\circ}; b=48^{\\circ}$). Assuming we have a similar motion relative to the CXB, the expected C-G dipole is estimated as $\\sqrt{C_{l=1}^{{\\rm CG}}}\\simeq 5.0 \\times 10^{-3}$.  On the other hand, it is well known that the shot noise fluctuation arises from the discreteness of the sources. Since this fluctuation is originated from the Poisson fluctuation of the discrete sources, the spectrum is white noise. The shot noise fluctuations in the CXB have been investigated by Lahav et al. (1997). Following their result, the amplitude of the fluctuation is estimated $\\sqrt{C_l^{\\rm SN}}\\simeq 1.2 \\times f_{\\rm m}^{1/4}$, where $f_{\\rm m}$ is a flux cut-off of bright removed sources in unit of ${\\rm erg~s^{-1}~cm^{-2}}$. If the flux cut-off level becomes lower, the amplitude of shot noise decreases. Thus the shot noise fluctuation depends on the flux cut-off in observational data and is important when comparing a theoretical model with the observed X-ray map (Treyer et al.~1998). The flux cut-off was $\\simeq 3\\times 10^{-11}~{\\rm erg~s^{-1}~cm^{-2}}$ in the HEAO-1, and the amplitude of shot noise is estimated as $\\sqrt{C_{l=1}^{{\\rm SN}}} \\simeq 2.8 \\times 10^{-3}$. Note also that the cut-off level is relevant to the parameter $z_{\\rm min}$ in the sense of bright source removability.  The dipole owing to the source clustering is shown in Fig.5 ($p=3, z_{max}=3$). In the case of $z_{\\rm min}=0$, the dipole anisotropy due to the clustering effect is comparable to the C-G effect. However, if $z_{\\rm min}$ is $O(0.1)$, the C-G dipole becomes well above the dipole due to the source clustering. If the flux cut-off becomes lower by the improvement of observation, the shot noise and the clustering effect could be sufficiently smaller than C-G effect to yield a good chance of measuring the C-G dipole in the CXB. This investigation also suggests that the careful treatment is required when comparing the observational map with the theoretical prediction. When the redshifts of the sources can not be determined, the subtraction of nearby X-ray sources from the observed map may contain a delicate problem because the nearby faint sources may contribute to the fluctuations sensitively. These problems are left as future problems."
    },
    "9805/astro-ph9805078_arXiv.txt": {
        "abstract": "We investigate the peculiar velocities predicted for galaxy clusters by theories in the cold dark matter family. A widely used hypothesis identifies rich clusters with high peaks of a suitably smoothed version of the linear density fluctuation field. Their peculiar velocities are then obtained by extrapolating the similarly smoothed linear peculiar velocities at the positions of these peaks. We test these ideas using large high resolution N--body simulations carried out within the Virgo supercomputing consortium. We find that at early times the barycentre of the material which ends up in a rich cluster is generally very close to a high peak of the initial density field. Furthermore the mean peculiar velocity of this material agrees well with the linear value at the peak. The late-time growth of peculiar velocities is, however, systematically underestimated by linear theory. At the time clusters are identified we find their {\\it rms} peculiar velocity to be about 40\\% larger than predicted. Nonlinear effects are particularly important in superclusters. These systematics must be borne in mind when using cluster peculiar velocities to estimate the parameter combination $\\sigma_8\\Omega^{0.6}$. ",
        "introduction": "\\label{intro} The motions of galaxy clusters are thought to result from gravitational forces acting over the very large scales on which superclusters are assembled. The {\\it rms} deviations from uniformity on such scales appear to be small, and so may be adequately described by the linear theory of fluctuation growth. For a linear density field of given power spectrum the {\\it rms} peculiar velocity is proportional to $\\sigma_8\\Omega^{0.6}$ where $\\Omega$ is the cosmic density parameter and $\\sigma_8$, the {\\it rms} mass fluctuation in a sphere of radius $8~h^{-1}$Mpc, is a conventional measure of the amplitude of fluctuations (e.g. Peebles 1993). (As usual the Hubble constant is expressed as $H_0 = 100\\,h\\,$km/sec/Mpc.)   Distance indicators such as the Tully-Fisher or $D_n$--$\\sigma$ relations allow the peculiar velocities of clusters to be measured, thus providing a direct estimate of this parameter combination (see, for example, Strauss \\& Willick 1996).  Essentially the same parameter combination can also be estimated from the {\\it abundance} of galaxy clusters (e.g. White et al.\\ 1993) and a comparison of the two estimates could in principle provide a check on the shape of the assumed power spectrum and on the assumption that the initial density field had gaussian statistics. In practice this is difficult because of the uncertainties in relating observed cluster samples to the objects for which quantities are calculated in linear theory or measured from N-body simulations. The standard linear model was introduced by Bardeen et al.\\ (1986; hereafter BBKS). It assumes that clusters can be identified with ``sufficiently'' high peaks of the linear density field after convolution with a ``suitable'' smoothing kernel. The peculiar velocity of a cluster is identified with the linear peculiar velocity of the corresponding peak extrapolated to the present day. In the present paper we study the limitations both of this model and of direct N-body simulations by comparing their predictions for clusters on a case by case basis.  In the next section we summarize both the linear predictions for the growth of peculiar velocities and the BBKS formulae for the values expected at peaks of the smoothed density field. Section 3 then presents our set of N-body simulations and outlines our procedures for identifying peaks in the initial conditions and clusters at $z=0$. Section 4 begins by studying how well peaks correspond to the initial barycentres of clusters; we then show that the smoothed linear velocity at a peak agrees well with the mean linear velocity of its cluster; finally we show that the growth of cluster peculiar velocities is systematically stronger at late times than linear theory predicts. A final section presents a brief discussion of these results. ",
        "conclusions": "\\label{summ}  We have investigated the peculiar velocities predicted for galaxy clusters by theories in the Cold Dark Matter family. A widely used hypothesis identifies rich clusters with high peaks of a smoothed version of the linear density fluctuation field. Their peculiar velocities are then obtained by extrapolating the similarly smoothed linear peculiar velocities at the positions of these peaks. We have tested this using a set of four large high--resolution N--body simulations. We identify galaxy clusters at $z=0$ and then trace the particles they consist of back to earlier times. In the initial density field, the barycenters of 70\\% and 80\\% of the clusters with masses exceeding $3.5 \\times 10^{14}\\,h^{-1}\\,M_{\\odot}$ lie within $4\\,h^{-1}$\\,Mpc (comoving) of a peak with $\\nu>1.5$ for the low and high $\\Omega$ models, respectively. Furthermore, the mean linear peculiar velocity of the material which forms a cluster at $z=0$ agrees well with the value at that peak.  However, the late--time growth of peculiar velocities is systematically underestimated by linear theory. At the time clusters are identified, i.e.\\ at $z=0$, we find that the {\\it rms} peculiar velocity is about 40\\% larger than predicted. Nonlinear effects are particularly important in superclusters; the {\\it rms} values for clusters which are members of superclusters are about 20\\% to 30\\% larger than those for isolated clusters."
    },
    "9805/astro-ph9805287_arXiv.txt": {
        "abstract": " ",
        "introduction": "The properties of the faint end of the luminosity function are poorly constrained outside the Local Group, both in terms of the numbers and characteristics of the galaxies. Some studies have found evidence for very steep luminosity functions in both cluster~\\cite{sdp97}\\cite{t98} and field~\\cite{lvdy97} environments. The importance of very low luminosity galaxies in understanding galaxy formation and evolution~\\cite{t98} demands that progress is made in identifying and studying low luminosity dwarfs in new environments~\\cite{cadcs98}.  This survey extends the study of extremely low luminosity galaxies to the Virgo Cluster (to M$_{\\rm R} = -11$~mag), into the regime of Local Group dwarf spheroidals. It extends the definitive study of the Virgo Cluster by Binggeli, Sandage~\\& Tammann~\\cite{bst85} by 3~mag beyond their completeness limits and complements the survey of large, low surface brightness galaxies of Impey, Bothun~\\& Malin~\\cite{ibm88}. By coadding multiple photographic exposures with the UK Schmidt Telescope, galaxies are detected having central surface brightnesses as faint as 25~R~mag~arcsec$^{-2}$, equivalent to 26.5~B~mag~arcsec$^{-2}$. We report here results from an initial survey~\\cite{ppsj98} covering~3.2~deg$^2$. ",
        "conclusions": ""
    },
    "9805/astro-ph9805308_arXiv.txt": {
        "abstract": "In this work we investigate the evolution of the mass function of the Galactic globular cluster system (GCMF) taking into account the effects of stellar evolution, two-body relaxation, disk shocking and  dynamical friction on the evolution of individual globular clusters. We have adopted a log-normal initial GCMF and  considered a wide range of initial values for the dispersion, $\\sigma$, and the mean value, $\\langle \\log M\\rangle$. We have studied in detail the dependence on the initial conditions of the final values of $\\sigma$, $\\mm$, of the fraction of the initial number of clusters surviving after one Hubble time, and of the difference between the properties of the GCMF of clusters closer to the Galactic center and the properties of those located in the outer regions of the Galaxy.  In most of the cases considered evolutionary processes alter significantly the initial population of globular clusters and the disruption of a significant number of globular clusters leads to a flattening in the spatial distribution of clusters in the central regions of the Galaxy.  The initial log-normal shape of the GCMF  is preserved in most cases and if a power-law in $M$ is adopted for the initial GCMF, evolutionary processes tend to modify it into a log-normal GCMF. The difference between  initial and final values of $\\sg$ and $\\mm$ as well as the difference between  the final values of these parameters for inner and outer clusters can be positive or negative depending on initial conditions. A significant effect of evolutionary processes does not necessarily give rise to a strong trend of $\\mm$ with the galactocentric distance. The existence of a particular initial GCMF able to keep its initial shape and parameters unaltered during the entire evolution through a subtle balance between disruption of clusters and evolution of the masses of those which survive, suggested in Vesperini (1997), is confirmed. ",
        "introduction": "Investigation of the luminosity function of the globular cluster system (hereafter GCLF) of our Galaxy and of globular cluster systems in external galaxies has been the subject of many observational works (see e.g. Secker 1992, McLaughlin 1994, Abraham \\& van den Bergh 1995, Kissler-Patig 1997, Harris 1991 and references therein) because of its relevance for a number of issues of great astrophysical interest, such as the formation of globular clusters, the role of the external galactic field on the evolution of their properties and the possibility of using the turnover of the GCLF as a standard candle calibrated on the values of globular cluster systems in the Local Group for the determination of the distance of external galaxies (see e.g. Jacoby et al. 1992).  The determination of the distances of external galaxies by means of the turnover of the GCLF relies on the assumed constancy of the properties of the GCLF for galaxies of different structure and type which is quite surprising: in fact, unless one advocates  a  scenario in which clusters in different galaxies have different initial conditions and different dynamical histories but all leading to the same final state, such common characteristics imply that the process of formation of globular clusters does not depend on the galactic environment and that evolutionary processes (tidal stripping, disk and bulge shocking, dynamical friction; see e.g. Meylan \\& Heggie 1997, for a recent review on the  dynamics of globular clusters) have not played a  relevant role in determining the present  properties of globular clusters.  While the present knowledge of the processes leading to the formation of globular clusters is still very uncertain (see e.g. Fall \\& Rees 1985, Harris \\& Pudritz 1994, Vietri \\& Pesce 1995, Elmegreen \\& Efremov 1997) many theoretical investigations (Aguilar, Hut \\& Ostriker 1988, Chernoff, Kochanek \\& Shapiro 1986, Chernoff \\& Shapiro 1987, Vesperini 1994, 1997, Okazaki \\& Tosa 1995, Murali \\& Weinberg 1997, Gnedin \\& Ostriker 1997, Baumgardt 1998) have clearly shown that  evolutionary processes should have altered significantly the initial properties of globular clusters in galaxies like the Milky Way by causing the complete disruption of a fraction of them and by altering the initial properties of the surviving ones. As it has been  shown in several works (see e.g. Caputo \\& Castellani 1984, Chernoff, Kochanek \\& Shapiro 1986, Vesperini 1994, 1997, Murali \\& Weinberg 1997, Ostriker \\& Gnedin 1997, Baumgardt 1998) the inner regions of the Galaxy are those where evolutionary effects are expected to be more efficient and where to look for traces of their effects. Indeed Chernoff \\& Djorgovski (1989) have provided the first observational evidence of this by showing that the fraction of clusters in the post-core collapse phase increases as the distance from the Galactic center decreases. In a subsequent work Bellazzini et al. (1996) have shown the existence for clusters located in the inner regions of the Galaxy of a significant correlation between concentration and galactocentric distance in the sense of more concentrated clusters being on average closer to the Galactic center. This is likely to result from evolution (Vesperini 1994, 1997, Bellazzini et al. 1996) occuring faster for clusters located in the inner regions of the Galaxy than for those  in the outer ones.  The situation concerning differences between the GCLF of inner and outer clusters is far from being clear. In a recent work Gnedin (1997) has carried out an analysis of the available observational data for clusters in the Galaxy, M31 and M87, and his results seem to point to the existence of some differences between the GCLF of inner and outer clusters for all these three  galaxies: inner clusters tend to be brighter and to have smaller dispersions than outer clusters. Kavelaars \\& Hanes (1997) have  addressed the same issue for the Milky Way and M31 and their conclusion is that, while there is no significant difference in the mean luminosity of inner and outer clusters, their distributions are actually different, the inner clusters being well described by a Gaussian in the magnitude with a dispersion significantly smaller than that of outer clusters; as for M87, Harris et al. (1998) in a recent analysis have not found any significant radial gradient in the properties of the GCLF for clusters with masses $M>10^5 M_{\\odot}$ except for a possible trend for the dispersion of the GCLF of the innermost region of M87 that they have considered to be smaller than the dispersion of the GCLF of clusters in the outer regions of the galaxy. As discussed in Gnedin (1997) the reason for the difference between  his analysis and that by Kavelaars and Hanes could reside in the different statistical methods adopted for deriving the parameters of the distribution. A trend for inner clusters to be brighter than the outer ones was previously suggested by  van den Bergh (1995) and Crampton et al. (1985) for clusters in the Milky Way and in M31 respectively.  From a theoretical point of view, as we said above, this trend is consistent with that expected to result from evolutionary processes, at least for some initial GCMFs, which are more efficient in the inner regions of the Galaxy where they can efficiently  disrupt low-mass clusters (see e.g. fig. 2 in Vesperini 1997). In Vesperini (1994, 1997) the evolution of the properties of a system of globular clusters located in a model of the Milky Way under the effects of relaxation, disk shocking and, in an approximate way, of dynamical friction, starting from three different initial GCMF, has been investigated. In all these cases a trend for inner clusters to be more massive than outer clusters was obtained as a result of evolutionary processes, with the extent of the difference depending on the initial GCMF chosen.  While in Vesperini (1994,1997), besides addressing some general issues on the evolution and the properties of the GCMF, we investigated the origin of some observed correlations between structural properties of individual globular clusters and between structural parameters of clusters and their position inside the host galaxy, in this work we will focus our attention on and investigate in larger detail the evolution of the properties of the GCMF of a globular cluster system located in a model for the Milky Way adopting some analytical formulae for the time evolution of the masses of individual clusters obtained by the results of a large set of $N$-body simulations carried out by Vesperini \\& Heggie (1997).  We will adopt a  log-normal distribution  for the initial GCMF and we will consider a wide range of different initial conditions largely spanning the space of the initial parameters (dispersion and mean) of the GCMF. Different functional forms for the initial GCMF have also been studied to investigate  the evolution of their shape  and in particular to establish if the current gaussian shape could result from an initial GCMF with a different functional form.  We will devote a section to the comparison of our results with observational data, but we point out that due to some assumptions made in our analysis, which will be discussed in sect.2 together with the description of the method adopted for our investigation, an exact comparison of our results with the available observational data is beyond the scope of our work. The main goal of our analysis is that of  providing general indications on the evolution of the properties of the GCMF, of the spatial distribution and the fraction of the initial number of clusters surviving after one Hubble time. The evolution of the shape of the GCMF for the whole sample of clusters and the possible development of differences between the GCMF of clusters located in the inner and in the outer regions of the Galaxy will be thoroughly investigated paying particular attention to the dependence of the final results on the initial conditions. The issue, raised in Vesperini (1997), of the  possible existence of  a dynamical  ``equilibrium'' GCMF able to preserve its initial shape and parameters for one Hubble time through a subtle balance between disruption of clusters and evolution of the masses of the surviving ones is further investigated.   The scheme of the paper is the following. In sect.2 the method adopted for our study is described; in section 3 we report the main results of the investigation: after a preliminary qualitative discussion on the evolution of the GCMF in section 3.1, in sections 3.2-3.6 we describe the results obtained not including the effects of disk shocking; in particular section 3.2 discusses the possible evolutionary paths  of  the parameters of the GCMF depending on the initial conditions (see e.g. figure 3) and the existence of a GCMF of dynamical equilibrium is shown and discussed in detail (see e.g. figure 4c), section 3.3 is focussed on the dependence of  the final  GCMF on the distance from the Galactic center (see e.g. figure 10 and 11), in section 3.4 the time evolution of some systems is followed in detail and some other aspects of  the GCMF able to stay in dynamical equilibrium are studied (see figure 13). Section 3.5 and 3.6 are devoted to the study of the fraction of surviving clusters and their spatial distribution in the Galaxy respectively; in section 3.7 we discuss the results obtained including the effects of disk shocking. In section 4 we describe the results obtained assuming a power-law initial GCMF and section 5 is devoted to the comparison of our results with observational data. Summary and conclusions are in section 6. ",
        "conclusions": "In this work we have investigated the evolution of the mass function of a globular cluster system located in a model for the Milky Way. The effects of stellar evolution, two-body relaxation, disk shocking, dynamical friction and  the presence of the tidal field of the Galaxy have been taken into account in the evolution of the mass of individual globular clusters in the system which is calculated on the basis of the results of the $N$-body simulations carried out by Vesperini \\& Heggie (1997).  A log-normal and a power-law initial GCMF have been considered. The main effort has been devoted to the investigation of systems starting with an initial log-normal GCMF spanning a wide range of values of the mean value and  the dispersion of the initial distribution. The gaussian shape has been shown to be preserved very well during the entire evolution until $t=15 $ Gyr  while for systems in which  the initial GCMF is a power-law  a bell-shaped GCMF resembling a Gaussian in $\\log M$ tends to be established in the  course of evolution.  Depending on the initial GCMF parameters, the mean value of the GCMF, $\\langle \\log M \\rangle $, can increase or decrease during the evolution according to whether  the dominant process  is that of disruption of low-mass clusters by evaporation of stars through  the tidal boundary (for initial GCMF dominated by low-mass clusters) or that of disruption by dynamical friction of high-mass clusters (for initial GCMF dominated by high-mass clusters). The regions in the space of initial parameters   $\\langle \\log M \\rangle_i $-$\\sigma_i$ corresponding to these two different regimes  as well as the corresponding regions for the evolution of the dispersion $\\sigma$ has been shown.  The differences between the final values of $\\langle \\log M \\rangle$, $\\Delta \\mm_{in-out}$,  and $\\sigma$, $\\Delta \\sigma_{in-out}$, of inner ($R_g<8 $ Kpc) and outer clusters ($R_g>8$ Kpc) have been investigated. Depending on the dominant evolutionary process ( disruption of low-mass clusters or dynamical friction) $\\Delta \\mm_{in-out}$ can be larger or smaller than zero. As for  $\\Delta \\sigma_{in-out}$, in most cases considered evolutionary processes tend to make the dispersion of inner clusters smaller than that of outer clusters. The formation of a gradient of $\\mm$ and $\\sigma$ with the galactocentric distance due to evolutionary processes has been investigated. The direction of the gradient of $\\mm$ depends on the initial GCMF: an increasing $\\mm$ as $R_g$ increases is common for systems initially containing many high-mass clusters while the opposite trend is typical of systems initially dominated by low-mass clusters. It has been shown that a significant effect of evolutionary processes does not necessarily imply the formation of a strong radial trend of $\\langle \\log M\\rangle$.  For most initial conditions considered, and in particular for those likely to be relevant for real systems, evolutionary processes give rise to a trend for $\\sigma$ to decrease at smaller galactocentric distances.  The existence of a particular GCMF able to stay in dynamical equilibrium keeping its initial shape and parameters unaltered during the entire evolution by means  a subtle balance of disruption of clusters and evolution of the masses of those surviving, first suggested by Vesperini (1997), has been confirmed.  The initial number density distribution of clusters in the Galaxy has been taken proportional to $R_g^{-3.5}$ and it has been shown that evolutionary processes tend to flatten this distribution close to the Galactic center. The extent of the flattening depends on the initial conditions and it has been estimated quantitatively by calculating the final core radius, $R_c$, of the distribution of survived clusters in the Galaxy. The range spanned by $R_c$ for the  initial conditions considered in our work is $0.4-2$ Kpc.  The fraction of the total initial number of clusters surviving after one Hubble time, $F_N$,  the fraction  of the total initial mass of all the clusters in the system, $F_M$,  and the current cluster disruption rate (defined as the fraction of the number of clusters at $t=15 $  Gyr  undergoing disruption within the next 1 Gyr), have been calculated and their dependence on the initial conditions investigated.   The exact comparison of our results with observational data is beyond the scope of our work both because of some simplifying assumptions we have made and because of the current lack of a precise knowledge of the initial properties of the Galactic globular cluster system; nevertheless assuming the current properties of outer clusters to be similar to the initial ones of the entire system we have calculated the values predicted from our analysis for some of the observed properties of the Galactic globular cluster system and we have found them to be in general in good agreement with the  observational values.  As for the fraction of the total initial number of cluster surviving at the current epoch, $F_N$, and the fraction of the total initial mass of all the globular clusters, $F_M$, the values predicted for the Galactic system are $F_N \\simeq 0.54$ (0.48 if the effects of disk shocking are included) and $F_M\\simeq 0.41$ (0.38 with disk shocking). These values imply that the initial population of Galactic globular clusters would consist of about 300 clusters with a total mass of about $9\\times 10^7 M_{\\odot}$ and that the contribution to the halo mass from disrupted clusters and stars escaped from survived clusters would be about $5.5\\times 10^7 M_{\\odot}$. The distribution of clusters disruption times has been calculated and shown to be similar to the distribution of disruption times for a sample of 119 Galactic clusters obtained by Gnedin \\& Ostriker (1997); we have shown that very different initial distributions of disruption timescales can lead to very similar final distributions and thus much caution is necessary in drawing any conclusion on the initial population of clusters from the present distribution of disruption timescales."
    },
    "9805/astro-ph9805372_arXiv.txt": {
        "abstract": "We report on the \\xray\\ spectrum of the 401~Hz \\xray\\ pulsar and type~I burst source \\saxj\\, during its 1998 April/May hard outburst. The observations were made with \\rxte\\ over a period of three weeks.  The spectrum is well-described by a power law with photon index $\\rm 1.86 \\pm 0.01$ that is exponentially cut off at high energies.  Excess soft emission above the power law is present as well as a weak Fe-K line.  This is the first truly simultaneous broad-band (2.5--250~keV) spectrum of a type~I burst source in the hard state. The spectrum is consistent with other hard state burster spectra which cover either only the soft (1--20\\,keV) or hard (\\gtsim 20\\,keV) bands, or cover both, but not simultaneously.  The cut-off power law spectrum resembles that of black hole candidates (BHCs) in their low states, observed with \\rxte.  We compare the \\saxj\\ spectrum to three BHCs and find that the power law is somewhat softer. This suggests that the photon index may provide a way to distinguish between low state emission from Galactic black holes and type~I bursters. ",
        "introduction": "\\saxj\\ is the first object thought to display both type~I \\xray\\ bursts and coherent \\xray\\ pulsations.  Its low implied magnetic field (B \\ltsim$\\rm 2 \\times 10^8$\\,G, Wijnands \\& van der Klis 1998b) and high spin frequency may make \\saxj\\ a missing link in the evolution of the millisecond radio pulsars. It was discovered in observations of the Galactic center region made during 1996 September 12--17 with the Wide Field Camera (WFC) on \\bsax\\ (\\cite{intZ98}).  During six days of observations, the flux level was \\aprx\\ 50--100\\,mCrab (2--10\\,keV). Earlier and later observations which did not detect the source limit the outburst duration to between 6 and 40 days. This duration was confirmed with data from the All Sky Monitor (ASM) on the {\\em Rossi X-ray Timing Explorer} (\\rxte), which detected \\saxj\\ for about 20 days beginning 1996 September 8 (\\cite{intZ98}).  Two type~I \\xray\\ bursts were also detected during the \\bsax\\ observations, making the identification of the source as a neutron star in a low mass \\xray\\ binary highly probable.  Assuming that these bursts reached the Eddington luminosity for a 1.4\\Msun neutron star implies a distance of \\aprx 4\\,kpc.  Following the 1996 outburst, \\saxj\\ remained undetected until a slew of the \\rxte\\ pointed instruments on 9 April 1998 serendipitously detected a source (designated XTE~J1808-3658) whose location is consistent with the \\bsax\\ error region (\\cite{Mar98}).  The flux level at this time was \\aprx 50\\,mCrab (2--10\\,keV), corresponding to a luminosity of $\\rm 1.5 \\times 10^{36}$\\,ergs/s at a distance of 4\\,kpc.  Twenty-one \\rxte\\ pointed observations over the next 4~weeks saw the flux increase to 60\\,mCrab (2.5-20\\,keV) and decrease approximately exponentially with a time constant of about 10~days (see Figure~\\ref{f_lc}).  After 26~May, the source dimmed rapidly by a factor of \\aprx 5 in 2 days \\nocite{Glf98}. \\begin{figure} \\caption{The light curve of the 1998 April outburst of \\saxj\\ in three energy bands from the ASM, PCA, and HEXTE on \\rxte. The ASM data are the publicly available daily averages provided by the ASM/RXTE team (http://heasarc.gsfc.nasa.gov/docs/xte/asm\\_products.html).\\label{f_lc}} \\centerline{\\epsfig{file=j1808_lcs.ps,width=5.4in}} \\end{figure}   Timing analyses of \\rxte/Proportional Counter Array (PCA) data from 11 April 1998 revealed that \\saxj\\ is an \\xray\\ pulsar with a frequency of 401\\,Hz, making it the first accretion-powered millisecond \\xray\\ pulsar (\\cite{Wij98a,Wij98b}).  The pulsed amplitude was quite low, only \\aprx 4\\% RMS (2--60\\,keV). Chakrabarty \\& Morgan (1998a)\\nocite{Cha98a}, also using PCA data, detected the binary orbit. They derived an orbital period of 7249.119(1)\\,s, a projected semimajor axis of $\\rm a_xsin{\\em i} = 62.809(1)$\\,lt-ms, and an \\xray\\ mass function of $\\rm 3.85 \\times 10^{-5}$\\Msun (\\cite{Cha98b}).  They also placed upper limits on the pulse frequency derivative and the eccentricity ($< 5\\times 10^{-4}$) of the orbit.  The very small mass function implies that the companion mass is $\\rm <0.18$\\,\\Msun\\ for a neutron star less massive than 2\\,\\Msun\\ (\\cite{Cha98b}).  During the recent \\xray\\ outburst, optical imaging of the \\saxj\\ field revealed an object with magnitude $\\rm V = 16.6$ that was not present in the Digitized Sky Survey (DSS) to a limiting magnitude of $\\rm V > 19 $ (\\cite{Roc98}).  This object was confirmed as the likely optical counterpart of \\saxj\\ when multiple V-band exposures covering the 2\\,hr binary orbit showed ``roughly sinusoidal'' variability of 0.12\\,mag (\\cite{Gil98}).  Early work with the \\rxte\\ data indicated that the \\saxj\\ spectrum was Crab-like and continued unbroken to energies greater than 100\\,keV (\\cite{Hei98,Gil98}.  In this letter, we perform detailed spectral studies and show that the spectrum is somewhat harder than the Crab at low energies (\\ltsim 30\\,keV) and is exponentially cut off at higher energies. ",
        "conclusions": "The spectrum of the unique 401~Hz pulsar and type~I burster \\saxj\\ during its recent outburst was quite hard. The photon spectral index was 1.86, with a slow cutoff at high energies.  There was clear evidence of excess soft emission and a weak Fe-K line.  These \\rxte\\ observations have provided the highest quality broad-band spectrum of a suspected \\xray\\ burster during a period of hard emission.  The spectrum is in good agreement with previous observations of other bursters made over more limited energy bands and with lower statistical significance. By comparing to observations of low state BHCs, we find that BHCs and bursters can be distinguished by the slope of their power law emission.  In particular, the photon indices of bursters are greater (i.e. softer) by about 0.3, even though the overall spectral shapes are quite similar."
    },
    "9805/astro-ph9805234_arXiv.txt": {
        "abstract": "Since Baschek \\& Slettebak (1988) drew attention to the similarity between the abundance pattern of $\\lambda$ Boo stars and that of Vega, there has been a long debate whether Vega should be listed among the chemically peculiar stars of $\\lambda$ Boo type. We performed an elemental abundance analysis using a high dispersion spectrum in the optical region, and confirmed its mild metal underabundance. In our discussion we reinforce the suggestion that Vega is a mild $\\lambda$ Boo star. ",
        "introduction": "Vega ($\\alpha$ Lyr = HD 172167 = HR 7001) is a Population I star of spectral type A0 V, with a projected rotational velocity $v \\sin i = 23$ km$\\,$s$^{-1}$. It has been extensively studied both for its role of primary spectrophotometric standard in the visual and in the UV, respectively, and for its role of comparison star in abundance studies of A-type stars.  Vega has a distinctly non--solar composition. Since Baschek \\& Slettebak (1988) drew attention to the similarity between the abundance pattern of $\\lambda$ Boo stars and that of Vega, there has been a long debate whether Vega should be listed among the chemically peculiar stars of $\\lambda$ Boo type.  Moreover, Baschek \\& Slettebak suggested that the $\\lambda$ Boo stars may be regarded as {\\em rotating Vegas}, or conversely Vega may be regarded as a {\\em mild non--rotating $\\lambda$ Boo star}. ",
        "conclusions": ""
    },
    "9805/astro-ph9805220_arXiv.txt": {
        "abstract": "It has been often considered that the dissipative collapse caused by a merger of two gas-rich galaxies is responsible for the intense nuclear starbursts or the nonthermal quasar activity in ultraluminous infrared galaxies due to the efficient fueling induced by it. It is also widely known that Ultraluminous Infrared Galaxies (ULIGs) are often found in merging systems. Some ULIGs, such as Arp 220, show two compact starburst regions which are considered to be associated with two galactic nuclei in the process of merging. However, since a merger between two galaxies may make only one compact starburst region, we suggest the possibility that double-nucleus ULIGs are composed of two merging nuclei, each of which contains a couple of galactic  nuclei. ",
        "introduction": "Ultraluminous Infrared Galaxies (hereafter ULIGs) have attracted much attention since their discovery by {\\it IRAS} in 1984 (Soifer et al. 1984; Wright, Joseph, \\& Meikle 1984; see for a review Sanders \\& Mirabel 1996). The multiple morphological studies published to date have shown that these objects tend to be found in galaxy mergers or in strongly interacting galaxies (Sanders et al. 1988a; Lawrence et al. 1991; Leech et al. 1994). Although the origin of their huge infrared luminosities is still not fully understood, it is considered to come from intense starbursts, central active galactic nuclei or a combination of both (Joseph \\& Wright 1985; Sanders et al. 1988a, 1988b; Solomon \\& Sage 1988; Scoville et al. 1991; Condon et al. 1991; Majewski et al. 1993; Lonsdale et al. 1994; Skinner et al. 1997). As numerical simulations have shown (see for a review Shlosman, Begelman, \\& Frank 1990; Barnes \\& Hernquist 1992), galaxy mergers cause efficient gas fueling toward the nuclear regions of the merging systems that ultimately can trigger and maintain any of the central activities mentioned above, either as a result of the piling of gas (Negroponte \\& White 1983; Barnes 1988; Barnes \\& Hernquist 1991; Olson \\& Kwan 1990a, 1990b; Noguchi 1991; Bekki \\& Noguchi 1994; Mihos \\& Hernquist 1994a, 1994b; Hernquist \\& Mihos 1995) or by the dynamical effect of supermassive binaries (Taniguchi \\& Wada 1996; Taniguchi 1997; Taniguchi, Wada, \\& Murayama 1997).  Current models for the origin of ULIGs only consider the merging of {\\it two} gas-rich galaxies (Sanders et al. 1988a; Kormendy \\& Sanders 1992). However, taking into account that there are a large number of nearby compact galaxy groups (e.g., Hickson 1982), formation of ULIGs due to a multiple merger cannot be ruled out. In fact, the presence of three OH maser components in the archetypical ULIG Arp 220 suggests the possibility that this ULIG originates from a multiple merger (Diamond et al. 1989). In this {\\it Letter}, we discuss the multiple merger scenario for the formation of ULIGs based on the observational properties of Arp 220. ",
        "conclusions": "\\subsection{Origin of the Nuclear Starbursts in ULIGs}  Numerical simulations of galaxy mergers between two gas-rich disk galaxies have shown that they induce efficient gas fueling into the central a few 100 pc region of the merging systems (e.g., Mihos \\&  Hernquist 1994b and references therein). However, it is known that the properties of the merger-driven starbursts are sensitive to the structure of the progenitor galaxies (Mihos \\& Hernquist 1994b). If the progenitor galaxies are bulgeless (i.e., late-type spirals), the successive close encounters during the merger strongly affect their respective gas disks. As the merger proceeds, gas clouds in each galaxy are channeled to each nuclear region and  if a dynamical instability occurs in the central region of each member, intense star formation would occur there (e.g., Noguchi 1991; Shlosman \\& Noguchi 1994). Since the gas contained in the progenitors would be used up while the merger is in progress, no intense starbursts would occur when the merger is complete. The well-known merger, NGC 7252, may be a good example for this case because only moderate star-forming regions can be seen in the central part of this galaxy (Whitmore et al. 1993). On the other hand, if the galactic bulge of each progenitor is massive enough to stabilize their nuclear gas disks, it will prevent bulge strong gas inflows until the galaxies merge, giving rise to a single intense starburst in the central part of the merger remnant (see also Bekki \\& Noguchi 1993). Since in the case of Arp 220, the projected separation between the two compact starburst regions is 350 pc, it is suggested that at least if  double-nucleus ULIGs come from mergers between two galaxies, their progenitors could be gas-rich galaxies without prominent bulges.  The double-nucleus (traced by radio continuum and NIR emission) nature of Arp 220 has often used as an example of this two-galaxy merger scenario (Baan \\& Haschick 1987; Norris 1988; Sanders et al. 1988a; Graham et al. 1990; Scoville et al. 1998) However, there is some observational evidences that Arp 220 may come in fact from the merger of more than two spirals. VLBI mapping of OH megamaser emission in Arp 220 has revealed the presence of at least three bright OH megamaser spots in its nuclear region (Diamond et al. 1989), which would imply the coexistence of three active galactic nuclei in the region. Therefore, given the fact that there exist a fairly large amount of nearby compact galaxy groups that could end up merging since their merging timescales are generally shorter than the Hubble time (Barnes 1989; Weil \\& Hernquist 1996), we cannot rule out the multiple merger scenario for the formation of ULIGs.  For simplicity, we consider the case of ULIG coming from a merger of four comparably nucleated disk galaxies. At an early stage of the merger it is expected that parings would occur; i.e., two pairs of galaxies merge first. Then the merging remnants would merge again into one final object (cf. Barnes 1989; Weil \\& Hernquist 1996). It is known that a merger between two {\\it nucleated} galaxies can form a single dense gaseous system in the merger remnant because of the dynamical disturbance of the binary potential to the gas clouds  (Bekki \\& Noguchi 1994; Taniguchi \\& Wada 1996). If double-nucleus ULIGs are on the way to the final merger, we can explain why they have two dense gaseous systems each of which can be associated with a merger remnant between two galaxy nuclei.   \\subsection{A Multiple Merger Model for Arp 220}  We now consider more carefully the possibility that Arp 220 comes from a merger of four comparably nucleated disk galaxies. In this scenario, the double nucleus of this object should correspond to the final stages of the merging of two pairs of nuclei. As mentioned above, there are at least three bright OH megamaser spots in the nuclear region of Arp 220 (Diamond et al. 1989). The eastern nucleus contains two bright OH maser spots with a projected separation of 47.6 pc while the western shows only one component. Recent VLBI measurements by Lonsdale et al. (1994) have shown that the western (i.e., the brightest) component of the OH maser originates from a very compact region whose size is less than 1 pc. This measurement suggests strongly that the western component is produced by the pumping by far-infrared continuum emitted by a dusty torus around an active galactic nucleus rather than by the luminous nuclear starbursts. If this is also the case for the two eastern OH maser components, Arp 220 would contain at least three active galactic nuclei (i.e., three supermassive black holes) as suggested by Diamond et al. (1989). It is of course possible that the single western OH megamaser component actually represents the accidental alignment along our line of sight of two nuclei, which would be resolved if observed in other conditions.  Another support for the multiple merger scenario comes from the observational fact that the two OH maser spots in the eastern nucleus show a velocity gradient which is almost perpendicular to the eastern-western nucleus axis, implying that the eastern nucleus is dynamically different from the western nucleus (Diamond et al. 1989). Furthermore, H$_2$CO maser observations of this object (Baan \\& Haschick 1995) show velocity gradients in maser emission associated with both the eastern and western nuclei that are more compatible with rotational motion around the individual nuclei rather than with the global rotation around the two nuclei (see Baan \\& Haschick 1995).  In our scenario it should also be expected that each pair of nuclei has a rotating gas disk settled roughly in the orbital plane of each of the two nuclei in each pair because nuclear gas will settle in a relatively short timescale there. In fact, the double-peaked nature of the HCN and HCO$^+$ (Solomon et al. 1992) and CO($J$=1-0) emissions (Scoville et al. 1997) suggest that the dense gaseous systems are associated with each pair of merging nuclei. At larger scales, a circumnuclear gas disk with a radius of $\\sim$ 300 pc surrounds the two pairs of nuclei (Scoville et al. 1997).  In summary, the multiple merger scenario that we proposed here explains almost all the observational properties of Arp 220 consistently. A schematic illustration of the nuclear region of Arp 220 is shown in Fig. 2. The rotation of the eastern black hole binary was determined from the velocity difference between the two OH megamaser components, IIa and IIb (Diamond et al. 1989) while that of the eastern one is from the rotation of the H$_2$CO masing gas (Baan \\& Haschick 1995). The global rotation of the E and W nuclei is from the NIR spectroscopy (Larkin et al. 1995). Although any current observational facilities may not be able to verify this model, we hope that the multiple merger scenario will be taken into account in the future study on the origin of ULIGs.  \\vspace{0.5cm}  We would like to thank Dave Sanders, Baltasar Vila-Vilaro, Sumio Ishizuki, Seiichi Sakamoto, and Neil Trentham for useful discussion and suggestions. We also thank John Hibbard and Dave Sanders for providing us with their CCD image of Arp 220. This work was financially supported in part by Grant-in-Aids for the Scientific Research (No. 0704405) of the Japanese Ministry of Education, Culture, Sports and Science."
    },
    "9805/astro-ph9805016_arXiv.txt": {
        "abstract": "The application of the Str\\\"omgren photometric luminosity calibrations  to  different types of CP stars is reexamined in the light of the new Hipparcos data. A first attempt is made to use the LM statistical parallax method (Luri {\\it et al.}, 1996) --  based on the maximum likelihood principle -- to  obtain a calibration of the absolute magnitude  as a function of two Str\\\"omgren colour indices,  thus reflecting effective temperature and evolution.  Its application to a sample of Si stars and to a sample of normal main sequence stars in the same spectral range allows us to compare the calibrations obtained and to discuss the position of Si stars in the HR diagram.  Additionally, a sample of {\\it bonafide}, spectroscopically  selected Am stars together with normal main sequence stars are used to derive a new absolute magnitude calibration for late A-type main sequence stars, taking into account the effects of evolution, metallicity and stellar rotation. ",
        "introduction": "To deal with luminosity calibrations and to obtain a good exploitation of the high precision parallax data obtained by Hipparcos, it is required to use both, robust statistical methods, capable to take into account physical characteristics of stars (i.e. evolution, metallicity, rotational effects, etc.), and a well defined spectroscopic sample. Taking advantage of the fact that Str\\\"omgren photometry has proved to be a powerful tool for characterizing the physics of main sequence early type stars, some preliminary work is presented here for application to Si and Am stars.  Our sample contains all CP stars included in the Hipparcos Catalogue (ESA, 1997) having good spectroscopic information on peculiarity and complete Str\\\"om\\-gren photometry in the Hauck and Mermilliod (1996) compilation. The selection of the stars according to their spectral types is fully explained in G\\'omez et al. (1998, this colloquium). For comparison, a sample of normal main sequence stars in the range B0-A9 V, selected from the Hipparcos Survey  (G\\'omez et al., 1997), has been  also used. Table~\\ref{t1} shows the stars in each CP group (the sample is reduced to less than 50 \\% when Str\\\"omgren photometry is required). \\begin{table}[t] \\small \\begin{center} \\caption{CP stars with $uvbyH_{\\beta}$ photometry.} \\label{t1} \\begin{tabular}{|c|c|c|c|} \\hline \\hline & Selection from & with $uvbyH_{\\beta}$ & Range of \\\\ & G\\'omez et al. (1997a) & Hauck \\& Mermilliod & SP \\\\ &  & (1996)   &   \\\\ \\hline \\hline Si             & 440 & 173 & B5-A3 \\\\ HgMn           &  76 & 69 & B7-A0 \\\\ Sr-Cr-Eu       & 378 & 172 &B8-A9  \\\\ Am             & 1059 & 533& A0-A9 \\\\ Normal B0-A9 V & 3460 & 1589 & B0-A9 \\\\ \\hline \\end{tabular} \\end{center} \\end{table} Using the different samples, a test of the capability of the $uvbyH_{\\beta}$ system to detect CP stars has been performed.  The $([m_1], [c_1])$ plane has been classically accepted as the most discriminant plane in the Str\\\"omgren system to separate CP from normal stars. Thus for example, Crawford (1979) established the criterion $\\delta m_1 \\lid -0.020$ to separate Am stars from A3-A9 normal main sequence stars. In contrast to Abt (1984), who concluded that about 75 \\% of the Am stars could be photometrically classified as such, we obtain that only 55 \\% of the classical Am (Sp(k)-Sp(m) $\\gid$ 5) and 33 \\% of the proto-Am  (Sp(k) - Sp(m) $<$ 5) can be detected using only Str\\\"omgren colours. Philip et al (1976) criteria $(E(b-y) \\lid$ -0.040) to detect hot CP stars is found to be unable to acomplish the objective (only 2 \\% of Si stars and 5 \\% of Sr stars has been detected). Taking into account that peculiar spectral features alter all Str\\\"omgren indices, Masana et al. (1998) established a new criterion of detection defining a $\\Delta p$ parameter -- a linear combination of several $uvbyH_{\\beta}$ indices obtained through Multiple Discriminant Analysis. Applying this criterion to the hot stars in the present sample (see Figure~\\ref{fig1}) we photometrically classify as peculiar 31 \\% of the Si stars and 56 \\% of hot Sr-Cr-Eu stars, whereas no HgMn star is detected. Furthermore, as indicated by Masana et al. (1998), the reddening decreases $\\Delta p$, so reddened hot CP stars may be seen as non-reddened normal stars, thus decreasing its capability for detection. We can conclude that  the Str\\\"omgren system is suitable but less powerful for a photometric detection of peculiarities than the $\\Delta a$ system (Maitzen, 1976; specifically designed to measure characteristic features on the spectra of peculiar stars) or even the  $\\Delta (V_1 - G)$ combination of the Geneva indices  (Masana et al., 1998).  A crucial point to attack the problem of the  $M_v$ calibration of CP stars is the reddening correction. For CP2 stars, the peculiarities in the spectra result in bluer colour indices, the $(b-y)$ and $c_1$ indices being smaller, thus leading to underestimate their reddening when treating them as normal stars. In agreement with Adelman (1980) and Maitzen (1980), Masana et al. (1998) derived a correction  to the $E(b-y)$ obtained when using standard relations valid for normal stars as a function of the $\\Delta p$ parameter ($\\Delta E(b-y) = -0.001 + 0.008 \\Delta p$). Even applying this correction to the stars with $\\Delta p \\geq 1.5$, we obtain that 57 \\% of Si stars and 38 \\% of Sr-Cr-Eu (early region) stars have $E(b-y) < 0$ , compared with only 9 \\% of normal main sequence stars in the same spectral range with negative excess (using Crawford's (1978) standard relations). \\begin{figure}[th] \\centerline{ \\psfig{figure=T10f1.eps,height=6.2cm}} \\caption{Detection of hot CP stars using the $\\Delta p$ parameter: $\\Delta p = p -p_o = 1.5$ (dashed line), $p_o = f([u-b])$ being the relation for normal stars (full line)} \\label{fig1} \\end{figure} The problem is maintained in the intermediate region (Grosb{\\o}l's (1978) standard relation): 56\\% of Sr-Cr-Eu stars have negative excess compared to 37 \\% of normal main sequence stars. In the late region (A3-A9), where Crawford's (1979) standard relation is considered, the overestimation of reddening is also clear for Sr-Cr-Eu (18 \\% with $E(b-y) <0$  compared with 37\\%  for normal stars). The same excess distribution is obtained for Am and normal A3-A9 stars (35 \\% and 37 \\% of stars with $E(b-y)<0$ respectively). Having in mind these biased photometric reddening determinations when peculiarities are present, we preferred to use the interstellar absorption model by Arenou et al. (1992) in the LM method, which gives reddening corrections as a function of the star position $(r,l,b)$. Improvement of the interstellar absorption models using the new Hipparcos data is needed. ",
        "conclusions": ""
    },
    "9805/astro-ph9805366_arXiv.txt": {
        "abstract": "We report the discovery of low frequency quasi-periodic oscillations centered at 0.11 Hz in the newly discovered 221 s X-ray pulsar XTE J1858+034. Among about 30 known transient X-ray pulsars this is the sixth source in which QPOs have been observed. If the QPOs are produced because of inhomogeneities in the accretion disk at the magnetospheric boundary, the low frequency of the QPOs indicate a large magnetosphere for this pulsar. Both the Keplerian frequency model and the beat frequency model are applicable for production of QPOs in this source. The QPOs and regular pulsations are found to be stronger at higher energy which favours the beat frequency model. The magnetic field of the pulsar is calculated as a function of its distance. The energy spectrum is found to be very hard, consisting of two components, a cut-off power law and an iron fluorescence line. ",
        "introduction": "Quasi periodic oscillations (QPOs) observed in X-ray binaries are generally thought to be related to the rotation of the inner accretion disk. When the accretion disk can reach very close to the compact object, like in the case of black hole candidates and low magnetic field neutron star sources, the rotation of the inhomogeneities or hot blobs of material in the inner disk are reflected in the light curve as QPOs. In X-ray pulsars, however, the disk is interrupted at a large distance by the strong magnetic field of the neutron star, and the inner transition zone of the disk, which is at a large distance from the neutron star, does not emit in X-rays. Hence strong QPOs are believed to be rare in X-ray pulsars.  The hard X-ray transient XTE J1858+034 was discovered with the RXTE All Sky Monitor (ASM) in 1998 February (Remillard \\& Levine 1998). The spectrum was found to be hard, similar to the spectra of X-ray pulsars. Observations were made immediately after this with the Proportional Counter Array (PCA) of the RXTE and regular pulsations with a period of $221.0 \\pm 0.5$ s were discovered (Takeshima et al. 1998). The pulse profile is found to be nearly sinusoidal with a pulse fraction of $\\sim 25\\%$. From the transient nature of this source and pulsations they suggested that this is a Be-X-ray binary. The position of the X-ray source was refined by scanning the sky around the source with the PCA (Marshall et al. 1998).  From the XTE target of opportunity (TOO) public archival data of the observations of XTE J1858+034, made in 1998 February 20 and 24, we have discovered the presence of low frequency QPOs. We also have obtained the pulse profile of this source in two energy bands and the energy spectrum in one of the observations. In the following sections we describe the archival data that has been used, the analysis and results and discuss some implications of the detection of QPOs in this source. ",
        "conclusions": "The transient X-ray pulsars in which QPOs have been detected, are the high mass X-ray binaries (HMXB) EXO 2030+375 (Angelini et al. 1989), A 0535+262 (Finger et al. 1996), 4U 0115+63 (Soong \\& Swank 1989) and V 0332+53 (Takeshima et al. 1994) and the LMXB GRO J1744-28 (Zhang et al. 1996; See Finger 1998 for a review of the QPO in transient X-ray pulsars). QPOs have also been observed in some of the persistent HMXB sources: Cen X-3 (Takeshima et al. 1991), SMC X-1 (Angelini et al. 1991), X Persei (Takeshima 1997) and 4U 1907+09 (in'tZand et al. 1998) and the LMXB 4U 1626-67 (Shinoda et al. 1990; Kommers et al. 1998). Both the Keplerian frequency model (in which the QPOs are produced because some inhomogeneous structure in the Keplerian disk attenuates the pulsar beam regularly) and the beat frequency model (in which the material influx to the pulsar from the disk is modulated at the Keplerian frequency) are in very good aggrement with the observations in EXO 2030+375 and A 0535+262. In 4U 0115+63, V 0332+52, Cen X-3, 4U 1626-67 and SMC X-1 however, the QPO frequency is found to be lower than the pulsation frequency hence the Keplerian frequency model is not applicable in these sources because if the Keplerian frequency at the magnetospheric boundary is less than the spin frequency, centrifugal inhibition of mass accretion will take place. For V 0332+52 the beat frequency model may also be inapplicable because the magnetospheric boundary calculated from the QPO properties and from observed luminosity are in disagreement in this source. In the LMXB transient pulsar GRO J1744-28, large change in X-ray flux was found to be associated with a very little change in the QPO frequency which ruled out both the Keplerian and the beat frequency models for QPOs in this source (Zhang et al. 1996). The beat frequency model is applicable in many sources though there is no convincing evidence of positive correlation between the QPO frequency and the X-ray luminosity in some of them.  According to the beat frequency model, the QPOs are a result of beat phenomena between the rotation of the innermost part of the disk and the spin of the neutron star. The Keplerian rotation frequency $\\nu_K$ of the disk at the magnetosphere boundary, the rotation frequency of the neutron star $\\nu_S$ and the QPO frequency $\\nu_{QPO}$ are related as $\\nu_{QPO}$ = $\\nu_K$ - $\\nu_S$. Assuming that the QPOs are produced as a result of this phenomena, the Keplerian rotational frequency of the innermost part of the disk is just sum of the QPO frequency and the rotation frequency of the pulsar. For an assumed mass of 1.4 M$_\\odot$, this can be related to the magnetospheric radius r$_M$ of the X-ray pulsar.  In XTE J1858+034, we find that $\\nu_{QPO}$ = 0.11 $\\pm$ 0.01 Hz, $\\nu_S$ = 0.0045 Hz and the radius of the magnetospheric boundary is calculated to be \\begin{equation} r_M = \\left({{GM}\\over { 4 \\pi^2 \\nu_K^2}}\\right)^{1 \\over 3} = 6.5~10^8 \\left({M \\over {M_\\odot}}\\right)^{1 \\over 3} {\\rm cm} \\end{equation} where M is the mass of the neutron star.  The pulse averaged X-ray flux in the 1.3-100 keV band is 6.5 10$^{-10}$ erg cm$^{-2}$ s$^{-1}$ which, for a distance of r$_{kpc}$, amounts to an X-ray luminosity L$_X$ of 7.9 10$^{34}$ r$_{kpc}^2$ erg s$^{-1}$. For a standard accretion disk with disk axis parallel to the magnetic field axis and dipole magnetic field structure of the neutron star, the radius of the inner transition zone can also be expressed as (Frank et al. 1992)  \\begin{equation} r_M = 2.9 \\times 10^{8} {\\left(M\\over M_\\odot\\right)}^{1\\over7} R_6^{-{2\\over7}} L_{37}^{-{2\\over7}} \\mu_{30}^{4\\over7} \\end{equation} where, R$_6$ is the radius of the neutron star in unit of 10$^6$ cm, L$_{37}$ is X-ray luminosity in unit of 10$^{37}$ erg and $\\mu_{30}$ is magnetic moment in unit of 10$^{30}$ cm$^3$ Gauss.  Combining the above two equations, and using M = 1.4 M$_\\odot$, R$_6$ = 1, the magnetic moment $\\mu_{30}$ of the pulsar is calculated to be $\\sim$ 0.4 $\\times$ 10$^{30}$ r$_{kpc}$, which for a neutron star radius of 10$^6$ cm, is equivalent to a magnetic field of 0.8 $\\times$ 10$^{12}$ r$_{kpc}$ Gauss.  If origin of the QPOs in this source is the magnetospheric boundary, the QPOs cannot arise from the modulation of X-rays emitted from the accretion disk because for a magnetospheric radius of 3 $\\times$ 10$^8$ cm the disk temperature is rather low to emit in X-rays. The X-ray modulation at the QPO frequency can arise either because some inhomogeneous structure in the Keplerian disk attenuates the pulsar beam regularly at its rotation frequency, or the material influx to the pulsar from the disk is modulated at the Keplerian frequency. The fact that the strength of the QPO is greater at higher energies indicates that the latter is likely to be the case for XTE J1858+034. A detailed analysis (which is currently in progress) of the QPO feature as a function of pulse phase and energy will help in firmly deciding one of the two alternatives for the QPO phenomenon."
    },
    "9805/astro-ph9805150_arXiv.txt": {
        "abstract": "We give concisely the formulae governing diffusion of chemical elements and their isotopes in quiescent stellar atmospheres, due to electrostatic, gravitational and radiation fields and to impacts between particles. Isotope segregation of heavy elements due to light-induced drift is emphasized. ",
        "introduction": "\\label{intr} The diffusive separation of chemical elements and their isotopes in stellar atmospheres  can occur only in the case of lacking macroscopic motions, i.e. if the stellar wind and the meridional circulation are extremely weak and there is no convective turbulence. These conditions hold only for CP stars, and  overabundances of heavy elements in their atmospheres can reach  several orders of magnitude. ",
        "conclusions": ""
    },
    "9805/astro-ph9805293_arXiv.txt": {
        "abstract": "We present new observational results on the kinematics of the damped \\lya systems.  Our full sample is now comprised of 31 low-ion profiles and exhibits similar characteristics to the sample from Paper I. The primary exception is that the new distribution of velocity widths includes values out to a maximum of nearly 300~\\kms, $\\approx$ 100~\\kms greater than the previous maximum. These high velocity width systems will significantly leverage models introduced to explain the damped \\lya systems. Comparing the characteristics from low-redshift and high-redshift sub-samples, we find no evidence for significant evolution in the kinematic properties of protogalaxies from $z = 2.0 - 3.3$.  The new observations give greater statistical significance to the main conclusions of our first paper. In particular, those models inconsistent with the damped \\lya observations in Paper I are ruled out at even higher levels of confidence. At the same time, the observations are consistent with a population of rapidly rotating, thick disks (the TRD model) at high redshift, as predicted by cosmologies with early structure formation.  Buoyed by the success of the TRD model, we investigate it more closely by considering more realistic disk properties.  Our goal is to demonstrate the statistical power of the damped \\lya observations by investigating the robustness of the TRD model. In particular, we study the effects of warping, realistic rotation curves, and photoionization on the kinematics of disks in the TRD model. The principal results are: (1) disk warping has only minimal effect on the kinematic results, primarily influencing the effective disk thickness, (2) the TRD model is robust to more realistic rotation curves; we point out, however, that the rotation curve derived from centrifugal equilibrium with HI gas alone does not yield acceptable results, rather flat rotation curves such as those generated by dark matter halos are required, and (3) the effects of photoionization require thicker disks to give consistent velocity width distributions. ",
        "introduction": "This paper marks the third in a series of papers on the kinematics of the damped \\lya protogalaxies.  These HI gas layers observed along sightlines to distant QSO's are widely believed to be the gaseous progenitors of modern galaxies (\\cite{wol95b}). Hence an examination of damped systems at high redshift provides insight into the process of galaxy formation.  For instance, identifying the physical nature of these systems may distinguish between the monolithic collapse model (\\cite{egg62}) and the hierarchical scenario favored by standard cosmogony.  In our first paper (\\cite{pro97b}, hereafter PW), we demonstrated that the kinematics of damped \\lya systems at high redshift are consistent with these systems being thick, rapidly rotating disks; it is a description not unlike that predicted by monolithic collapse formation scenarios.  At the same time, we found damped \\lya systems cannot be simple exponential disks in a cluster normalized Standard Cold Dark Matter cosmology (e.g.\\ \\cite{kau96}). Subsequently, Jedamzik and Prochaska (1998) tightened this conclusion by considering a range of disk characteristics and CDM normalizations. They found that only a finely tuned disk model within the framework of CDM could be made marginally consistent with the damped \\lya observations. Recently, Haehnelt et al.\\ (1997) have offered an alternative model for damped systems as gaseous protogalactic clumps undergoing infall within dark matter halos which may be consistent with the kinematic characteristics of the damped \\lya systems.  Such a description lends itself naturally to the hierarchical cosmologies where merging plays a vital role. A future paper will address this model in greater detail. For the present work, we will focus on the interpretation of damped \\lya systems as thick rotating disks at high redshift.  In PW, we analyzed the low-ion profiles from 17 damped \\lya systems and compared their kinematic characteristics with those of simulated profiles derived from several physical models.  Of the models tested, we found the thick rotating disk model to be the only model consistent with the observations. The basic assumptions of the model are a flat rotation curve and an exponential gas distribution, both chosen to roughly correspond with the observations of local spiral galaxies. In this paper we present low-ion profiles for 14 additional damped systems.  Therefore, our full kinematic sample consists of 31 low-ion profiles.  We interpret the sample with disk models containing more physically realistic characteristics, e.g. disk warping and photoionization.  We have two primary goals in mind: (1) to test the robustness of the interpretation of damped \\lya systems as disks and (2) to determine the effects on our conclusions regarding the thickness and rotation speed of these disks.  In $\\S 2$ we review the terminology and methodology introduced in PW.  The new data are presented and tested against the models of PW in $\\S 3$.  We investigate the effects of more realistic disk properties in $\\S 4$ and in $\\S 5$ we present a summary. ",
        "conclusions": "We have presented new observations on the low-ion kinematics of the damped \\lya systems.  The full sample of 31 profiles confirms the primary conclusions of PW, in particular, (i) models with kinematics dominated by random or symmetric velocity fields are inconsistent with the damped \\lya kinematics, (ii) the TRD model, which consists of a population of thick, rapidly rotating disks at high $z$, naturally reproduces both the observed edge-leading asymmetry of the empirical profiles as well as the distribution of velocity widths, and (iii) models incorporating centrifugally supported disks within the framework of the standard CDM cosmology are ruled out at high levels of confidence. In addition, a comparison of the kinematic properties of profiles of the highest redshift systems ($\\bar z = 3.24$) with the lower redshift systems ($\\bar z = 2.06$) reveals no significant evolution in the kinematics of the damped \\lya systems.  This last observation may place strong constraints on scenarios of galaxy formation which predict significant evolution over this epoch.  Presently there are two working models which explain the kinematic characteristics of the damped \\lya systems:  (1) the TRD model and (2)  merging protogalactic clumps in numerical simulations of the standard Cold Dark Matter cosmology (\\cite{hae97}). In this paper we have focused on the TRD model.  In particular, we have investigated the robustness of the model to including more realistic disk properties, specifically disk warping, physical rotation curves and photoionization. Given the prevalence of warping in local disk galaxies, we considered its effects on the kinematics of the disks in the TRD model. We found that the results of warping are dominated by two competing effects.  Sightlines which penetrate both the unwarped inner disk and the warped outer disk yield moderately higher $\\delv$ than those simply intersecting an unwarped disk.  At the same time, however, some warped disks have significantly larger cross-section to sightlines with large impact parameters which tend to yield small $\\delv$. Having considered a number of warped disks with a broad range of properties, we find: (i) in extreme cases, warping mimics disks with up to $50\\%$ larger or smaller effective thickness ($h/R_d$ value), (ii) warping leads to very few extra large $\\delv$ values in the $\\f{\\delv}$ distribution and therefore has little consequence on the acceptable values for $v_{rot}$, and (iii) the lower limit to $h/R_d$ is nearly unchanged as we find $h$ must be $> 0.1 R_d$ for both warped and unwarped disks.  In PW, we assumed a flat rotation curve, $v_\\phi = v_{rot}$, extending from $R = 0 \\to \\infty$ and $Z = 0 \\to \\infty$. In this paper we adopted rotation curves derived from specific bulge, halo, and disk components. Assuming an exponential profile is a good description of the density profile for the damped \\lya systems, we find the rotation curves derived from gravity generated by the HI gas {\\it alone} cannot reproduce the empirical $\\f{\\delv}$ distribution. If the rotation curve is dominated by the disk, one must introduce another mass component (e.g.\\ stars, molecules) to establish consistency. At the same time we find that the rotation curve derived from a massive halo with core radius $R_h \\lesssim R_d$ also yields a $\\f{\\delv}$ distribution consistent with the observations. We believe that this latter explanation is more plausible. We also find the presence of the bulge to be largely inconsequential.  Lastly, we studied the effects of the intergalactic photoionizing background radiation on the disk kinematics.  We made two separate approximations to model the photoionization of the disks: (a) an HI edge model where the disk is photoionized at radii $R > R_{ph}$ with $R_{ph}$ set by the intensity of the photoionizing background and the disk properties and (b) a Critical Density model where all gas with volume density $n \\leq n_{ph}$ is presumed ionized.  Contrary to our expectations, we find that photoionization tends to worsen the agreement between the TRD model and the damped \\lya observations.  The effect, however, is not large ($ < 30 \\%$) for the favored value of $J_{912} = 10^{-21.5}$, but for $J_{912} = 10^{-21}$ a substantial $(> 50 \\%)$ increase in the thickness of the disks would be required.  In summation, then, we find the TRD model is very robust to tests against the damped \\lya observations. The challenge remains, however, to consistently incorporate this model within a cosmological framework. While the clump model fits naturally within the SCDM cosmology, it must be demonstrated that the clump model exhibits similar robustness to comparisons with the damped \\lya kinematics. While Haehnelt et al.\\ (1997) did show that the clump model could explain the damped \\lya observations from PW for a single set of parameters, a formal investigation of the full physical parameter space with meaningful statistics has yet to be performed.  In addition, it is not clear how that model will change given different cosmological parameters, e.g.\\ an Open Universe where merging plays a smaller role at $z \\approx 2.5$.  The model must also be tested against the new observations, in particular the new $\\f{\\delv}$ distribution which extends to $\\approx 300 \\mkms$. Finally, the fact that the numerical simulations do not reproduce the observed properties of modern galaxies when evolved to the present universe (\\cite{nav95}) suggests the model may have serious inconsistencies in the early universe.  In future papers we will introduce observations of the high-ion transitions (e.g.\\ CIV~1548) with the aim of further constraining the two working models as well as advancing our understanding of the ionized gas associated with the damped \\lya systems. This gas is presumed to reside in the halo of these protogalaxies and therefore may give more direct indications of the dark matter associated with the damped \\lya systems. We also intend to consider effects (e.g.\\ multiple disks) which would improve the agreement between the semi-analytic models of standard cosmology (\\cite{kau96,mmw97})."
    },
    "9805/astro-ph9805037_arXiv.txt": {
        "abstract": "Element abundances of three roAp stars, HD\\,166473, HD\\,203932, and HD\\,217522, were determined using Kurucz model atmospheres with metal abundances scaled to solar ones and the results were compared with data from the literature concerning three further roAp stars, normal B and A stars and two $\\lambda$ Bootis stars. Up to 38 elements could be identified and therefore, this work represents the most complete chemical investigation hitherto published, which can be summarized as follows:  \\begin{description} \\item{$\\bullet$} all investigated roAp stars have a similar abundance pattern, \\item{$\\bullet$} the overabundances of rare earth and other heavy elements are comparable to cool non-pulsating Ap-stars, \\item{$\\bullet$} iron belongs to the most deficient and cobalt to the most enhanced elements in the group of the iron peak elements, and \\item{$\\bullet$} the light elements carbon, nitrogen, and oxygen are less abundant than in atmospheres with abundances scaled to the Sun. \\end{description}  \\noindent Beside an unexpected possible relation between effective temperature and metallicity of roAp stars, no outstanding differences from non-pulsating Ap stars could be detected. This statement, however, suffers from the lack of comparably detailed investigations of the latter. ",
        "introduction": "Rapidly oscillating Ap (roAp) stars are a subgroup of the CP2 stars, which oscillate with non-radial, high overtone, low order acoustic $p$-modes with the axis of oscillation aligned with the axis of the magnetic field (Kurtz 1982). A still open question is concerned with the excitation mechanism for roAp stars and other physical parameters which distinguish them from non-pulsating CP stars. ",
        "conclusions": "This study shows that roAp stars have similar abundances for up to 38 elements identified in the sample. These abundances are being used to compute stellar atmosphere models based on individual opacities. Within the significantly smaller number of analyzed elements in the literature, a comparison to non-roAp stars do not reveal large abundance differences. The derived fundamental stellar parameters and abundances allow to locate the stars in the HR-diagram and provide important boundary values for pulsation models. Stellar structure and evolutionary parameters can be derived from such models.   A comparison of the iron peak abundances in the six roAp stars shows an unexpected relation between effective temperature and metallicity. Fig.\\,\\ref{relation} shows this tendency for the total metallicity of all iron peak elements \\logZ\\ versus the spectroscopically derived effective temperature \\Teff. However, since all stellar parameters -- including also effective temperature -- are determined mainly with iron lines, it is not clear whether this tendency is astrophysically significant or an artifact due to the applied analyzing algorithm. The derived stellar parameters and abundances are resulting from optimization routines. Inadequate model atmospheres, limited signal-to-noise ratio of the spectra, errors in atomic parameters and problems in defining the continuum can result in fairly large errors. Therefore, the correlation between metallicity and effective temperature has to be further investigated.  \\begin{figure} \\centerline{ \\psfig{figure=T11f2.eps,height=6.5cm}} \\caption{Relation between effective temperature and total metallicity of all iron peak elements.} \\label{relation} \\end{figure}  Beside this possible relation, no outstanding differences between pulsating and non-pulsating Ap stars could be detected."
    },
    "9805/astro-ph9805201_arXiv.txt": {
        "abstract": "We present spectral and photometric observations of 10 type Ia supernovae (SNe Ia) in the redshift range 0.16 $\\leq z \\leq$ 0.62. The luminosity distances of these objects are determined by methods that employ relations between SN Ia luminosity and light curve shape. Combined with previous data from our High-Z Supernova Search Team (Garnavich et al. 1998; Schmidt et al. 1998) and Riess et al. (1998a), this expanded set of 16 high-redshift supernovae and a set of 34 nearby supernovae are used to place constraints on the following cosmological parameters: the Hubble constant ($H_0$), the mass density ($\\Omega_M$), the cosmological constant (i.e., the vacuum energy density, $\\Omega_\\Lambda$), the deceleration parameter ($q_0$), and the dynamical age of the Universe ($t_0$). The distances of the high-redshift SNe Ia are, on average, 10\\% to 15\\% farther than expected in a low mass density ($\\Omega_M=0.2$) Universe without a cosmological constant. Different light curve fitting methods, SN Ia subsamples, and prior constraints unanimously favor eternally expanding models with positive cosmological constant (i.e., $\\Omega_\\Lambda > 0$) and a current acceleration of the expansion (i.e., $q_0 < 0$). With no prior constraint on mass density other than  $\\Omega_M \\geq 0$, the spectroscopically confirmed SNe Ia are statistically consistent with $q_0 <0$ at the 2.8$\\sigma$ and 3.9$\\sigma$ confidence levels, and with $\\Omega_\\Lambda >0$ at the 3.0$\\sigma$ and 4.0$\\sigma$ confidence levels, for two different fitting methods respectively. Fixing a ``minimal'' mass density, $\\Omega_M=0.2$, results in the weakest detection, $\\Omega_\\Lambda>0$ at the  3.0$\\sigma$ confidence level from one of the two methods.  For a flat-Universe prior ($\\Omega_M+\\Omega_\\Lambda=1$), the spectroscopically confirmed SNe Ia require $\\Omega_\\Lambda >0$ at  7$\\sigma$ and 9$\\sigma$ formal significance for the two different fitting methods.  A Universe closed by ordinary matter (i.e., $\\Omega_M=1$) is formally ruled out at the 7$\\sigma$ to 8$\\sigma$ confidence level for the two different fitting methods.  We estimate the dynamical age of the Universe to be 14.2 $\\pm 1.5$ Gyr including systematic uncertainties in the current Cepheid distance scale. We estimate the likely effect of several sources of systematic error, including progenitor  and  metallicity evolution, extinction, sample selection bias, local perturbations in the expansion rate, gravitational lensing, and sample contamination.  Presently, none of these effects reconciles the data with $\\Omega_\\Lambda=0$ and $q_0 \\geq 0$. ",
        "introduction": "This paper reports observations of 10 new high-redshift type Ia supernovae (SNe Ia) and the values of the cosmological parameters derived from them.   Together with the four high-redshift supernovae previously reported by our High-Z Supernova Search Team (Schmidt et al. 1998; Garnavich et al. 1998) and two others (Riess et al. 1998a), the sample of 16 is now large enough to yield interesting cosmological results of high statistical significance. Confidence in these results depends not on increasing the sample size but on improving our understanding of systematic uncertainties.  The time evolution of the cosmic scale factor depends on the composition of mass-energy in the Universe.  While the Universe is known to contain a significant amount of ordinary matter, $\\Omega_M$, which decelerates the expansion, its dynamics may also be significantly affected by more exotic forms of energy.  Pre-eminent among these is a possible energy of the vacuum ($\\Omega_\\Lambda$), Einstein's ``cosmological constant,'' whose negative pressure would do work to accelerate the expansion (Carroll, Press, \\& Turner 1992; Schmidt et al. 1998).  Measurements of the redshift and apparent brightness of SN Ia of known intrinsic brightness can constrain these cosmological parameters.  \\subsection{The High-Z Program}  Measurement of the elusive cosmic parameters $\\Omega_M$ and $\\Omega_\\Lambda$ through the redshift-distance relation depends on comparing the apparent magnitudes of low-redshift SNe Ia with those of their high-redshift cousins.  This requires great care to assure uniform treatment of both the nearby and distant samples.  The High-Z Supernova Search Team has embarked on a program to measure supernovae at high redshift and to develop the comprehensive understanding of their properties required for their reliable use in cosmological work. Our team pioneered the use of supernova light curve shapes to reduce the scatter about the Hubble line from $\\sigma$ $\\approx$ 0.4 mag to $\\sigma$ $\\approx$ 0.15 mag (Hamuy et al. 1996a,b, 1995; Riess, Press \\& Kirshner 1995, 1996a).  This dramatic improvement in the precision of SNe Ia as distance indicators increases the power of statistical inference for each object by an order of magnitude and sharply reduces their susceptibility to selection bias.  Our team has also pioneered methods for using multi-color observations to estimate the reddening to each individual supernova, near and far, with the aim of minimizing the confusion between effects of cosmology and dust (Riess, Press, \\& Kirshner 1996a; Phillips et al. 1998).  Because the remaining scatter about the Hubble line is so small, the discussion of the Hubble constant from low-redshift SNe Ia has already passed into a discussion of the best use of Cepheid distances to galaxies that have hosted SNe Ia (Saha et al. 1997; Kochanek 1997; Madore \\& Freedman 1998; Riess, Press, \\& Kirshner 1996a; Hamuy et al. 1996a; Branch 1998).  As the use of SNe Ia for measuring $\\Omega_M$ and $\\Omega_\\Lambda$ progresses from its infancy into childhood, we can expect a similar shift in the discussion from results limited principally by statistical errors to those limited by our depth of understanding of SNe Ia.  Published high-redshift SN Ia data is a small fraction of the data in hand both for our team and for the Supernova Cosmology Project (Perlmutter et al. 1995, 1997, 1998).  Now is an opportune time to spell out details of the analysis, since further increasing the sample size without scrupulous attention to photometric calibration, uniform treatment of nearby and distant samples, and an effective way to deal with reddening will not be profitable. Besides presenting results for four high-z supernovae, we have published details of our photometric system (Schmidt et al. 1998) and stated precisely how we used ground-based photometry to calibrate our {\\it Hubble Space Telescope (HST)} light curves (Garnavich et al. 1998). In this paper, we spell out details of newly-observed light curves for 10 objects, explain the recalibration of the relation of light curve shape and luminosity for a large low-redshift sample, and combine all the data from our team's work to constrain cosmological parameters. We also evaluate how systematic effects could alter the conclusions.  While some comparison with the stated results of the Supernova Cosmology Project (Perlmutter et al. 1995, 1997, 1998) is possible, an informed combination of the data will have to await a similarly detailed description of their measurements.  \\subsection{A Brief History of Supernova Cosmology}  While this paper emphasizes new data and constraints for cosmology, a brief summary of the subject may help readers connect work on supernovae with other approaches to measuring cosmological parameters.  Empirical evidence for SNe I presented by Kowal (1968) showed that these events had a well-defined Hubble diagram whose intercept could provide a good measurement of the Hubble constant. Subsequent evidence showed that the original spectroscopic class of Type I should be split (Doggett \\& Branch 1985; Uomoto \\& Kirshner 1985; Wheeler \\& Levreault 1985; Wheeler \\& Harkness 1986; Porter \\& Filippenko 1987).  The remainder of the original group, now called Type Ia, had peak brightness dispersions of 0.4 mag to 0.6 mag (Tammann \\& Leibundgut 1990; Branch \\& Miller 1993; Miller \\& Branch 1990; Della Valle \\& Panagia 1992; Rood 1994; Sandage \\& Tammann 1993; Sandage et al. 1994).  Theoretical models suggested that these ``standard candles'' arose from the thermonuclear explosion of a carbon-oxygen white dwarf that had grown to the Chandrasekhar mass (Hoyle \\& Fowler 1960; Arnett 1969; Colgate \\& McKee 1969).  Because SNe Ia are so luminous ($M_B \\approx -19.5$ mag), Colgate (1979) suggested that observations of SNe Ia at $z \\approx 1$ with the forthcoming Space Telescope could measure the deceleration parameter, $q_0$.  From a methodical CCD-based supernova search that spaced observations across a lunation and employed prescient use of image-subtraction techniques to reveal new objects,  Hansen, N\\o rgaard-Nielsen, \\& Jorgensen (1987) detected SN 1988U, a SN Ia at $z=0.31$ (N\\o rgaard-Nielsen et al. 1989).    At this redshift and distance precision ($\\sigma \\approx 0.4 $ to 0.6 mag), $\\sim 100$ SNe Ia would have been needed to distinguish between an open and closed Universe.  Since the Danish group had already spent two years to find one object, it was clear that larger detectors and faster telescopes needed to be applied to this problem.  Evidence of systematic problems also lurked in supernova photometry so that merely increasing the sample would not be adequate. Attempts to correct supernova magnitudes for reddening by dust (Branch \\& Tammann 1992) based on the plausible (but incorrect) assumption that all SNe Ia had the same intrinsic color had the unfortunate effect of increasing the scatter about the Hubble line or alternately attributing bizarre properties to the dust absorbing SN Ia light in other galaxies.  In addition, well-observed supernovae such as SN 1986G (Phillips et al. 1987; Cristiani et al. 1992), SN 1991T (Filippenko et al. 1992a; Phillips et al. 1992; Ruiz-Lapuente et al. 1992), and SN 1991bg (Filippenko et al. 1992b; Leibundgut et al. 1993; Turatto et al. 1996) indicated that large and real inhomogeneity was buried in the scatter about the Hubble line.  Deeper understanding of low-redshift supernovae greatly improved their cosmological utility.    Phillips (1993) reported that the observed peak luminosity of SNe Ia varied by a factor of 3.  But he also showed that the decrease in $B$ brightness in the 15 days after peak ($\\Delta m_{15}(B)$) was a good predictor of the SN Ia luminosity, with slowly declining supernovae more luminous than those that fade rapidly.  A more extensive database of carefully and uniformly observed SNe Ia was needed to refine the understanding of SN Ia light curves. The Cal\\'{a}n/Tololo survey (Hamuy et al. 1993a) made a systematic photographic search for supernovae between cycles of the full moon. This search was extensive enough to guarantee the need for scheduled follow-up observations, which were supplemented by the cooperation of visiting observers, to collect well-sampled light curves.  Analysis of the Cal\\'{a}n/Tololo results generated a broad understanding of SNe Ia and demonstrated their remarkable distance precision (after template fitting) of $\\sigma \\approx 0.15$ mag (Hamuy et al. 1995, 1996a,b,c,d).  A parallel effort employed data from the Cal\\'{a}n/Tololo survey and from the Harvard-Smithsonian Center for Astrophysics (CfA) to develop detailed empirical models of SN Ia light curves (Riess, Press, \\& Kirshner 1995; Riess 1996).  This work was extended into the Multi-Color Light Curve Shape (MLCS) method which employs up to 4 colors of SN Ia photometry to yield excellent distance precision ($\\approx 0.15$ mag) and a statistically valid estimate of the uncertainty for each object with a measurement of the reddening by dust for each event (Riess, Press, \\& Kirshner 1996a; Appendix of this paper).  This work has also placed useful constraints on the nature of dust in other galaxies (Riess, Press, \\& Kirshner 1996b).  The complete sample of nearby SNe Ia light curves from the Cal\\'{a}n/Tololo and CfA samples provides a solid foundation from which to extend the redshift-distance relation to explore cosmological parameters.  The low-redshift sample used here has 34 SNe Ia with $z < 0.15$.  Since the high-redshift observations reported here consumed large amounts of observing time at the world's finest telescopes, we have a strong incentive to find efficient ways to use the minimum set of observations to derive the distance to each supernova.  A recent exploration of this by Riess et al. (1998a) is the ``Snapshot\" method which uses only a single spectrum and a single set of photometric measurements to infer the luminosity distance to a SN Ia with $\\sim$ 10\\% precision. In this paper, we employ the snapshot method for six SNe Ia with sparse data, but a shrewdly designed program that was intended to use the snapshot approach could be even more effective in extracting useful results from slim slices of observing time.  Application of large-format CCDs and sophisticated image analysis techniques by the Supernova Cosmology Project (Perlmutter et al. 1995) led to the discovery of SN 1992bi ($z=0.46$) followed by 6 more SNe Ia at $z \\approx 0.4$ (Perlmutter et al. 1997).  Employing a correction for the luminosity/light-curve shape relation (but none for host galaxy extinction), comparison of these SNe Ia to the Cal\\'{a}n/Tololo sample gave an initial indication of ``low'' $\\Omega_\\Lambda$ and ``high'' $\\Omega_M$: $\\Omega_\\Lambda=0.06^{+0.28}_{-0.34}$ for a flat Universe and $\\Omega_M=0.88^{+0.69}_{-0.60}$ for a Universe without a cosmological constant ($\\Omega_\\Lambda \\equiv 0$).  The addition of one very high-redshift ($z=0.83$) SN Ia observed with {\\it HST} had a significant effect on the results: $\\Omega_\\Lambda=0.4\\pm0.2$ for a flat Universe, and $\\Omega_M=0.2\\pm0.4$ for a Universe with $\\Omega_\\Lambda \\equiv 0$. (Perlmutter et al. 1998). This illustrates how young and volatile the subject is at present.  \\subsection{This Paper}  Our own High-Z Supernova Search Team has been assiduously discovering high-redshift supernovae, obtaining their spectra, and measuring their light curves since 1995 (Schmidt et al. 1998).  The goal is to provide an independent set of measurements that uses our own techniques and compares our data at high and low redshifts to constrain the cosmological parameters.   Early results from 4 SNe Ia (3 observed with {\\it HST}) hinted at a non-negligible cosmological constant and ``low'' $\\Omega_M$, but were limited by statistical errors: $\\Omega_\\Lambda=0.65 \\pm 0.3$ for a flat Universe, $\\Omega_M=-0.1\\pm0.5$ when $\\Omega_\\Lambda \\equiv 0$ (Garnavich et al. 1998).  Our aim in this paper is to move the discussion forward by increasing the data set from four high-redshift SNe to 16, to spell out exactly how we have made the measurement, and to consider various possible systematic effects.  In \\S 2 we describe the observations of the SNe Ia including their discovery, spectral identification, photometric calibration, and light curves.  We determine the luminosity distances (including $K$-corrections) via two methods, MLCS and a template fitting method ($\\Delta m_{15}(B)$), as explained in \\S 3.  Statistical inference of the cosmological parameters including $H_0$, $\\Omega_M$, $\\Omega_\\Lambda$, $q_0$, $t_0$, and the fate of the Universe is contained in \\S 4.  Section 5 presents a quantitative discussion of systematic uncertainties which could affect our results: evolution, absorption, selection bias, a local void, weak lensing, and sample contamination.  Our conclusions are summarized in \\S 6. ",
        "conclusions": "The results of \\S 4 suggest an eternally expanding Universe which is accelerated by energy in the vacuum. Although these data do not provide independent constraints on $\\Omega_M$ and $\\Omega_\\Lambda$ to high precision without ancillary assumptions or inclusion of a supernova with uncertain classification, specific cosmological scenarios can still be tested without these requirements.  {\\it High-redshift SNe Ia are observed to be dimmer than expected in an empty Universe (i.e., $\\Omega_M=0$) with no cosmological constant.}  A cosmological explanation for this observation is that a positive vacuum energy density accelerates the expansion.  Mass density in the Universe exacerbates this problem, requiring even more vacuum energy.  For a Universe with $\\Omega_M=0.2$, the MLCS and template fitting distances to the well-observed SNe are 0.25 and 0.28 mag farther on average than the prediction from $\\Omega_\\Lambda=0$.  The average MLCS and template fitting distances are still 0.18 and 0.23 mag farther than required for a 68.3\\% (1$\\sigma$) consistency for a Universe with $\\Omega_M=0.2$ and without a cosmological constant.  Depending on the method used to measure all the spectroscopically confirmed SN Ia distances, we find $\\Omega_\\Lambda$ to be inconsistent with zero at the 99.7\\% (3.0$\\sigma$) to $>$99.9\\% (4.0$\\sigma$) confidence level. Current acceleration of the expansion is preferred at the 99.5\\% (2.8$\\sigma$) to $>$99.9\\% (3.9$\\sigma$) confidence level. The ultimate fate of the Universe is sealed by a positive cosmological constant.  Without a restoring force provided by a surprisingly large mass density (i.e., $\\Omega_M > 1$) the Universe will continue to expand forever.  How reliable is this conclusion?  Although the statistical inference is strong, here we explore systematic uncertainties in our results with special attention to those that can lead to overestimates of the SNe Ia distances.  \\subsection{Evolution}  The local sample of SNe Ia displays a weak correlation between light curve shape (or luminosity) and host galaxy type.  The sense of the correlation is that the most luminous SNe Ia with the broadest light curves only occur in late-type galaxies.  Both early-type and late-type galaxies provide hosts for dimmer SNe Ia with narrower light curves (Hamuy et al. 1996c).  The mean luminosity difference for SNe Ia in late-type and early-type galaxies is $\\sim 0.3$ mag (Hamuy et al. 1996c).  In addition, the SN Ia rate per unit luminosity is almost twice as high in late-type galaxies as in early-type galaxies at the present epoch (Cappellaro et al. 1997).  This suggests that a population of progenitors may exist in late-type galaxies which is younger and gives rise to brighter SNe Ia (with broader light curves) than those contained in early-type galaxies or within pockets of an older stellar population in the late-type galaxies.  Such observations could indicate an evolution of SNe Ia with progenitor age.  H\\\"{o}flich, Thielemann,  \\& Wheeler (1998) calculate differences in the light curve shape, luminosity, and spectral characteristics of SNe Ia as a function of the initial composition and metallicity of the white dwarf progenitor.  As we observe more distant samples, we expect the progenitors of SN Ia to come from a younger and more metal-poor population of stars.  H\\\"{o}flich, Thielemann, \\& Wheeler (1998) have shown that a reduction in progenitor metallicity by a factor of 3 has little effect on the SN Ia bolometric luminosity at maximum.  For their models, such a change in metallicity can alter the peak luminosity by small amounts ($\\sim 0.05$ mag) in rest-frame $B$ and $V$, accompanied by detectable spectral signatures.  These spectral indicators of evolution are expected to be most discernible in the rest-frame $U$ passband where line blanketing is prevalent.  Future detailed spectral analyses at these short wavelengths might provide a constraint on a variation in progenitor metallicity.  The effect of a decrease in SN Ia progenitor age at high redshift is predicted to be more significant than metallicity (H\\\"{o}flich, Thielemann, \\& Wheeler 1998). Younger white dwarfs are expected to evolve from more massive stars with a lower ratio of C/O in their cores. The lower C/O ratio of the white dwarf reduces the amount of $^{56}$Ni synthesized in the explosion, but an anticipated slower rise to maximum conserves more energy for an increased maximum brightness. By reducing the C/O ratio from 1/1 to 2/3, the $B-V$ color at maximum is expected to become redder by 0.02 mag and the post-maximum decline would become steeper.  This prediction of a brighter SN Ia exhibiting a faster post-maximum decline is opposite to what is seen in the nearby sample (Phillips 1993; Hamuy et al. 1995; Hamuy et al. 1996a,b,c,d; Riess, Press, \\& Kirshner 1996a; Appendix) and will be readily testable for an enlarged high redshift sample.  Specifically, a larger sample of distant SNe Ia (currently being compiled) would allow us to determine the light curve shape relations at high-redshift and test whether these evolve with look-back time.  Presently, our sample is to small to make such a test meaningful.  We expect that the relation between light curve shape and luminosity that applies to the range of stellar populations and progenitor ages encountered in the late-type and early-type hosts in our nearby sample should also be applicable to the range we encounter in our distant sample. In fact, the range of age for SN Ia progenitors in the nearby sample is likely to be {\\it larger} than the change in mean progenitor age over the 4 to 6 Gyr look-back time to the high-redshift sample. Thus, to first order at least, our local sample should correct our distances for progenitor or age effects.  We can place empirical constraints on the effect that a change in the progenitor age would have on our SN Ia distances by comparing subsamples of low redshift SNe Ia believed to arise from old and young progenitors.  In the nearby sample, the mean difference between the distances for the early-type (8 SNe Ia) and late-type hosts (19 SNe Ia), at a given redshift, is 0.04 $\\pm$ 0.07 mag from the MLCS method.  This difference is consistent with zero.  Even if the SN Ia progenitors evolved from one population at low redshift to the other at high redshift, we still would not explain the surplus in mean distance of 0.25 mag over the $\\Omega_\\Lambda=0$ prediction. For the template fitting approach, the mean difference in distance for SNe Ia in early-type and late-type hosts is 0.05 $\\pm$ 0.07 mag.  Again, evolution provides an inadequate explanation for the 0.28 mag difference in the template fitting SNe Ia distances and the $\\Omega_\\Lambda=0$ prediction.  However, the low-redshift sample is dominated by late-type hosts and these may contain a number of older progenitors.  It is therefore difficult to assess the precise effect of a decrease in progenitor age at high redshift from the consistency of distances to early-type and late-type hosts (see Schmidt et al. 1998).  If, however, we believed that young progenitors give rise to brighter SNe Ia with broader light curves (Hamuy et al. 1996c) as discussed above, we could more directly determine the effect on distance determinations of drawing our high-redshift sample from an increasingly youthful population of progenitors.  The mean difference in the Hubble line defined by the full nearby sample and the subsample of SNe Ia with broader than typical light curves ($\\Delta < 0$) is 0.02 $\\pm$ 0.07 for the MLCS method.  For the template fitting method, the difference between the full sample and those with broader light curves ($\\Delta m_{15}(B) < 1.1$) is 0.07 $\\pm$ 0.07.  Again, we find no indication of a systematic change in our distance estimates with a property that may correspond to a decrease in progenitor age.  Another valuable test would be to compare low-redshift distances to starburst and irregular type galaxies which presumably are hosts to progenitors which are young and metal-poor.  Such a nearby sample may yield the closest approximation to the SNe Ia observed at high redshift. Future work will be needed to gather this informative sample which would be composed of objects such as SN 1972E in NGC 5253 (which we does fit the luminosity light curve shape relations; Hamuy et al 1996b).  Another check on evolutionary effects is to test whether the distribution of light curve decline rates is similar between the nearby sample of supernovae and the high-redshift sample. Figure 10 shows the observed distribution of the MLCS light-curve shape parameters, $\\Delta$, and the template fitting parameters, $\\Delta m_{15}(B)$, with redshift. A Kolmogorov-Smirnov test shows no significant difference in the distributions of the low and high-redshift samples, but the sample is too small to be statistically significant. The actual difference in mean luminosity between the low-redshift and high-redshift samples implied by the light curve shapes is 0.02 mag by either method.   We conclude that there is no obvious difference between the shapes of SNe~Ia light curves at $z \\approx 0$ and at $z \\approx 0.5$.  It is reassuring that initial comparisons of high-redshift SN Ia spectra appear remarkably similar to those observed at low-redshift. This can be seen in the high signal-to-noise ratio spectra of SN 1995ao ($z=0.30$) and SN 1995ap ($z=0.23$) in Figure 1. Another demonstration of this similarity at even higher redshift is shown in Figure 11 for SN 1998ai ($z=0.49$; IAUC 6861) whose light curve was not used in this work.  The spectrum of SN 1998ai was obtained at the Keck telescope with a 5 x 1800 s exposure using LRIS and was reduced as described in \\S 2.2 (Filippenko et al. 1998). The spectral characteristics of this SN Ia appear to be indistinguishable from the range of characteristics at low redshift to good precision. In additon, a time sequence of spectra of SN Ia 1997ex ($z$=0.36; Nugent et al. 1998a) compared with those of local SNe Ia reveals no significant spectral differences (Filippenko et al. 1998).  We expect that our local calibration will work well at eliminating any pernicious drift in the supernova distances between the local and distant samples. Until we know more about the stellar ancestors of SNe Ia, we need to be vigilant for changes in the properties of the supernovae at significant look-back times. Our distance measurements could be particularly sensitive to changes in the colors of SNe Ia for a given light curve shape.  Although our current observations reveal no indication of evolution of SNe Ia at $z \\approx 0.5$, evolution remains a serious concern which can only be eased and perhaps understood by future studies.  \\subsection{Extinction} Our SNe Ia distances have the important advantage of including corrections for interstellar extinction occurring in the host galaxy and the Milky Way.  The uncertainty in the extinctions is a significant component of error in our distance uncertainties.  Extinction corrections based on the relation between SN Ia colors and luminosity improve distance precision for a sample of SNe Ia that includes objects with substantial extinction (Riess, Press, \\& Kirshner 1996a). Yet, in practice, we have found negligible extinction to the high-redshift SNe Ia.  The mean $B-V$ color at maximum is $-0.13 \\pm 0.05$ from the MLCS method and $-0.07 \\pm 0.05$ from the template fitting approach, consistent with an unreddened $B-V$ color of $-0.10$ to $-0.05$ expected for slowly declining light curves as observed in the high-redshift sample (Riess, Press, \\& Kirshner 1996a; Appendix).  Further, the consistency of the measured Hubble flow from SNe Ia with late-type and early-type hosts (\\S 5.1) shows that the extinction corrections applied to dusty SNe Ia at low redshift do not alter the expansion rate from its value measured from SNe Ia in low dust environments.  The conclusions reached in \\S 4 would not alter if low and high-redshift SNe with significant extinction were discarded rather than included after a correction for extinction.  The results of \\S 4 do not depend on the value of the ratios between color excess and selective absorption used to determine the extinctions of the high-redshift sample because the mean observed reddening is negligible.   Some modest departures from the Galactic reddening ratios have been observed in the Small and Large Magellanic Clouds, M31, and the Galaxy, and they have been linked to metallicity variations (Walterbos 1986; Hodge \\& Kennicutt 1982; Bouchet et al 1985; Savage \\& Mathis 1979).  Although our current understanding of the reddening ratios of interstellar dust at high redshift is limited, the lack of any significant color excess observed in the high-redshift sample indicates that the type of interstellar dust which reddens optical light is not obscuring our view of these objects.   Riess, Press, \\& Kirshner (1996b) found indications that the Galactic ratios between selective absorption and color excess are similar for host galaxies in the nearby ($z$ $\\leq 0.1$) Hubble flow.  Yet, what if these ratios changed with look-back time?  Could an evolution in dust grain size descending from ancestral interstellar ``pebbles'' at higher redshifts cause us to underestimate the extinction?  Large dust grains would not imprint the reddening signature of typical interstellar extinction upon which our corrections rely. However, viewing our SNe through such grey interstellar grains would also induce a {\\it dispersion} in the derived distances.  To estimate the size of the dispersion, we assume that the grey extinction is distributed in galaxies in the same way as typical interstellar extinction.  Hatano, Branch, \\& Deaton (1997) have calculated the expected distribution of SN Ia extinction along random lines of sight in the host galaxies. A grey extinction distribution similar to theirs could yield differing amounts of mean grey extinction depending on the likelihood assigned to observing an extinction of $A_B$=0.0 mag.  In the following calculations we vary only the likelihood of $A_B$=0.0 mag to derive new extinction distributions with varying means. These different distributions also have differing dispersions of extinction. A mean grey extinction of 0.25 mag would be required to explain the measured MLCS distances without a cosmological constant.  Yet the dispersion of individual extinctions for a distribution with a mean of 0.25 mag would be $\\sigma_{A_B}$=0.40 mag, significantly {\\it larger} than the 0.21 mag dispersion observed in the high-redshift MLCS distances. Grey extinction is an even less likely culprit with the template fitting approach; a distribution with a mean grey extinction of 0.28 mag, needed to replace a cosmological constant, would yield a dispersion of 0.42 mag, significantly higher than the distance dispersion of 0.17 mag observed in the high-redshift template fitting distances.  Furthermore, most of the observed scatter is already consistent with the estimated statistical errors as evidenced by the $\\chi^2_\\nu$ (Table 8), leaving little to be caused by grey extinction. Nevertheless, if we assumed that {\\it all} of the observed scatter were due to grey extinction, the mean shift in the SNe Ia distances would only be 0.05 mag. With the observations presented here, we cannot rule out this modest amount of grey interstellar extinction.  This argument applies not only to exotic grey extinction but to any interstellar extinction not accounted for which obscures SNe Ia.  Any spotty interstellar extinction which varies with line-of-sight in a way similar to the Hatano, Branch, \\& Deaton (1997) model of galaxies will add dispersion to the SN Ia distances.  The low dispersion measured for the high-redshift sample places a strong limit on any mean spotty interstellar extinction.  Grey intergalactic extinction could dim the SNe without either telltale reddening or dispersion, if all lines of sight to a given redshift had a similar column density of absorbing material. The component of the intergalactic medium with such uniform coverage corresponds to the gas clouds producing Lyman-$\\alpha$ forest absorption at low redshifts.  These clouds have individual H I column densities less than about $10^{15} \\, {\\rm cm^{-2}}$ (Bahcall et al. 1996).  However, these clouds display low metallicities, typically less than 10\\% of solar. Grey extinction would require larger dust grains which would need a larger mass in heavy elements than typical interstellar grain size distributions to achieve a given extinction.  Furthermore, these clouds reside in hard radiation environments hostile to the survival of dust grains.  Finally, the existence of grey intergalactic extinction would only augment the already surprising excess of galaxies in high-redshift galaxy surveys (Huang et al. 1997).  We conclude that  grey extinction does not seem to provide an observationally or physically plausible explanation for the observed faintness of high-redshift SNe Ia.  \\subsection{Selection Bias}   Sample selection has the potential to distort the comparison of nearby and distant supernovae. Most of our nearby ($z < 0.1$) sample of SNe~Ia was gathered from the Cal\\'an/Tololo survey (Hamuy et al. 1993a) which employed the blinking of photographic plates obtained at different epochs with Schmidt telescopes and from less well-defined searches (Riess et al. 1998b). Our distant ($z>0.16$) sample was obtained by subtracting digital CCD images at different epochs with the same instrument setup.  If they were limited by the flux of the detected events, both nearby and distant SN Ia searches would preferentially select intrinsically luminous objects because of the larger volume of space in which these objects can be detected.  This well-understood selection effect could be further complicated by the properties of SNe Ia; more luminous supernovae have broader light curves (Phillips 1993; Hamuy et al. 1995, 1996c; Riess, Press, \\& Kirshner 1995, 1996a).  The brighter supernovae remain above a detection limit longer than their fainter siblings, yet also can fail to rise above the detection limit in the time interval between successive search epochs.   The complex process by which SNe Ia are selected in low and high-redshift searches can be best understood with simulations (Hamuy \\& Pinto 1998).  Although selection effects could alter the ratio of intrinsically dim to bright SNe Ia in our samples relative to the true population, our use of the light curve shape to determine the supernova's luminosity should correct most of this selection bias on our distance estimates. However, even after our light-curve shape correction, SNe~Ia still have a small dispersion as distance indicators ($\\sigma \\approx 0.15$ mag), and any search program would still preferentially select objects which are brighter than average for a particular light curve shape and possibly select objects whose light curve shapes aid detection.  To investigate the consequence of sample selection effects, we used a Monte Carlo simulation to understand how SNe Ia in our nearby and distant samples were chosen. For the purpose of this simulation we first assumed that the SN Ia rate is constant with look-back time.  We assembled a population of SNe Ia with luminosities described by a Gaussian random variable $\\sigma_{M_B}=0.4$ mag and light-curve shapes which correspond to these luminosities as described by the MLCS vectors (see the Appendix).  A Gaussian random uncertainty of $\\sigma = 0.15$ mag is assumed in the determination of absolute magnitude from the shape of a supernova's light curve.  The time interval between successive search epochs, the search epoch's limiting magnitudes, and the apparent light-curve shapes were used to determine which SNe Ia were ``discovered'' and included in the simulation sample. A separate simulation was used to select nearby objects, with the appropriate time interval between epochs and estimates of limiting magnitudes.  The results are extremely encouraging, with recovered values exceeding the simulated value of $\\Omega_M$ or $\\Omega_\\Lambda$ by only 0.02 for these two parameters considered separately. Smoothly increasing the SN Ia rate by a factor of 10 by $z=1$ doubles this bias to 0.04 for either parameter. There are two reasons we find such a small selection bias in the recovered cosmological parameters.  First, the small dispersion of our distance indicator results in only a modest selection bias.  Second, both nearby and distant samples include an excess of brighter than average SNe, so the {\\it difference} in their individual selection biases remains small.   As discussed by Schmidt et al. (1998), obtaining accurate limiting magnitudes is complex for the CCD-based searches, and essentially impossible for the photographic searches. Limiting magnitudes vary from frame to frame, night to night, and film to film, so it is difficult to use the actual detection limits in our simulation. Nevertheless, we have run simulations varying the limiting magnitude, and this does not change the results significantly.  We have also tried increasing the dispersion in the SN~Ia  light curve shape vs. absolute magnitude correlation at wavelengths shorter than $5000$~\\AA. Even doubling the distance dispersion of SNe~Ia (as may be the case for rest-frame $U$) does not significantly change the simulation results.  Although these simulations bode well for using SNe~Ia to measure cosmological parameters, there are other differences between the way nearby and distant supernova samples are selected which are more difficult to model and are not included in our present simulations.  Von Hippel, Bothun, \\& Schommer (1997) have shown that the selection function of the nearby searches is not consistent with that of a strict magnitude-limited search. It is unclear whether a photographic search selects SNe Ia with different parameters or environments than a CCD search or how this could affect a comparison of samples. Future work on quantifying the selection criteria of the samples is needed. A CCD search for SNe Ia in Abell clusters by Reiss et al. (1998) will soon provide a nearby SN Ia sample with better understood selection criteria. Although indications from the distributions of SN Ia parameters suggest that both our searches have sampled the same underlying population (see Figure 10), we must continue to be wary of subtle selection effects which might bias the comparison of SNe~Ia near and far.   \\subsection{Effect of a Local Void}  It has been noted by Zehavi et al. (1998) that the SNe Ia out to 7000 km s$^{-1}$ exhibit an expansion rate which is 6\\% greater than that measured for the more distant objects.  The significance of this peculiar monopole is at the 2$\\sigma$ to 3$\\sigma$ confidence level; it is not inconsistent with the upper limit of $\\sim$ 10\\% for the difference between the local and global values of $H_0$ found by Kim et al. (1997).  The implication is that the volume out to this distance is underdense relative to the global mean density.  This effect appears as an excess redshift for a given distance modulus (within 7000 km s$^{-1}$) and can be seen with both the MLCS method and the template fitting method in Figures 4 and 5 .  If true, what effect would this result have on our conclusions? In principle, a local void would increase the expansion rate measured for our low-redshift sample relative to the true, global expansion rate.  Mistaking this inflated rate for the global value would give the false impression of an increase in the low-redshift expansion rate relative to the high-redshift expansion rate.  This outcome could be incorrectly attributed to the influence of a positive cosmological constant. In practice, only a small fraction of our nearby sample is within this local void, reducing its effect on the determination of the low-redshift expansion rate.  As a test of the effect of a local void on our constraints for the cosmological parameters, we reanalyzed the data discarding the seven SNe Ia within 7000 km s$^{-1}$ (108 Mpc for $H_0=65$).  The result was a reduction in the confidence that $\\Omega_\\Lambda > 0$ from 99.7\\% (3.0$\\sigma$) to 98.3\\% (2.4$\\sigma$) for the MLCS method and from $>$99.9\\% (4.0$\\sigma$) to 99.8\\% (3.1$\\sigma$) for the template fitting approach.  The tests for both methods excluded the unclassified SN 1997ck and included the snapshot sample, the latter without two SNe Ia within 7000 km s$^{-1}$.  As expected, the influence of a possible local void on our cosmological conclusions is relatively small.  \\subsection{Weak Gravitational Lensing}  The magnification and demagnification of light by large-scale structure can alter the observed magnitudes of high-redshift supernovae (Kantowski, Vaughan, \\& Branch 1995).  The effect of weak gravitational lensing on our analysis has been quantified by Wambsganss et al. (1997) and summarized by Schmidt et al. (1998).  SN Ia light will, on average, be demagnified by $0.5$\\% at $z=0.5$ and $1$\\% at $z=1$ in a Universe with a non-negligible cosmological constant. Although the sign of the effect is the same as the influence of a cosmological constant, the size of the effect is negligible.  Holz \\& Wald (1997) have calculated the weak lensing effects on supernova light from ordinary matter which is not smoothly distributed in galaxies but rather clumped into stars (i.e., dark matter contained in MACHOS). With this scenario, microlensing by compact masses becomes a more important effect further decreasing the observed supernova luminosities at $z=0.5$ by 0.02 mag for $\\Omega_M$=0.2 (Holz 1998).  Even if most ordinary matter were contained in compact objects, this effect would not be large enough to reconcile the SNe Ia distances with the influence of ordinary matter alone.   \\subsection{Light Curve Fitting Method}  As described in \\S 3.2, two different light curve fitting methods, MLCS (Riess, Press, \\& Kirshner 1996a; Appendix) and a template fitting approach (Hamuy et al. 1995, 1996d), were employed to determine the distances to the nearby and high-redshift samples.  Both methods use relations between light curve shape and luminosity as determined from SNe Ia in the nearby Hubble flow.  Both methods employ an extinction correction from the measured color excess using relations between intrinsic color and light curve shape.  In addition, both the MLCS and template fitting methods yield highly consistent measurements for the Hubble constant of $H_0$=65.2 $\\pm 1.3$ and $H_0$=63.8 $\\pm 1.3$, respectively not including any uncertainty in the determination of the SN Ia absolute magnitude which is the dominant uncertainty.  It is also worth noting that both methods yield SN Ia distance dispersions of $\\sim$ 0.15 mag when complete light curves in $B,V,R$, and $I$ are employed.  For the purpose of comparing the same data at high and low redshifts, the use of SN Ia observations at low redshift were restricted to only $B$ and $V$ within 40 days of maximum light.  Although the conclusions reached by the two methods when applied to the high-redshift SNe are highly consistent, some differences are worth noting. There are small differences in the distance predictions at high redshift.  For the distant sample, the template fitting distances exhibit a scatter of 0.17 mag around the best fit model as compared to 0.21 mag for the MLCS method.  In addition, the template fitting distances to the high-redshift SNe Ia are on (weighted) average 0.03 mag farther than the MLCS distances relative to the low-redshift sample.  These differences together result in slightly different confidence intervals for the two methods (see Figures 6, 7, and 8 and Table 8).  For the set of 10 well-observed SNe Ia, a sample with scatter 0.17 mag or less is drawn from a population of scatter 0.21 mag 25\\% of the time. The chance that 10 objects could be drawn from this same population with a mean difference of 0.03 mag is 66\\%.   Future samples of SNe Ia will reveal if the observed differences are explained by chance.  Until then, we must consider the difference between the cosmological constraints reached from the two fitting methods to be a systematic uncertainty.  Yet, for the data considered here, both distance fitting methods unanimously favor the existence of a non-negligible, positive cosmological constant and an accelerating Universe.   \\subsection{Sample Contamination}   The mean brightness of SNe Ia is typically 4 to 40 times greater than that of any other type of supernova, favoring their detection in the volume of space searched at high redshift. Yet in the course of our high-redshift supernova search (and that of the Supernova Cosmology Project; Perlmutter et al. 1995) a small minority of other supernova types have been found and we must be careful not to include such objects in our SN Ia sample.   The classification of a supernova is determined from the presence or absence of specific features in the spectrum (Wheeler \\& Harkness 1990; Branch, Fisher, \\& Nugent 1993; Filippenko 1997).  The spectra of Type Ia supernovae show broad Si II absorption near 6150 \\AA\\, Ca II (H\\&K) absorption near 3800 \\AA\\, a S II absorption doublet near 5300 \\AA\\ and 5500 \\AA\\, and numerous other absorption features with ionized Fe a major contributor (Filippenko 1997). For supernovae at high redshift, some of these characteristic features shift out of the observer's frequency range as other, shorter wavelength features become visible.  Classification is further complicated by low signal-to-noise ratio in the spectra of distant objects. The spectra of SNe Ia evolve with time along a remarkably reliable sequence (Riess et al. 1997). Final spectral classification is optimized by comparing the observed spectrum to well-observed spectra of SNe Ia at the same age as determined from the light curves.  For most of the spectra in Figure 1, the identification as a SN Ia is unambiguous.  However, in three of the lowest signal-to-noise ratio cases -- 1996E, 1996H, and 1996I -- the wavelengths near Si II absorption (rest-frame 6150 \\AA\\ ) were poorly observed and their classification warrants closer scrutiny.  These spectra are inconsistent with Type II spectra which show H$\\beta$ (4861 \\AA\\ ) in emission and absorption and lack Fe II features shortly after maximum.  These spectra are also inconsistent with Type Ib spectra which would display He I~$\\lambda$5876 absorption at a rest wavelength of $\\sim$ 5700 \\AA .  The most likely supernova type to be misconstrued as a Type Ia is a Type Ic, as this type comes closest to matching the SN Ia spectral characteristics. Although SN Ic spectra lack Si II and S II absorption, the maximum-light spectra at blue wavelengths can resemble those of SNe Ia $\\sim$ 2 weeks past maximum when both are dominated by absorption lines of Fe II with P Cygni profiles.  Type Ic events are rare and one luminous enough to be found in our search would be rare indeed, but not without precedent.  An example of such an object is SN 1992ar (Clocchiatti et al. 1998), which was discovered in the course of the Cal\\'{a}n/Tololo SN survey and which reached an absolute magnitude, uncorrected for host galaxy dust extinction, of $M_V = -19.3$ ($H_0$ = 65 km s$^{-1}$ Mpc$^{-1}$).  For both SN 1996H and SN 1996I, the spectral match with a Type Ia at rest wavelengths less than 4500 \\AA\\ is superior to the fit to a Type Ic spectrum (see Figure 1).  In both cases the spectra rise from deep troughs at the 3800 \\AA\\ Ca II break (rest-frame) to strong peaks at 3900 to 4100 \\AA\\ (rest-frame) as observed in SNe Ia.  Type Ic spectra, by comparison, tend to exhibit a much weaker transition from trough to peak redward of the Ca II break (see Figure 12).  For SN 1996E, the spectral coverage does not extend blueward of a rest wavelength of 4225 \\AA\\ rendering this diagnostic unusable. The absence of pre-maximum observations of SN 1996E makes it difficult to determine the age of the spectrum and that of the appropriate comparison spectra. As shown in Figure 12, the spectroscopic and photometric data for SN 1996E are consistent with a SN Ia caught $\\sim$ 1 week after maximum light, or a luminous SN Ic discovered at maximum. There is a weak indication of S II absorption at $\\sim$ 5375 \\AA\\ which favors classification as a Type Ia (see Figures 1 and 12), but this alone does not provide a secure classification. Note that the $K$-corrections for a SN Ia or SN Ic at this redshift ($z=0.43$) would be nearly identical due to the excellent match of the observed filters ($B45$ and $V45$) to the rest-frame ($B$ and $V$) filters.  We have reanalyzed the cosmological parameters discarding SN 1996E as a safeguard against the possible contamination of our high-redshift sample.  We also excluded SN 1997ck which, for lack of a definitive spectral classification, is an additional threat to contamination of our sample.  With the remaining ``high-confidence'' sample of 14 SNe Ia we find the statistical likelihood of a positive cosmological constant to be 99.8\\% (3.1 $\\sigma$) from the MLCS method, a modest increase from 99.7\\% (3.0 $\\sigma$) confidence when SN 1996E is included.  For the template fitting approach, the statistical confidence in a positive cosmological constant remains high at $>$99.9\\% (4.0 $\\sigma$), the same result as with SN 1996E. We conclude that for this sample our results are robust against sample contamination, but the possible contamination of future samples remains a concern. Even given existing detector technology, more secure supernova classifications can be achieved with greater signal-to-noise ratios for observed spectra, with optimally timed search epochs which increase the likelihood of pre-maximum discovery, and with an improved empirical understanding of the differences among the spectra of supernova types.  \\subsection{Comparisons}  The results reported here are consistent with other reported observations of high-redshift SNe Ia from the High-z Supernova Search Team (Garnavich et al. 1998; Schmidt et al. 1998), and the improved statistics of this larger sample reveal the potential influence of a positive cosmological constant.  These results are inconsistent at the $\\sim 2\\sigma$ confidence level with those of Perlmutter et al. (1997), who found $\\Omega_M=0.94 \\pm 0.3 \\, (\\Omega_\\Lambda=0.06)$ for a flat Universe and $\\Omega_M=0.88 \\pm 0.64$ for $\\Omega_\\Lambda \\equiv 0$.  They are marginally consistent with those of Perlmutter et al. (1998) who, with the addition of one very high redshift SN Ia ($z=0.83$), found $\\Omega_M=0.6 \\pm 0.2 \\, (\\Omega_\\Lambda=0.4)$ for a flat Universe and $\\Omega_M=0.2 \\pm 0.4$ for $\\Omega_\\Lambda \\equiv 0$.  Although the experiment reported here is very similar to that performed by Perlmutter et al. (1997, 1998), there are some differences worth noting.  Schmidt et al. (1998), Garnavich et al. (1998), and this paper explicitly correct for the effects of extinction evidenced by reddening of the SNe Ia colors.  Not correcting for extinction in the nearby and distant sample could affect the cosmological results in either direction since we do not know the sign of the difference of the mean extinction. In practice we have found few of the high-redshift SNe Ia to suffer measurable reddening.  A number of objects in the nearby sample display moderate extinction for which we make individual corrections.  We also include the Hubble constant as a free parameter in each of our fits to the other cosmological parameters. Treating the nearby sample in the same way as the distant sample is a crucial requirement of this work.  Our experience observing the nearby sample aids our ability to accomplish this goal.  The statistics of gravitational lenses provide an alternate method for constraining the cosmological constant (Turner 1990; Fukugita, Futamase, \\& Kasai 1990).  Although current gravitational lensing limits for the cosmological constant in a flat Universe ($\\Omega_\\Lambda \\leq 0.66$ at 95\\% confidence; Kochanek 1996) are not inconsistent with these results, they are uncomfortably close.  Future analysis which seeks to limit systematic uncertainties affecting both experiments should yield meaningful comparisons. The most incisive independent test may come from measurements of the fluctuation spectrum of the cosmic microwave background.  While the supernova measurements provide a good constraint on $\\Omega_M - \\Omega_\\Lambda$, the CMB measurements of the angular scale for the first Doppler peak, referring to much earlier epochs, are good measures of $\\Omega_M + \\Omega_\\Lambda$ (White \\& Scott 1996).  Since these constraints are nearly orthogonal in the coordinates of Figure 6 and 7, the region of intersection could be well defined.  Ongoing experiments from balloons and the South Pole may provide the first clues to the location of where that intersection.  {\\it Our detection of a cosmological constant is not limited by statistical errors but by systematic ones.}  Further intensive study of SNe Ia at low ($z$ $<$ 0.1), intermediate ($0.1 \\leq z \\leq 0.3$), and high ($z >$ 0.3) redshifts is needed to uncover and quantify lingering systematic uncertainties in this striking result."
    },
    "9805/astro-ph9805171_arXiv.txt": {
        "abstract": "Spectroscopic orbits have been reported for nine unseen companions orbiting solar-type stars with minimum possible masses in the range 0.5 to 10 Jupiter masses.  We compare the mass distribution of these nine planet candidates with the distribution of low-mass secondaries in spectroscopic binaries. Although we still have only a very small number of systems, the two distributions suggest two distinctive populations. The transition region between the two populations might be at the range of 10--30 Jupiter masses.  \\subjectheadings{binaries: spectroscopic --- planetary systems} ",
        "introduction": "Eight candidates for extrasolar planets have been announced over the past two years (e.g., Marcy \\& Butler 1998).  In each case, very precise stellar radial-velocity measurements, with a precision of about 10~m~s$^{-1}$ or better, indicated the presence of a low-mass unseen companion orbiting a nearby solar-type star.  The individual masses of the eight companions are not known, because the inclination angles of their orbital planes relative to our line of sight could not have been measured. The minimum masses for the eight candidates, attained for an inclination angle of $90^{\\circ}$, are in the range 0.5 to 7.4 Jupiter masses (\\MJ). These findings render the eight companions to be giant planets or at least `planet candidates'.  The detections of these eight companions were announced seven to nine years after a companion of HD 114762 was discovered (Latham et al. 1989), based on measurements with a lower precision (Latham 1985).  Mazeh, Latham, \\& Stefanik (1996) have shown that the minimum mass for the companion of HD~114762 is 9.4 \\MJ. Therefore, when considering the emerging population of planet candidates, HD~114762 should be considered together with the eight new candidates. Table 1 lists the minimum mass, period and discovery date of the nine objects.  For random orbital orientations, the expectation value for sin $i$ is 0.76, so the actual masses of the nine companions are expected to be in the range of 0.6--12 \\MJ.  \\begin{deluxetable}{lcrlc} \\tablewidth{0pt} \\tablecaption{The Planet-Candidates} \\tablehead{ \\colhead{Name} &\\colhead{$M_2 \\sin i$} & \\colhead{P\\phantom{.0}} & \\colhead{Discovery} & \\colhead{Ref.}\\nl & \\colhead{$(M_{J})$} & \\colhead{(days)}&\\colhead{Date}  \\nl} \\startdata HD 114762      &   9.4    &  84{\\phantom{.0}}    &  1989  & 1,2 \\nl 51 Peg         &   0.5    &  4.2                 &  1995  & 3 \\nl 47 UMa         &   2.5    &  1090{\\phantom{.0}}  &  1996  & 4 \\nl 70 Vir         &   7.4    &  117{\\phantom{.0}}   &  1996  & 5 \\nl 55 Cnc         &   0.8    &  14.7                &  1996  & 6 \\nl $\\tau$ Boo     &   3.9    &  3.3                 &  1996  & 6 \\nl $\\upsilon$ And &   0.7    &  4.6                 &  1996  & 6 \\nl 16 Cyg B       &   1.6    &  804{\\phantom{.0}}   &  1996  & 7 \\nl $\\rho$ CrB &   1.1    &   39.6               &  1997  & 8 \\nl \\enddata \\tablecomments{$^1$Latham et al. 1989; $^2$Mazeh, Latham \\& Stefanik 1996; $^3$Mayor \\& Queloz 1995; $^4$Butler \\& Marcy 1996; $^5$Marcy \\& Butler 1996; $^6$Butler et al. 1997; $^7$Cochran et al. 1997; $^8$Noyes et al. 1997.} \\end{deluxetable}   The nature of the newly discovered low-mass companions is not yet clear. They could be planets, as suggested by various authors (e.g., Marcy \\& Butler 1998), or many could just be brown-dwarf secondaries, formed in binary stars (Black 1997).  With the small, but not insignificant number of spectroscopic orbits implying planetary minimum masses, we can now begin to study the distribution of their orbital parameters, in order to address this very basic question.  In this paper we discuss the emerging difference between the {\\it mass} distribution of planet candidates and the low-mass end of the distribution of binary secondaries.  This point has been already discussed by previous studies (Basri \\& Marcy 1997; Mayor, Queloz \\& Udry 1998; Mayor, Udry \\& Queloz 1998; Marcy \\& Butler 1998), but in those papers the mass distribution was binned linearly. Here we choose to use a logarithmic scale to study the mass distribution, because of the large range of masses, 0.5--300 \\MJ, involved.  The logarithmic scale has also been used by Tokovinin (1992) to study the secondary mass distribution in spectroscopic binaries, and was suggested by Black (1998) to study the mass distribution of the planetary-mass companions.  This work is based on an extremely small sample, and the validity of our results will need to be verified by many more detections. However, if verified, the difference in mass distributions that we find might provide an important clue for how to distinguish between planets and low-mass stellar companions.  A preliminary version of this work was presented at the meeting ``Physical Processes in Astrophysical Fluids'', in Haifa, January 1998 (Mazeh 1998). ",
        "conclusions": "The corrected combined histogram might suggest that we see here two populations.  At the high-mass end of the histogram we see a distribution which drops steeply when we move from 200 to 20 \\MJ. At the planetary range of masses we see a flat distribution, which might even rise very mildly when we move from, say, 20 to 0.6 \\MJ.  Unfortunately, the number of systems in each bin is small. However, the two different slopes in the two parts of the diagram seem real, as they are based on more than one bin.  The derived diagram depends on two parameters --- \\Kmin\\ and the number of bins of the histogram. We got the same gross features, namely two opposite slopes in the two parts of the diagram, when we changed the values of these two parameters. Dividing the total range of the diagram into 5 or 4 bins instead of 6 bins shifted the transition region between the two slopes somewhat to the right. Changing \\Kmin\\ from 20 to 50 \\ms\\ made the slope at the left hand side of the diagram steeper. We conclude therefore that the overall shape of the diagram does not depend strongly on the specific values of the parameters of the derivation.  The transition region between the two populations is at about 10--30~\\MJ. Unfortunately, the relative error of this bin is very large. Nevertheless, it seems that this is the bin with the smallest number of systems. The very low count estimate in this bin is supported by the fact that the very sensitive searches for planets, which yielded the discovery of the eight new planet candidates, did not find any companions with minimum masses between 10 and 30 \\MJ.  With \\Kmin\\ of 20~m~s$^{-1}$ these searches could detect more than 99\\% of the binaries in this bin. The fact that no precise search discovered any binary in this bin indicates that the number of systems with secondary masses between 10 and 30 \\MJ\\ is very small.  The drop of the secondary mass distribution we find when moving from 200 to 20~\\MJ\\ is consistent with the finding of Halbwachs, Mayor \\& Udry (1998), who studied the mass {\\it ratio} distribution of spectroscopic binaries in the samples of G and K stars of Mayor et al. (1997).  Halbwachs, Mayor \\& Udry found a flat histogram of the mass ratio, although they could not exclude increasing or decreasing power laws of the form $q^{\\alpha}$, where $q$ is the mass ratio and $-0.82\\leq \\alpha\\leq 0.87$. The flat distribution of $q$ yields constant $dN/dm_2$, if all primary masses are similar.  This corresponds to $dN/d\\log(m_2)\\propto m_2$, consistent with our findings. The drop we find is also consistent with the findings of Mayor, Queloz \\& Udry (1998; see also Mayor, Udry \\& Queloz 1998) who found that $dN/dm_2\\propto m_2^{-0.4}$. Their result corresponds to $dN/d\\log(m_2)\\propto m_2^{+0.6}$. Figure 2 of this work suggests a steeper drop, but the difference is within the errors.  The transition region between the two populations, or between the two slopes, that we find here is, however, different from the findings of Mayor, Queloz \\& Udry (1998) and Mayor, Udry \\& Queloz (1998). They find a borderline at 7 \\MJ, while our logarithmic treatment of the data suggests a transition region at the range of 10--30 \\MJ. Another difference is the shape of the distribution in the planetary mass range. They find a very steep rising distribution when moving down towards the range of 1--5 \\MJ.  We find an almost flat logarithmic distribution, with perhaps a mild rise towards lower masses, depending on the exact value of \\Kmin.  Let us {\\it assume} that Figure 2 indeed shows two distinctive slopes in the two parts of the diagram.  Let us further {\\it assume} that this reflects the fact that we see here two different populations, one below 10--30 \\MJ, and one with masses larger than this transition region.  One possible interpretation of the diagram, if indeed we see here two different populations, is that the two populations were formed differently. Maybe the lower-mass population was formed like planets, out of an accretion disc, while the higher-mass population was formed like binary stars, in a mechanism which probably involves large-scale gravitational collapse (e.g. Boss 1996; Black 1986).  If this is the case then the binary secondaries include stars and brown dwarfs together.  Figure 2 suggests that the transition region between the two populations is at about 10--30~\\MJ. This is of astrophysical significance, if indeed the lower-mass population is composed of planets, as it might tell us about the lower limit and upper limit of the formation of secondaries and planets, respectively (Marcy \\& Butler 1995, 1998; Mayor, Queloz \\& Udry 1998; Mayor, Udry \\& Queloz 1998).  The upper limit of the planetary masses is set by the conditions in the accretion disc, and most probably by the interaction between the planet and the gas and dust in the disc. Boss (1996) already noted that Lin and Papaloizou (1980) theoretically predicted that the maximum mass for the formation of a planet in the disc of the Solar nebula is about 1 \\MJ. As the maximum mass depends on the mass of the early nebula, we can get somewhat higher masses in different cases.  The lower limit for secondary masses in binaries is set by the binary formation mechanism, whatever that mechanism might be. Boss (1988), for example, noted that the theory of opacity limited cloud fragmentation predicts that the minimum mass for a companion is about 10 \\MJ.  In fact, Low \\& Lynden-Bell (1976) estimated already 20 years ago that the minimum Jeans mass for fragmentation of a molecular cloud is 7 \\MJ. Silk (1977), in a contemporaneous study, came up with minimum masses between 10 and 100 \\MJ, depending on the shape of the collapse. If we indeed see the transition region between the two populations at about 10--30 \\MJ, this is not too far from the predictions of the theories.  Duquennoy \\& Mayor (1991; see also Mayor, Queloz \\& Udry 1998) have suggested that the observed orbital eccentricities can be used to distinguish between planets and stellar companions. However, Mazeh, Mayor \\& Latham (1996), when discussing the eccentricity versus mass of the known planet candidates, pointed out that the planet-disc interaction (e.g., Goldreich \\& Tremaine 1980) is a possible mechanism for generating a strong dependence of eccentricity versus mass (Artymowicz 1992; Lubow \\& Artymowicz 1996), at least for moderate eccentricities. This possibility can undermine the potential of the eccentricity-mass dependence to distinguish between planets and secondaries. Furthermore, Black (1997) analyzed the eccentricity as a function of {\\it period} and concluded that the eccentricities observed are consistent with the assumption that all the planet candidates are actually low-mass brown dwarfs formed like binary stars. It seems therefore that it might be too early to distinguish between brown dwarfs and planets solely on the basis of their orbital eccentricity.  Mazeh, Mayor \\& Latham (1996) speculated that ``The 10--40 \\MJ\\ mass gap may prove to be critical for the interpretation of'' the eccentricity-mass dependence.  We confirm here that the transition region between the two populations could be at this range of masses.  Obviously, the left hand side of the histogram and the transition region between the two slopes derived in this paper are based on a very small number of objects all together, and these features need to be verified by many more detections. Further, one still needs to make sure that the different slope in the planetary-mass range is not due to some selection effects. For example, there might be a correlation between the orbital period and the secondary mass, which might make the small-mass secondaries easier to detect. Such an effect could cause the histogram to appear to rise towards smaller mass. However, if the shape of the histogram can be verified, and if the planetary-mass objects prove to be extrasolar planets, the shape of the histogram might give us the long-sought clue for how to distinguish planets from low-mass stellar companions.  We express our thanks to J.-L. Halbwachs for his very useful comments on the manuscript. We thank M. Mayor, S. Udry and J.-L. Halbwachs for letting us use their unpublished results.  We thank the referee, Dr. D. Black, for a critical reading of the manuscript and for his comments that led to significant improvement of the paper.  This work was supported by US-Israel Binational Science Foundation grant 94-00284 and by the Israeli Science Foundation."
    },
    "9805/astro-ph9805347_arXiv.txt": {
        "abstract": "A magnetically--structured accretion disk corona, generated by buoyancy instability in the disk, can account for observations of flare--like events in Active Galactic Nuclei.  We examine how Petschek magnetic reconnection, associated with MHD turbulence, can result in a violent release of energy and heat the magnetically closed regions of the corona up to canonical X-ray emitting temperatures.  X-ray magnetic flares, the after effect of the energy released in slow shocks, can account for the bulk of the X-ray luminosity from Seyfert galaxies and consistently explain the observed short--timescale variability. ",
        "introduction": "Active Galactic Nuclei (AGN) are believed to be powered by accretion. UV and X-ray observations of AGN, particularly in Seyfert 1 galaxies, indicate that the gravitational binding energy of a massive black-hole is dissipated partly in a cold accretion disk and partly in a hot, tenuous corona above it.  Thermal Comptonization of soft UV-radiation in the corona leads to the production of a hard X-ray continuum, some of which is reprocessed by the cold, dense disk (Haardt \\& Maraschi 1991, 1993) giving rise to the observed reflexion features in the spectra.  Recent work, both on the observational and theoretical side, (Sincell \\& Krolik 1997, Nandra 1997, Stern et al. 1995) indicates that the UV continuum in AGN may be mostly produced by reradiation of energy absorbed from X-rays irradiating the accretion disk. Also, in order to explain the different ratios of X-ray and UV luminosities in different objects it has been suggested that the corona consists of localized active regions (e.g. Haardt, Maraschi \\& Ghisellini 1994). Popular models for the production of the X-rays, therefore, suppose that a large part of the disk's dissipation takes place in a small amount of mass in the hot corona.  Variability timescales of the order of few hours observed in AGN also put upper limits on the size of the emitting regions, and imply that enormous amounts of energy need to be released on very short timescales. It is therefore important to understand what processes can efficiently channel significant amounts of energy into the hot coronal medium and the way in which energy can be released in localized flare--type events.  In a differentially rotating disk the evolution of a magnetically structured corona is very hard to suppress.  The strong magnetic fields, continuously generated by the dynamo action in an accretion disk, are strongly buoyant and are forced to invade the region sandwiching the disk itself. Once outside the disk, the magnetic flux tubes can reconnect efficiently and dissipate part of the accretion energy in localized active flares. Buoyancy, therefore, constitutes a mechanism that channels part of the energy released in the accretion process directly into the corona outside the disk. Magnetic reconnection can be responsible for the rapid dissipation of magnetic energy though a field-aligned electric potential in thin current sheets.  We, here, describe how the observationally required large heating rates per unit mass, in the optically thin gas of a hot coronal region, can be accounted for by an ion--acoustic instability in the context of slow shocks associated with Petschek type reconnection and give rise to flare--like events. We estimate that such a process may be responsible for the heating of the coronal plasma in the active regions of an AGN to a level where it emits X-rays at a temperature of $\\sim 10^9 \\K$ even in the presence of Inverse Compton and synchrotron cooling processes.  Magnetic buoyancy, which drives magnetic flux out of the disk, naturally decreases the plasma density to the point where the active region can reconnect very efficiently. X-ray observations imply that the coronal plasma is very tenuous with a density much lower than that of the underlying disk.  We show that the energetics of such flares are consistent with the X-ray luminosities of typical Seyfert galaxies. The onset of magnetic flares through slow shocks associated with Petschek reconnection, should occur over timescales comparable to those required to explain X-ray variability observed in AGN.  This picture is also, by analogy, validated by X-ray observations of the solar corona.  Recent {\\it Yohkoh} observations (Tsuneta 1996, Yokoyama \\& Shibata 1995 and references therein) of solar flares show clear evidence for magnetic reconnection taking place in magnetically confined loop-like volumes. There is evidence that reconnection serves as a highly efficient engine for converting magnetic energy into kinetic and thermal energy with standing slow shocks.  Moreover, it is now well established that in the Sun active regions are formed from the emergence of magnetic flux. It has been shown (Shibata et al. 1989) that magnetic reconnection can be driven self-consistently by the magnetic buoyancy instability (the Parker instability) between emerging flux and the overlying coronal magnetic field. ",
        "conclusions": "We have presented a simple model that takes into account the impulsive dissipation of magnetic energy during coronal flares in AGN. According to standard models, the X-rays are produced by Inverse Compton scattering of lower energy photons on energetic electrons. We show that the energy flow to the electrons that is needed for this process is guaranteed by the release of coronal magnetic energy. The onset of an ion-acoustic instability associated with slow MHD shocks and Petschek reconnection, heats the flare plasma to X-ray temperatures $\\sim 10^9 \\K$, as required by observations. The luminosity produced in a magnetically structured, flaring corona is completely consistent with typical power outputs of AGN if at least $N \\sim 10$ flares are triggered at any given time.  The plasma reaches low enough densities, (high Alfv\\'en speeds), when reconnection is driven by the Parker instability, such that flares are typically triggered at $h \\sim 8 H$ above the accretion disk (coronal structures can have scale-heights greater than the disk). Energetic flare--type events naturally explain the observations of short timescale variability in AGN.  The buoyantly unstable magnetic flux tubes, once outside the disk, rise through the coronal atmosphere at their local Alfv\\'en speed. The timescale for them to rise up to a few scale heights therefore is much shorter than the disk dynamical timescale (and less than the shear timescale in the disk $\\sim 2/3\\Omega$). This implies that tubes will not be disrupted even when flares are triggered a few scale-heights above the disk.  An important problem faced by any model of AGN coronae, is to determine the population of non-thermal particles produced by the slow MHD shock and the effects of the impulsive electric field.  According to theoretical arguments for the kind of electric fields of interest, the turbulence is characterized by an highly anisotropic distribution of ion-acoustic waves.  In these circumstances it has been shown (Heyvaerts et al. 1977) that electron heating occurs with practically no acceleration. Because electrons travel much faster than ion-acoustic waves, the resonance occurs mainly with waves propagating normal to the particle velocity: this leads to angular diffusion with little change in particle energy. This implies that electrons can be characterized by a mean increase of energy or 'temperature'.  In this sense we can treat the hot current sheet as a 'thermal' source of electrons as required by the observations of spectral cut-offs in the X-ray spectra of AGNs.  Note, though, that the presence of an electric field and plasma turbulence in the sheet would inevitably cause acceleration of charged particles to a certain degree. This is particularly relevant in the case of Galactic black hole candidates where recent high energy X-ray observations require the presence of non-thermal electron tails (e.g. Poutanen \\& Coppi 1988). Further investigations are needed but they are beyond the scope of this paper.  The geometry of reconnection described here does not provide the only viable way for the process to occur. In an accretion disk corona, the closed magnetic loops will get twisted by the rotation of the accretion disk. As the twist accumulates, the magnetic loops expand and finally approach the open field configuration. A current sheet is formed inside the expanding loops and, in the presence of resistivity, magnetic reconnection will take place. Open field lines anchored to the disk, may be the region where winds or jets blow from the disk; these could play an important role in the dynamics of the accretion disk itself. The luminosity of flares will then vary with respect to the different values of $B$ in different morphologies in which reconnection takes place, and with the size and the total number of flares.  All of these factors can give rise different covering fractions of the X-ray emitting regions which can result in very different variability timescales as observed in AGN.  Finally, magnetic buoyancy is not strictly necessary in order to maintain a certain level of activity. Once a flare has been triggered, reconnection can be maintained by positive feedback (the fast outflow due to reconnection rarefies the reconnection region, thinning the current channel in this way maintaining the anomalous resistivity at the neutral sheet)."
    },
    "9805/astro-ph9805084_arXiv.txt": {
        "abstract": "Pairs of quasi-periodic oscillations (QPOs) at kilohertz frequencies are a common phenomenon in several neutron-star low-mass X-ray binaries. The frequency separation of the QPO peaks in the pair appears to be constant in many sources and directly related to the neutron star spin frequency. However, in Sco~X-1 and possibly in 4U~1608$-$52, the frequency separation of the QPOs decreases with increasing inferred mass accretion rate. We show that the currently available {\\em Rossi X-ray Timing Explorer\\/} data are consistent with the hypothesis that the frequency separations in all sources vary by amounts similar to the variation in Sco~X-1. We discuss the implications for models of the kilohertz QPOs. ",
        "introduction": "Quasi-periodic X-ray brightness oscillations at kilohertz frequencies (hereafter kHz QPOs) have recently been discovered in many neutron-star low-mass X-ray binaries (LMXBs) with the {\\em Rossi X-ray Timing Explorer\\/} (RXTE; see, e.g., van der Klis et al.\\markcite{vdketal96} 1996; Strohmayer et al.\\markcite{Setal96} 1996). These are strong, often relatively coherent ($\\nu/\\delta \\nu$ up to $\\sim 200$) oscillations that occur commonly in pairs (see van der Klis\\markcite{vdk98} 1998 for a recent review).  The frequencies of the kHz QPOs are comparable to the dynamical timescale near the neutron star surface and depend on the mass accretion rate as inferred from the observed countrates and the spectral properties of the sources (van der Klis et al.\\ 1996; Strohmayer et al.\\ 1996\\markcite{Setal96}; Ford et al.\\markcite{Fetal97a}\\markcite{Fetal97b} 1997a, 1997b;  van der Klis et al.\\markcite{vdketal97} 1997). The peak separation between the lower-frequency (hereafter the lower kHz QPO) and the upper-frequency kHz QPO (hereafter the upper kHz QPO) in a given source is generally consistent with a constant value, independent of the mass accretion rate (Strohmayer et al.\\ 1996; Ford et al.\\ 1997a, 1997b; Wijnands et al.\\markcite{Wetal97b}\\markcite{Wetal98a}\\markcite{Wetal98b} 1997b, 1998a, 1998b). In 4U~1728$-$34 and in 4U~1702$-$43, this peak separation is closely equal to the frequency of the nearly coherent oscillations observed during type~I X-ray bursts that are thought to be produced at the spin frequencies of the neutron stars (Strohmayer et al.\\ 1996;  Strohmayer, Zhang, \\& Swank\\markcite{SZS97}\\markcite{SZS98} 1997, 1998); in 4U~1636$-$536 and in KS~1731$-$26, the peak separation is closely equal to half the frequency of the nearly coherent oscillations observed during type~I X-ray bursts (Smith, Morgan, \\& Bradt\\markcite{SMB97} 1997; Wijnands \\& van der Klis\\markcite{WK97} 1997; Strohmayer et al.\\ 1998).  The above observations offer strong evidence in favor of beat-frequency models, in which the frequency of the lower kHz QPO is the beat frequency between the upper kHz QPO and the neutron star spin (Strohmayer et al.\\ 1996; Miller, Lamb, \\& Psaltis\\markcite{MLP98} 1998). In Sco~X-1, however, which is a luminous LMXB, the peak separation of the kHz QPOs is {\\em not\\/} constant, but decreases with increasing inferred mass accretion rate (van der Klis et al.\\ 1997). In 4U~1608$-$52, which is a less luminous LMXB, there is also evidence for a peak separation that is not constant (M\\'endez et al.\\markcite{Metal98} 1998).  In this paper, we use previously published {\\em RXTE\\/} data on several low-mass X-ray binaries to critically discuss the evidence in favor of a constant frequency separation between the kHz QPO peaks required in any simple beat frequency interpretation. We find that the current data on all sources except Sco~X-1 are insufficient, when used individually, to distinguish between a constant peak separation and a peak separation that varies by amounts similar to those seen in Sco~X-1. When we use the combined dataset of all sources, we find a remarkable correlation between the frequencies of the lower and upper kHz QPOs, which suggests that the peak separation may be varying in all sources. ",
        "conclusions": "In \\S2 we showed that, using current kHz QPO data of neutron-star low-mass X-ray binaries, we cannot reject the hypothesis that the frequency separations of the two kHz QPO peaks in the pair are constant in each source besides Sco~X-1, nor the hypothesis that they vary in a way similar to Sco~X-1. In Sco~X-1, which is the only source with very precisely measured centroid frequencies for the kHz QPOs, the data are inconsistent with a constant peak separation (see also van der Klis et al.\\ 1997). Furthermore, measurements of the peak separation in all other sources in which two simultaneous kHz QPOs have been detected are consistent with change of the kind observed in Sco~X-1. This is the case for the Z sources, which are thought to be accreting at near-Eddington mass accretion rates (see, e.g., Hasinger \\& van der Klis\\markcite{HK89} 1989), as well as for the atoll sources, which are thought to be accreting at substantially lower rates.  These results hint that the frequencies of the upper and lower kHz QPOs in all sources considered individually, including Sco~X-1, are consistent with following a simple (but not necessarily the same) relation, such as a power-law, which is nevertheless very similar even for sources with very different mass accretion rates. Obeying such a relation would contradict any beat-frequency interpretation of the pairs of kHz QPOs in LMXBs, in which the frequency separation of the QPOs is exactly constant. However, the nearly coherent oscillations observed during type~I X-ray bursts in two sources with frequencies closely equal to the peak separations of the kHz QPOs and the fact that the oscillation amplitudes evolve systematically during the bursts are very strong evidence that the peak separations in these sources are similar to the spin frequencies of the neutron stars (Strohmayer et al.\\ 1996, 1997b). Most importantly, in 4U~1728$-$34 the frequencies of the oscillations in the tails of bursts separated by about 20 months are consistent with being constant, implying a timescale for the frequency change of $\\gtrsim 10^3-10^4$~yr (Strohmayer\\markcite{S97} 1997). The only conceivable frequency in these systems that is stable to the degree inferred from the observations of 4U~1728$-$34 is the spin frequency of the neutron star. Therefore, the frequency separation of the two kHz QPO peaks in this source appears to be {\\em closely equal but perhaps not identical\\/} to the spin frequency of the neutron star. In 4U~1636$-$536 and in KS~1731$-$26, the peak separation of the kHz QPOs also appears to be directly related to the neutron star spin frequency.  In current beat-frequency models for the pair of kHz QPOs (see, e.g., Strohmayer et al.\\ 1996; Miller et al.\\ 1998), the upper kHz QPO peak is produced at the Keplerian orbital frequency at a characteristic radius in the accretion disk; the lower kHz QPO peak is then produced at the beat frequency of the upper kHz QPO with the neutron star spin. In order for a beat-frequency model to account for the varying peak separation between the kHz QPOs, one of the above assumptions would need to be relaxed. For example, the frequency that is beating with the neutron star spin to produce the lower kHz QPO may not be the frequency of the upper kHz QPO, i.e., the two frequencies could correspond to different radii in (or heights above) the disk plane (see Miller et al.\\ 1998).  The fact that the variation in the peak separation of the two kHz QPOs in Sco~X-1 is larger than the FWHM of either the lower or upper kHz QPOs implies that the two annuli or regions in the accretion disk responsible for the two kHz QPOs are not overlapping. Alternatively, the frequency that is beating with the upper kHz QPO to produce the lower kHz QPO may be nearly but not strictly equal to the neutron star spin frequency (see, e.g., White \\& Zhang\\markcite{WZ97} 1997).  In conclusion, the data from both the Z and atoll sources are consistent with a varying peak separation between the kHz QPOs. If future data support this conjecture, they will pose interesting new constraints on beat-frequency models for these QPOs: the peak separation should correlate more strongly with the frequency of the upper kHz QPO than with the mass accretion rate or the magnetic field strength."
    },
    "9805/astro-ph9805165_arXiv.txt": {
        "abstract": "\\noindent Selected results from the HEGRA experiment on charged Cosmic Rays and on very high energy gamma-rays are presented. The MAGIC Teles\\-cope is presented as an outlook to the future of Gamma-Ray astronomy. ",
        "introduction": "As a general rule the dynamic range of precision detectors is limited to roughly 2 to 3 orders of magnitude in energy. In case of the charged Cosmic Rays (CR) with an energy spectrum extending over more than 13 orders of magnitude, this neccesitates a large number of different experimental setups in order to cover the full spectrum. Space-borne, i.e. direct, experiments, cover the spectrum from $\\approx $ 10$^{7}$ eV to $\\approx $ 10$^{15}$ eV/nucleon, and ground-based experiments operate above total energies of a few 10$^{12}$ eV up to more than 10$^{20}$ eV.  In the following we will concentrate on the ground-based measurements. Here various experiments which are sensitive in the energy region around 10$^{15}$ eV consistently show a significant steepening of the all-particle spectrum around this energy. When studying the data more closely, however, the agreement between the experiments turns out to be not so good, i.e., differences well above the fluctuations given by the individual errors. This is shown in fig.~1 where the data on the 'knee' in the all-particle spectrum are collected \\cite{bws}. From these data one must conclude that the absolute position, the 'sharpness' of the knee, and also the absolute flux in this energy region are more uncertain than expected from the individual errors. % \\leavevmode \\begin{figure}[tbp] \\centering \\leavevmode \\epsfxsize=10cm \\epsffile{all_spec.ps} \\caption{{\\protect\\small The charged Cosmic Rays all-particle spectrum around the 'knee' as measured by a number of ground-based experiments (taken from \\protect\\cite{bws}).}} \\label{fig-1} \\end{figure} Ground-based experiments use a detector, i.e. the atmosphere as absorber with some added readout elements, like scintillators, Cherenkov detectors, etc., which can only be calibrated in the laboratory to a very limited degree. The calibration therefore has to rely very heavily on MC simulations of the development of the extensive air showers and of the performance of the detectors. One possible reason for the deviations in the measured spectra might thus be the use of different Monte Carlo (MC) generators in the data analyses. This will be briefly discussed below for the most recent data. ",
        "conclusions": ""
    },
    "9805/astro-ph9805023_arXiv.txt": {
        "abstract": "We observed 21 polarized background radio sources in the field of M31 at 1.365 GHz and 1.652 GHz, and determined their rotation measures (RMs). The RM data show that the regular magnetic field of the disk probably extends from about 5 kpc to 25 kpc from the center with similar structure. The RMs obtained from the polarized emission from M31 at $\\lambda$6 cm and $\\lambda$11 cm indicate that M31 might have a  weak poloidal field in its tenuous halo. Observational features of the odd and even dynamo modes in a galaxy are discussed. An even mode (S0) dynamo may operate in M31. ",
        "introduction": "The bright ``ring'' of nonthermal radio emission (Pooley \\cite{poo69}; Beck et al. \\cite{becket98}. See also Fig.~\\ref{fig1}) at a radius of about 10 kpc in the disk of M31 (NGC 224, the Andromeda Nebula) is often referred to as basic evidence for the lowest mode of an axisymmetric dynamo (eg. Beck et al. \\cite{becket96}). The regular magnetic field is aligned  along the ``ring'' (Beck et al. \\cite{becket80}; Beck \\cite{beck82}; Beck et al. \\cite{becket89}). Outside the ``ring'', however, very little is known about the magnetic field, mainly because the polarized radio emission is very weak and superimposed onto extended emission from a foreground Galactic spur (Berkhuijsen \\cite{ber72}; Gr\\\"ave et al. \\cite{graet81}).  The radio disk in galaxies often is  radially  more extended than the optical disk, sometimes also vertically. For example, radially extended radio emission was detected in M51 (see Fig.1 of Berkhuijsen et al. \\cite{beret97}). Radio emission extending far away from the galactic plane was observed from NGC 4631 (Hummel et al. \\cite{humet91}), NGC 891 (Sukumar \\& Allen \\cite{sa91}) and a few other galaxies (eg. NGC 3432 by English \\& Irwin \\cite{ei97}). In M31, extended {\\sc Hi} and optical emission was detected up to more than $0.5\\degr$ from the center along the minor axis (Emerson \\cite{eme74}; Innanen et al. \\cite{innet82}). However, the radio emission seems not that extended and the existence of a radio halo in M31 is still uncertain (eg. Wielebinski \\cite{wie76};  Volodin \\& Dagkesamanskii \\cite{vd78}; Gr\\\"ave et al. \\cite{graet81}; Berkhuijsen et al. \\cite{beret91}). Our Galaxy has a vertically extended thick radio disk (Beuermann et al. \\cite{beuet85}) and rotation measures of extragalactic radio sources have revealed the existence of an extended magneto-ionic disk (Clegg et al. \\cite{cleet92}; Han \\& Qiao \\cite{hq94}).  In this paper, we will investigate the extended magneto-ionic disk in M31.  The detection of the regular magnetic field outside the ``ring'', either interior to the ``ring'' or in the outer spiral arms or halo, will constrain the theoretical models for the type and origin of the field (eg. Poezd et al. \\cite{poeet93}; Howard \\& Kulsrud \\cite{hk97}). A field residing just in the ``ring'' is very difficult to understand in the frame-work of a primordial field origin. More seriously, a dynamo cannot generate a field that is limited to a given small range of radius. The field should be much more extended (Moss et al. \\cite{moset98}), and should have another weaker ring exterior (Ruzmaikin et al. \\cite{ruzet88}) or interior to the observed ``ring'' (Moss et al. \\cite{moset98}). The emission possibly detected interior to the ``ring'' from the inner arms (eg. Beck \\cite{beck82}; Berkhuijsen et al. \\cite{beret91}) is too weak for a measurement of the regular field.  M31 is a nearby spiral galaxy which optically extends more than $4\\degr$ along the major axis on the sky. There are a number of polarized background radio sources in the field of M31, which can be used as probes of the magneto-ionic medium in M31. If the magnetic structure in the halo or extended disk is somewhat ordered, the rotation measures (RMs) of these sources should have systematic deviations from the average. We observed 21 polarized radio sources with the VLA in 6 fields in the direction of M31 at two frequencies, and determined their RMs from the observed position angles (PAs). We present the observations and data reduction in Sect.2, the results in Sect.3, and discuss them in Sect.4. \\begin{table}\t\t% \\caption{Positions of the observed fields} \\begin{tabular}{cccl} \\hline \\hline Field & RA(1950) & Dec(1950) & \\multicolumn{1}{c}{Polarized Sources}\\\\ No.  & h!m~!s~ & ~!$\\degr$!~$'$~!$''$ & \\multicolumn{1}{c}{(37W-No.)}\\\\ \\hline B1. & 00~38~00.0 & +41~45~00 & 45,50,57,91\\\\ B2. & 00~39~30.0 & +41~10~00 & 89,94,115,144\\\\ B3. & 00~40~30.0 & +41~40~00 & 131,152,175\\\\ B4. & 00~40~40.0 & +40~40~00 & 168,172\\\\ B5. & 00~42~00.0 & +41~23~00 & 188\\\\ B6. & 00~42~30.0 & +41~00~00 & 205,207B,211,219\\\\ \\hline \\hline \\end{tabular} \\end{table}  \\begin{table*}\t\t% \\begin{minipage}{180mm} \\caption{The polarized discrete sources in the field of M31} \\begin{small} \\begin{tabular}{lccrccccrrr} \\hline \\hline Object & RA(1950) & Dec(1950) & \\multicolumn{1}{c}{I$_{\\rm 1652}$} & \\multicolumn{1}{c}{PI$_{\\rm 1652}$} &  p & \\multicolumn{1}{c}{ PA$_{\\rm 1652}$} & \\multicolumn{1}{c}{PA$_{\\rm 1365}$} &\\multicolumn{1}{c}{RM} & \\multicolumn{1}{c}{PA$_{\\rm intri}$}& Notes % \\\\ 37W-- & h!m!~s! & $\\degr$!~~$'$~~~$''$ & \\multicolumn{1}{c}{(mJy)}&\\multicolumn{1}{c}{(mJy)}&(\\%)& \\multicolumn{1}{c}{($\\degr$)} & \\multicolumn{1}{c}{($\\degr$)} & \\multicolumn{1}{c}{rad$\\cdot$m$^{-2}$} & \\multicolumn{1}{c}{($\\degr$)}\\\\ \\hline !45a& 00~37~14.0&40~55~18.0& 15.12$\\pm$0.07 & 0.55$\\pm$0.06 & !4& 91$\\pm$5! & 168$\\pm$5!! &$-$113$\\pm$6!& 134$\\pm$10 & 1,2,3 \\\\ !50a& 00~37~30.2&40~52~09.1& 16.00$\\pm$0.34 & 0.71$\\pm$0.06 & !4& 16$\\pm$5! & 133$\\pm$4!! & $-$76$\\pm$5!& 163$\\pm$11 & 1,2,3 \\\\ !50b& 00~37~29.7&40~52~35.0&  7.60$\\pm$0.16 & 1.70$\\pm$0.07 & 22& 25$\\pm$2!& 127$\\pm$1!! & $-$90$\\pm$3!&  15$\\pm$7! & 1,2,3 \\\\ !57 & 00~37~40.7&40~50~45.6& 22.06$\\pm$0.20 & 0.56$\\pm$0.06 & !3& 118$\\pm$5!!&  42$\\pm$3! & $-$86$\\pm$6!& 101$\\pm$16 & 1,3   \\\\ !74A/B& 00~38~25.0&41~08~31.6&              &               &      & &                &$-$105$\\pm$5!& 140$\\pm$12 & 1,2,4 \\\\[1mm] !89a& 00~38~55.2&41~14~04.8& 13.99$\\pm$0.17 & 0.52$\\pm$0.06 & !4& 34$\\pm$5! &139$\\pm$5!! & $-$86$\\pm$9!&  18$\\pm$20 & 1,2,3 \\\\ !89b& 00~38~55.3&41~14~12.7& 13.45$\\pm$0.17 & 1.08$\\pm$0.07 & !8& 23$\\pm$3! &123$\\pm$3!! & $-$91$\\pm$5!&  15$\\pm$11 & 1,2,3 \\\\ !91 & 00~38~57.2&40~47~08.0& 41.54$\\pm$0.77 & 1.35$\\pm$0.07 & !3& 137$\\pm$2!! & 89$\\pm$1! & $-$55$\\pm$2!& 66$\\pm$5! & 1,2,3 \\\\ !94a& 00~39~04.1&41~02~20.9& 18.62$\\pm$0.17 & 0.54$\\pm$0.06 & !3& 143$\\pm$5!! & 36$\\pm$13 &$-$122$\\pm$9!& 14$\\pm$32 & 1,5   \\\\ 115 & 00~39~34.5&41~13~01.0&319.99$\\pm$1.10 & 8.58$\\pm$0.06 & !3& 99$\\pm$1! & 19$\\pm$1! & $-$92$\\pm$1!& 94$\\pm$2! & 1,2,3 \\\\[1mm] 131 & 00~39~51.2&41~41~20.6& 65.05$\\pm$0.56 & 1.88$\\pm$0.06 & !3& 70$\\pm$2! & 168$\\pm$1!! & $-$93$\\pm$2!& 68$\\pm$5! & 3,5   \\\\ 144 & 00~40~07.3&41~10~09.3& 21.26$\\pm$0.15 & 0.39$\\pm$0.05 & !2& 50$\\pm$7! & 18$\\pm$13 & $-$37$\\pm$17&120$\\pm$35 & 3,5   \\\\ 152 & 00~40~23.5&41~38~43.4&  1.86$\\pm$0.16 & 0.17$\\pm$0.07 & !9& 80$\\pm$13 & 159$\\pm$7!! &$-$115$\\pm$17&118$\\pm$45 & 3,5   \\\\ 168 & 00~40~56.8&40~38~04.8& 52.59$\\pm$4.53 & 2.90$\\pm$0.05 & !5& 177$\\pm$1!! & 118$\\pm$1!! & $-$67$\\pm$1!&124$\\pm$1! & 3,5   \\\\ 172 & 00~41~10.0&40~30~10.2&143.89$\\pm$2.26 & 2.70$\\pm$0.14 & !2& 81$\\pm$2! &  21$\\pm$2! & $-$68$\\pm$3!& 31$\\pm$6! & 3,5   \\\\[1mm] 175a& 00~41~15.2&41~40~53.0& 39.69$\\pm$0.35 & 1.18$\\pm$0.06 & !3& 43$\\pm$4! & 123$\\pm$2!! &$-$111$\\pm$5!& 74$\\pm$12 & 3,5   \\\\ 175b& 00~41~13.0&41~40~52.7& 45.82$\\pm$0.60 & 2.09$\\pm$0.06 & !5& 12$\\pm$2! &  80$\\pm$2! &$-$127$\\pm$3!& 73$\\pm$6! & 3,5   \\\\ 188 & 00~41~39.6&41~14~17.7&  5.80$\\pm$0.28 & 0.29$\\pm$0.05 & !5& 178$\\pm$4!! &  44$\\pm$3! &$-$152$\\pm$5!&106$\\pm$13 & 3,5   \\\\ 205 & 00~42~16.9&41~08~30.1& 31.33$\\pm$1.99 & 0.81$\\pm$0.06 & !3& 90$\\pm$1! &  11$\\pm$13 & $-$91$\\pm$9!& 83$\\pm$18 & 3,5   \\\\ 207B& 00~42~21.0&41~06~17.0&  3.72$\\pm$0.33 & 0.55$\\pm$0.10 & 15& 42$\\pm$3! & 110$\\pm$5!! &$-$129$\\pm$7!&108$\\pm$15 & 3,5   \\\\[1mm] 211 & 00~42~27.0&40~55~09.0& 20.89$\\pm$0.34 & 4.20$\\pm$0.06 & 20& 38$\\pm$1! & 143$\\pm$1!! & $-$86$\\pm$1!& 21$\\pm$2! & 3,5   \\\\ 219 & 00~42~54.6&40~56~03.4& 11.73$\\pm$0.20 & 0.50$\\pm$0.05 & !4& 42$\\pm$3! & 148$\\pm$7!! & $-$83$\\pm$9!& 20$\\pm$18 & 5     \\\\ \\hline \\hline \\end{tabular}\\\\ Notes: 1. PA value at 21 cm (1490 MHz) from Beck et al. (\\cite{becket89}) is also considered; 2. As note 1 but at 6.3 cm (4760 MHz); 3. PA value at 1400 MHz from NVSS catalog (Condon et al. \\cite{conet98}) is considered; 4. No flux desities are given because of no detection in our high-resolution observations. The RM is taken from Beck et al. (\\cite{becket89}); 5. PA value at 1465 MHz from Beck et al. (\\cite{becket98}) is considered. \\end{small} \\end{minipage} \\end{table*} ",
        "conclusions": "\\begin{figure*}\t% \\centering \\begin{tabular}{ccc} \\mbox{\\psfig{file=0867.f3a,height=45mm,width=82mm}} & & \\mbox{\\psfig{file=0867.f3b,height=45mm,width=82mm}} \\\\[5mm] \\mbox{\\psfig{file=0867.f3c,rotate=90,height=45mm,width=70mm}} & & \\mbox{\\psfig{file=0867.f3d,rotate=90,height=45mm,width=70mm}} \\end{tabular}\\\\[-1mm] \\caption{The field configurations of the odd (A0) and the even (S0) dynamo modes in a galaxy, with thick lines indicating the toroidal field and thin lines the poloidal field, are illustrated in the two plots above; the RM of polarized background sources from the poloidal field is shown in the two plots below (with arbitrary scale for RM). The background radio sources are supposed to be located in the plane of the azimuthal angles of $90\\degr$ and $270\\degr$ (the minor axis). The RM of the extended polarized emission from a galaxy, eg. from the ``ring'' of M31, near lines of sight No.1 and No.2 have smaller amplitudes than those shown in the lower plots. } \\label{fig3} \\end{figure*}  If the magnetic field in a galaxy is generated or maintained by a dynamo, theoretical simulations show that the lowest mode dynamo is excited most easily (eg. Beck et al. \\cite{becket96}). The two lowest modes, the odd mode A0 and the even mode S0, could exist in galaxies. Observational proofs of these modes are obviously important, not only to further theoretical studies but also to discrimination between the dynamo theory and its alternatives, eg. primordial origin (eg. Kulsrud et al. \\cite{kulet97}), local origin (eg. Daly \\& Loeb \\cite{dl90}), or MHD waves (eg. Fan \\& Lou \\cite{fl96}).  As illustrated by Fig.19 in Wielebinski \\& Krause (1993), the toroidal field in the A0 mode is antisymmetric with respect to the galactic plane, and the poloidal field has the structure of a dipole (see our Fig.~\\ref{fig3}). The S0 mode field, however, has the same toroidal field structure above and below the plane, but an antisymmetric poloidal field, like two opposite dipoles.  Our Galaxy is an edge-on galaxy. Han et al. (\\cite{hanet97}) analysed the RM distribution of several hundred background sources, and showed that our Galaxy possibly has an A0 odd mode field, eg. an antisymmetric toroidal field with respect to the Galactic plane. For M31, the well-known ``ring'' is thought to be an indication of a toroidal field, but it is not clear yet whether it is of odd or even mode.  Polarization observations of M31's extended emission show that there is a systematic RM variation around the observed ``ring'' located at a galacto-centric radius of about 10 kpc, as shown in Fig.~\\ref{fig2}, which is caused by the regular magnetic field inside the ``ring'' (Beck \\cite{beck82}; Berkhuijsen et al. \\cite{beret87}; Beck et al. \\cite{becket89}; Berkhuijsen et al. in preparation). The gas layer causing the observed RM variations is primarily the near half, in contrast to the full thickness of the disk where the emission comes from. The RM variation with azimuthal angle of such a continuum ``ring'' should have the same shape for magnetic fields of the A0 odd mode and of the S0 even mode, but they differ in amplitude by a factor of 2 according to Beck et al. (\\cite{becket96}). This difference cannot be used as a criterion to distinguish A0 or S0 mode, because it is not known to which mode the observed amplitude of the RM curve refers.  The RMs of the background sources reveal the average magnetic field across the full thickness of the disk. If the toroidal magnetic field of an S0 even dynamo mode is uniform in the disk of a galaxy, the RMs of background sources should be twice that of the extended emission from M31. On the other hand, for an A0 odd dynamo field the RM contribution from the toroidal field in the near half and the far half of the disk of M31 will cancel each other and therefore there will be no net RM contribution when observing background sources.  Observations of the RMs near the minor axis, either of background sources or of the extended emission from the ``ring'', could provide evidence for a symmetric poloidal field of an even dynamo mode or an antisymmetric poloidal field of an odd dynamo mode. The toroidal field near these azimuthal regions is almost perpendicular to the line of sight and therefore contributes a negligible RM. Because of the poloidal field,  the RMs of the sources along the lines of sight No.1 and No.5 (or No.2 and No.4) should similarly deviate from the average foreground RM in the odd mode, while opposite deviations should be found in the even mode (Fig.~\\ref{fig3}).  In the following we will consider that the deviating RMs of the observed sources are caused by the magneto-ionic medium in either the disk or the halo of M31, or both. \\begin{table}\t% \\caption{The geometric parameters of polarized sources in order of $\\theta_{\\rm sky}$} \\begin{tabular}{lccccc} \\hline \\hline Object   & RM &  $\\theta_{\\rm sky}$ $^1$ & $\\theta_{\\rm M31}$ $^{1,2}$  & $R_{\\rm M31}$ $^2$\\\\ 37W--   & (rad m$^{-2}$) & ($\\degr$) & ($\\degr$)  & (kpc) \\\\ \\hline 188 & $-152\\pm5!$ & !15 & !52 & !7.3  \\\\ 205 & $!-91\\pm9!$ & !34 & !73 & 15.1  \\\\ 207B& $-129\\pm7!$ & !39 & !76 & 16.8  \\\\ 219 & $!-83\\pm9!$ & !59 & !83 & 27.1  \\\\ 211 & $!-86\\pm1!$ & !62 & !84 & 23.7  \\\\ 168 & $!-67\\pm1!$ & 116 & !96 & 20.6  \\\\ 172 & $!-68\\pm3!$ & 118 & !96 & 26.0  \\\\ !91 & $!-55\\pm2!$ & 188 & 213 & !4.9  \\\\ !57 & $!-86\\pm6!$ & 215 & 253 & 15.8  \\\\ !50 & $!-76\\pm5!$ & 219 & 255 & 18.3  \\\\ & $!-90\\pm3!$ &     &     &       \\\\ !45 & $-113\\pm6!$ & 225 & 258 & 22.3  \\\\ !94 & $-122\\pm9!$ & 246 & 265 & 10.0  \\\\ !74 & $-105\\pm5!$ & 259 & 268 & 18.7  \\\\ !89 & $!-86\\pm9!$ & 282 & 273 & 17.6  \\\\ & $!-91\\pm5!$ &     &     &       \\\\ 115 & $!-92\\pm1!$ & 302 & 278 & 11.4  \\\\ 131 & $!-93\\pm2!$ & 321 & 284 & 26.4  \\\\ 144 & $!-37\\pm17$ & 329 & 289 & !5.3  \\\\ 152 & $-115\\pm17$ & 329 & 289 & 20.3  \\\\ 175 & $-127\\pm3!$ & 342 & 302 & 15.4  \\\\ & $-111\\pm5!$ &     &     &       \\\\ \\hline \\hline \\end{tabular}\\\\ Notes: (1) Azimuthal angles $\\theta = 0\\degr$, $90\\degr$, $180\\degr$ and $270\\degr$ are shown in Fig.1. Just for these values, they are the same in the plane of sky and in the plane of M31. (2) We assumed an inclination of $78\\degr$ and a distance to M31 of 690 kpc. \\end{table}  \\subsection{On the toroidal field}  The toroidal magnetic field in M31 results in a systematic RM variation around the observed ``ring'' (Fig.2). The slight phase shift of the RM curve (RM$\\sim 0$ at $\\theta_{\\rm M31} \\simeq  80\\degr$ and $250\\degr$) is due to the pitch angle of the field in the ``ring''. A detailed discussion of this RM variation, including the big jumps (Beck 1982) near $\\theta_{\\rm M31} = 150\\degr$ and $0\\degr$/$360\\degr$, will be given by Berkhuijsen et al. (in preparation).  The field in the ``ring''  contributes to RMs of the discrete sources if they are located behind the ``ring''. However, if a source has a large angular separation from the ``ring'', then the field in the ``ring'' will not affect its RM unless the field extends to that large radius. In Table 3, we give the apparent radial distances $R_{\\rm M31}$ of the sources from the M31 center and their azimuthal angle $\\theta_{\\rm M31}$, both in the galactic plane of M31. These two parameters indicate which part of M31 affects the observed RM of the background sources. With an inclination angle of $78\\degr$ (eg. Braun \\cite{bra91}; Ma et al. \\cite{maet97}) and a distance\\footnote{ Recently, Feast \\& Catchpole (\\cite{fc97}) published the M31 modulus $24.77\\pm0.11$, implying a larger distance to M31 of $900\\pm45$ kpc. For comparison with earlier work, we use 690 kpc.} to M31 of 690 kpc (Baade \\& Swope \\cite{bs63}), $1'$ corresponds to 200 pc in the plane of M31 along the major axis. The azimuthal angle increases counter-clockwise from $0\\degr$ on the northern major axis to east (see Fig.~\\ref{fig1}). We plotted the observed RMs of discrete sources onto Fig.~\\ref{fig2} with different symbols indicating the range of apparent distances to the center of M31.  Most of the observed sources are near the azimuthal angles $\\theta_{\\rm M31} = 90\\degr$ and $270\\degr$ (Fig.~\\ref{fig2}), and therefore have a RM not deviating much from the average RM or from the RM variation of the ``ring'', except for a few.  Three sources interior to the ``ring'' (see Fig.~\\ref{fig1}) with quite different azimuthal angles $\\theta_{\\rm M31}$ have apparent distances less than 8.0 kpc (Tab.2) and are located far from the minor axis of M31. Note that two of them, 37W188 (RM =$-152\\pm5$) and 37W91 (RM =$-55\\pm2$), have RMs quite consistent with those of the continuum ``ring''. First of all, this indicates that the field with the same structure as in the ``ring'' also extends to the inner part of M31. Second, the field very probably has an even mode, eg. no field reversal below and above the galactic plane of M31. Otherwise there should be no net RM from M31. Third, this means that the strength of the regular magnetic field has about one half of the strength compared to that in the ``ring'', if the electron density were constant. The third source, 37W144 (RM =$-37\\pm17$), has a large positive RM after the foreground RM is subtracted, opposite to that of the continuum ``ring''. This source is exactly located behind a dust lane (Walterbos \\& Kennicutt \\cite{wk88}) and an {\\sc Hi} spiral arm (Brinks \\& Shane \\cite{bs84}). Comparison with the spiral structure of M31 (Braun \\cite{bra91}) suggests some strong local perturbation of the magnetic field from the {\\sc Hii} regions in this direction, or possibly a field reversal in the dust arm.  Two sources appear on the ring, 37W94 (RM =$-122\\pm9$) and 37W115 (RM =$-92\\pm1$). Both of them probably are background sources, as indicated by the {\\sc Hi} absorption lines (Braun \\& Walterbos \\cite{bw92}). The former one has an RM deviating from the average twice as much as the continuum emission, though with large error bars, indicating that the field there might be of an even mode as discussed above. 37W115 is the strongest source detected. This source has soft X-ray emission (Supper et al. \\cite{supet97}), and its X-ray hardness (Supper et al. \\cite{supet97}) suggests that it is a background source. However, there seems to be no net RM contribution from M31 so that the RM of this source is equal to the average (eg. the foreground RM). This could happen if the 3 spiral arms (Braun \\cite{bra91}), through which the line of sight to 37W115 passes, have reversed field directions. The RM of the extended polarized emission from the ``ring'' is not cancelled because it mainly emerges from one arm (Berkhuijsen et al. \\cite{beret93}). Another sources, 37W89 (RM =$-86\\pm9$ and $-91\\pm5$), not on the ring, is located at a similar $\\theta_{\\rm M31}$, but this source may be a supernova remnant within M31 (Dickey \\& Brinks \\cite{db88}), which explains the small RM contribution from M31.  Sources exterior to the ``ring''  are interesting as well. Compared with the spiral structure given by Braun (\\cite{bra91}), we found that only 37W131 (RM =$-93\\pm2$) is far away from the spiral arms, and has an RM equal to the average, which means that the magneto-ionic medium does not extend to $R_{\\rm M31} = 26.4$ kpc in this part of M31. However, in the lower part of Fig.1, most sources are in the spiral region outlined by Braun (\\cite{bra91}) and their RMs are consistent with the RM variation of the ``ring''. This suggests that the magnetic field in the disk extends as far as $R_{\\rm M31} = 26$kpc in this region, as indicated by 37W168 (RM =$-67\\pm1$) and 37W172 (RM =$-68\\pm3$). Furthermore, the toroidal field at this radius has the same direction as that in the ``ring'', and a significant strength.  The RMs of 37W175 (=$-127\\pm3$ and $-111\\pm5$) suggest that the field with the same structure as in the ``ring'' extends to $R_{\\rm M31} = 15$ kpc but with about one half or one third of the field strength in the ``ring'', assuming constant electron density. Three sources at the upper right corner of Fig.1 have about average RMs, similar to the RM of continuum ``ring'' at these azimuthal angles. Only the farthest one, 37W45 (RM =$-113\\pm6$), has a somewhat negative deviation, which suggests that a weak field extends even to $R_{\\rm M31} = 22$ kpc. This also holds for 37W74 (RM =$-105\\pm5$).  In summary, our results suggest that (1) magnetic field exists exterior  as well as interior to the ``ring''; (2) the magneto-ionic disk of M31 is extended, probably out to 25 kpc in the plane of the disk; and (3) the field in the extended disk generally has a  well-ordered structure similar to that in the ``ring'', but the field strength is smaller than in the ``ring'', in agreement with the very weak synchrotron emission observed from the extended disk (Berkhuijsen \\& Wielebinski \\cite{bw74}).  \\subsection{On the poloidal field}  The RM data of the continuum ``ring'' near the azimuthal angles of $90\\degr$ and $270\\degr$ seem to give some indication for the poloidal magnetic field (in the halo and disk). As is shown in Fig.3, the RMs along lines of sight No.2 and No.4 are expected to have the largest opposite deviations from the mean curve caused by RMs originating in the poloidal field which are added to those of the toroidal field. Indeed, two opposite RM bumps, though with  large error bars, appear in Fig.2 on the RM variation of the ``ring'' near $\\theta_{\\rm M31} = 100\\degr$ and $280\\degr$, which could be  caused by the poloidal field (Fig.3). Since the part of M31 at $\\theta =270\\degr$ is the near one, we see a smaller bump. This indicates that the poloidal field of M31 is of even mode as is shown for the S0 dynamo field in Fig.3. The field is directed inwards near the galactic plane, but is oppositely going outwards when it is far from plane. If this opposite deviation could be confirmed by more RM data of background sources, then the magnetic field in the halo of M31 is most probably generated by an even mode dynamo,  consistent with the ring field in the disk. The toroidal field in the M31 disk should then be mirror-symmetric with respect to its galactic plane without a vertical field reversal.  No matter what kind of dynamo operates in the halo of M31, if we attribute the deviating RMs of the bumps, $\\Delta RM \\sim 15$ rad~m$^{-2}$, to the poloidal field, we can estimate an upper limit of the field strength. Assuming a thermal electron density of  0.003 cm$^{-3}$ and a line of sight of  20 kpc through the thick disk or halo, then this upper limit is about 0.3$\\mu$G, almost the same as the halo field of our Galaxy near the sun (Han \\& Qiao 1994; Han et al. \\cite{hanet98}).   \\subsection{Concluding remarks}  Our RM observations of background radio sources suggest that the regular magnetic field in the bright ``ring'' of M31 extends from a radius of about 5 kpc interior to the ``ring'' to as far as 25 kpc from the center, probably with  similar structure. We presented evidence that the magnetic field of M31 has the structure of an even mode dynamo field (S0). The poloidal field of even symmetry has a strength $\\le0.3\\mu$G."
    },
    "9805/astro-ph9805336_arXiv.txt": {
        "abstract": "We have used innovative features of the Taurus Tunable Filter instrument on the 3.9-m Anglo-Australian Telescope to obtain nearly-continuous, high-throughput, linear photometry of V2116~Oph in a 7 \\AA\\ bandpass at the center of the O\\,I $\\lambda$8446 emission line. This instrumental technique shows promise for applications requiring precise, rapid, narrowband photometry of faint objects.  The spectrum of V2116~Oph, the counterpart of GX\\,1+4 (=X1728--247), is exotic, even among the unusual spectra of other optical counterparts of compact Galactic X-ray sources.  The second strongest emission line is an unusual one, namely extremely prominent O\\,I $\\lambda$8446, which is likely to result from pumping by an intense Ly$\\beta$ radiation field. As the X-radiation from GX\\,1+4 is steadily pulsed, with typical pulsed fractions of 0.4, the O\\,I~$\\lambda$8446 emission in V2116~Oph may also be strongly modulated with the current 127~s period of the X-ray source.  If so, this may well allow us to obtain high signal-to-noise radial velocity measurements and thus to determine the system parameters.  However, no such pulsations are detected, and we set an upper limit of $\\sim1$\\% (full-amplitude) on periodic $\\lambda$8446 oscillations at the X-ray frequency.  This value is comparable to the amplitude of {\\it continuum} oscillations observed on some nights by other workers.  Thus we rule out an enhancement of the pulsation amplitude in O\\,I emission, at least at the time of our observations. ",
        "introduction": "GX\\,1+4 (=X1728--247), a classical, luminous X-ray binary observed for 25 years, is projected very close to the galactic center and probably is at a distance of $\\sim8$~kpc, as the inferred X-ray luminosity is then $6\\times10^{37}$ erg~s$^{-1}$, near the Eddington limit.  The spectrum of the $V\\sim19$ optical counterpart, first identified by Glass \\& Feast (1973) and now known as V2116~Oph, has at times shown higher excitation emission lines than any other known X-ray star; for example, [Fe\\,X] (I.~P. = 235 eV) and [Ar\\,XI] appeared (Davidsen et al. 1977).  The spectrum appears to be markedly time variable on scales from minutes to years; in recent epochs, the highest excitation lines have disappeared (Chakrabarty \\& Roche 1997), but enormously strong H$\\alpha$ emission remains.  The symbiotic-like optical spectrum of V2116~Oph directly shows the presence of a red giant, of type near M5\\,III (Chakrabarty \\& Roche 1997), and suggests that we may be viewing the system at a very special, short-lived stage, when the normal primary is passing through a quite brief phase of its evolution.  This point is made even more vivid by the X-ray behavior.  GX\\,1+4 is an X-ray pulsar with a coherent X-ray period of about 130~s, a pulse amplitude \\squig 0.4, and an enormous X-ray period {\\it derivative} of $\\dot P = dP/dt\\sim-3$~s~yr$^{-1}$ (Laurent et al. 1993 and references therein). While the period is slow, although not inordinately slow for X-ray binaries, the spin-up rate is the fastest for any known X-ray pulsar. The characteristic age, $P/\\dot P \\sim40$~yr, confirms that this is an amazingly rapidly evolving object.  More recent X-ray observations show that the X-ray $\\dot P$ has reversed sign, although the modulus remains very large (Laurent et al. 1993, Chakrabarty et al. 1997).  It seems clear that in GX\\,1+4 we have the chance to observe an X-ray pulsar undergoing rapid evolution.  Yet even the orbital parameters of this system remain unknown, and the detection of an optical analogue to the X-ray pulses could yield an elegant and accurate radial velocity solution (\\S 3).  Recently broadband optical flickering and pulsations at the X-ray period have indeed been reported from V2116~Oph, at an amplitude of a few percent (Jablonski et al. 1997, Jablonski \\& Pereira 1997).  These data seem to indicate a complex and/or erratic dependence of pulse amplitude on wavelength, optical brightness of the system, etc., and further observations will clearly be needed to sort out the situation.  The second strongest emission line after H$\\alpha$ in the optical spectrum of V2116~Oph is a remarkable one, namely extremely prominent O\\,I $\\lambda$8446, which can be readily seen in our low-resolution spectrum in Fig. 1.  This line has been reported in a small number of interesting objects ranging from Seyfert galaxies to occasional odd stars.  In some objects, the great observed strength relative to other common species must be explained by some type of preferential emission mechanism.  Grandi (1975, 1976) suggested that pumping by an intense Ly$\\beta$ radiation field, in a wavelength coincidence (Ly$\\beta\\ \\lambda1025.72\\approx {\\rm O\\,I} \\ \\lambda1025.76$) not dissimilar to the famous Bowen mechanism in which He II Ly $\\alpha$ pumps O\\,III and N\\,III in nebulae and mass-exchange binaries, may be responsible.  O\\,I $\\lambda$8446 has been reported at great strength in the symbiotic star V1016~Cyg (Rudy et al. 1990), and the presence there of the OI $\\lambda11287$ line at the expected strength confirms that Ly$\\beta$ pumping is the mechanism.  As V2116 Oph is exposed to $\\sim10^{38}$ erg s$^{-1}$ of pulsed ionizing radiation, it seems possible that pulsed Lyman photons may trigger coherent 127~s pulsations in O\\,I $\\lambda$8446, possibly at quite large amplitudes. Here we report on a search for these pulsations using an innovative technique with applicability to a variety of other problems. ",
        "conclusions": "With a giant primary and any reasonable assumption for the mass of the secondary (presumably a neutron star), the system orbital period must be of order one year.  Although we would very much like to understand even the basic parameters of this unusual system, simply extracting even the period will not be easy.  Searches for periodic variations in the X-ray pulse timing residuals are foiled by large, irregular torque derivatives (Chakrabarty et al. 1997, Bildsten et al. 1997).  Normal radial velocity spectroscopy of the optical lines will be difficult; the expected variations are very small compared at least to the very broad H$\\alpha$ emission.  Searches for periodic variations in the H$\\alpha$ intensity (Greenhill et al. 1995) and H$\\alpha$ profile (e.g., Sood et al. 1995) have not revealed a significant orbital period component in the large variations, presumably due to accretion rate fluctuations.  In principle the detection of an optical analogue to the X-ray pulses could yield an accurate radial velocity solution.  If the ionizing radiation is reprocessed to visible light at a location in the system fixed with respect to the barycenter, and the recombination times  are short compared to the pulse period, the resulting optical pulses can be searched for periodic modulation due to orbital motion. If the reprocessing surface is physically larger than the light travel distance in one pulse cycle, phase mixing may dilute the pulse amplitude greatly, but working in the Fourier domain provides far more potential sensitivity than standard radial velocity spectroscopy.  This technique has of course been applied decades ago, with great success, to the X-ray pulsar HZ~Her/Her X-1 (Middleditch \\& Nelson 1976).  We have conducted a first search for pulsations of O\\,I $\\lambda$8446 in GX\\,1+4, in the hope of establishing if measurement of the periodic modulation of this strong emission line may be used to obtain high signal-to-noise radial velocity measurements, and thus to determine the system parameters.  Despite less than optimal observing conditions, we are able to rule out an enhancement of the pulsation amplitude in O\\,I at the time of our observations; our combined data do not reveal any pulsation at the X-ray period, with a $3\\sigma$ upper limit of 1\\% (full-amplitude).  Krzeminski \\& Priedhorsky (1978) report similar limits to H$\\alpha$ periodic variations.  Chakrabarty et al. (1998) report $2\\sigma$ upper limits of 5\\% and 9\\% for pulsation in infrared He I and Pa\\,$\\beta$, respectively, at the time of their observations. Broadband optical pulsations up to $\\sim5$\\% have been observed, but only on certain occasions, and broadband pulsation upper limits of 0.1\\% have also been documented (Jablonski et al. 1997).  Those authors have reported a correlation where these pulsations are stronger during brighter optical states.  However, our data do not allow us to determine the absolute brightness of V2116~Oph during the observations.  The lack of pulsation enhancement in O\\,I $\\lambda$8446 allows several possible interpretations.  If the Ly~$\\beta$ photons responsible for pumping the O\\,I are indeed created by reprocessed X-rays, the reprocessing region must be sufficiently large to completely dilute the pulse amplitude.  A circumstellar nebula such as discussed by Chakrabarty \\& Roche (1997) is one such possibility.  A second possible interpretation follows a suggestion of Chakrabarty \\& Roche (1997) based on other optical emission lines: the Ly~$\\beta$ photons are excited by the thermal emission of the accretion disk instead of reprocessed X rays.  Substantial further observations of this complex system are needed to unravel its nature.  Optical pulsations in HZ Her are well-known to be detected only during a highly restricted subset of orbital and precessional phases in the system.  Nonetheless, the unique utility of TTF in charge shuffle mode for narrow band time-series photometry of faint objects is clear."
    },
    "9805/astro-ph9805100_arXiv.txt": {
        "abstract": " ",
        "introduction": "Dwarf galaxies have, on the whole, been overlooked in the grand scheme of things, often considered either insignificant or irrelevant for cosmological purposes. Recently though problems in the interpretation of ultra-deep images have brought dwarf galaxies to the fore. This conundrum, known as the faint blue galaxy problem\\cite{kk}\\cite{ellis}, could be readily explained if the local space density of dwarf galaxies was higher than previously supposed\\cite{dpdmd} and/or dwarf galaxies underwent dramatic recent evolution\\cite{pd}. Certainly the constraints on the space density of dwarf galaxies are weak. This stems from the fact that bright magnitude limited redshift surveys fail to probe representative volumes for dwarf systems\\cite{dp} and that faint redshift surveys are too small, too incomplete, model dependent and still not faint enough\\cite{dc}. An additional and probably more fundamental constraint is that dwarf galaxies are typically of low surface brightness, making spectroscopic redshift determination itself problematical. If their space density is high could they, (a) constitute a significant number of baryons (and/or cold dark matter), and (b) present a foreground screen of objects contaminating our window into the distant Universe ? ",
        "conclusions": ""
    },
    "9805/astro-ph9805270_arXiv.txt": {
        "abstract": "In this paper we revise and complete  the photometric survey of the instability strip of the southern open cluster NGC 2516 published by Antonello and Mantegazza (1986). No variable stars with amplitudes larger than $0^m.02$ were found. However by means of an accurate analysis based on a new statistical method  two groups of small amplitude variables have been disentangled: one with periods $< 0^d.25$ (probably $\\delta$ Scuti stars) and one with periods $>0^d.025$. The position in the HR diagram and the apparent time-scale may suggest that the stars of the second group  belong to a recently discovered new class of variables, named $\\gamma$ Dor variables. They certainly deserve further study. We also present a comparison between the results of the photometric survey and the available pointed ROSAT observations of this cluster. ",
        "introduction": "The observation of pulsating variables in homogeneous samples such as stellar clusters is a good way to collect information on the effect of age, chemical composition and rotation on pulsation.  The studies of open clusters in the northern sky made by several authors (see Slovak, 1978) have shown that cluster variable stars of $\\delta$ Sct type, i.e. the typical low-instability strip pulsators, have an incidence of around 30\\% apparently independent of age.  The southern open cluster NGC 2516, located at right ascension $\\alpha(2000)=7^h\\: 58^m.3$ and declination $\\delta(2000)=-60^o\\: 52'$ is a young cluster believed to have a common origin with $\\alpha$ Per, the Pleiades and IC 2606 (Eggen 1983), but unlike these clusters it has an unusually high abundance of Ap stars. An extensive study by Mermillod (1981) on young open clusters identifies it as the prototype of a group, named the NGC 2516 cluster group, that also includes NGC 2168, NGC 2301, NGC 3114, NGC 5460, NGC 6025, NGC 7243. They have almost the same age (log age =8.3) and are all characterized by a peculiar gap near the turn  off point and by the presence of numerous Ap stars in the extreme blue region of the main sequence.  A first survey for the detection of variable stars in the lower part of the instability strip was carried out by Antonello and Mantegazza (1986, hereinafter paper I) and provided evidence, besides a normal incidence of shorter period variables ($P\\leq .25$ d), of a number of longer period variables ($P\\geq .25$ d), lying on the cool border of the instability strip.  In the last years the presence of variable stars showing small amplitudes with longer time scales than the typical $\\delta$ Sct periodicities has been often reported. Mantegazza et al (1993) first gathered a sample of 8 field stars and 5 cluster stars with very similar location in the HR diagram, on or just beyond the cool border of the instability strip, apparently showing a common behavior. They have been named $\\gamma$ Dor stars and an updated list has been recently published by Krisciunas and Handler (1995).  In the present paper we present previously unpublished observations of NGC 2516 that complete the survey for variability inside the instability strip of the cluster. The survey is complete in the sense that all the stars in the instability strip have been searched for variability.  The absence of a physical lower limit to the possible variability of stars implies the need of a statistical method of data analysis in order to identify the variable stars according to a probability criterion. A method was developed and described in paper I.  In this paper we present and use a different statistical approach with respect to paper I, less sensitive to the photometric quality of the night. A comparison between the two approaches is also reported. ",
        "conclusions": "In this paper we have reported on a survey for variability of a selected sample of A- and early F-type stars in the southern open cluster NGC 2516. The sample is complete in the sense that all dwarfs and subgiants with $0.00<(b-y)_o<0.25$ belonging to the cluster have been observed.  The new results confirm the normal incidence of shorter period variables (likely $\\delta$ Scuti) and the anomalous incidence of longer period variables previously found. The results have been also compared to the survey in the X band obtained by ROSAT but no relation between variables and X-ray sources could be found.  The statistic used is designed to be less sensitive to spurious effects such as air transparency variations. Observational effects can therefore be excluded and the variability ascribed to variations of the luminosity of the sources.  While the hypothesis of binarity for explaining longer period variables can not be ruled out through our data, the position occupied by them suggests their possible belonging to the newly discovered class of $\\gamma$ Dor variables.  If confirmed, the $\\gamma$ Dor nature of the longer period variables in NGC 2516 would be an important result because it is unique among clusters, and the 8 objects disentangled in our survey certainly deserve further investigation.  \\newpage"
    },
    "9805/astro-ph9805358_arXiv.txt": {
        "abstract": "MHD turbulence is generally believed to have two important functions in accretion disks: it transports angular momentum outward, and the energy in its shortest wavelength modes is dissipated into the heat that the disks radiate.  In this paper we examine a pair of mechanisms which may play an important role in regulating the amplitude and spectrum of this turbulence: photon diffusion and viscosity.  We demonstrate that in radiation pressure-dominated disks, photon damping of compressive MHD waves is so rapid that it likely dominates all other dissipation mechanisms. ",
        "introduction": "Turbulence is widely thought to be central to the dynamics of accretion disks.  A combination of magnetic and Reynolds turbulent stresses may be responsible for the outward transport of angular momentum without which no accretion could occur (\\cite{sha73}, \\cite{bal94}). % The energy put into this turbulence is ultimately deposited as heat, and is therefore the energy source for the radiation by which we observe accretion disks.  Although much effort has gone into identifying mechanisms which excite turbulence (\\cite{bal91}), far less attention in the literature has been given to how the turbulence dissipates.  In most instances, it is simply assumed that nonlinear couplings transfer energy from long wavelengths to short, and that some dissipative mechanism eventually damps very short wavelength motions.  One reason why little thought has been given to the specifics of dissipation is that, as matter drifts inward through an accretion disk, if the disk is in a time-steady state its lost potential energy is transformed into heat and kinetic energy at a rate which is entirely fixed by global properties.  If the gravitational potential is dominated by the mass $M$ of the central object, the heating rate per unit area is \\begin{equation} Q = {3 \\over 4\\pi}{ GM \\dot M \\over r^3} R_R (r). \\end{equation} $R_R$ ($\\simeq 1$ at large radii) describes the reduction of the local heating due both to the kinetic energy carried outward with the angular momentum flux, and relativistic effects should the central object be a neutron star or black hole (\\cite{nov73}).  It is a great simplification to calculations of disk equilibria that the heating rate should depend only on global quantities.  However, this fact leaves open the question of how exactly the energy lost by the accretion flow is transformed into heat, and there are strong observational consequences that depend on just how this happens.   For example, the existence of weakly-radiative disks (\\cite{ich77}; \\cite{ree82}; \\cite{nar95}) depends critically on the assumption that most of the heat goes to the ions, not the electrons.  There have been other suggestions that a significant part of the heat goes into non-thermal particle distributions (e.g. \\cite{fer84} or \\cite{ste91}). Alternatively, the energy can be lost in magnetic fields which escape the disk, forming a corona or outflow (Galeev et al. 1979).  Balbus \\& Hawley (1991) pointed out that MHD fluctuations should be linearly unstable in weakly-magnetized accretion disks.  Fully nonlinear simulations (Hawley, Gammie \\& Balbus 1995; Brandenburg et al. 1995;  Stone et al. 1996) have shown that these fluctuations grow until the field energy density approaches the pressure in the disk, and that nonlinear couplings create fluctuations on shorter and shorter wavelengths. Most recent work on how the energy in these fluctuations is dissipated has concentrated on plasma physics effects that work on modes of very short wavelength (e.g. Bisnovatyi-Kogan \\& Lovelace 1997; Quataert 1997; Blackman 1997; Gruzinov 1997), especially in low-density, high temperature disks.  Although this focus is well-grounded in reality in the context of MHD turbulence in laboratory plasmas, it ignores the fact that accretion disks are often extremely bright, and can contain such high photon densities that radiation dominates the total pressure.  In this paper we point out that photon diffusion and viscosity can, {\\it in radiation-dominated accretion disks} dominate all other mechanisms of dissipation.  When that is so, compressive modes whose wavelengths are almost as great as a disk thickness can be rapidly damped.  Significant consequences follow for the amplitude of MHD turbulence, the rate at which angular momentum may be transported, and the way in which the energy associated with the turbulence is dissipated into heat.  The structure of this paper is as follows: we first extend (\\S 2) the theory of MHD modes interacting with a background photon gas by substituting a time-dependent radiation transfer solution for the conventional description in terms of a photon viscosity.  Our procedure is similar in character to the one adopted to treat photon diffusion damping of perturbations in the early Universe (``Silk damping\": Silk 1968, Hu \\& Sugiyama 1996).  We then apply this improved theory to conventional accretion disk models (\\S 3).  In \\S 4 we discuss the impact of photon damping on both advection-dominated accretion disks and disks in which the dissipation is segregated into a corona.  Finally, in \\S 5 we summarize our results and discuss their significance.  We close this introduction with some notes of distinction.  There were earlier suggestions by \\cite{loe92} and \\cite{tsu97} that photon viscosity due to an external radiation field might explain the radial angular momentum transport in some accretion disks.  We do {\\it not} make that claim; in this paper we consider only how photon kinetic effects help regulate the amplitude of the MHD turbulence that is responsible for angular momentum transport.  The effects of photon damping we consider are for a scattering-dominated plasma, and thus the relations we derive are different from those found by \\cite{bog89} and \\cite{mih83}, who derived the dispersion relation for a radiation field in LTE, ignoring both scattering and radiation viscosity, which we include. Our problem also differs from that treated by \\cite{cas96}, who considered only optically thick spiral density waves that lose angular momentum through radiation.  Our equations are very similar to those of \\cite{jed98} and \\cite{sub97} in the diffusion and free-streaming limits; however, we have bridged the two regimes by truncating the radiation field moment expansion above quadrupole moment.  We also note that Thompson and Blaes (1998) have considered radiation damping for waves in the context of gamma ray bursts. ",
        "conclusions": "We have shown in the previous sections that the effectiveness of radiation in damping fluid motions depends strongly on the ratio of radiation to gas pressure.  As noted in equation (\\ref{rtrans}), radiation tends to be most important in the inner parts of accretion disks, which are, of course, the most important for energy release.  At least some part of the disk is radiation-dominated when \\begin{equation} \\label{mdcrit} \\dot m > 1.0 \\times 10^{-3} x_{min}^{21/16} \\alpha^{-1/8} M_{8}^{-1/8}, \\end{equation} where $x_{min}$ is the inner radius of the disk.  If the central object is a black hole or a weakly-magnetized neutron star, we may expect $x_{min}$ to be the radius of the marginally stable orbit, $ = 6$ in the limit of a spinless black hole, and $\\rightarrow 1$ as the spin of the black hole approaches its maximum possible value.  However, if the disk does not extend in so far, whether because the central mass is a strongly-magnetized neutron star, or a larger object such as a white dwarf, the minimum accretion rate for which at least part of the disk is radiation-dominated rises, and may become impossibly high.  The remainder of this section, in which we outline the consequences of radiation damping in accretion disks, is divided according to consequences applicable to radiation-dominated disks and those applicable to the gas pressure-dominated case.  Whether one set or the other is relevant to a given disk depends on how it fares according to the criterion of equation (\\ref{mdcrit}).  \\subsection{Radiation pressure-dominated disks}  Two qualitative physical consequences follow from the strength of radiation damping in photon pressure-dominated disks.  First, dissipative heating is delivered to the electrons and photons through radiation scattering, and not to the ions.  Because it is the electrons that cool the gas through the creation and upscattering of photons, the only energy exchange process involving the ions is Coulomb scattering.  This mechanism should keep the ion temperature very close to the electron temperature. If the average energy of photons is less than $\\beta^2m_ec^2/3+4k_BT_e$, where $T_e$ is the electron temperature, then the photons will receive most of the energy from scattering (\\cite{psa97}).  The $\\beta^2m_ec^2$ term represents a modification of the Compton temperature due to bulk Comptonization.  Second, the process by which these disks shine may be thought of as a sort of ``bootstrap\":  if the disk were initially free of radiation, any initial photon creation by the electrons would lead to wave dissipation that heats the electrons, and therefore leads to more radiation.  The question of what makes near-Eddington accretion disks shine has a tautological answer: bright accretion disks shine because they are so bright.  That MHD fluctuations should be present at all is likely due to the operation of the magneto-rotational instability identified by Balbus \\& Hawley (1991).  This instability grows at a rate $\\sim k v_A$ for wavenumbers $k \\leq \\sqrt{3} \\Omega/v_A$ when the magnetic field is weak (i.e. $v_A < c_s$).  The compressibliity of the growing modes is slight, so the corresponding radiation damping rate should be a fraction of the pure compressive rate, as given by equation (\\ref{icrat}).  If most of the torque in the disk is due to magnetic fluctuations, the ratio between the magneto-rotational growth rate and the radiation damping rate is then at least $\\sim 10 (c_s/v_A) /(kh)$.  We therefore expect the linear growth of MHD fluctuations to proceed unaffected by radiation damping.  However, shorter wavelength waves are {\\it not} amplified by the magneto-rotational instability.  Instead, they are pumped by nonlinear coupling with the longer wavelength, growing modes.  Because the radiation damping rate is $\\propto k^2$ in the diffusive regime, compressive modes excited by nonlinear coupling will be strongly damped.  In other words, {\\it provided only that the nonlinear coupling between incompressible and compressible modes is reasonably strong}, the ``inner scale\" of the MHD turbulence will be not much shorter than its ``outer scale.\"  Any turbulent ``inertial range\" will be severely limited.  This fact leads to several other results.  At a purely technical level, if short wavelengths are all severely damped, the life of the numerical simulator is made much easier, for there is no need to strive for very fine spatial resolution.  More physically, radiation damping may play an important role in regulating the value of the ``viscosity\" parameter $\\alpha$. The magnetic part of the stress causing angular momentum transport may be written in the form \\begin{equation} T_{r\\phi} = {-1 \\over 4\\pi} \\int \\, d^3 k \\, \\delta \\hat B_r (\\vec k) \\delta \\hat B_{\\phi}^* (\\vec k) . \\end{equation} If there is little power in the fluctuations at wavenumbers much more than $\\sim 1/h$, the angular momentum transport is reduced below what it would otherwise be.  Disks in this situation would then maintain rather larger surface densities.  Increased optical depth also leads to greater radiation pressure for fixed emergent flux.  Another consequence for turbulence in radiative disks is that the ratio of the sound speed to the Alfv\\'en speed changes with wavelength. In the diffusive regime, $c_s \\sim c_r \\gg v_A$, leading to a large plasma $\\beta \\equiv P_{tot}/P_{mag}$.  When the plasma $\\beta$ is large, MHD fluctuations are generally close to incompressible because pressure waves can travel rapidly enough to smooth out density disturbances. However, for short wavelengths, the radiation field decouples from the fluid, and $c_s \\sim c_g \\ll v_A$, which means the plasma $\\beta$ becomes effectively quite small.  For these short wavelengths, then, we can expect the turbulence to exhibit much greater compressibility. In the compressible regime, the speeds of the magnetosonic and Alfv\\'en waves are comparable, so they may couple much more easily.  A similar effect happens for Alfv\\'en waves near recombination, as discussed by \\cite{sub97}.  The slope and inertial range of the turbulent spectrum will also be affected by the plasma $\\beta$ parameter, which is usually held fixed in compressive MHD simulations (\\cite{mat96}).  Analytic theory and simulations show that for compressible MHD, $\\delta \\rho/\\rho \\sim (\\delta v_A/c_s)^2$ (where $\\delta v_A \\equiv |\\delta {\\bf B}|/\\sqrt{4\\pi\\rho}$), so when $c_s$ drops dramatically in the non-diffusive regime, compressive damping will become very effective.  Simulations of turbulent cascades with small $\\beta$ but with incompresible stirring will show how much energy can be transferred to compressible modes. Current simulations of compressible turbulence in the ISM (Charles Gammie, private communication) show that shocks form when $v_A \\gg c_g$, so that if the incompressible cascade does not transfer energy to compressible modes before reaching the non-diffusive scale, the energy may be dissipated in shocks at that scale.  The dissipation in these shocks may be partly due to ordinary plasma processes, and partly due to radiation scattering. Thus, we expect that $k_{max}$ will never be much greater than $k_D$.  As the radiation pressure varies with disk radius, $k_D$ changes and thus $k_{max}$ changes, so the value of $\\alpha$ may become a function of radius.  Although certain consequences of radiation damping are relatively clear (at least qualitatively), consideration of this process also raises a number of questions:  1) What is the nature of the coupling between compressive and incompressive modes?   Is it large enough to allow the radiation damping rate to compete with the nonlinear frequency?  Are the analytic estimates we have made useful in the nonlinear regime?  2) In the simulations done to date, in which radiation pressure and transport are equally ignored, the magnetic energy density is an interesting fraction of the pressure and the associated fluctuations lead to a stress which is also proportional to the pressure.  The question naturally arises whether, in radiation pressure-dominated disks, the $r-\\phi$ stress and the energy in the magnetic field scale with the total pressure, or just with the gas pressure.  The photon bubble instability (\\cite{aro92}) will likely affect the disk structure and stress (Gammie 1998).  With explicit consideration of the quality of dynamical coupling between radiation fluctuations and fluid fluctuations, as outlined here, simulations should now be able to answer these questions.  3) Can thermal or viscous instabilities be suppressed by radiation damping?  Or does the dependence of dissipation on the radiation pressure exacerbate these instabilities?  In both cases, the most important modes have radial wavenumbers $< h^{-1}$, so the calculation here does not directly bear on them.  However, one might expect that some of the same effects will qualitatively carry over.  4)  The relativistic portions of accretion disks may trap a number of long wavelength (i.e. $kh < 1$) normal modes (Nowak \\& Wagoner 1991, 1992).  Some of these {\\it grow} in amplitude due to viscous dissipation (Nowak \\& Wagoner 1992).  Modulo the caveat of point 3), will radiation damping enhance (or destroy) these modes?  5) Many seek the origin of disk coronal heating in the dissipation of rising MHD waves (e.g. Rosner, Tucker, \\& Vaiana 1978; Heyvaerts \\& Priest 1989; Tout \\& Pringle 1996).  If radiation damping quenches short wavelength fluctuations, will this affect the rate at which magnetic flux rises to the disk surface?   \\subsection{Gas-pressure dominated disks}  When gas pressure dominates over radiation pressure, radiation damping does not compete with the nonlinear frequency. The question of what causes the heating of the disk therefore remains open. This conclusion is equally true of conventional gas pressure-dominated disks and unconventional ones like ADAFs.  Finally, the contrast between the radiation pressure-dominated and gas pressure-dominated regimes may mean that interesting observable effects occur in disks whose accretion rate fluctuates around the critical value of equation (\\ref{mdcrit}).  If the value of $\\alpha$ and the radiative efficiency depend on whether radiation damping plays a role, there could be significant modulations in the luminosity and spectrum on a viscous timescale."
    },
    "9805/astro-ph9805028_arXiv.txt": {
        "abstract": "We will summarize results of calculations of the modes of oscillation trapped within the inner region of accretion disks by the strong-field gravitational properties of a black hole (or a compact, weakly-magnetized neutron star). Their driving and damping will also be addressed. The focus will be on the most observable class: the analogue of internal gravity modes in stars. Their frequencies which corrrespond to the lowest mode numbers depend almost entirely upon only the mass and angular momentum of the black hole. Such a feature may have been detected in the X-ray power spectra of two galactic `microquasars', allowing the angular momentum of the black hole to be determined in one case. ",
        "introduction": "In this review, dedicated to one of my oldest and best friends, Giora Shaviv, on the occasion of his 60th birthday, a potentially powerful probe of black holes will be described. But first we may ask to what extent their existence has been verified, keeping in mind our definition of a black hole: a region of spacetime described by the Kerr metric. There are several lines of evidence.  A) The determinations that a certain amount of mass is contained within a certain radius (typically many orders of magnitude greater than the horizon size of a black hole of that mass) are based mainly upon (a) the time dependence of the Doppler velocity of a companion for binary systems and (b) the Doppler velocity and orbital radius of gas or stars for galactic nuclei. There is strong evidence for a mass greater than that of standard neutron stars in roughly six Galactic binaries (McClintock 1998), and of supermassive objects in roughly twice as many galactic nuclei. The rapid variability and high energies of much of the radiation provides evidence that the emission region is indeed compact ($GM/Rc^2\\aproxgt 0.1$).  B) The energy dependence of emission line profiles, when interpreted as Doppler and gravitational shifts of excited line emission from accretion disks, provides values of its extent, in terms of $R/M$ (e.~g., Fabian et al.~1989; Tanaka et al.~1995). However, a dependence of the line emissivity on radius must be assumed. In one case, it has been claimed that a rapidly rotating black hole ($a\\equiv cJ/GM^2\\approx 1$ is required to explain inferred values of inner radius $r_i < 2GM/c^2$ (Iwasawa et al.~1996).  C) Another approach involves fits to the energy dependence of the continuum (X-ray) spectrum (Zhang et al.~1997). A disk luminosity and maximum temperature is obtained for sources of known distance in which the non power-law component of the energy spectrum can be fit with a standard (diluted blackbody) optically thick accretion disk model. This provides an effective area. After radiative transfer and general relativistic corrections, an inner radius of the disk is extracted. For luminosities sufficiently below the maximum (Eddington luminosity), $r_i$ is mainly a function of $M$ and $a$.  D) Recently, Narayan et al.~(1997,1998) have claimed evidence for an event horizon, based upon the advectively dominated class of accretion disk models that have been developed (by Abramowicz, Lasota, Narayan, and others). These flows carry most of the energy generated into the black hole, rather than being radiated. Without a horizon, the energy would appear when the flow hit the surface of the central object (such as occurs with a neutron star). From spectral fits and the determination of the range of luminosities of systems whose accretion rate varies, they have identified various Galactic X-ray black hole candidates, as well as our and other galactic nuclei, as being in this state. An accretion rate less than about $8 \\%$ of the maximum is required. However, there is still no criterion to determine whether a disk with such an accretion rate will instead choose to be in the standard (radiating) state. The low luminosity would then be due to a lower accretion rate.  E) The frequencies of narrow features in the power spectra of various black hole candidates have been ascribed to persistent structures in the associated accretion disk. There have been various proposals for an identification with the orbital frequency of hot `blobs', usually at the innermost radius of the disk. However, it is not clear (a) how such `blobs' form, (b) how they survive as coherent structures, and (c) why they are confined to a particular radius. Another identification is with inertial-acoustic traveling waves, which we shall comment upon later. Instead, we focus on the spectrum of normal modes of oscillation, which must exist at some level (determined by the driving and damping processes in the disk). In the same spirit with which helioseismology probes the interior of the Sun, this probe of the Kerr metric (and its accretion disk) has been dubbed (relativistic) diskoseismology. We now analyze this approach. [For an up-to-date survey of black hole accretion disk theory, see the monograph by Kato, Fukue, and Mineshiga (1998).]  Since 1990, our group has been investigating consequences of the realization [initially by Kato and Fukue (1980)] that general relativity can trap normal modes of oscillation near the inner edge of accretion disks around black holes. The strong gravitational fields that are required can also be produced by neutron stars that are sufficiently compact (requiring a soft equation of state) and weakly magnetized that there is a gap between the surface of the star and the innermost stable orbit of the accretion disk. Although we shall not explicitly consider such neutron stars here, the results obtained will also apply to them to first order in the dimensionless angular momentum parameter $a=cJ/GM^2$, since their exterior metric is identical to that of a black hole to that order. It should be noted that $a\\aproxlt 0.5$ for almost all models of rotating neutron stars (Friedman \\& Ipser 1992; Cook, Shapiro, \\& Teukolsky 1994).  These modes of oscillation provide a potentially powerful probe of both strong gravitational fields and the physics of accretion disks, since:  $\\bullet$ They do not exist in Newtonian gravity  $\\bullet$ Their frequencies depend upon the angular momentum as well as the mass of the black hole  $\\bullet$ The fractional frequency spread of each mode depends upon $\\alpha$, the elusive viscosity parameter of the accretion disk. ",
        "conclusions": "Returning to our initial question, we have seen that the evidence for compact, nonstellar masses is strong, while the evidence for horizons is less strong but intriguing. The ultimate evidence of black holes (the Kerr metric) and measurement of their only other property (angular momentum) is still elusive. However, we have seen that the `diskoseismic' probe, as well as spectroscopic (line and continuum) methods, have great promise. For instance, we note the fact that the coincidence between the g--mode and Stefan-Boltzmann determinations of the GRO J1655-40 black hole angular momentum would have disappeared if the metric was significantly different from Kerr.  It is clear that continued X-ray timing observations at the times scales sufficiently short to  see the high frequency modes are crucial to their identification. These are being carried out by RXTE (although power spectra to higher frequency are critical) and will soon be carried out by the USA satellite. In addition, we should not lose sight of the fact that supermassive black hole accretion disks should exhibit similar signatures, merely scaled by the mass. For instance, the g--mode period for a $10^8$ solar mass slowly rotating black hole is 1.6 days. Long-term monitoring with at least 1\\% accuracy photometry would be required.  \\medskip Most of the results reported here were obtained in collaboration with Dana Lehr, Michael Nowak, and Alexander Silbergleit. The research was supported in part by NASA through ATP grant NAG 5-3102 to R.V.W. and grant NAS8-39225 to Gravity Probe B."
    },
    "9805/astro-ph9805052_arXiv.txt": {
        "abstract": "The results of the abundance analyses of six \\LB\\ stars are presented. For three stars, the impact of individual ODFs and various treatments of convection in the calculactions of the model atmospheres are investigated. ",
        "introduction": "\\label{intr} As has been discussed in the course of this workshop, the classification of \\LB\\ stars is not as simple and straightforward as for other groups of chemically peculiar stars. When only low resolution classification spectra are used, it is likely that stars with different astrophysical properties but similar characteristics in the narrow spectral range are included. Therefore a detailed spectroscopic investigation for a large number of confirmed \\LB\\ stars (e.g. Paunzen et al. 1997) covering the whole range of atmospheric parameters is required to establish a common abundance pattern for these stars. Up to now we have analyzed the chemical composition of six \\LB\\ stars where two of them turned out to be spectroscopic binaries (Paunzen et al. 1998).  In order to put our results on more solid ground we investigated the impact of using different programmes for the calculation of the model atmospheres on the abundances of three stars with different temperatures, gravities and metallicities (Heiter et al. 1998). The two issues we examined concerned the inclusion of individual line opacities and the calculation of the convective flux in the model atmospheres. ",
        "conclusions": ""
    },
    "9805/astro-ph9805114_arXiv.txt": {
        "abstract": "In a class of models designed to solve the cosmological constant problem by coupling scalar or tensor classical fields to the space-time curvature, the universal scale factor grows as a power law in the age, $a \\propto t^\\alpha$, regardless of the matter content or cosmological epoch.  We investigate constraints on such ``power-law cosmologies\" from the present age of the Universe, the magnitude-redshift relation, and from primordial nucleosynthesis. Constraints from the current age of the Universe and from the high-redshift supernovae data require ``large\" $\\alpha$ ($\\approx 1$), while consistency with the inferred primordial abundances of deuterium and helium-4 forces $\\alpha$ to lie in a very narrow range around a lower value ($\\approx 0.55$). Inconsistency between these independent cosmological constraints suggests that such power-law cosmologies are not viable. ",
        "introduction": "According to General Relativity all mass/energy gravitates, including the energy density of the vacuum.  In modern quantum field theory the vacuum is the lowest energy -- but not necessarily the zero energy -- state.  From this perspective a cosmological constant ($\\Lambda$) may be associated with the vacuum energy density, $\\rho_{\\rm vac} = \\Lambda/8\\pi G \\equiv \\Omega_\\Lambda \\rho_{\\rm c}$, where $\\rho_{\\rm c} \\equiv 3{{\\rm H}_0}^2/8\\pi G \\sim 10^{-48} {\\rm GeV}^4$ is the critical density.  Although some recent data favor a non-zero value of $\\Lambda$ \\cite{Ia}, observations do limit $\\Omega_\\Lambda \\la 1$\\cite{lambda}  corresponding to a vacuum energy density that is very small when compared to that expected from physics at the Planck scale ($\\sim 10^{19}$ GeV).    This is because although we may wish to set $\\Lambda = 0$ in the Einstein equations, quantum fluctuations in the fields present in the Universe can establish a non-zero vacuum energy and, hence, a non-zero effective cosmological constant.  We may associate the vacuum energy density with an energy scale $M$ which might be the scale associated with the spontaneous symmetry breaking from one vacuum state to another, $\\rho_{\\rm vac}\\sim M^4$.  In some sense the only ``natural\" scale in cosmology is the Planck scale, $M \\sim 10^{19}$~GeV.  In this case the observations require that the present vacuum energy density is some 120 orders of magnitude smaller than its ``natural\" value.  The smallness of $\\rho_{\\rm vac}$ is a key problem in modern cosmology: the ``$\\Lambda$\" or ``cosmological constant problem''\\cite{swein}.   One class of attempts to solve the $\\Lambda$-problem considers the evolution of classical fields which are coupled to the curvature of the space-time background in such a way that their contribution to the energy density self-adjusts to cancel the vacuum energy\\cite{regulate}. Although the dynamical framework in these approaches is well defined, the addition of the special fields is unmotivated but for solving the cosmological constant problem.    The common result of these approaches is that the vacuum energy may be nearly cancelled and the expansion of the Universe is governed by the uncompensated vacuum energy density.  In such models the expansion is a power-law in time, independent of the matter content or cosmological epoch (see Ford, ref\\cite{regulate}).  That is, in such models the scale factor varies according to $a(t) \\propto t^\\alpha$, where $\\alpha$ is determined solely by the parameters of the model and can be anywhere in the range $0\\le\\alpha\\le \\infty$. In addition, there are  models designed to solve other cosmological fine-tuning problems (\\eg, flatness \\cite{Allen}) which also result in power-law cosmologies.  In this {\\sl Letter} we explore the constraints on $\\alpha$ from the age-expansion rate data, from the magnitude-redshift relation of type Ia supernovae (SN~Ia) at redshifts 0.4 -- 0.8, and from the requirement that primordial nucleosynthesis produce deuterium and helium-4 in abundances consistent with those inferred from observational data. ",
        "conclusions": "Can the evolution of the Universe -- from very early epochs to the present -- be described by a simple power law relation between the age and the scale factor (temperature)?  In standard cosmology the early Universe is radiation dominated (RD) and the expansion is a power law with $\\alpha_{\\rm RD} = 1/2$.  But, in standard cosmology the Universe switched from RD to matter dominated (MD) at a redshift between 10$^{3}$ and 10$^{4}$.  Thereafter the Universe expanded (for a while at least) according to a power law with a different power: $\\alpha_{\\rm MD} = 2/3$.  If the present Universe has a low density (compared to the critical density) and lacks a significant cosmological constant, it is ``curvature\" dominated (CD) and its expansion may be well approximated by a power law with $\\alpha_{\\rm CD} = 1$.  Thus in standard cosmology, although power law expansion may provide a good description for some epochs, there is no single power which can describe the entire evolution from, for example, BBN to the present.  The question then is, can a ``compromise\" $\\alpha$ be found which is consistent with BBN as well as with observations of the present/recent Universe?  We have explored this question and answered it in the negative.  The present age/expansion rate (Hubble parameter) constraint $\\alpha = H_{0}t_{0} = 1.0 \\pm 0.2$ and the SN~Ia magnitude-redshift relation require $\\alpha \\approx 1$ (or, $\\alpha$ $\\ga$ 0.6), while production of primordial helium and deuterium force $\\alpha$ to be smaller.  The extreme sensitivity of the helium yield to $\\alpha$ (see Fig. 3), precludes raising the upper bound on $\\alpha$ from BBN.  Unless the Universe is much younger ($\\la $~10~Gyr) and/or the Hubble parameter much smaller ($\\la~$50~km~s$^{-1}$ Mpc$^{-1}$) than currently believed and the SN~Ia magnitude-redshift relation plagued by systematic errors, or there was substantial entropy release after BBN, power law cosmologies are not the solution to the cosmological constant problem."
    },
    "9805/astro-ph9805322_arXiv.txt": {
        "abstract": "We describe a simple efficient algorithm that allows one to construct Monte-Carlo realizations of merger histories of dark matter halos. The algorithm is motivated by the excursion set model (Bond et al. 1991) for the conditional and unconditional halo mass functions.  The forest of trees constructed using this algorithm depends on the underlying power spectrum.  For Poisson or white-noise initial power-spectra, the forest has exactly the same properties as the ensemble of trees described by Sheth (1996) and Sheth \\& Pitman (1997).  In this case, many ensemble averaged higher order statistics of the tree distribution can be computed analytically.  For Gaussian initial conditions with more general power-spectra, mean properties of the ensemble closely resemble the mean properties expected from the excursion set approach. Various statistical quantities associated with the trees constructed using our algorithm are in good agreement with what is measured in numerical simulations of hierarchical gravitational clustering. ",
        "introduction": "It is widely believed that the massive dark matter halos which exist today have grown from small initially Gaussian density fluctuations. Many questions of astrophysical interest can be addressed using Monte--Carlo realizations of the merger histories of such massive dark matter halos.  An efficient method for generating Monte-Carlo realizations of the clustering process is essential for addressing such questions (e.g. Kauffmann \\& White 1993). Ideally, such an algorithm should produce results that are consistent with other known properties of the clustering process.  In this paper, we develop an algorithm which is motivated by known excursion set results (Bond et al. 1991; Lacey \\& Cole 1993). For example, these earlier results provide expressions for the mean number of halos of mass $m$ that are later incorporated into a larger halo of mass $M>m$.  For the special case of clustering from Poisson, or white-noise initial conditions, a model for the higher order moments of this subclump distribution exists (Sheth 1996; Sheth \\& Pitman 1997).  Appendix~\\ref{ppf} of this paper shows that these higher order moments are consistent with the fact that mutually disconnected regions in Poisson or white-noise Gaussian distributions are mutually independent. Section~\\ref{pics} describes an algorithm which uses this fact to generate a forest of merger trees.  The algorithm reproduces the mean and higher-order Poisson and white-noise results exactly.  Although the algorithm is not of binary split type, trees constructed using binary splits provide good approximations to ours (Section~\\ref{bsplit}).  Unfortunately, previous work does not provide expressions for the higher order moments of the subclump distribution for other initial power spectra (e.g. Sheth \\& Pitman 1997). Section~\\ref{mctree} shows that, to describe the trees associated with arbitrary Gaussian initial conditions, a simple modification of our white-noise algorithm produces good approximations to the known (excursion set) mean values. Section~\\ref{scfsim} shows that, for a range of power spectra of current interest, our algorithm provides higher order distributions that are similar to those measured in numerical simulations of hierarchical gravitational clustering.  A final section summarizes our results. It argues that, in addition to allowing one to generate a forest of merger history trees, our results are also useful for studying the spatial distribution of dark matter halos. ",
        "conclusions": "\\label{concl} We described an algorithm which allowed us to partition halos at a given time into subhalos at an arbitrary earlier time.  The algorithm is exact for Poisson and white-noise initial conditions.  In these cases, we provided analytic expressions for the higher order moments of the subclump distribution.  We discussed two possible ways of modifying the algorithm to partition halos which form from more general Gaussian initial conditions. We then showed that the one-step partition algorithm can be embedded into a loop over many time steps to generate a forest of merger history trees.  This also showed why binary split algorithms previously used to construct the merger trees may be reasonably accurate.  We then embedded the second of our partition algorithms into a loop over time steps to generate the forest of merger trees. We compared this forest of trees with the ensemble of trees measured in numerical simulations of hierarchical gravitational clustering from scale-free initial conditions. For the initial conditions we studied, the agreement between our trees and those of the simulations was fair.  Furthermore, this comparison showed that our analytic formulae for the higher order moments of the subclump distribution provided reasonable fits to the numerical simulation results even when the initial conditions were different from white noise.  This result is particularly useful for studying the spatial distribution of dark matter halos. Mo \\& White (1996) argued that the higher order moments of the subclump distribution associated with the forest of merger history trees of dark matter halos can be related to the higher order moments of the spatial distribution of these halos. Since the higher order moments associated with our merger tree algorithm, and also of the merger trees in the numerical simulations, were reasonably well fit by our equation~(\\ref{c120}), it is possible to write down analytic approximations to spatial quantities like the halo--halo correlation function that are also reasonably accurate. This is the subject of ongoing work (Sheth \\& Lemson 1998).  Before concluding, we note that Sheth (1998) describes a model for clustering from compound Poisson initial conditions. For these more general initial conditions, the analogue of equation~(\\ref{fjk}) factors similarly to how it does for the Poisson case.  This means that a similar algorithm for generating the associated forest of merger trees can be used there also."
    },
    "9805/astro-ph9805186_arXiv.txt": {
        "abstract": "In order to find out if regularities and systematic trends found to be apparent among experimental Stark line shifts allow the accurate interpolation of new data and critical evaluation of experimental results, the exceptions to the established regularities are analysed on the basis of critical reviews of experimental data, and reasons for such exceptions are discussed.  We found that such exceptions are mostly due to the situations when: (i) the energy gap between atomic energy levels within a supermultiplet is equal or comparable to the energy gap to the nearest perturbing levels; (ii) the most important perturbing level is embedded between the energy levels of the supermultiplet; (iii) the forbidden transitions have influence on Stark line shifts. ",
        "introduction": "Wiese and Konjevi\\'c (1982) established that for experimental Stark widths of non-hydrogenic lines, there are similarities (see as well references in Wiese \\& Konjevi\\'c 1982 and Dimitrijevi\\'c 1982) of line widths within a multiplet, a supermultiplet and a transition array, as well as for analogous transitions of homologous atoms and ions. They found as well a systematic behaviour of Stark line widths along spectral series. The exceptions to these similarities and systematic trends have been analyzed by Dimitrijevi\\'c (1982), who found that the reasons for such exceptions may be divided in two categories: (i) irregular atomic energy level structure and (ii) inadequacy of the model used for the emitter structure. He emphasized as well, that the simple analysis of Grotrian diagrams for corresponding radiator energy levels, may be useful for prediction of mutual relations among Stark widths within multiplets, supermultiplets and transition arrays. Extending their work of 1982 on Stark widths, Wiese \\& Konjevi\\'c (1992) carried out the same kind of research on experimental Stark line shifts, and showed numerous examples where the same regularities and systematic trends hold. Similarly as in Dimitrijevi\\'c (1982) for widths, we want to analyze here the exceptions to the established regularities and systematic trends for Stark line shifts. ",
        "conclusions": "The exceptions to the established regularities have been analysed on the basis of critical reviews of experimental data (Konjevi\\'c \\& Roberts 1976; Konjevi\\'c \\& Wiese 1976; Konjevi\\'c et al. 1984ab; Konjevi\\'c \\& Wiese 1990). The complete analysis will be published elsewhere. We found that such exceptions are mostly due to the situations when: (i) the energy gap between atomic energy levels within a supermultiplet is equal or comparable to the energy gap to the nearest perturbing levels; (ii) the most important perturbing level is embedded between the energy levels of the supermultiplet; (iii) the forbidden transitions have influence on Stark line shifts.  The example of Stark line shifts from F I 3s - 3p (quartets) supermultiplet, illustrates the case when the energy gap between upper atomic energy levels for particular members of a supermultiplet is not negligible in comparison to the energy gap to the most important perturbing levels. For the 3p$^4$S$^o$ energy level for instance, the influence of the upper perturbing levels 4s and 3d, is larger in comparison with this influence for the 3p$^4$P$^o$ energy level, and the contribution of the 3s energy level is smaller. The effect of such an energy structure is larger on the shift than on the width, since all partial contributions to the width are positive while the contribution of the level 3s as a perturbing level of 4p to the shift is negative. Consequently, the shift of lines within the 3s$^4$P - 3p$^4$S$^o$ multiplet is larger than the shifts within the 3s$^4$P - 3p$^4$P$^o$ multiplet."
    },
    "9805/astro-ph9805009_arXiv.txt": {
        "abstract": "\\bigskip We present evolutionary models for low mass stars from 0.075 to 1 $\\msol$ for solar-type metallicities [M/H]= 0 and -0.5. The calculations include the most recent interior physics and the latest generation of {\\it non-grey} atmosphere models. We provide mass-age-color-magnitude relationships for both metallicities. The mass-M$_V$ and mass-M$_K$ relations are in excellent agreement with the empirical relations derived observationally. The theoretical color-magnitude diagrams are compared with the sequences of globular clusters (47 Tucanae) and open clusters (NGC2420 and NGC2477) observed with the  Hubble Space Telescope. Comparison is also made with field star sequences in $\\mv$-$(V-I)$, $\\mk$-$(I-K)$ and $\\mk$-$(J-K)$ diagrams. These comparisons show that the most recent improvements performed in low-mass star atmosphere models yield now reliable stellar models in the near-infrared. These models can be used for metallicity, mass, temperature and luminosity calibrations. Uncertainties still remain, however, in the optical spectral region below $\\te \\sim 3700K$, where predicted (V-I) colors are too blue by 0.5 mag for a given magnitude. The possible origins for such a discrepancy, most likely a missing source of opacity in the optical and the onset of grain formation are examined in detail.   \\bigskip ",
        "introduction": "The numerous data obtained within the past few years with ground-based and space-based near infrared projects provide nowadays a wealth of low-mass star observations from 1 $\\msol$ down to the brown dwarf regime. Observations cover a wide range of stellar populations, belonging to young, open or globular clusters and to halo and disk fields. Their analysis requires accurate theoretical models spanning a large range of ages, masses and metallicities. Important progress has been realized recently on the theoretical side, which emphasizes the complex physics involved in the modeling of these cool and dense objects. Recent work has demonstrated the necessity to use accurate internal physics and outer boundary conditions based on non-grey atmosphere models to describe correctly the mechanical and thermal properties of low mass objects (Burrows et al. 1993; Baraffe et al. 1995, 1997; Chabrier and Baraffe 1997, CB97). The tremendous efforts accomplished recently in the modeling of atmosphere models and the derivation of synthetic spectra (see the review by Allard et al. 1997), combined with interior models, now provide synthetic colors and magnitudes which can be compared directly to observed quantities, avoiding the use of uncertain empirical $\\te$ and bolometric correction scales.  In a recent paper (Baraffe et al. 1997, BCAH97), we have derived evolutionary models for metal-poor low mass stars based on the stellar interior physics described in CB97 and on the Allard and Hauschildt (1998) \"NextGen\" atmosphere models. Comparison with the lower Main Sequence of globular clusters observed with the HST has assessed the validity of the models in the metallicity-range $-2.0\\le [M/H]\\le -1.0$. The success of these models has been confirmed recently by new observations of NGC6397 down to $\\sim 0.1 \\,\\msol$ (King et al., 1998), but more importantly with the observations realized with the NICMOS camera. Indeed, at the time of the BCAH97 analysis, only optical (V-I) colours were available for the clusters. Recent observations performed with NICMOS for the first time provide colour-magnitude diagrams (CMDs) in the near-infrared domain for $\\omega$Cen (Pulone et al. 1998) and NGC6397 (Paresce, priv. com.). The agreement with the models is excellent and the observations confirm in particular the predicted blueshift in IR colors near the bottom of the main sequence, which stems from ongoing collision-induced  absorption of molecular hydrogen (see BCAH97 Fig. 7). This, we believe, assesses the reliability of our metal-poor models down to the bottom of the main sequence.  The natural continuation of this work is the extension to solar-like metallicities. This is the aim of the present paper. The present calculations are based on the same microphysics, described in CB97, and are confronted to available observations in the range $-0.5\\le [M/H]\\le 0$. The calculation of atmosphere models for solar metal-composition is rendered more complex by the importance of molecular metal-bands, which shape the emergent spectrum. In this range of metallicity, the stellar spectra and atmospheric structures become very sensitive to the treatment of molecular opacity, dominated by H$_2$O in the IR and TiO and, to a less extent, VO in the optical. It is thus essential to confront theory with observations at these wavelengths to determine the remaining uncertainties in the models for solar-metallicity. We first summarize the physics entering specifically the solar models (\\S2). In \\S 3, we compare the theoretical mass-magnitude relationships in the V- and K-bands with the observationally-derived relationships. In \\S 4, we compare the results with observed CMDs of (i) the globular cluster 47 Tucanae with [M/H] $\\sim$ -0.5, (ii) two open clusters observed with the HST, $NGC2477$ and $NGC2420$ with [M/H] $\\sim$ 0, and (iii) disk field stars in optical and near-infrared colors. Section 5 is devoted to the conclusion. ",
        "conclusions": "We have presented solar-type metallicity evolutionary models from 1 $\\msol$ down to the hydrogen burning minimum mass. These models include the most recent interior physics and non-grey atmosphere models and rely on fully consistent calculations between the stellar interior and the atmosphere, with {\\it not a single} adjustable parameter. Any discrepancy between theory and observation thus reflects remaining limitations in the physics entering the theory and not internal inconsistency in the models. In order to carefully examine these limitations, special attention has been paid to the comparison with observed mass-magnitude relationships and color-magnitude diagrams.  Regarding the optical spectral region, the theoretical mass-$\\mv$ relationship is in excellent agreement with observational data of Henry and McCarthy (1993) all the way down to the bottom of the main sequence. However, the analysis of the $\\mv$-$(V-I)$ CMDs highlights the limitation of the present models for colors redward of (V-I) $\\sim 2$, i.e. $\\te \\simle 3600$ K ($m\\simle 0.4\\,\\msol$), predicting (V-I) colors substantially bluer than the observed ones. This suggests a possible underestimate of opacity below 1 $\\mu$m. We have tested such an hypothesis by increasing arbitrarily the opacity over the V-band in a couple of test atmosphere models and obtained the required effect i.e redder (V-I) color {\\it and} unaffected near infrared fluxes and atmosphere profiles. Metal-poor models, which are less sensitive to metallic molecular absorbers, do not suffer from this shortcoming, as illustrated by the remarkable agreement obtained for globular star clusters with [M/H] $\\le -1$ (BCAH97). The discrepancy begins to appear at [M/H]$\\sim$ -0.5, as suggested by the comparison with 47 Tuc, although the observational error bars remain large, and becomes obvious for solar metallicity.  This shortcoming is observed also in the comparison between synthetic and observed M-dwarf (Leggett et al., 1996) and cool giant (Alvarez \\& Plez, 1998) spectra and thus stems most likely from a still incomplete description of the {\\it atmosphere} of cool objects, rather than from substantial modifications of their  structural and transport properties. This shortcoming is inherent to all currently available atmosphere models and represents the next challenge for cool star theorists.  In the near-infrared, the results are very satisfactory. Contrarily to models based on previous generations of atmosphere models, the present mass-$\\mk$ relationship is in excellent agreement with the Henry and Mc Carthy (1993) observational data. The analysis of $\\mk - (I-K)$ and $\\mk - (J-K)$ CMDs for young open clusters (Zapatero et al., 1997; Pinfield et al. 1997) and for field disk stars down to the bottom of the main sequence confirms the significant improvement of the present models over previous generations.  As for sub-solar metallicities, the photometric signature of the bottom of the main sequence and of the substellar domain in the near-IR is a large blueshift due in that case to H$_2$ CIA absorption but also to the onset of CH$_4$ formation, shifting the peak of the flux to short wavelengths ($\\sim 1\\,\\mu$m), as observed e.g. in Gl229B (Matthews et al. 1995; Geballe et al. 1996). The very behaviour of this shift at the high-mass end of the sub-stellar domain, however, is likely to be affected by grain formation and remains to be characterized precisely in this region.  At last, we want to stress the following point: comparison between theory and observation in {\\it theoretical} $\\, \\te$-M$_{bol}$ diagrams should be avoided. Such comparisons, except for the seldom cases where the exact bolometric magnitude is determined, are unreliable since they are based on empirical color-$\\te$ or color-bolometric correction relations which do not take into account effects of age, gravity or metallicity among the sample of objects used to derive them. Although, as stressed in our previous and present calculations, shortcomings are still present in the theory, yielding still slightly inaccurate bolometric magnitudes, discrepancies arising from comparisons in $\\te$-M$_{bol}$ diagrams reflect primarily uncertainties or inconsistencies in the various transformations. Such dubious comparisons, which used to be the only possible ones a few years ago when no {\\it synthetic colors} were available, lead in general to incorrect and misleading conclusions both for the theory and the observations. Meaningful, consistent comparisons, which avoid external sources of errors, must be done in the various {\\it observational} color-magnitude and color-color diagrams. These latter represent much more stringent constraints for the theory than a global $\\te$-M$_{bol}$ diagram. Models aimed at describing cool low-mass object structural and thermal properties must rely on such a general {\\it parameter-free, consistent} theory and must first be proven to be valid in {\\it every} available passband. With as a holy grail the accurate description of the physical properties of the star in {\\it all} characteristic passbands.   As shown in this paper, the present stellar models, although based on updated physics and consistent interior-atmosphere calculations, still suffer from uncertainties at optical wavelengths, at least for solar metallicity. This reflects our still uncomplete knowledge of the complex physics characteristic of cool star-like objects. It is our aim to solve these shortcomings in a near future and to derive fully reliable models with completely accurate color- effective temperature relationships and bolometric correction scales.  \\medskip Tables 1-3 are available by anonymous ftp (including a larger grid in ages): \\par \\hskip 1cm ftp ftp.ens-lyon.fr \\par \\hskip 1cm username: anonymous \\par \\hskip 1cm ftp $>$ cd /pub/users/CRAL/ibaraffe \\par \\hskip 1cm ftp $>$ get BCAH98\\_models \\par \\hskip 1cm ftp $>$ quit \\bigskip"
    },
    "9805/astro-ph9805073_arXiv.txt": {
        "abstract": "We present preliminary results of our search for circumstellar absorption features in the Ca~K lines based on high S/N observations obtained with the ESO~CAT/CES system. ",
        "introduction": "\\label{intr} Main-sequence A stars have shallow surface convection zones, therefore their composition responds sensitively to any `contamination' by processes of diffusion or accretion. For example, the metal deficiency of the $\\lambda$\\,Bootis stars indicates accretion of depleted gas after separation of gas and dust in the stellar environment. Yet not all A stars with circumstellar (CS) matter show chemical anomalies indicative of accretion. A prominent example is $\\beta$\\,Pictoris.  However, in most cases nothing is known about their composition and the presence of CS gas. Accurate surface abundances of A stars that are positive or negative IRAS detections and a sensitive search for CS lines permit to trace the signature of accretion differentially and with a high sensitivity. ",
        "conclusions": ""
    },
    "9805/astro-ph9805303_arXiv.txt": {
        "abstract": " ",
        "introduction": "For years researchers have realized that portions of the region of a post-bounce core collapse supernova interior to a stalled prompt shock are convectively unstable.  The origin of this convective instability is fairly well understood.  After the ``bounce'' of the collapsed stellar core the prompt shock wave is formed near the sonic point which separates the inner homologous core from the supersonicly infalling outer core.  As the shock begins to propagate outward from this point (at an enclosed baryon mass of about $0.5-0.7 M_\\odot$) it weakens as energy is expended in dissociating the heavy nuclei into free nucleons.  This weakening of the shock produces a negative entropy gradient that is convectively unstable.  Until recently, researchers had numerically modeled the prompt phase of a core collapse supernova via 1-dimensional (1-D) radiation hydrodynamic models.  While these models, in many cases, included fairly sophisticated microphysics and radiation transport techniques they were unable to accurately model the convection which is inherently multi-dimensional in nature.  This situation changed radically with the pioneering work of \\cite{hbc92} who were the first to conduct multi-dimensional simulations of this convection.  Within the the past few years there have been a spate of models \\cite{mwm93,bhf95,jm96,mcbbgsu96a,mcbbgsu96b} that have begun to carry out 2-D simulations of the evolution of this convectively unstable region in the post-bounce epoch.  Most of these models have shown convection to take place in the region between the neutrinosphere and the stalled prompt shock.   However, the advancement to multi-dimensional numerical models of post-bounce supernovae has forced some compromises in the way that neutrino transport is modeled.  In this paper we report on some consequences of the use of the gray approximation to radiation transport in 2-D supernova models.  In the context of supernova models the gray approximation has taken on a unique form.  The neutrinos in the interior high density regions of the supernova are assumed to be in LTE with matter while in outer regions of the core the neutrinos are assumed to be thermally and chemically decoupled from the matter \\cite{cvb86} (hereafter CVB).  In the high density regions the assumption of LTE is adequate.  However, the fact that the neutrinos are decoupled in the exterior regions of the core require an assumption of spectral properties for each species of neutrino.  The use of the gray approximation to describe the evolution of the neutrinos in a core collapse supernova should be contrasted with the multi-group treatment that has been employed in a plethora of 1-D calculations \\cite{mbhlsv87,mb89,bru85,bruenn89a,bruenn89b,slm94}. The multi-group approach does not assume a spectral energy distribution for the neutrinos, rather the spectrum is explicitly modeled.  This is accomplished by solving a monochromatic transport equation \\cite{mm84} for a set of discrete energy ``groups'' which span the spectral range of interest.   From 1-D simulations it has been known for some time that for the problem of modeling neutrino transport in supernovae the gray methods and the multi-group methods of neutrino transport can yield substantially different results.  In particular we have discovered that the results obtained by the use of the gray approximation are extremely sensitive to the {\\em ad hoc} choice of parameters needed to describe the decoupling of neutrinos and matter and to the choices made in the implementation of the weak interaction rates that describe the coupling of neutrinos to matter.  Our purpose in this paper is to describe these sensitivities and discuss how they effect the multi-dimensional models in the convective epoch of supernovae. ",
        "conclusions": ""
    },
    "9805/astro-ph9805135_arXiv.txt": {
        "abstract": "Solar neutrino fluxes and sound speeds are calculated using a  systematic reevaluation of nuclear fusion rates. % The largest uncertainties are identified  % and their effects on the solar neutrino fluxes are estimated. % ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805298_arXiv.txt": {
        "abstract": "Accurate CCD observations of three Cepheids in the SMC were made with the purpose of confirming their nature of second overtone mode Cepheids. The stars were suspected pulsating in the second overtone mode owing to the unusual light curve and short period reported by Payne-Gaposchkin \\& Gaposchkin (\\cite{pgg}). The analysis of the new data shows that for two stars the previous periods are wrong, and in the three cases the new light curves are normal. According to the new observations, HV 1353 is a fundamental mode pulsator with small amplitude, and HV 1777 and HV 1779 are first overtone mode pulsators.  Also the star HV 1763, whose nature was unknown, was observed in the field of HV 1777. The new data show that it is a first overtone mode Cepheid with $P=2^d.117$. ",
        "introduction": "One of the main by-product results of MACHO and EROS projects is the discovery of many double-mode Cepheids (DMCs) pulsating in both the fundamental- and first-overtone mode (F/1O), and the first- and second-overtone (1O/2O) modes, in the LMC and SMC (e.g. Alcock et al. \\cite{a1}; Alcock et al. \\cite{a2}; Beaulieu et al. \\cite{be}). The presence of many 1O/2O DMCs raises immediately the question of the possible existence of Cepheids pulsating purely in the second overtone mode, but up to now none have been found, probably because it is difficult to identify them unambiguously. They should have short periods and should be found preferentially in low metallicity galaxies (see e.g. Alcock et al. \\cite{a2}). In fact, while many  1O/2O DMCs have been discovered in the Magellanic Clouds, only one of these stars (CO Aur; Mantegazza \\cite{ma}) was observed in our Galaxy.  The importance of second overtone mode Cepheids relies on their being, among Cepheids, the third possible benchmark for the stellar interior and evolution theory beside fundamental and first overtone mode Cepheids. Recently  Antonello \\& Kanbur (\\cite{ak}) studied the characteristics of these stars predicted by nonlinear pulsation models, and remarked in particular the effects of the resonance $P_2/P_6=2$ at $P_2 \\sim 1$ d ($P$ is the period) between the second and sixth overtone mode. Resonances represent a powerful comparison tool between observations and theoretical model predictions, because they affect the shape of the curves of stars in a specific period range, for example $P_0/P_2=2$ at $P_0 \\sim 10$ d in fundamental mode Cepheids (Simon \\& Lee, \\cite{sl}) and $P_1/P_4=2$ at $P_1 \\sim 3$ d in first overtone Cepheids (Antonello et al. \\cite{apr}). The close comparison allows to probe the stellar interior and to put constraints on the stellar physical parameters. Recently, Beaulieu (\\cite{be1}) mentioned some possible second overtone candidates in the SMC, and it would be interesting to observe them accurately in order to confirm their nature.  The Magellanic Cloud variables were extensively studied about thirty years ago by C. Payne--Gaposchkin and S. Gaposchkin. One of us (Antonello, \\cite{a93}) used their results concerning the asymmetry parameter of Cepheid light curves for studying the differences between fundamental and first overtone mode Cepheids, and, in the case of the SMC (Payne-Gaposchkin \\& Gaposchkin \\cite{pgg}), he noted four stars with short period and unusual asymmetry parameter (that is unusual light curve) and indicated them as possible second overtone mode candidates. In the present note we report about the results of new observations of three of these stars: HV 1777, HV 1779 and HV 1353.  \\begin{table*} \\caption[]{$V$ Photometric Observations.} \\begin{flushleft} \\begin{tabular}{ll|ll|lll} \\hline Hel.J.D. & HV 1353 & Hel.J.D. & HV 1779 & Hel. J.D. & HV1777 & HV 1763 \\\\ 2450300+ &         & 2450300+ &         & 2450300+  &        &          \\\\ &&&&&&\\\\ \\hline 70.036 &  16.330 &  70.021 &  15.871 &  70.042 &  16.138 &  16.312\\\\ 70.183 &  16.355 &  70.178 &  15.923 &  70.186 &  16.163 &  16.331\\\\ 71.029 &  16.441 &  71.025 &  16.278 &  71.034 &  15.845 &  15.972\\\\ 71.160 &  16.440 &  71.156 &  16.253 &  71.164 &  15.845 &  15.972\\\\ 71.301 &  16.444 &  71.296 &  16.191 &  71.304 &  15.819 &  16.014\\\\ 72.168 &  16.094 &  72.163 &  16.035 &  72.172 &  16.066 &  16.342\\\\ 72.298 &  16.134 &  72.294 &  16.119 &  72.303 &  16.105 &  16.349\\\\ 73.055 &  16.311 &  73.051 &  16.216 &  73.060 &  16.116 &  15.997\\\\ 73.186 &  16.328 &  73.182 &  16.101 &  73.190 &  16.039 &  15.974\\\\ 73.310 &  16.386 &  73.305 &  15.964 &  73.318 &  15.940 &  16.004\\\\ 73.325 &  16.351 &  74.034 &  16.075 &  74.045 &  15.824 &  16.285\\\\ 74.040 &  16.404 &  74.158 &  16.143 &  74.164 &  15.859 &  16.308\\\\ 74.170 &  16.412 &  74.285 &  16.204 &  74.291 &  15.893 &  16.330\\\\ 74.297 &  16.439 &  75.031 &  16.031 &  75.037 &  16.139 &  16.070\\\\ 75.042 &  16.133 &  75.153 &  15.914 &  75.159 &  16.150 &  16.004\\\\ 75.164 &  16.036 &  75.289 &  15.860 &  75.299 &  16.166 &  15.996\\\\ 75.304 &  16.049 &  76.032 &  16.219 &  76.038 &  15.899 &  16.224\\\\ 76.044 &  16.250 &  76.158 &  16.251 &  76.163 &  15.850 &  16.298\\\\ 76.168 &  16.299 &  76.289 &  16.259 &  76.294 &  15.821 &  16.335\\\\ 76.300 &  16.315 &  77.027 &  15.860 &  77.032 &  16.040 &  16.134\\\\ 77.037 &  16.437 &  77.155 &  15.870 &  77.161 &  16.060 &  16.059\\\\ 77.166 &  16.429 &  77.292 &  15.909 &  77.298 &  16.093 &  16.009\\\\ 77.303 &  16.422 &  78.026 &  16.260 &  78.032 &  16.114 &  16.205\\\\ 78.037 &  16.348 &  78.139 &  16.286 &  78.144 &  16.054 &  16.257\\\\ 78.150 &  16.237 &  78.290 &  16.297 &  78.295 &  15.990 &  16.286\\\\ 78.300 &  16.101 &  79.017 &  15.916 &  79.023 &  15.806 &  16.237\\\\ 79.029 &  16.237 &  79.154 &  15.959 &  79.171 &  15.836 &  16.137\\\\ 79.177 &  16.287 &  79.275 &  16.040 &  79.286 &  15.880 &  16.063\\\\ 79.292 &  16.279 &  79.280 &  16.029 \\\\ 79.297 &  16.285 \\\\ \\hline \\end{tabular} \\end{flushleft} \\end{table*}  \\section {Observations}  The observations were obtained with the Dutch 0.9m telescope at La Silla Observatory (ESO) during ten consecutive nights (Oct. 13--22, 1996) by means of a TEK512 CCD with 580 columns and 520 rows (ESO chip \\#33). Each frame covers a field of view of 3.8' square. The characteristics of the instrumentation are described by Schwartz et al. (1995). A total of 28, 29 and 30 $V$ frames were obtained for HV 1777, HV 1779 and HV 1353, respectively, with a typical exposure time of 5 minutes. During the last night some $R$ frames were obtained besides the $V$ and $R$ frames at different airmasses of the two standard CCD fields Rubin 149A and SA 98-650 (Landolt, \\cite{lan}) in order to derive standard $V$ and $R$ magnitudes. The frames were reduced in the usual way by means of IRAF packages and using sky flat fields obtained both at sunset and dawn. Measurements were then performed by means of the aperture photometry package APPHOT.  From the observed colours of the 13 standard stars in the two fields we obtained  transformation equations which allow to fit the standard colours with rms scatters of 0.009 and 0.007 mag in $V$ and $R$ respectively. Since our aim was to perform differential photometry between our Cepheids and some suitable comparison stars, and to detect other possible unknown Cepheids in the fields, all the brightest objects were measured, that is almost all the stars with $V \\la 17.5$. A total of 19, 9 and 21 stars in the fields of HV 1777, HV 1779 and HV 1353, respectively, were measured. The identification maps are reported in Figures 1, 2, 3; each side is 3.8'. \\begin{figure} \\epsfysize=8truecm \\epsffile{p3_fig1.ps} \\caption[ ]{Field of HV 1353. Circles indicate the measured stars. \"Var\" is the Cepheid while \"C\" is the star adopted for computing differential magnitudes. North is up and Right Ascension increases towards right. } \\end{figure} \\begin{figure} \\epsfysize=8truecm \\epsffile{p3_fig2.ps} \\caption[ ]{Field of HV 1777. Symbols as in previous figure. \"Var b\" is HV 1763. } \\end{figure} \\begin{figure} \\epsfysize=8truecm \\epsffile{p3_fig3.ps} \\caption[ ]{Field of HV 1779. Symbols as in Fig. 1} \\end{figure}  The three panels of Fig. 4 show for each field the standard deviations of the differential magnitudes with respect to the brightest star versus the standard $V$ magnitude. This figure allows to evaluate the intrinsic accuracy of the measurements. Apart from the three Cepheids there is also a strongly deviating object in the field of HV 1777. It corresponds to the variable HV 1763, detected by Leavitt (\\cite {lea}), but with unknown period. \\begin{figure} \\epsfysize=9truecm \\epsffile{p3_fig4.ps} \\caption[ ]{Standard deviations of the magnitude differences between measured objects and the brightest one for each of the three investigated fields} \\end{figure} Table 1 contains times and $V$ magnitudes of the 4 variable stars.  \\begin{table} \\caption[]{Period and relevant data of observed Cepheids} \\begin{flushleft} \\begin{tabular}{llllr} \\hline Star & Period & $ <V> $ &$A_V$ & Maximum \\\\ &   [d] &    [mag] & [mag] &[Hel.J.D.]\\\\ \\hline HV 1779 & 1.784 & 16.09 & 0.42  &2450375.797\\\\ HV 1763 & 2.117 & 16.18 & 0.37  & 75.227\\\\ HV 1777 & 2.515 & 15.99 &  0.35 & 75.476\\\\ HV 1353 & 3.232 & 16.31 & 0.40 & 76.573 \\\\ \\hline \\end{tabular} \\end{flushleft} \\end{table} ",
        "conclusions": "Second overtone mode Cepheids could be discriminated from fundamental and first overtone mode Cepheids by taking into account their period, luminosity, low amplitude  and Fourier parameters (Antonello \\& Kanbur \\cite{ak}; Alcock et al. \\cite{a2}). However, none of the stars analyzed in the present study, and which are characterized by low amplitude, can be discriminated from the other known Cepheid types on the basis of the light curve shape. Only HV 1779 is relatively bright for its period, but this is not a sufficient criterium, since the relatively large luminosity could be explained by other reasons (e.g. a companion or a background star). We conclude that the three suspected second overtone candidates are fundamental (HV 1353) and first overtone (HV 1777, HV 1779) mode pulsators. HV 1763, whose nature was previously unknown, resulted to be a short--period  Cepheid pulsating in the first overtone mode."
    },
    "9805/astro-ph9805251_arXiv.txt": {
        "abstract": "The magnetic field of the CP star HD~14437 was discovered by Glagolevskij et al. (1985) using the 6-m telescope of the Special Astrophysical Observatory. No polarity changes have been found during 2 years of observations. A very long ($>$~3--4 years) period of rotation was proposed to explain the measurements. To check this hypothesis, we made a new series of magnetic field observations for this star 10 years later with the 6-m telescope. The polarity of the longitudinal magnetic field is still negative and has not shown any change during more than 10 years of observations. We found the most probable periods to be several days. It means that we have observed a star where the magnetic and rotation axes are inclined at a small angle, and the negative pole is not far from the line of sight. ",
        "introduction": "HD~14437 is a poorly studied peculiar A--star with a magnitude of $7^m.4$, effective temperature $T_{\\rm eff}=10800$ (Glagolevskij, 1995). Observations of the 1980s with the 6-m telescope showed a variability of the longitudinal magnetic field with an amplitude of about 2 kG and a constant negative sign. The description of the observations and data reductions are presented by Glagolevskij et al. (1985) in more details.  In this paper we present new observations of HD~14437 which were carried out in 1996--97 using the CCD detector (Chuntonov \\& Glagolevskij, 1997). We have observed the spectra of the star on the 6-m telescope with a spectral resolution $R=30000$. The context NICE (Knyazev \\& Shergin, 1995) in the MIDAS system was used. The new and old observational data are presented in Table~\\ref{observations} (JD~2444655--2445900: with photographic plates, JD~2449555 and JD~2449556: with the hydrogen lines magnetometer, from JD~2450000 on: with the CCD detector).  We have carried out the line identification using the Vienna Astrophysical Line Data--Base (VALD) (Piskunov et al., 1995) and Moore's tables (Moore, 1945). In view of determining the individual Land\\'e factor of each spectral line used in our measurements, we have identified the lines in the following spectral ranges: $\\lambda$5955--6385 and $\\lambda$4460--4680. Si and Cr are observed to be overabundant. We have found an underabundance of O by the lack of the lines OI~$\\lambda$6155.97, 6556.78 and 6158.18 which are usually strong at spectral class A2.  We have determined the effective magnetic field using Babcock's standard formula with the individual Land\\'e factor~$z$ for each line. The measurements are given in Table~\\ref{observations}.  \\begin{table}[t] \\small \\begin{center} \\caption{The observations of the effective magnetic field.} \\label{observations} \\begin{tabular}{lr|lr} \\hline\\hline JD 2400000&$B_{\\rm e}\\pm\\sigma$,G&JD 2400000&$B_{\\rm e}\\pm\\sigma$,G\\\\ \\hline 44655.208&$-1620\\pm170$& 50415.240&$-1940\\pm180$\\\\ 44656.308&$-1030\\pm170$& 50415.269&$-2570\\pm180$\\\\ 44659.188&$-440\\pm170$&  50499.196&$-2650\\pm260$\\\\ 44660.196&$-800\\pm140$&  50499.221&$-2600\\pm160$\\\\ 44860.541&$-2280\\pm110$& 50500.161&$-2460\\pm220$\\\\ 45303.354&$-1230\\pm120$& 50500.188&$-2050\\pm250$\\\\ 45303.362&$-1020\\pm110$& 50617.507&$-1820\\pm260$\\\\ 45476.541&$-2080\\pm280$& 50643.521&$-1360\\pm110$\\\\ 45900.493&$-2040\\pm130$& 50705.451&$-1330\\pm210$\\\\ 49555.472&$-1340\\pm380$*&50706.410&$-960\\pm110$\\\\ 49556.420&$-1710\\pm460$*&50707.412&$-830\\pm100$\\\\ 50413.523&$-2040\\pm300$& 50709.602&$-1370\\pm110$\\\\ 50414.125&$-2310\\pm280$& 50710.432&$-2110\\pm120$\\\\ 50415.217&$-2460\\pm270$&&\\\\ \\hline\\hline \\multicolumn{4}{l}{* --- observations with the magnetometer} \\end{tabular} \\end{center} \\end{table}  The effective magnetic field has preserved its sign for about ten years. The suggestion about the long period of the star contradicts our data, because we have observed the longitudinal component of the magnetic field to undergo significant changes within a few days only. So, we can draw the conclusion that the axis of rotation, the magnetic axis and the line of sight are inclined at small angles. ",
        "conclusions": ""
    },
    "9805/hep-th9805100_arXiv.txt": {
        "abstract": "The randomly driven Navier-Stokes equation without pressure in $d$-dimensional space is considered as a model of strong turbulence in a compressible fluid. We derive a closed equation for the velocity-gradient probability density function. We find the asymptotics of this function for the case of the gradient velocity field (Burgers turbulence), and provide a numerical solution for the two-dimensional case. Application of these results to the velocity-difference probability density function is discussed.    ~\\\\ \\noindent PACS Number(s): 47.27.Gs, 03.40.Kf, 52.35.Ra. ",
        "introduction": "The Burgers equation with a random external force is considered to be the first exactly solvable model of $1d$~turbulence and has been extensively studied in recent years \\cite{Polyakov,Y-Ch,Boldyrev,Boldyrev2,Bouchaud,Sinai,G-M,Bouchaud1,Balkovsky,Gotoh,Ivashkevich}. Though rather simplified, this model can serve as a test model for some general ideas within the theory of strong turbulence. In~1995, methods of quantum field theory were applied to this problem by A.~Polyakov~\\cite{Polyakov} which enabled a qualitative explanation of velocity-difference probability density functions (PDFs) measured numerically by A.~Chekhlov and V.~Yakhot~\\cite{Y-Ch}. In~\\cite{Boldyrev} it was shown that the approach~\\cite{Polyakov} allows one to obtain quantitatively correct results. Extensive numerical simulations published recently by T.~Gotoh and R.~Kraichnan~\\cite{Gotoh} show that the predictions of~\\cite{Polyakov,Boldyrev} are quite accurate, and coincide with the numerical simulations to within about~5$\\%$. V.~Yakhot has shown in~\\cite{Yakhot} that the ideas introduced in~\\cite{Polyakov} can have much wider application, and can also work for incompressible velocity fluctuations.  We believe that the operator product expansion (OPE), introduced in~\\cite{Polyakov} to take into account the viscous term, is an adequate language to treat compressible turbulence in higher dimensions as well, where shock structures and associated local dissipation persist. In the present paper we find a closed equation for the velocity-gradient probability-density function (PDF) for compressible turbulence in any number of dimensions. We investigate the asymptotics of the PDF and present the numerical solution for the~$2d$ case.   The basic equation we will study is the following:  \\begin{eqnarray} {\\bf u}_t + ({\\bf u}\\cdot \\nabla){\\bf u} = \\nu \\Delta {\\bf u} + {\\bf f}\\,\\,. \\label{Eq.1} \\end{eqnarray}  The force~$f$ is chosen to be Gaussian with zero mean and white in time variance: \\begin{eqnarray} \\langle f^i({\\bf x}, t)f^k({\\bf x}^{\\prime}, t^{\\prime}) \\rangle = \\delta(t-t^{\\prime}) \\kappa^{ik}({\\bf x}-{\\bf x}^{\\prime})\\,\\,, \\label{Eq.0} \\end{eqnarray}  \\noindent where the $\\kappa$~function is concentrated at some large scale~$L$, and can be expanded as follows  \\begin{eqnarray} \\kappa^{ik}({\\bf y})=\\kappa_0\\delta^{ik}-\\kappa_1\\left(y^2\\delta^{ik}+2\\alpha y^iy^k\\right) \\label{force} \\end{eqnarray} \\noindent for $y \\ll L$. We assume that the steady states for velocity gradient and velocity difference exist; for this we can require, for example, that periodic boundary conditions on a scale much larger that~$L$ be imposed, and that the zero harmonic in the $\\kappa$~function be absent. These assumptions are usually used in numerical simulations \\cite{Y-Ch,Gotoh}.   In this paper we appeal to the results obtained for the~$1d$ Burgers turbulence without pressure in~\\cite{Polyakov,Y-Ch,Boldyrev,Boldyrev2,Sinai,Gotoh}. In particular, we are interested in the velocity-gradient~PDF $P(\\partial u^i/\\partial x^k)$ and the velocity-difference~PDF $P_v({\\bf u}({\\bf x}_1)-{\\bf u}({\\bf x}_2))$, where the velocities are taken at the same time at some fixed points~${\\bf x}_1$ and~${\\bf x}_2$. The physical picture presented in these papers allows us to consider a general phenomenon such as intermittency on a rigorous basis; it is related to the spontaneous breakdown of the Galilean invariance of the forced equation and to the algebraic decay of the PDFs. We will not repeat these arguments here; instead, we will concentrate on the main ideas which allow us to consider the multi-dimensional case.  We will be interested in the case of small dissipation $\\nu$, and will consider distances $|x_1-x_2|\\ll L$. The following order of the limits should be considered to get the steady state: we first set $t\\rightarrow \\infty$, and then consider the limit $\\nu \\rightarrow 0$. ",
        "conclusions": "The crucial assumption in our treatment of the dissipative anomaly is the assumption that only smooth parts of the velocity field contribute to the velocity-{\\em difference}~PDF. After the velocity-gradient $Z$~function~(\\ref{Zfunction}) is found, the velocity-difference $Z$~function can be constructed as follows:  \\begin{eqnarray} Z_v(\\zeta_i, y^k)=Z(\\zeta_i y^k)\\equiv\\langle \\exp(i\\zeta_i y^ku^i_k) \\rangle \\,\\,\\,, \\label{Zvfunction} \\end{eqnarray}  \\noindent i.e., we simply changed $\\sigma_{ik}\\rightarrow\\zeta_i y_k$ in~(\\ref{Zfunction}). The Fourier transform with respect to $\\zeta$ will then give the velocity-difference~PDF. As an example, let us consider the longitudinal velocity-difference PDF:  \\begin{eqnarray} & P_l(\\Delta u, y)\\propto \\nonumber \\\\ &  \\int \\mbox{d}u_{ik}\\mbox{d}\\zeta P(\\lambda)\\exp \\{ i\\zeta y n_i n_k u_{ik}-i\\zeta \\Delta u \\}, \\label{P_l-definition} \\end{eqnarray} \\noindent where~$n_i$ is a unit vector in the direction of~$y_i$. Since~$P_l(\\Delta u, y)$ does not depend on~$n_i$, one can average with respect to all possible directions of this vector and get the following result:  \\begin{eqnarray} \\label{P_l-integral} & P_l(\\Delta u, y)\\propto \\nonumber \\\\ & \\frac{1}{y}\\int P(\\lambda)|\\Delta(\\lambda)| \\delta \\left( \\lambda_i n_i^2 - \\frac{\\Delta u}{y} \\right )\\delta (1-{\\bf n}^2)\\prod\\limits^d_{i=1} \\mbox{d}\\lambda_i \\mbox{d} n_i . \\end{eqnarray}   In the two-dimensional case this integral can be simplified further:  \\begin{eqnarray} P_l(\\Delta u, y) \\propto \\frac{1}{y} \\int\\limits^{\\Delta u/y}_{-\\infty } \\int\\limits^{\\infty }_{\\Delta u/y} \\frac{\\mbox{d} \\lambda_1 \\mbox{d} \\lambda_2 \\vert \\lambda_1-\\lambda_2 \\vert P(\\lambda)} {\\left[ \\left( \\frac{\\Delta u}{y}-\\lambda_1 \\right)\\left(\\lambda_2 - \\frac{\\Delta u}{y} \\right)\\right]^{1/2} } \\end{eqnarray}     \\noindent In general,~$P_l(\\Delta u, y)$ can be represented as~$P_l(\\Delta u, y)=w(\\Delta u/y)/y$.  If we assume that for large negative~$\\Delta u/y $ the integral~(\\ref{P_l-integral}) is contributed to by the force-free asymptotic~(\\ref{left}), we immediately get the left tail asymptotic for the longitudinal velocity-difference PDF:  \\begin{eqnarray} w(z)\\propto z^{-(d+1)(d+2)/2}, \\,\\,\\, z \\to -\\infty. \\end{eqnarray} Analogously, one can obtain a PDF for $\\nabla \\cdot u$. For this purpose one should set $\\sigma_{ik}\\rightarrow\\delta_{ik}\\zeta$. Such a PDF was investigated numerically in~\\cite{Gotoh1}, though the~$Re$ number was not large enough to obtain the inertial range.  Finally, we would like to note that the absence of the $\\beta$~anomaly, that we assumed in our consideration, can be not a universal fact. It was conjectured in~\\cite{Boldyrev} that different dissipative regularizations (e.g. hyper-dissipation $(-1)^p\\partial^{2p}/\\partial x^{2p}$) can lead to different steady states. This assumption is natural for the language of the OPE: different dissipative operators should have different expansion coefficients~$a$ and~$b$ (we use the notation of \\cite{Polyakov}). Moreover, some analog of the $\\beta$ anomaly can also be present in Eq.~(\\ref{FP1}), since it describes a general velocity field, without ``gradient\" restriction~(\\ref{factorization}).    These questions are under consideration, the results will be reported elsewhere.   \\vskip5mm   I am indebted to A.~Polyakov and V.~Yakhot for many important discussions and comments. I would also like to thank T.~Gotoh, V.~Gurarie, and R.~Kraichnan for useful conversations, D.~Uzdensky for helpful discussions on both the physics and the numerics of the problem, and T.~Munsat for valuable remarks on the style of the paper.   This work was supported by U.S.D.o.E. Contract No. DE--AC02--76--CHO--3073.   \\vskip5mm"
    },
    "9805/astro-ph9805082_arXiv.txt": {
        "abstract": "We present the results of Smoothed Particle Hydrodynamics (SPH) simulations of the formation of a massive counterrotating disk in a spiral galaxy.  The current study revisits and extends (with SPH) previous work carried out with sticky particle gas dynamics, in which adiabatic gas infall and a retrograde gas-rich dwarf merger were tested as the two most likely processes for producing such a counterrotating disk.  We report on experiments with a cold primary similar to our Galaxy, as well as a hot, compact primary modeled after NGC~4138.  We have also conducted numerical experiments with varying amounts of prograde gas in the primary disk, and an alternative infall model (a spherical shell with retrograde angular momentum).  The structure of the resulting counterrotating disks is dramatically different with SPH.  The disks we produce are considerably thinner than the primary disks and those produced with sticky particles.  The time-scales for counterrotating disk formation are shorter with SPH because the gas loses kinetic energy and angular momentum more rapidly.  Spiral structure is evident in most of the disks, but an exponential radial profile is not a natural byproduct of these processes.  The infalling gas shells that we tested produce counterrotating bulges and rings rather than disks.  The presence of a considerable amount of preexisting prograde gas in the primary causes, at least in the absence of star formation, a rapid inflow of gas to the center and a subsequent hole in the counterrotating disk.  For a normal counterrotating disk to form, there must either be little or no preexisting prograde gas in the primary, or its dissipative influence must be offset by significant star formation activity. The latter scenario, along with the associated feedback to the ISM, may be necessary to produce a counterrotating disk similar in scale length and scale height to the primary disk.  In general, our SPH experiments yield stronger evidence to suggest that the accretion of massive counterrotating disks drives the evolution of the host galaxies towards earlier (S0/Sa) Hubble types. ",
        "introduction": "There are only a handful of known cases of {\\em massive} counterrotating disks in spiral galaxies to date, and yet counterrotation in spirals cannot be deemed a rare phenomenon by any standards.  There are hints that it may be quite common, in fact, in early-type spirals, particularly S0s (\\cite{kfm96}). The origin of any counterrotating mass (gas or stars) within a spiral disk is an important unsolved problem with profound implications for the formation and evolution of all spiral galaxies, but the existence of a significant retrograde mass component (comprising anywhere from $\\sim$10-50\\% of the total mass of the disk system) is a particularly intriguing question that threatens to radically alter our view of the evolution of spiral galaxies. The rogues' gallery of spirals with such massive counterrotating disks currently boasts the following members: NGC~4550 (\\cite{rgk92}), NGC~7217 (\\cite{mk94}), NGC~4826 (\\cite{bwk92}), NGC~3626 (\\cite{cbg95}), NGC~3593 (\\cite{bcc96}), and NGC~4138 (\\cite{jbh96}).  Apart from the challenge they present to the traditional view of spiral galaxies that has evolved over the last few decades, counterrotating disks raise several questions about other astrophysical processes, such as the role and fate of gas in galaxy interactions, star formation in galaxies that contain counterrotating gas, the accretion rates and star formation histories of spiral galaxies in general, and the impact of counterrotating populations on the overall stability of the disk system.  Even though a recent survey of S0s (\\cite{kfm96}) found counterrotating gas in almost a quarter of the sample, none of these galaxies have counterrotating stars.  Why is the counterrotating gas not forming stars?  If star formation is inhibited in counterrotating disks, how does one explain NGC~4550 and others with stellar counterrotating disks?  It is very unlikely that counterrotating systems can be produced indigenously or as a byproduct of the galaxy formation process.  The theory of formation of a spiral galaxy from a spinning protogalactic cloud does not admit the possibility of bidirectional spin being imparted to the disk system. Subsequent accretion or merger events are a much more plausible explanation, and even these are severely constrained by the observed coldness of the counterrotating galaxies.  Dissipationless mergers, especially between progenitors with comparable masses, can be ruled out in the general case, although \\cite{pf98} has recently been able to produce a remnant resembling NGC~4550 with a collisionless merger of two spirals with special initial conditions.  This leaves minor, gas-rich mergers, and gas accretion or infall, as the most promising candidates.  The most puzzling aspect of massive counterrotating disks in spiral galaxies is that the host galaxies appear quite normal in every other respect, and there is no evidence of excessive thickening of the primordial disks due to the accretion of the counterrotating disk.  This suggests that the accretion process must not be a rapid or violent one, but there may be deeper implications here for the interaction histories of all spiral galaxies.  A recent study of the tidal thickening of galaxy disks (\\cite{rc96}) indicates that the ratio of scale length to scale height, $h/z_{\\circ}$, is 1.5-2 times lower for interacting disks.  But if any galaxy can be assumed to have undergone an interaction in the past, the thinness of the non-interacting sample proves that this ratio returns to its higher value after a certain amount of time (order of 1 Gyr).  If it can be proved that a spiral can double its disk mass without destroying the primordial disk in the process, then the present-day appearance of spiral disks can no longer preclude such interactions in their past.  The claim that the thinness of spiral disks places a stringent limit on past accretion (\\cite{to92}) is further challenged by recent simulations showing that the halo absorbs a good portion of the orbital energy and angular momentum of the satellite in spiral-dwarf mergers (\\cite{wmh96}; \\cite{hc97}), and analytical results indicating that an isothermal halo may even shield the disk from an external tidal field due to a satellite (\\cite{mt97}).  A study of the lopsidedness of the disks of field spirals suggests that the accretion rate of spiral galaxies may be as high as one small ($\\sim$10\\% mass) companion every 4 Gyr (\\cite{zr97}).  This is still a very uncertain estimate, and a more accurate estimate of the accretion rate, and a knowledge of how much accretion a disk galaxy can withstand, are important questions from a cosmological point of view.  Observations of counterrotating disks combined with a good understanding of how they form provide a striking new way to independently constrain these estimates, because the accreted matter can be easily distinguished in a counterrotating galaxy.  With the aim of understanding the origin of massive counterrotating disks in spirals and S0s, we have developed a numerical code and run hydrodynamical simulations to investigate the processes that are most likely to produce such bizarre systems.  We have combined an N-body gravity solver with a gas dynamics particle code for this purpose.  To study the basic parameters of the processes involved, we first adopted a quick-and-dirty ``sticky particle'' gas dynamics approach that allowed us to test various scenarios with relatively small investments of CPU time.  In this way we were able to test massive counterrotating disk formation in a fiducial cold primary (\\cite{tr96}, hereafter TR) as well as model the formation of the recently discovered counterrotating disk in the early-type spiral NGC~4138 (\\cite{tr97}, hereafter TRJB).  For each type of primary modeled, we have tested two theories of origin: adiabatic (secular) gas infall and a gas-rich dwarf merger.  In order to produce a disk with opposite spin, both of these processes require a retrograde orbit of accretion for the infalling gas or satellite galaxy with respect to the primary's spin.  Gas infall works well for both late and early-type primaries, but a dwarf merger, especially with a substantial amount of dissipationless matter in the dwarf, is not viable for a cold primary because it plays havoc with the primary's disk.  An accurate representation of the gas in astrophysical systems typically requires that the hydrodynamical conservation equations be solved and the effects of physical processes such as shocks and viscosity be included.  This requires the gas to be modeled as a fluid, but a good compromise can be achieved with a particulate representation if each gas particle is smeared over a finite volume and its physical properties averaged or smoothed over that volume.  This Smoothed Particle Hydrodynamics (SPH) approach (\\cite{lu77}; \\cite{gm77}; \\cite{mo92}) meshes well with an N-body particle code, but is much more costly in terms of CPU resources than the sticky particle approach.  We therefore reserved it for a more detailed and restricted look at the structure of the counterrotating disk formed by reexamining (with SPH) a carefully chosen subset of the simulations we presented in TR and TRJB.  We present the results of these simulations here. Although most of the simulations are reruns of those presented before with sticky particles, a couple of new runs are also discussed.  Our SPH code currently does not include star formation.  We intend to incorporate this in our simulations in the future, but as we discuss in \\S{4} below, we do not expect it to have a profound impact on the structure of the counterrotating disks formed.  Our computational method is described in the following section.  Results of simulations are presented in \\S{3}, followed by discussion and conclusions in \\S{4} and \\S{5} respectively. ",
        "conclusions": "There are some notable differences between the characteristics of the counterrotating disks resulting from SPH and those obtained with sticky particle simulations by TR and TRJB.  The SPH disks are very thin compared to their sticky particle counterparts and compared to the primary disks.  They also show evidence of spiral structure, and their size and radial mass distribution are quite sensitive to the input parameters, particularly those that affect the initial angular momentum of the gas.  Other differences in the SPH results include the lack of clumping of the infalling gas, a problem that was quite severe with our sticky particle simulations, and the shorter time-scales for disk formation.  Although it is easy to produce thin counterrotating disks with gas infall, it is not so easy to obtain exponential radial profiles.  The initial angular momentum of the gas has to be low enough, and some combination of other processes such as prograde gas, star formation and energy feedback from massive stars may be necessary to produce counterrotating disks that are very similar to the primary disks.  Currently there is no evidence to indicate that counterrotating disks have predominantly exponential profiles, so this is not necessarily a problem.  In general, the process that dumps a massive counterrotating disk in a cold primary spiral, especially if it is a minor gas-rich merger but even if it is gas infall that occurs over a few dynamical times ($\\lesssim10 t_{\\rm dyn}$), is likely to heat up the primary substantially and change its type to an S0/Sa galaxy.  On the other hand, if the primary is already an S0/Sa galaxy to begin with, then it can acquire a counterrotating disk without changing its type significantly.  The fact that most of the currently known instances of massive spiral counterrotating disks are in S0/Sa galaxies is therefore a selection effect rather than an accident.  The presence of primordial prograde gas in the primary has a drastic effect on the retrograde gas that comes in contact with it.  Neutralization of the angular momenta is rapid, with both the prograde and retrograde gas particles ending up in the center of the primary within a few dynamical times.  This may be an indication of a problem with SPH that causes over-dissipation in counterstreaming gas flows, at least in the absence of star formation.  The inclusion of star formation and energy feedback from supernovae will most likely yield significantly different results in such situations.  A retrograde-rotating, infalling gas shell produces a counterrotating bulge and flat outer ring, but is unable to produce a counterrotating disk in the proper sense.  The size of the ring is well correlated with the angular momentum of the shell.  The formation of the ring is consistent with previous studies of collapsing isothermal gas clouds.  These studies also suggest that significantly hotter gas with lower angular momentum is necessary to produce a counterrotating disk with this model.  We hope to test our results further in the near future with the addition of thermal effects and star formation to our SPH code."
    },
    "9805/astro-ph9805177_arXiv.txt": {
        "abstract": "We present and discuss visible-wavelength long-slit spectra of four low redshift 3C galaxies obtained with the STIS instrument on the Hubble Space Telescope. The slit was aligned with near-nuclear jet-like structure seen in HST images of the galaxies, to give unprecedented spatial resolution of the galaxy inner regions. In 3C 135 and 3C 171, the spectra reveal clumpy emission line structures that indicate outward motions of a few hundred km s$^{-1}$ within a centrally illuminated and ionised biconical region. There may also be some low-ionisation high-velocity material associated with 3C 135. In 3C 264 and 3C 78, the jets have blue featureless spectra consistent with their proposed synchrotron origin. There is weak associated line emission in the innermost part of the jets with mild outflow velocity. These jets are bright and highly collimated only within a circumnuclear region of lower galaxy luminosity, which is not dusty. We discuss the origins of these central regions and their connection with relativistic jets. ",
        "introduction": "The HST snapshot program to obtain broad-band images of 3C radio sources revealed much of the small-scale optical structure of these extraordinary objects (De Koff et al 1996). We noticed particularly that a number of them have small bright structures which are approximately radially oriented from the nuclei. These jet-like phenomena lie within an arcsec of the nucleus, are usually obvious only on one side, and are approximately aligned with the large-scale radio structure. The original snapshots are only in one colour (F702W filter) so little more information was initially available. Since then, many of the objects have been observed (by SAB and collaborators) with the WFPC2 ramp filters to isolate the emission lines of [O III] or H$\\alpha$.  There are several possible origins of these `jets': they may be true synchrotron jets like 3C 273 or M87; extended line emission regions activated by the AGN like NGC 4151; or line emission from gas flows or star-formation triggered by radio jets. In order to distinguish these possibilities, and learn more about the nuclei of radio galaxies, we are carrying out a program of long-slit observations with the Space Telescope Imaging Spectrograph of the Hubble Space Telescope.  The observations consist of low dispersion optical spectra taken with a fairly wide (0.5 arcsec) slit placed along the direction of the jets as seen in the HST snapshots. The slit length is the full $\\sim$50 arcsec field of the CCD, which greatly exceeds the size of the central regions of interest. The width of the slit means that the emission line resolution is determined by the spectrograph and the size of features in the galaxies. It also means that the spectra are essentially slitless for the inner emission-line region, and thus similar to the NGC 4151 slitless spectra of Hutchings et al (1998).  Table 1 gives the journal of observations and basic information on the first four program objects. Exposure details were different for each, depending on the redshift and the flux levels measured from the snapshot images. The slit angle was positioned to within 5$^o$ of the `jet' and an acquisition image used to confirm the orientation.  The STIS pixel size in the visible wavelength region is 0.051 arcsec, and the spectral resolution is $\\sim$7\\AA~ for G450L and $\\sim$14\\AA~ for G750L for a point source. Each pixel samples 2.7 and 4.9\\AA ~respectively, in dispersion. All spectra were taken with several readouts, to eliminate cosmic rays. `Superdark' frames for the days of observation were derived and used to remove the hot pixels that are a significant noise source in these faint targets. Wavelength calibration exposures were taken with each observation and applied using CALSTIS routines in STSDAS and the STIS team IDL software. Current flux calibrations were applied. Using the orientation from the STIS data, the WFPC2 snapshots were rotated and resampled to match the STIS images, and used as undispersed templates to measure differential velocities of emission line regions within the jets. This is described more fully by Hutchings et al (1998).  The sections below describe each of the sources individually, as they are all significantly different. The measurements are described in more detail in the first object only, to avoid repetition. ",
        "conclusions": ""
    },
    "9805/astro-ph9805341_arXiv.txt": {
        "abstract": "We have calculated synchrotron spectra of relativistic blast waves, and find predicted characteristic frequencies that are more than an order of magnitude different from previous calculations.  For the case of an adiabatically expanding blast wave, which is applicable to observed gamma-ray burst (GRB) afterglows at late times, we give expressions to infer the physical properties of the afterglow from the measured spectral features.  We show that enough data exist for GRB\\,970508 to compute unambiguously the ambient density, $n=0.03\\un{cm}{-3}$, and the blast wave energy per unit solid angle, ${\\cal E}=3\\times10^{52}\\un{erg}{}/4\\pi\\un{sr}{}$.  We also compute the energy density in electrons and magnetic field. We find that they are 12\\% and 9\\%, respectively, of the nucleon energy density and thus confirm for the first time that both are close to but below equipartition.  For GRB\\,971214, we discuss the break found in its spectrum by Ramaprakash et al. (1998)\\nocite{rkfkk:98}. It can be interpreted either as the peak frequency or as the cooling frequency; both interpretations have some problems, but on balance the break is more likely to be the cooling frequency. Even when we assume this, our ignorance of the self-absorption frequency and presence or absence of beaming make it impossible to constrain the physical parameters of GRB\\,971214 very well. ",
        "introduction": "\\label{intro}  Explosive models of gamma-ray bursts, in which relativistic ejecta radiate away some of their kinetic energy as they are slowed down by swept-up material, naturally lead to a gradual softening of the emission at late times. This late-time softer radiation has been dubbed the `afterglow' of the burst, and its strength and time dependence were predicted theoretically (\\Mesz\\ and Rees, 1997)\\nocite{mr:97}. Soon after this prediction, the accurate location of GRB\\,970228 by the BeppoSAX satellite's Wide Field Cameras (Piro et al.\\ 1995, Jager et al.\\ 1995)\\nocite{psb:95,jhzb:95} enabled the detection of the first X-ray and optical afterglow (Costa et al.\\ 1997, Van Paradijs et al.\\ 1997)\\nocite{cfhfz:97,pggks:97}. Its behavior agreed well with the simple predictions (Wijers et al.\\ 1997, Waxman 1997a, Reichart 1997)\\nocite{wrm:97,waxma:97,reich:97}.  The basic model is a point explosion with an energy of order $10^{52}\\un{erg}{}$, which expands with high Lorentz factor into its surroundings. As the mass swept up by the explosion begins to be significant, it converts its kinetic energy to heat in a strong shock. The hot, shocked matter acquires embedded magnetic fields and accelerated electrons, which then produce the radiation we see via synchrotron emission. The phenomenon is thus very much the relativistic analogue of supernova remnant evolution, played out apparently in seconds due to the strong time contractions resulting from the high Lorentz factors involved.  Naturally, the Lorentz factor of the blast wave decreases as more matter is swept up, and consequently the power output and typical energy decrease with time after the initial few seconds of gamma-ray emission. This produces the X-ray afterglows, which have been detected up to 10 days after the burst (Frontera et al. 1998)\\nocite{fgacf:98}, and the optical ones, which have been detected up to a year after the burst (Fruchter et al.\\ 1997, Bloom et al.\\ 1998a, Castro-Tirado et al.\\ 1998)\\nocite{flmps:97,bdkf:98,cgggp:98}.  The burst of May 8, 1997, was bright for a relatively long time and produced emission from gamma rays to radio. This enabled a detailed analysis of the expected spectral features of a synchrotron spectrum, confirming in great detail that we are indeed seeing synchrotron emission, and that the dynamical evolution of the expanding blast wave agrees with predictions if the blast wave dynamics are adiabatic (Galama et al.\\ 1998a,b)\\nocite{gwbgs:98,gwbgs2:98}. In principle, one can derive the blast wave properties from the observed synchrotron spectral features. The problem is that the characteristic synchrotron frequencies and fluxes are taken from simple dimensional analysis in the published literature, so they are not suitable for detailed data analysis. Since there are now enough data on the afterglows of a few GRBs to derive their physical properties, we amend this situation in \\Sect{equa}, correcting the coefficients in the equations for the break frequencies by up to a factor 10. We then use our theoretical results to infer the physical properties of the afterglows of GRB\\,970508 (\\Sect{970508}) and attempt the same for GRB\\,971214 (\\Sect{971214}). We conclude with a summary of results and discuss some prospects for future improvements in observation and analysis (\\Sect{conclu}). ",
        "conclusions": "\\label{conclu}  We have calculated the synchrotron spectra from the blast waves causing GRB afterglows and derive improved expressions for the relations between measured break frequencies and the intrinsic properties of the blast wave. These allow us to relate the blast wave properties to observable quantities more accurately. We correct the expression for the blast wave energy by almost two orders of magnitude. Our expressions are exact for an undecelerated, uniform medium.  Deceleration and radial structure of the shock are expected to change the expressions for the final parameters by another factor few, much less than the corrections found here but still of interest. Combined with the uncertainties in the measured values of the spectral breaks and fluxes, this means that the blast wave parameters derived here are still uncertain by an order of magnitude (see the solution by Granot, Piran, and Sari 1998\\nocite{gps:98} as an illustration of the possible differences).  There is enough data on GRB\\,970508 to compute all intrinsic parameters of the blast wave. The energy in the blast wave is $3\\times10^{52}\\un{erg}{}/4\\pi\\un{sr}{}$. The ambient density into which the blast wave expands is 0.03$\\un{cm}{-3}$, on the low side for a disk of a galaxy.  The fraction of post-shock energy that goes into electrons is 12\\%, and that into magnetic field, 9\\%.  We also estimate the fraction of energy transferred to electrons during the gamma-ray phase, and find this to be 15\\%. The agreement with the later blast wave value suggests that the fraction of energy given to electrons is constant from 10\\,s to $10^6$\\,s after the trigger.  For GRB\\,971214 there is ambiguity in the interpretation of the break seen in the optical half a day after the burst. We argue that the break is most likely to be the cooling break, but the argument is not watertight. Assuming it is the cooling break, we still lack the self-absorption frequency, but radio limits constrain this to be in excess of about 5\\,GHz.  The limits on parameters that follow from this indicate that the afterglow properties of GRB\\,971214 are different from those of GRB\\,970508. GRB\\,971214 must either have had more narrow beaming in gamma rays than in optical or have radiated its initial energy with more than 99\\% efficiency in the gamma-ray phase, according to the parameters we derive. Also its magnetic field was far below equipartition, $\\epsB\\lsim10^{-4}$. However, beaming or other additions could ease the constraints, and allow parameter values similar to those of GRB\\,970508, so the physical parameters of GRB\\,971214 are very poorly constrained.  Our analysis emphasizes the importance of early measurements covering a wide range of wavelengths. The full identification of the cooling frequency $\\nuc$ in GRB\\,970508 hinged on abundant photometry, including colors, being available soon after the burst, since the break passed $R$ after 1.5 days (Galama et~al. 1998b\\nocite{gwbgs2:98}). In $H$ and $K$, the action lasted a week (Galama et~al. 1998b\\nocite{gwbgs2:98}), and this is the general trend: there is more time in IR, since all breaks pass later there. However, our revised coefficient for the peak frequency, $\\num$, shows that the peak can only be caught in the IR within hours of the trigger (or much later in the radio).  A case in point are the very early $K'$ band measurements of GRB\\,971214 by Gorosabel et al.\\ (1998)\\nocite{gcwhg:98}, which provide an invaluable constraint on this afterglow as they may have caught the passage of $\\num$ through $K'$.  Therefore, we encourage first and foremost early long-wavelength coverage, including searches for afterglows in IR, as a method of effectively constraining afterglow parameters. Two of the three crucial break frequencies in an afterglow can pass the optical and IR within hours and days, respectively.  There is no time to first search and only then attempt broad coverage.  Instantaneous alerts from HETE2 and SWIFT will therefore greatly advance our understanding of afterglow physics. For HETE2, and to some extent for SWIFT, this will require an enormous amount of work from a network of ground-based observatories with good coverage in longitude and latitude, so that always at least one observatory is well-placed for immediate response."
    },
    "9805/astro-ph9805031_arXiv.txt": {
        "abstract": "For the first time we have performed a detailed study of the X-ray, optical and infra-red light curves and an intercomparison between these wavelengths (including radio) of the 1975/1976 outburst of the famous black-hole transient A\\,0620$-$00 (Nova Mon 1975, V616\\,Mon). We also investigated the optical behaviour up to a year after the main outburst. This study enabled us to find some new features, which have not been discussed before.  During the various stages of the outburst of A\\,0620$-$00 we found the X-rays lag the optical on the order of $\\sim$5 to $\\sim$20~days. Moreover, we found evidence that the activity associated with the secondary maximum started even earlier in the infra-red. This suggests that most of the processes associated with the outburst occur in the outer parts of the accretion disk.  Although various drops in intensity (lasting on the order of a day or more) in the optical and X-ray outburst light curves of A\\,0620$-$00 have been reported before, we identified some new ones. One such X-ray `dip' only appeared in the soft X-rays (1.5--6\\,keV) whereas at higher energies ($\\gtrsim$6\\,keV) the intensity slightly increases. This shows that the X-ray spectrum pivots near $\\sim$6\\,keV around that time. In the optical we found evidence for another local maximum around that time (so-called `intermediate' maximum). The `intermediate' maximum appears $\\sim$30~days after the secondary maximum, which is close to the viscous time scale of an irradiated accretion disk. We suggest this feature to be an `echo' of the secondary maximum. At the end of the outburst another local maximum occurs. Since the time difference between the peak of the tertiary maximum and this local maximum is also on the order of the viscous time scale we suggest that this feature is an `echo' of the tertiary maximum. We find that drops in the optical and X-ray intensity before and during the rise to tertiary maximum are also present in various other short period soft X-ray transients (SXTs). They always occur $\\sim$150~days after the start of the outburst.  Although the X-ray spectrum of A\\,0620$-$00 gradually softens during the decay from outburst, we find for the first time that it starts to harden again $\\sim$100--150 days after the start of the outburst, similar to that seen in GS\\,2000+25 and GS\\,1124$-$68. This suggests we witness the transition from the so-called high to low state identified in black-hole X-ray binaries. We note that the time of spectral hardening in A\\,0620$-$00, GS\\,2000+25 and GS\\,1124$-$68 is simultaneous with the occurrence of the drops in optical and X-ray intensity.  We also show that the optical outburst amplitude and the shape of the optical outburst light curve of A\\,0620$-$00 closely resembles those of the cataclysmic variable AL\\,Com (where the compact star is a white dwarf). This strengthens the similarity in outburst and quiescent properties of the SXTs and `Tremendous Outburst Amplitude Dwarf novae' (TOADs) or WZ\\,Sge stars, and shows that in general the optical outburst light curves of both groups are governed by the disk properties and not by the compact object. Since irradiation provides a natural mechanism to prolong the outburst of SXTs, we suggest this could be of influence as well during TOAD outbursts. ",
        "introduction": "In 1975 a powerful X-ray transient, \\astrobj{A0620-00}{A\\,0620$-$00}, emerged in the constellation Monoceros (Elvis et al.\\ 1975). \\astrobj{A0620-00}{A\\,0620$-$00} was the first X-ray transient to be identified with an optical brightening of a star, \\astrobj{V616Mon}{V616\\,Mon}, at the same position (Boley \\&\\ Wolfson 1975; Boley et al.\\ 1976). It was therefore also designated as Nova Mon 1975. It is still the brightest extra-solar X-ray source seen to date, and many of the exciting results during the first months of its outburst were summarized by Maran (1976).  Because of its brightness the source has been studied in considerable detail at various wavelengths. The results have appeared in a large numbers of papers in the literature. There exists, however, no thorough comparison of the optical (UBV) light curves together with the X-ray light curves from various experiments. Either authors have presented (part of) the optical light curve (with sometimes the inclusion of data points from other observers) in one or two optical passbands and sometimes compared it with the X-ray light curve (e.g.\\ Robertson, Warren \\&\\ Bywater 1976; Shugarov 1976; Lloyd, Noble \\&\\ Penston 1977; Ciatti, Mammano \\&\\ Vittone 1977; Tsunemi, Matsuoka \\&\\ Takagishi 1977; Chen, Livio \\&\\ Gehrels 1993; Goranskii et al.\\ 1996). The first rough comparison of optical and X-ray light curves was undertaken by Whelan et al.\\ (1977), together with a crude comparison with radio, infra-red and ultra-violet measurements.  Fortunately, almost all the optical (plus infra-red and ultra-violet) observations have been compiled by Ron Webbink in his unpublished report in 1978. In this report he showed (only) the overall B band light curve, and the corresponding B$-$V, and U$-$B measurements, whenever they were simultaneous. Subsequently, only a few authors have shown his B band light curve (e.g., Chevalier \\&\\ Ilovaisky 1990; Van Paradijs \\&\\ McClintock 1995). Since the outburst of \\astrobj{A0620-00}{A\\,0620$-$00} more (soft) X-ray transients (SXTs) have been identified, some of them with considerable optical coverage (see Chen, Shrader \\&\\ Livio 1997). We note that most of these optical observations were only in one or two passbands, and so far \\astrobj{A0620-00}{A\\,0620$-$00} is still the only transient with a fairly complete optical light curve in more than 2 passbands (largely due to its brightness). Several of the X-ray transients have been either found to contain a black hole, including \\astrobj{A0620-00}{A\\,0620$-$00} (McClintock \\&\\ Remillard 1986), or a neutron star (e.g.\\ Van Paradijs 1998). Moreover, the comparison of the various outburst light curves (see White, Kaluzienski \\&\\ Swank 1984; Van Paradijs \\&\\ Verbunt 1984; Chen et al.\\ 1993); Van Paradijs \\&\\ McClintock 1995; Tanaka \\&\\ Lewin 1995; Tanaka \\&\\ Shibazaki 1996; Chen et al.\\ 1997), and the short term ($\\lesssim$100\\,s) (X-ray) variability (e.g.\\ Tanaka \\&\\ Lewin 1995; van der Klis 1995) of the X-ray transients has led to a general picture of their behaviour, in which the main driver for these characteristics is the mass accretion rate through the accretion disk. Since no thorough study of the outburst light curve of \\astrobj{A0620-00}{A\\,0620$-$00} has been performed to date, and the fact that we can now place the results of its 1975/1976 outburst within the general framework of X-ray transient behaviour, we decided to perform, for the first time, an analysis of all the available optical, infra-red and X-ray light curves of \\astrobj{A0620-00}{A\\,0620$-$00} and undertake a detailed comparison between them.  In the next Sections we will first give an overview of what is already known of \\astrobj{A0620-00}{A\\,0620$-$00}, then present the X-ray, optical and infra-red outburst light curves with newly identified features, and then compare it with behaviour seen in other X-ray transients. Finally, we will compare the properties of \\astrobj{A0620-00}{A\\,0620$-$00} with a special unstable class of the cataclysmic variables, i.e., the `Tremendous Outburst Amplitude Dwarf novae' (TOADs), or the \\astrobj{WZSge}{WZ\\,Sge} stars. ",
        "conclusions": "\\label{discussion}  In this paper we have compared the outburst light curves at various wavelengths of \\astrobj{A0620-00}{A\\,0620$-$00}. Although we confirm several of the features reported earlier, we found various new features not seen before.  \\subsection{X-ray delay}  During several stages of the outburst we found evidence for delays between the optical and X-ray. The peak of the secondary maximum was reached $\\sim$5 days earlier in the optical. The infra-red activity associated with the secondary maximum started even a few days earlier with respect to the optical and X-ray. During the trough of the outburst, drops in intensity are consistent with a similar delay of $\\sim$5~days. Near the end of the outburst a `hiccup' occurred in the optical, which was followed $\\sim$5 days later in the X-rays. The broad tertiary maximum started about two weeks earlier in the optical with respect to X-rays; also the peak of this broad maximum was consistent with a similar delay.  A delay between the optical and X-ray was also reported for \\astrobj{GS1124-68}{GS\\,1124$-$68} by Ebisawa et al.\\ (1994). They found that the peak of the X-ray outburst was reached $\\sim$4~days later than the peak of the optical outburst, i.e.\\ a very similar timescale to \\astrobj{A0620-00}{A\\,0620$-$00}. We note that we cannot say with certainty if a similar delay is present at the peak of the outburst of \\astrobj{A0620-00}{A\\,0620$-$00}. More recently, a delay of $\\sim$6~days between the optical and X-ray was found at the start of the outburst of the superluminal SXT \\astrobj{GROJ1655-40}{GRO\\,J1655$-$40} (Orosz et al.\\ 1997).  The delays seen between the optical and X-rays indicates that most of the processes associated with the outburst occur in the outer parts of the accretion disk. The outburst starts somewhere in the outer disk, because of the presence of an inner advection-dominated accretion flow (ADAF) which prevents forming an inner disk region during quiescence (see Narayan, McClintock \\&\\ Yi 1996; Lasota, Narayan \\&\\ Yi 1996; Narayan, Barret \\&\\ McClintock 1997; Hameury et al.\\ 1997). Additional X-rays appear only when the disk is able to extend inwards and therefore causes a delay. Renewed activity in the outer parts of the disk (e.g.\\ Ichikawa \\&\\ Osaki 1994; King \\&\\ Ritter 1998) may explain the delays seen between the optical (and infra-red) and X-rays (not all material has been accreted and it may already have returned to its cool state, or fresh material has been provided by the companion star). Such renewed activity might be induced by the intense irradiation causing the outer parts of the accretion disk to return to a high state again (e.g.\\ King \\&\\ Ritter 1998). We note that activity in the outer disk is also revealed by the onset of superhumps in the SXTs, which always occur just after the secondary maximum (see O'Donoghue \\&\\ Charles 1996). The fact that we see optical superhumps despite the luminous accretion disk has also been attributed to irradiation (O'Donoghue \\&\\ Charles 1996, see also Billington et al.\\ 1996).  \\subsection{Outburst decay and local X-ray maxima}  As already noted, the first exponential decay phase of the outburst of \\astrobj{A0620-00}{A\\,0620$-$00} is slower than its second exponential decay phase. This means that in the case of \\astrobj{A0620-00}{A\\,0620$-$00} the secondary maximum is a temporary enhancement on the first exponential X-ray decay light curve. Also in the optical the secondary (and `intermediate') maxima decay more rapidly in order to resume the same exponential decay as seen before the secondary maximum. In the optical this occurs much earlier than in X-rays ($\\sim$1 week in the optical versus $\\sim$2 months in X-rays). For the other (short-period) SXTs with comparable outburst light curves this seems not to be the case, i.e.\\ the decay timescales before and after the secondary maxima are similar (see e.g.\\ Augusteijn et al.\\ 1993; Chen et al.\\ 1993; 1997; see also Fig.~9).  It is also evident from the overall outburst light curve of \\astrobj{A0620-00}{A\\,0620$-$00} that the time difference between the main maximum and the secondary maximum and that of the secondary maximum and tertiary maximum are not of the same order, as seems to be the case for \\astrobj{GS2000+25}{GS\\,2000+25} (Augusteijn et al.\\ 1993, see also Chen et al.\\ 1993). However, the inferred third maximum by Augusteijn et al.\\ (1993) is not the real tertiary maximum; the tertiary maximum is probably reached $\\sim$200~days after the peak of the outburst (see Kitamoto et al.\\ 1992; Terada et al.\\ 1998; see also Tanaka \\&\\ Shibazaki 1996). Also for \\astrobj{GS1124-68}{GS\\,1124$-$68} the time differences between the different maxima are not similar (see Ebisawa et al.\\ 1994; see also Fig.~9), and its tertiary maximum is $\\sim$200~days after peak of the outburst (see Ebisawa et al.\\ 1994; see also Tanaka \\&\\ Shibazaki 1996), i.e.\\ similar to A\\,0620$-$00 and GS\\,2000+25. It has recently been shown that the time of the secondary maximum in SXTs is related to the viscous time scale of an irradiated disk (Shahbaz et al.\\ 1998b; see also King \\&\\ Ritter 1998). The time of the tertiary maximum seems to be unrelated to this.  It is interesting to note, however, that the times of the expected tertiary maxima of the outbursts of \\astrobj{A0620-00}{A\\,0620$-$00}, \\astrobj{GS2000+25}{GS\\,2000+25} and \\astrobj{GS1124-68}{GS\\,1124$-$68} in the model of Augusteijn et al.\\ (1993) are consistent with the times of start of the spectral hardening (see also next Section; \\astrobj{A0620-00}{A\\,0620$-$00}: this paper; \\astrobj{GS2000+25}{GS\\,2000+25}: Kitamoto et al.\\ 1992; Terada et al.\\ 1998); \\astrobj{GS1124-68}{GS\\,1124$-$68}: Kitamoto et al.\\ 1992; Ebisawa et al.\\ 1994). Moreover, the timescale between the primary and secondary maximum of the {\\it GRO} BATSE hard X-ray outburst light curve of \\astrobj{GROJ0422+32}{GRO\\,J0422+32} as derived by the model of Augusteijn et al.\\ (1993) is similar to the time of appearance of the first `minioutburst' with respect to the end of the X-ray outburst and the time between the two `minioutbursts' (see Callanan et al.\\ 1995, Chevalier \\&\\ Ilovaisky 1995).  \\subsection{X-ray spectral hardening}  For the first time we have demonstrated that \\astrobj{A0620-00}{A\\,0620$-$00} exhibited considerable hardening $\\sim$100~days after the start of the outburst. A similar hardening at nearly the same time after outburst maximum has also been seen in \\astrobj{GS1124-68}{GS\\,1124$-$68} and \\astrobj{GS2000+25}{GS\\,2000+25} (Kitamoto et al.\\ 1992; Ebisawa et al.\\ 1994; Terada et al.\\ 1998). The X-ray spectral and power spectral behaviour in \\astrobj{GS1124-68}{GS\\,1124$-$68} and \\astrobj{GS2000+25}{GS\\,2000+25} just before the hardening is consistent with canonical black-hole high-state behaviour (and maybe also for \\astrobj{A0620-00}{A\\,0620$-$00}, see Section~\\ref{a0620_intro}), whereas after the hardening it is consistent with the canonical black-hole low-state behaviour (Ebisawa et al.\\ 1994; Miyamoto et al.\\ 1994; Terada et al.\\ 1998). We therefore suggest that the power-spectral behaviour after the start of the spectral hardening in \\astrobj{A0620-00}{A\\,0620$-$00} might have shown low-state like behaviour as well.  The time of maximum X-ray spectral hardening is close to when the $\\sim$7.8-day modulations in the X-ray light curve of \\astrobj{A0620-00}{A\\,0620$-$00} are strongest. Similarly, the time of maximum X-ray spectral hardening in \\astrobj{GS1124-68}{GS\\,1124$-$68} and \\astrobj{GS2000+25}{GS\\,2000+25} also occur simultaneously with the drops in intensity, i.e.\\ $\\sim$150~days after the start of the outburst. This suggests a connection between the X-ray spectral hardening and the occurrence of periodic modulations or drops in intensity (see also Section~\\ref{dips}).  Recently, a self-consistent model of accretion flows around black holes with various \\.M has been put forward by Esin, McClintock \\&\\ Narayan (1997; see also Esin et al.\\ 1998). Their accretion flow consists of an ADAF and an outer (thin) accretion disk. In addition above the orbital plane there is a hot corona. \\.M determines the size of the ADAF region and the density of the corona. In this way they could explain the different states seen in black-hole binaries, the off-state, low state, high state and (possibly) the very-high state. The spectral hardening $\\sim$150 days after the outburst is identified with the transition from the high state to low state, i.e.\\ the intermediate state in their model.  We note that such hardening is probably not confined to black-hole SXTs. It has recently been suggested that the hardening near the end of the outburst of the neutron star SXT Aql\\,X-1 (Zhang, Yu \\&\\ Zhang 1998) after a small secondary maximum is a similar phenomenon (Shahbaz et al.\\ 1998a). If this is true, this should manifest itself in a change in the power spectral shapes, possibly resembling black-hole low-state behaviour during the hardening.  \\subsection{Dips and intermediate maxima}\\label{dips}  During the outburst the X-ray light curves show that various drops in intensity occurred, lasting from about one day to several days. They occurred $\\sim$87, $\\sim$103, and $\\sim$228~days after the start of the outburst. Possibly, also the precursor may in fact represent a dip. In addition to these dips, periodic drops in intensity have been seen in X-ray and optical with a period of about 7.8~days (see Section~\\ref{a0620_intro}), and possibly also in the infra-red (this paper). The occurrence of the dips are consistent with the $\\sim$7.8~day modulation period.  It has been suggested that the $\\sim$7.8-day modulation is the beat period between the orbital period and the (possible) superhump period (Zhang \\&\\ Chen 1992), i.e.\\ the disk precession period (see e.g.\\ Priedhorsky \\&\\ Holt 1987). In the SXTs where superhumps have been seen the period excess was found to be between $\\sim$1--2\\%\\ (O'Donoghue \\&\\ Charles 1996). This has also been found for the TOADs (see Kuulkers, Howell \\&\\ van Paradijs 1996, and references therein), and has been related to the extreme mass ratio in these systems. If this also holds for \\astrobj{A0620-00}{A\\,0620$-$00}, the estimated superhump period would be 7.83--7.90\\,hrs, which leads to a beat period between 17--32~days. This is consistent with estimates of the beat period period from its relation with the orbital period and the mass ratio (see e.g.\\ Warner 1995), i.e.\\ $\\sim$18--19~days. Hence the $\\sim$7.8-day modulation cannot be related to the beat period between the orbital period and the (possible) superhump period.  Alternative models include e.g.\\ intrinsic oscillations of the companion star, modulating the accreted matter onto the black hole (Ciatti et al.\\ 1977). We note that near the time of the strong $\\sim$7.8-day modulations the X-ray spectra harden (see previous Section), which suggest a connection between the two. It might be that the transition radius between the outer thin disk and inner ADAF flow in the model of Esin et al.\\ (1997) is oscillating around that time, i.e.\\ the transition radius moves inwards and outwards on a time scale of $\\sim$8~days, which therefore modulates the accretion rate and subsequently the amount of X-rays. Another speculation might be that the heating/cooling waves oscillate inwards and outwards between certain radii, as a result of a modulation in the strength of X-ray irradiation. It remains to be seen if such mechanisms do indeed exist. Moreover, the delay in optical and X-rays should also be explained in such models.  The dip in X-rays $\\sim$87~days after the start of the outburst has not been noted before. As we have shown this corresponds to a short {\\it rise} in intensity or flare at hard energies ($\\gtrsim$6\\,keV), which indicates the X-ray spectrum suddenly pivots for a short time. In the optical we see that there may be a depression in the light curve, but shortly afterward the brightness increased above the expected exponential decline, which resulted in another local maximum, a so-called `intermediate' maximum, $\\sim$30~days after the secondary maximum. A similar `intermediate' maximum may be present in the infra-red, whereas there is {\\it no} indication for such a maximum in X-rays.  The $\\sim$30~days interval between the secondary maximum and `intermediate' maximum is comparable to the viscous time scale in an irradiated accretion disk with parameters appropriate for A\\,0620$-$00 (King \\&\\ Ritter 1998; see also Shahbaz et al.\\ 1998b). If irradiation triggers a thermal instability in the outer accretion disk, new material will accrete one viscous time scale later. This has been put forward as the explanation for the secondary maximum as an echo of the primary maximum by King \\&\\ Ritter (1998; we refer to e.g.\\ Cannizzo 1998 and references therein for other models on the cause of the secondary maximum). We suggest that the secondary maximum induces again enhanced irradiation which may then show up one viscous time scale later. It is not clear to us, however, why this would cause the X-ray spectrum to pivot around that time, and why it does not show up as an `intermediate' maximum in the X-ray light curve.  The only other SXT where another local maximum has been reported which did not coincide with a local maximum in X-rays is \\astrobj{GROJ0422+32}{GRO\\,J0422+32}. Shrader et al.\\ (1994) reported a secondary maximum in the ultra-violet, which occurred $\\sim$38~days after the maximum of the {\\it GRO} BATSE outburst. However, the {\\it GRO} BATSE light curve is that for high ($\\gtrsim$20\\,keV) energies. For GS\\,1124$-$68 the soft X-ray ($\\sim$1--10\\,keV) light curves differs from that in hard X-rays ($\\sim$10--40\\,keV light curves (see Ebisawa et al.\\ 1994), and this could also apply to \\astrobj{GROJ0422+32}{GRO\\,J0422+32}. The local ultra-violet maximum might therefore have been simultaneous with a local soft X-ray maximum, which is expected to exist $\\sim$35~days after outburst maximum (Shahbaz et al.\\ 1998b).  Although the time of the `hiccup' at the end of the outburst is consistent with the minima expected in the $\\sim$7.8~days modulation, interestingly the time difference between the optical maximum of the tertiary peak and the `hiccup' is also on the order of 30~days. This may suggest a similar cause for the `hiccup' as that proposed for the `intermediate' maximum, where now the `hiccup' is a response to the tertiary maximum. We note that a similar `hiccup' might have been seen at the end of the outburst of \\astrobj{GS2000+25}{GS\\,2000+25} (Chevalier \\&\\ Ilovaisky 1990).  \\subsubsection{TOADs and SXTs}\\label{TOADs}  We have shown here that the outburst light curve of the SXT \\astrobj{A0620-00}{A\\,0620$-$00} is very similar in shape to that of the TOAD \\astrobj{ALCom}{AL\\,Com}. Also the structure of the 1978 outburst light curve of \\astrobj{WZSge}{WZ\\,Sge} (Patterson et al.\\ 1981) is very similar to that of \\astrobj{ALCom}{AL\\,Com} (see Howell et al.\\ 1996) and \\astrobj{A0620-00}{A\\,0620$-$00}, having a similar exponential decay, drops in magnitude near the end of the outburst and a `bump' at the end of the outburst. The timescales in the outbursts of \\astrobj{A0620-00}{A\\,0620$-$00} and \\astrobj{ALCom}{AL\\,Com} differ by a factor of $\\sim$5.3. This is close to the ratio of the orbital periods of \\astrobj{A0620-00}{A\\,0620$-$00} and \\astrobj{ALCom}{AL\\,Com} ($\\sim$84~min, see Howell et al.\\ 1996), i.e.\\ $\\sim$5.5. We note that the mass ratio of \\astrobj{ALCom}{AL\\,Com} is close to \\astrobj{A0620-00}{A\\,0620$-$00}, i.e.\\ $q$$\\lesssim$$0.15$, with likely values between 0.033--0.075 (Howell, Hauschildt \\&\\ Dhillon 1998). This shows that the optical outburst light curve is not governed by the mass of the compact object, but related to the (similar) disk properties.  Interestingly, the morphology of the orbital light curve of \\astrobj{ALCom}{AL\\,Com} in quiescence is rather unstable (Abbott et al.\\ 1992; Howell et al.\\ 1996, and references therein), which has also been reported for \\astrobj{A0620-00}{A\\,0620$-$00} by Haswell (1996) and Leibowitz, Hemar \\&\\ Orio (1998). We note that the optical spectra of \\astrobj{A0620-00}{A\\,0620$-$00} in quiescence have also been reported to change with time (Murdin et al.\\ 1980; Orosz et al.\\ 1994).  Masetti \\&\\ Reg\\H os (1997) suggested that SXTs share properties with another \\astrobj{SUUMa}{SU\\,UMa} subclass, i.e.\\ the so-called \\astrobj{ERUMa}{ER\\,UMa} stars. This was based solely on the comparison of the shape and  appearance of (super) humps in the light curves of SXTs and \\astrobj{ERUMa}{ER\\,UMa} stars. \\astrobj{ERUMa}{ER\\,UMa} stars are at the other extreme of the \\astrobj{SUUMa}{SU\\,UMa} class, i.e.\\ during quiescence they have highest mass transfer rates compared to other \\astrobj{SUUMa}{SU\\,UMa} stars in quiescence. They therefore show very frequent outbursts with short quiescent periods, 40--50~days, while their outbursts are of relatively small amplitude, $\\Delta$V$\\sim$3 (Kato \\&\\ Kunjaya (1995). These are properties clearly not shared by the SXTs.  The outburst and quiescence properties of SXTs have been shown, however, to be very similar to those of TOADs, having a fast rise, large outburst amplitude, a slow outburst decay, drops in intensity near the end of the main outburst, and/or post-outburst brightenings (Kuulkers et al.\\ 1996). Our comparison between \\astrobj{A0620-00}{A\\,0620$-$00} and \\astrobj{ALCom}{AL\\,Com} gives additional support for their similarity. The similarities between the TOADs and SXTs reflect that both have low mass ratios and very low mass transfer rates, \\.{M} (Kuulkers et al.\\ 1996).  It is clear that these light curves represent some very generic behaviour. This may be due to a pure viscous evolution, i.e.\\ the outburst disk just evolves under a hot-state viscosity, without the intervention of cooling fronts, until possibly the end of the outburst. Hence any mechanism that prevents the cooling front travelling in from the outer edge of the disk for a sufficiently long time will produce these light curves. In the SXTs irradiation naturally provides that mechanism (e.g.\\ Chen et al.\\ 1993; Van Paradijs 1996; King 1998; King \\&\\ Ritter 1998), whose presence is suggested (at least near maximum of the outburst) by observations (see e.g.\\ van Paradijs \\&\\ McClintock 1994; 1995; Shahbaz \\&\\ Kuulkers 1998)\\footnote{The amount of optical radiation reprocessed from X-ray irradiation is still a subject of debate (see e.g.\\ Lasota \\&\\ Hameury 1998). Current versions of the disk-instability model suggest that {\\it only} $\\sim$1~mag of the optical light during the first months of the outburst is due to X-ray irradiation (see Cannizzo 1998).}. For the outburst of AL\\,Com irradiation may prevent the cooling front from propagating as well, but it is more complicated in this case (see King 1997). If confirmed by more detailed modelling, it may be possible to provide a unified explanation of both cases."
    },
    "9805/astro-ph9805207_arXiv.txt": {
        "abstract": " ",
        "introduction": "For the last twenty years it has been established that the $^{3}$He isotope is present in the spectra of some peculiar stars, which ocupy a narrow range of effective temperatures. An explanation of these results has been attempted in the framework of the diffusion theory (Michaud et al., 1979). On the other hand, the He\\,{\\sc i} lines used for detecting the $^{3}$He isotope are burdened with blends, which were unknown at the times of the first studies of $^{3}$He; therefore, careful reanalysis of the $^{3}$He isotope is needed before making comparisons with the predictions of helium diffusion .  Three of the stars studied here were checked for the presence of $^{3}$He by Hartoog \\& Cowley(1979). They searched for $^{3}$He on photographic spectra by accurately measuring the isotopic shifts of the He\\,I lines. We performed our work as a continuation of a systematic study of the $^{3}$He isotope in stellar atmospheres, on the basis of high dispersion, high S/N spectra and using the spectral synthesis technique. ",
        "conclusions": ""
    },
    "9805/astro-ph9805213_arXiv.txt": {
        "abstract": "We present an abundance analysis of light elements in He-rich stars. The analysis is based on both low and high resolution observations collected at ESO, La Silla, Chile in the optical region and includes 6 standards and 21 He-rich stars. Light-element abundances display a diverse pattern: they range from under-solar up to above-solar values. ",
        "introduction": "The He-rich stars are the most massive CP stars. Their helium abundance ranges from nearly solar up to larger than unity with respect to hydrogen ($n(He)$=1.). Spectroscopic and photometric variability is explained by an abundance distribution across the stellar surface. Given standard atmospheric conditions in B type stars (assuming no wind), diffusion is unable to support helium in the stellar atmosphere and helium sinks. However, Vauclair (1975) showed that diffusion could lead to He overabundance in the presence of mass-loss. Models of abundance anomalies with selective mass loss for He-rich stars suggest normal CNO abundances as a test (Michaud et al. 1987). So in this contribution, we analyse both low and high resolution ESO spectra to obtain light element abundances for He-rich stars and to put constraints on the theoretical model. He and preliminary CNO abundances have already been examined by Zboril et al. (1994, 1997). ",
        "conclusions": "The observed CNO abundances do not entirely fulfill the predictions of the model proposed by Michaud et al. (1987). In particular, C appears underabundant in most He-rich stars (in agreement with Hunger \\& Groote 1993), especially the hotter ones. More detailed models including the magnetic and wind geometry would be welcome."
    },
    "9805/astro-ph9805025_arXiv.txt": {
        "abstract": "We discuss properties of thermal and hybrid (thermal/non-thermal) electron-positron plasmas in the pair and energy equilibria. Various accretion disc-corona models, recently proposed to explain properties of galactic  as well as extragalactic accreting black holes, are confronted with the observed broad-band X-ray and $\\gamma$-ray spectra. ",
        "introduction": "It was realized quite early that broad-band X/$\\gamma$-ray spectra of Galactic black holes (GBHs) can be explained in terms of successive Compton scatterings of soft photons (Comptonization) in a hot electron cloud. The Comptonizing medium was assumed to be thermal with a given temperature, $\\Te$, and a Thomson optical depth, $\\tau_T$. The theoretical spectra were computed by analytical (Shapiro, Lightman \\& Eardley 1976; Sunyaev \\& Titarchuk 1980) and Monte-Carlo methods (Pozdnyakov, Sobol' \\& Sunyaev 1983).  The problem with such an approach is that in any specific geometry arbitrary combinations of $(\\tau_T, \\Te)$ are not possible. Both GBHs and Seyfert galaxies show a hardening of the spectra at $\\sim 10$ keV, which is attributed to Compton reflection (combined effect of photo-electric absorption and Compton down-scattering) of hard radiation from a cold material (White, Lightman \\& Zdziarski 1988; George \\& Fabian 1991). Hard radiation, reprocessed in the cold matter, can form a significant fraction of the soft seed photons for Comptonization. The energy balance of the cold and hot phases determine their temperatures and the  shape of the emerging spectrum (Haardt \\& Maraschi 1991, 1993; Stern \\etal\\/ 1995b; Poutanen \\& Svensson 1996).  The situation becomes more complicated when a notable fraction of the total luminosity escape at energies above $\\sim 500$ keV. Then hard photons can produce $e^{\\pm}$ pairs which will be added to the background plasma. Electrons (and pairs) Comptonize soft photons up to $\\gamma$-rays and produce even more pairs. Thus, the radiation field, in this case, has an influence on the optical depth of the plasmas, which in its turn produces this radiation. This makes the problem very non-linear.  Another complication appears when the energy distribution of particles starts to deviate from a Maxwellian. In the so called non-thermal models, relativistic electrons are injected to the soft radiation field. The steady-state electron distribution should be computed self-consistently, balancing electron cooling (e.g., by Compton scattering and Coulomb interactions) and acceleration, together with the photon distribution. The pioneering steps in solving this problem were done by Stern (1985, 1988) using Monte-Carlo techniques and by \\cite{f86,lz87,coppi92} using the method of kinetic equations (see Stern \\etal\\/ 1995a; Pilla \\& Shaham 1997; Nayakshin \\& Melia 1998, for recent developments). Non-thermal model have been used extensively in the end of 1980s and beginning of 1990s for explaining the X-ray spectra of active galactic nuclei (see, e.g., Zdziarski \\etal\\/ 1990), while recently pure thermal model were preferred, since the data show spectral cutoffs at $\\sim$ 100 keV in both GBHs and Seyferts (Grebenev \\etal\\/ 1993, 1997; Johnson \\etal\\/ 1997). However, power-law like spectra extending without a cutoff up to at least $\\sim 600$ keV, observed in some GBHs in their soft state (Grove \\etal\\/ 1997a,b), give new strength to the undeservedly forgotten non-thermal models.  Spectral fitting with multi-component models following simultaneously energy balance and electron-positron pair balance, give stronger constraints on the physical condition in the X/$\\gamma$-ray source, its size and geometry, presence of $e^{\\pm}$ pairs, and give a possibility to discriminate between various accretion disc models. In this review, we first describe thermal as well as non-thermal pair models that have been used recently for spectral fitting of GBHs and Seyferts. We discuss spectral properties of $e^{\\pm}$ plasmas in energy and pair equilibria for various geometries of the accretion flow. Separately for GBHs and Seyferts, we briefly review X/$\\gamma$-ray observations. Then, we consider physical processes responsible for spectral formation and confront phenomenological models of the accretion discs with data. We restrict our analysis to ``radiative'' models where radiative processes  and radiative transfer in realistic geometries are considered in details while heating and acceleration mechanisms are not specified. ",
        "conclusions": ""
    },
    "9805/astro-ph9805355_arXiv.txt": {
        "abstract": "We consider the merging of compact binaries consisting of a high-mass black-hole and a neutron star. From stellar evolutionary calculations which include mass loss we estimate that a ZAMS mass of $\\gsim 80\\msun$ is necessary before a high-mass black hole can result from a massive O-star progenitor.  We first consider how Cyg X-1 with its measured orbital radius of $\\sim 17 R_\\odot$ might evolve. Although this radius is substantially less than the initial distance of two O-stars, it is still so large that the resulting compact objects will merge only if an eccentricity close to unity results from a high kick velocity of the neutron star in the final supernova explosion. We estimate the probability of the necessary eccentricity to be $\\sim 1\\%$; i.e., $99\\%$ of the time the explosion of a Cyg X-1 type object will end as a binary of compact stars which will not merge in a Hubble time (unless the orbit is tightened in common envelope evolution which we discuss later).  Although we predict $\\sim 7$ massive binaries of Cyg X-1 type, we argue that only Cyg X-1 is narrow enough to be observed and that only it has an appreciable chance of merging in a Hubble time. This gives us a merging rate of $\\sim 3\\times 10^{-8} {\\rm yr}^{-1}$ in the galaxy, the order of magnitude of the merging rate found by computer driven population syntheses, if extrapolated to our mass limit of $80\\msun$ ZAMS mass for high-mass black hole formation. Furthermore, in both our calculation and in those of population syntheses, almost all of the mergings involve an eccentricity close to unity in the final explosion of the O-star. >From this first part of our development we obtain only a negligible contribution to our final results for mergers, and it turns out to be irrelevant for our final results.  In our main development, instead of relying on observed binaries, we consider the general evolution of binaries of massive stars. The critical stage is when the more massive star A has become a black hole and the less massive star B is a giant, reaching out to A. We then have a common envelope, and we expect hypercritical accretion to A. A will accept a small fraction of the mass of the envelope of B but will plunge deep into B while expelling B's envelope.  We expect that star B can at least be in the mass range $15 \\sim 35 \\msun$ while the black hole A has a mass of $10\\msun$. About 20 percent of the binaries of this type are found to end up in a range of orbital radii favorable for merging; i.e., outside of the relevant Roche Lobes, but close enough so that these final binaries of compact objects will merge in a Hubble time.   The narrow black-hole, O-star orbits do not seem to be found in population syntheses because in them mergers happen almost completely as a result of kick velocities. In the exception, Case H of Portegies Zwart \\& Yungelson (1998) which includes hypercritical accretion, common envelope evolution is more effective and we are in agreement with their results.  We find that the high-mass black-hole, neutron-star systems contribute substantially to the predicted observational frequency of gravitational waves.  We discuss how our high mass for high-mass black hole formation can be reconciled with the requirements of nucleosynthesis and indicate that a bimodal distribution of masses of black holes in single stars can account, at least qualitatively, for the many transient sources which contain high-mass black holes. ",
        "introduction": "The supernova community used to believe that stars above a certain mass, about $30-40 \\msun$ ZAMS (zero age main sequence) will collapse into a massive black hole (MBH) of mass of order $10\\msun$.  The argument was that in these stars the mantle was bound with a binding energy well above $10^{51}$ erg so that the supernova shock was not strong enough to expel it.  Whereas this may be true for single stars, Woosley, Langer \\& Weaver (WLW,1995) showed that in binaries, where the hydrogen envelope of the primary star has been transferred to the companion in RLOF (Roche Lobe Overflow), the evolution of the resulting ``naked\" He star; \\ie the star without hydrogen envelope, led to a substantially smaller presupernova core than that of a single star with hydrogen envelope.  A comparison of compact core masses from naked He stars and those evolved by Woosley \\& Weaver (1995) for single stars is shown in Figure~1, taken from Brown, Weingartner \\& Wijers (1996).  Detailed reasons for the great difference in the evolution of ``clothed\" and naked He cores are given in WLW (1995).  Stars with ZAMS masses $\\gsim 40\\msun$ lose their masses by strong winds, whether in binaries or not, and become Wolf--Rayet stars.  In an earlier paper WLW (1993) investigated ZAMS masses of 35, 40, 60 and $85\\msun$.  In those up through $60 \\msun$ the hydrogen envelope was blown off early enough for the He cores to evolve as naked ones and compact core masses were around $1.5 \\msun$ (gravitational). In fact, with inclusion of extensive mass loss the lower line in Fig.~\\ref{fig1}, which heads just above $1.5 \\msun$ for the higher ZAMS masses, gave the WLW (1993) correspondence of Fe core mass to ZAMS mass for single stars of masses $35-60 \\msun$. Thus, rapid mass loss by wind in this mass region which removes the H-envelope before appreciable He core burning begins leaves a He core which burns as a ``naked\" one.  (In the case of the massive stars the relation shown in Fig.~\\ref{fig1} between Fe core mass and He core mass no longer holds because of large wind losses in the latter.) In the case of the $85\\msun$ star some hydrogen envelope remained during an appreciable part of the He burning, so the He core burned as a (partially) clothed one and the compact core was more massive, in the range $1.7 - 2.0\\msun$, depending upon $^{12}C(\\alpha,\\gamma)^{16}O$ burning rate. For the very massive stars, the He core burns, at least part of the time, as if clothed.  The mass loss rate used by WLW (1993) has been shown to be too high.  Such rates were obtained from the free--free fluxes and modelling of infrared spectral lines which involve the density quadratically. Measurements from the polarization of the radiation (St-Louis et al. 1993; Moffat \\& Carmelle 1994) obtain a mass loss rate of $\\lsim$ 50\\% of that used by WLW.  The polarization depends linearly on the density, provided that the wind is optically thin to electron scattering.  Thus, estimates based on this are independent of the inhomogeneities which are known to be present.  The lower mass loss rates come into agreement with the $\\dot{M}$ from dynamical arguments.  Certainly the WLW calculations should be redone with the lower rates, which may change our conclusions.   \\setcounter{equation}{0} ",
        "conclusions": ""
    },
    "9805/astro-ph9805163_arXiv.txt": {
        "abstract": "We present Australia Telescope Compact Array observations of the supernova remnant (SNR) \\snr. In a 1.3-GHz continuum image the remnant appears as a near-circular shell, but with two brightened and distorted arcs of emission on opposite sides. \\HI\\ absorption against the SNR yields a distance in the range 5.4 to 14.1~kpc, corresponding to an age $(1-20) \\times 10^3$~yr.  On the basis of the SNR's morphology we argue that it is a younger analogue of the W~50 / SS~433 system, and that its unusual appearance is a result of opposed jets or outflows from a central source. A jet-like feature and breaks in the shell can both be seen along the axis of proposed outflow, providing further support for this interpretation; the central source itself is not detected.  The SNR may be interacting with the adjacent \\HII\\ region RCW~80 through an extension of the proposed outflow beyond its shell.  This would put the SNR at the lower limit of its distance range and would imply an age $\\la$4000~yr.  We consider other SNRs similar to \\snr, and propose remnants whose shells are affected by jets as one of several classes of SNR from which the presence of a central source can be inferred. ",
        "introduction": "\\label{sec_g309_intro}  Radio observations of supernova remnants (SNRs) demonstrate a vast range of shapes (e.g.\\ Whiteoak \\& Green 1996\\nocite{wg96}). While most SNRs have a distorted and complicated appearance reflecting their interaction with an inhomogeneous interstellar medium (ISM), some SNRs have striking symmetry properties which require other explanations (e.g.\\ Manchester 1987; Roger et al. 1988; R\\'{o}\\.{z}yczka et al. 1993; Gaensler 1998\\nocite{man87,rmk+88,rtfb93,gae98}).  \\snr\\ was first identified as a SNR on the basis of its non-thermal spectrum \\cite{gre74,ccg75}. Subsequent higher resolution observations \\cite{cmw81,kc87,wg96} have shown a distorted shell with two opposed, symmetric bright ends, and a weak compact source in the interior. Continuing a programme to study unusual southern SNRs (Gaensler, Manchester \\& Green 1998a\\nocite{gmg98}, hereafter Paper~I), we present high resolution 1.3-GHz continuum and \\HI\\ absorption observations of \\snr, as well as observations of the region in H$\\alpha$, in X-rays and in the near-infrared. In Section~\\ref{sec_g309_obs} we briefly describe our observations and analysis, before presenting our results in Section~\\ref{sec_g309_results}. In Section~\\ref{sec_g309_discuss} the morphology of SNR~\\snr\\ is discussed, and is compared to that of other SNRs. ",
        "conclusions": "\\label{sec_g309_conclusion}  We have presented \\HI\\ and 1.3-GHz continuum observations of SNR~\\snr, as well as H$\\alpha$, {\\em ROSAT}\\ PSPC and {\\em IRAS}\\ 60~$\\mu$m data on the region. We put a lower limit on linear polarization from the SNR of 1.4 per cent, a low level which we attribute to beam depolarization. We find a rotation measure towards most of the SNR of --930~rad~m$^{-2}$, but a distinctly different RM of \\mbox{--570~rad~m$^{-2}$} towards one component.  This difference is best explained in terms of ISM differences rather than by conditions within the SNR itself.  \\HI\\ absorption puts lower and upper limits on the SNR's systemic velocity of --50 and +40~\\kms\\ respectively, putting it at a distance between 5.4$\\pm$1.6 and 14.1$\\pm$0.7~kpc and implying an age in the range $1-20\\times10^3$~yr.  The nearby \\HII\\ region G309.548--00.737 shows absorption out to the tangent point, consistent with its recombination line velocity and putting it at a distance 5.4$\\pm$1.6~kpc.  SNR~\\snr\\ appears to be a typical shell SNR but with two brightened and distorted `ears' at opposed position angles, which have a similar spectral index to the rest of the shell.  No emission corresponding to the remnant is apparent in the infrared or in H$\\alpha$, while diffuse emission can be seen in X-rays. The compact X-ray source 1WGA~J1346.5--6255 within the SNR is probably associated with the foreground open cluster NGC~5281.  We consider various explanations for the morphology of SNR~\\snr, and argue that the remnant's appearance is best explained by the presence of opposed jets from a central source which collide with and distort the surrounding shell. We propose \\snr\\ as a possible younger analogue to the X-ray binary SS~433 and its associated SNR~W~50.  A faint jet-like structure oriented along the symmetry axis of \\snr\\ may correspond to the outflow itself, while  breaks in the ears along this axis may represent this outflow travelling beyond the shell. The weak source \\src\\ in the SNR's interior is unlikely to be associated with the remnant. We do not detect any other central source in either X-rays or in radio. The former can be attributed to a lack of sensitivity in the observations and to absorption along the line of sight, while the latter may indicate a binary system in a quiescent state or a pulsar with radio beams directed away from us.  To the SNR's north is an unusual column of radio emission, which at one end may connect with the proposed outflow from the SNR's centre and, at the other end, with the \\HII\\ region RCW~80.  Such an association puts the SNR at a distance 5.4$\\pm$1.6~kpc and corresponds to an age of less than 4000~yr.  The details of the physical process behind such an interaction are unclear, but we note that a similar combination of outflow, distortion and termination in a thermal region has been claimed for both G332.4+00.1 (Kes~32) and G320.4--01.2 (MSH 15--5{\\em2}).  Further observations of SNR~\\snr\\ will be required to determine whether our interpretation for its appearance is valid. Higher frequency radio observations can be used to provide higher resolution images of the `ear' and `jet' regions and the interaction between them, to better study the polarimetric properties of the SNR and, together with lower frequency data, to better constrain any spectral index differences between the different components of the remnant. If \\snr\\ is similar to W~50, X-ray observations of greater sensitivity  and at higher energies should be able to detect both a central source and evidence for outflow from it.  Apart from \\snr, we find at least eight other SNRs in which the shell may be affected in some way by jets or outflows from an associated compact source, and suggest G290.1--01.8 (MSH~11--6{\\em 1}A) as a possible further example. While the characteristic morphology associated with such outflow may become another means of determining which supernovae have massive star progenitors, there is good reason to believe that a significant fraction of SNRs harbour compact remnants which, for various reasons, we still have not detected."
    },
    "9805/hep-ph9805383_arXiv.txt": {
        "abstract": "{We calculate the cross section for $s$-wave neutralino annihilation to three-body final states below the $W^{+} W^{-}$ and $t\\bar{t}$ thresholds. Such three-body channels may dominate the annihilation cross section if the neutralino mass is not too much less than $m_{t}$ and $m_{W}$ respectively. Furthermore, because neutrinos produced in these channels are much more energetic than those from the $b\\bar{b}$ or $\\tau^{+} \\tau^{-}$ channels, they can dominate the energetic-neutrino fluxes from neutralino annihilation in the Sun or Earth far below these thresholds and significantly enhance the neutrino signal in certain regions of the supersymmetric parameter space. }  \\begin{document}  \\long\\def\\comment#1{} \\def\\VEV#1{{\\left\\langle #1 \\right\\rangle}} ",
        "introduction": "It has long been well established that the observed luminous matter in Galactic halos cannot account for their total mass.  Determination of the identity of this unseen dark matter has become one of the most important problems in modern cosmology. Perhaps the most promising dark-matter candidate is the neutralino $\\chi$ \\cite{report}, a linear combination of the supersymmetric partners of the $Z$, $\\gamma$, and Higgs bosons.  The existence of neutralinos in our halo could be inferred by observation of energetic neutrinos from annihilation of neutralinos that have accumulated at the core of the Sun and/or Earth in detectors such as AMANDA, super-Kamiokande, MACRO, and HANUL \\cite{energeticneutrinos}. These neutrinos are produced by decays of neutralino annihilation products such as $\\tau$ leptons, $c$, $b$, and $t$ quarks, and gauge and Higgs bosons if the neutralino is heavy enough.  In all cases where the signal is expected to be observable by these or subsequent-generation detectors, accretion of neutralinos from the Galactic halo onto the Sun or Earth comes into equilibrium with their depletion through annihilation \\cite{kamionmodind}. Therefore, the annihilation rate depends on the rate of capture of neutralinos from the halo.  Although the total annihilation cross section is not needed for flux predictions, the branching ratios into the various annihilation products are:  The rate for observation of neutrinos from decays of various annihilation products may differ considerably. Energetic neutrinos are much more easier to detect than low-energy ones. For example, the flux of energetic neutrinos from decays of $b$ quarks is much smaller than that from gauge bosons or top quarks with the same injection energy.  The best technique for inferring the existence of these neutrinos is to observe an upward muon produced by a charged-current interaction in the rock below the detector. The rate for observation of energetic neutrinos is proportional to the second moment of the neutrino energy spectrum, so it is this neutrino energy moment weighted by the corresponding branching ratio that determines the detection rate.  By now, the cross sections for annihilation have been calculated for all two-body final states that arise at tree level (fermion-antifermion and gauge- and Higgs-boson pairs). The precise branching ratios depend on numerous coupling constants and superpartner masses.  However, roughly speaking, among the two-body channels the $b\\bar{b}$ and $\\tau^{+}\\tau^{-}$ final states usually dominate for $m_\\chi < m_W$. Neutralinos that are mostly higgsino annihilate primarily to gauge bosons if $m_\\chi>m_W$,  because there is no $s$-wave suppression mechanism for this channel. Neutralinos that are mostly gaugino continue to annihilate primarily to  $b\\bar{b}$ pairs until the neutralino mass exceeds the top-quark mass, after which the $t\\bar t$ final state dominates, as the cross section for annihilation to fermions is proportional to the square of the fermion mass.  Three-body final states arise only at higher order in perturbation theory and are therefore usually negligible. However, as we already noted, some two-body channels easily dominates the cross section when they are open because of their large couplings, for example the $W^{+} W^{-}$ for the higgsinos and $t \\bar{t}$ for gauginos. This suggests that their corresponding three-body final states can be important just below these thresholds.  More importantly, as mentioned above, the rate for indirect detection is also proportional to the second moment of the energy of the muon neutrino. The neutrinos produced in these three-body final states are generally much more energetic than those produced in $b$ and $\\tau$ decays, so even when the cross sections to these channels are small compared with others, they could dominate the indirect-detection rate.  In this paper, we calculate the cross section for the processes $\\chi \\chi \\to W^{+} W^{-*} \\to W f \\bar{f'}$ and $\\chi \\chi \\rightarrow t\\bar t^* \\rightarrow t W^{-} \\bar{b}$, and their charge conjugates (heretofore referred as $tWb$ and $WW^*$ states) in the $v_{\\rm rel} \\to 0$ limit, where the star denotes virtual particles. Our calculation for the $W W^* $ is applicable for generic neutralinos, but it is mostly significant for neutralinos that are primarily made of higgsino. This is because the gauginos have small couplings to $W^{+} W^{-}$ pairs. As for the $t W b$ final state, our calculation is only applicable to the gaugino.  If the neutralino is primarily higgsino, then annihilation to this final state could also proceed through a $WW^*$ intermediate state, which we have not included in our calculation.  However, if the neutralino is primarily higgsino with $m_{W}<m_{\\chi}<m_{t}$, then it will annihilate primarily to $W^{+} W^{-}$ pairs below the $t \\bar{t}$ threshold. Furthermore, the neutrino yield from $W^{+} W^{-}$ pairs is similar to that for $t \\bar{t}$ pairs. Therefore, neutrino rates are affected only weakly if the process $\\chi\\chi\\to W W^{*} \\to t W b$ is neglected.  Annihilation in the Sun and Earth takes place only when the annihilating neutralinos have relative velocities $v\\sim 10^{-3} \\ll 1$.  Therefore, if we expand the annihilation cross section as \\begin{equation} \\sigma v = a + b v^2\\,, \\label{vpower} \\end{equation} then we consider only the $a$ (the $s$-wave) term. The relic abundance also depends on the annihilation cross section.  However, annihilation in the early Universe takes place when $v\\sim0.5$.  In the early Universe, thermal averaging of the cross section smoothes out the jump in the annihilation rate near particle thresholds \\cite{kimpaolo}.  Therefore, three-body final states should have only a negligible effect on the relic abundance, so we do not consider it further.  We calculate the cross section and neutrino signal for the $t t^*$ final state in Section II and those for the $W W^*$ final state in Section III, and discuss these results in Section IV. Some lengthy matrix elements needed for the calculation are given in the Appendices. ",
        "conclusions": "We made several approximations in our calculation. For example, in the $tWb$ calculation, we neglected the contributions from diagrams involving production of the $tWb$ final state through virtual $W$ bosons. However, in those regions of parameter space where such omissions become unacceptable, the process we are considering is not very important. In the $Wf\\bar{f'}$ case, our calculation is generally applicable, but it is most useful for a higgsino, since in that case the $W^{+} W^{-}$ channel becomes large above threshold.  Although the contributions of these three-body channels are important only in a limited region of parameter space, it may be a large effect. In fact, our calculation shows that although the cross sections of these annihilations are significant only just below the two-body channel threshold, due to the high energy of the neutrinos they produce, they can enhance the neutrino signal by many times and actually dominate the neutrino signal far below the two-body threshold. Furthermore, the regions in question may be of particular interest. For example, motivated by collider data, Kane and Wells proposed a light higgsino \\cite{light higgsino} (but see \\cite{Ellis98}). There are also arguments that the neutralino should be primarily gaugino with a mass somewhere below but near the top-quark mass \\cite{leszek}.  There are many parameters in the minimal supersymmetric model. The results shown in Figs.~\\ref{Nz2figureEarth} and \\ref{W3Earth} are of course model dependent, and these effects might be more or less important in models with different parameters. For a generic neutralino, these effects would be small. However, for the regions of parameter spaces we have focussed on it seems that the increase of the indirect-detection rate is fairly robust. For these regions of parameter space these effects should be considered for calculations of rates for energetic-neutrino detection."
    },
    "9805/astro-ph9805353.txt": {
        "abstract": "We have used the Very Large Array(VLA) in C configuration to carry out a sensitive 20cm radio survey of regions of sky that have been surveyed in the Far Infra-Red over the wavelength range 5-200 microns with ISO as part of the European Large Area ISO Survey(ELAIS).  As usual in surveys based on a relatively small number of overlapping VLA pointings the flux limit varies over the area surveyed. The survey has a flux limit that varies from a 5$\\sigma$ limit of 0.135mJy over an area of 0.12deg$^2$ to a 5$\\sigma$ limit of 1.15mJy or better over the whole region covered of 4.22 deg$^2$. In this paper we present the radio catalogue of 867 sources. These regions of sky have previously been surveyed to shallow flux limits at 20cm with the VLA as part of the VLA D configuration NVSS(FWHM=45 arcsec) and VLA B configuration FIRST(FWHM=5 arcsec) surveys. Our whole survey has a nominal 5 sigma flux limit a factor of 2 below that of the NVSS; 3.4 deg$^2$ of the survey reaches the  nominal flux limit of the FIRST survey and 1.5 deg$^2$ reaches to 0.25 mJy, a factor of 3 below the nominal FIRST survey limit. In addition our survey is at resolution intermediate between the two surveys and thus is well suited for a comparison of the reliability and resolution dependent surface brightness effects that affect interferometric radio surveys. We have carried out a a detailed comparison of the reliability of our own survey and these two independent surveys in order to assess the reliability and completeness of each survey.  % These comparisons %  between our maps and those of FIRST and NVSS will be generally useful. ",
        "introduction": "The Infrared Space Observatory (ISO, Kessler et al. 1996), launched in November 1995 was the successor of the Infrared Astronomical Satellite (IRAS) and provided unparallel sensitivity in mid to far infrared wavelengths ($i.e.$ 5--200 $\\mu m$).  The European Large-Area ISO Survey (ELAIS, Oliver et al. 1997, Oliver et al. 1998 in preparation) is a project that used ISO to carry out a deep wide angle survey at wavelengths of 6.7, 15, 90 and 175 $\\mu m$. The 6.7 and 15 $\\mu m$ surveys were carried out with the ISO-CAM camera (Cesarsky et al. 1996) with the aim to reach a 5$\\sigma$ sensitivity of $\\sim$2mJy at 15microns. The $90\\mu m$ and 175$\\mu m$ surveys used the ISO-PHOT camera (Lemke et al. 1994)  with the aim to reach a 5$\\sigma$ sensitivity of $\\sim$25mJy. At these limits, we expect ISO to be confusion limited at 90 $\\mu m$ and 175 $\\mu m$ by galaxies and galactic cirrus emission and hence this survey should be the deepest FIR survey possible with the satellite.  The area covered in the ELAIS survey is $\\sim 13$ square degrees at 15 and 90 microns, $\\sim 7$ square degrees at 6.7 microns and $\\sim 3$ square degrees at 175microns.  The ELAIS survey is $\\sim$50 times deeper at 5-20$\\mu m$ than IRAS. Thus our survey will allow us to explore IRAS-like populations to higher redshift and possibly unveil new classes of objects or unexpected phenomena. We expect to detect thousands of galaxies, many of which will be at high redshifts and undergoing vigorous star formation. The expected large number of high-z IR galaxies should provide vital information about the star formation rate out to z=1 and possibly earlier.  The spatial resolution of ISO will be insufficient to properly identify optically faint objects. At 15 microns, the survey resolution is $\\sim$ 10 arcsec and at 90 microns it will be about one arc minute. Complementary radio data will play a crucial role in identifying many of the most interesting objects, as they did in the early days of X-ray astronomy (e.g. Cyg X-1) and in more recent times for IRAS (e.g. IRAS F10214+4714 (Rowan-Robinson et al. 1991).  In this paper we report the description of the radio observations obtained in the three ISO-ELAIS survey regions in the northern celestial hemisphere (N1\\_1610+5430, N2\\_1636+4115 and N3\\_1429+3306). The observations are made with the Very Large Array (VLA) radio telescope at 1.4GHz (20cm) in the VLA C-configuration (maximum baseline 11km) with a resolution (FWHM) of $\\sim15$ arcsec.  The aim of these VLA observation was to obtain an uniform covering of the ELAIS regions with a rms noise limit of $\\sim$50 $\\mu$Jy. These VLA observations will be  essential in the optical identification phase of the ELAIS sources and in assessing the reliability of the ELAIS source lists.  Moreover, with a radio survey it will be possible to investigate the radio/far--infrared correlation in star forming galaxies to  flux densities deeper than those reached by IRAS. Helou, Sofier \\& Rowan-Robinson (1985) noted a strong correlation between radio and far infrared flux for star forming galaxies, valid over a very wide range of infrared luminosities, and this has been confirmed in many other studies (e.g. Wunderlich, Klein \\& Wielebinski 1987; Condon, Anderson \\& Helou 1991). The radio emission is interpreted as a synchrotron radiation from relativistic electron which have leaked out of supernova remnants. It is expected that this correlation should extend below the IRAS flux level since the majority of the sub-mJy radio sources have been identified with faint blue galaxies whith spectra similar to those of star forming objects (Benn et al. 1993).  In addition, combining deep radio and optical data with the ISO survey fluxes will provide information on the trivariate IR-radio-optical luminosity function and its evolution and the contribution of starburst galaxies to the sub-mJy radio source counts.  The ratio of the FIR emission and radio emission will also allow is to investigate the physical origin and spatial distribution of the energy sources in the detected objects in the same way that VLA maps have been central to our understanding of the origin of IRAS sources.  Finally, this survey, due to its depth and extension, is very important also as radio survey in its own right. In fact, the selected sample is large and deep enough to constitute a statistically significant sample of sub-mJy radio sources, whose nature and characteristics are still a major topic in observational cosmology (see Windhorst, Mathis \\& Neuschaefer 1990, Fomalont et al. 1991, Rowan-Robinson et al. 1993, Gruppioni et al. 1997) ",
        "conclusions": "Using the Very Large Array (VLA) radio telescope, we observed at 1.4 GHz a total area of 4.222 deg$^2$ in the ISO/ELAIS regions N1 N2 and N3. The lower flux density limit reached by our observation is 0.135 mJy (at 5 $\\sigma$ level) on an area of 0.118  deg$^2$, while the bulk of the observed regions are mapped with a flux density limit of 0.250 mJy (5 $\\sigma$). The data were analyzed using the  NRAO {\\tt AIPS} reduction package. The source extraction has been carried out with the {\\tt AIPS} task {\\tt SAD}. The reliability of  {\\tt SAD} has been tested using the maps of the radio surveys $FIRST$ and $NVSS$.  Considering all the available observations, we detected a total of 867 sources at 5 $\\sigma$ level, 44 of which have multiple components. These sources were used to calculate the normalized differential source counts. They provide a check on catalogue completeness and reliability plus information about source evolution. A comparison with other surveys shows a very good agreement, confirming the presence of the well-know flattening of the counts below 1 mJy, the completeness of our catalogue and  the reliability of our procedure for the source extraction.  A comparison with the $FIRST$ and $NVSS$ radio surveys has confirmed the incompleteness of these two surveys near their flux limits, while a flux comparison between the three surveys has shown that our survey with the VLA array in C configuration is the best compromise between high and low resolution radio surveys. The positional errors of the radio sources are $\\sim$ 2 arcsec for the fainter sources ($\\sim$0.13 mJy) and $\\sim$ 0.6  arcsec for the brighter sources ($>$ 10 mJy). This small value will enable us to obtain an accurate and fast optical/infrared identification of the radio sources."
    },
    "9805/astro-ph9805119_arXiv.txt": {
        "abstract": "We investigate the importance of several numerical artifacts such as lack of resolution on spectral properties of the \\lya-forest as computed from cosmological hydrodynamic simulations in a standard cold dark matter universe. We use a new simulation code which is based on a combination of a hierarchical particle-particle--particle-mesh (P3M) scheme for gravity and smoothed particle hydrodynamics (SPH) for gas dynamics. We have performed extensive comparisons between this new code and a modified version of the \\hydra code of Couchman \\etal and find excellent agreement. We have also rerun the \\tree simulations of Hernquist \\etal using our new codes and find very good agreement with their published results. This shows that results from hydro dynamical simulations that include cooling are reproducible with different numerical algorithms. We then use our new code to investigate several numerical effects such as resolution on spectral statistics deduced from Voigt profile fitting of lines by running simulations with gas particle masses of $1.4\\times 10^{8}$, $1.8\\times 10^{7}$, $2.2\\times 10^{6}$ and $2.1\\times 10^{5} M_\\odot$. When we increase the numerical resolution the mean effective hydrogen optical depth converges and so does the column density distribution. However, higher resolution simulations produce narrower lines and consequently the $b$-parameter (velocity width) distribution has only marginally converged in our highest resolution run. Obtaining numerical convergence for the mean \\Hep transmission is demanding. When progressively smaller halos are resolved at better resolution, a larger fraction of low density gas contracts to moderate over densities in which \\Hep is already optically thick, and this increases the net transmission, making it difficult to simulate \\Hep reliably. Our highest resolution simulation gives a mean effective optical depth in \\Hep 5\\% lower than the simulation with eight times lower mass resolution, illustrating the degree to which the \\Hep optical depth has converged. In contrast, the hydrogen mean optical depth for these runs is identical. Since many properties of the simulated \\lya-forest depend on resolution, one should be careful when deducing physical parameters from a comparison of the simulated forest with the observed one. We compare predictions from our highest resolution simulation in a cold dark matter universe with a photo-ionising background inferred from quasars as computed by Haardt \\& Madau (1996), with observations. The simulation reproduces both the \\H column density and $b$-parameter distribution when we assume a high baryon density, $\\Omega_B h^2 \\gtsima 0.028 $. In addition we need to impose a higher IGM temperature than predicted within our basic set of assumptions. We argue that such a higher temperature could be due to differences between the assumed and true reionization history. The simulated \\H optical depth is in good agreement with observations but the \\Hep optical depth is lower than observed. Fitting the \\Hep optical depth requires a larger jump $\\sim 14$ between the photon flux at the \\H and \\Hep edge than is present in the Haardt \\& Madau spectrum. ",
        "introduction": "Sight lines to distant quasars intersect many cosmological structures containing neutral hydrogen and \\lya scattering by the \\H in these structures produces a forest of lines blueward of the quasar's \\lya emission line (Lynds 1971). This \\lq\\lya forest\\rq~ contains unbiased information on the temperature, density, velocity and ionization structure of the intergalactic medium (IGM) along the line of sight to the quasar, making the structures responsible for the \\lya forest a useful probe for studying the high-redshift universe. In addition, it is likely that the absorbing gas retains a memory of its state at even higher redshifts, enabling us to study its initial conditions (Croft \\etal 1998) and previous history. Since these structures are of moderate density contrast, they are easier to simulate numerically than galaxies, and consequently the high redshift universe can be studied efficiently and accurately by comparing simulations of the \\lya forest with observations.  Recent hydrodynamic simulations of hierarchical structure formation in a universe dominated by cold dark matter (CDM) have been shown to be remarkably successful in reproducing a variety of statistics of \\lya absorption lines (Cen \\etal 1994, Zhang, Anninos \\& Norman 1995, Miralda-Escud\\'e \\etal 1996, Hernquist \\etal 1996, Wadsley \\& Bond 1996, Zhang \\etal 1997), including the number of lines per unit redshift per unit column density and the number of lines with given width (\\lq $b$\\rq~ parameter), as well as its evolution at low-redshift (Theuns, Leonard \\& Efstathiou 1998). This is quite encouraging for the hierarchical picture of structure formation since the underlying cosmological models were designed with galaxy formation in mind, hence their \\lya properties can be considered to be a genuine and successful prediction. Most simulations to date have assumed a critical density, cold dark matter model, in which a photo-ionising background close to that inferred from quasars as computed by Haardt \\& Madau (1996) is required to explain the properties of the \\lya forest. However, other variants of the CDM model still provide acceptable fits, with only minor modifications to the required photo-ionization background (Cen \\etal 1994, Miralda-Escud\\'e \\etal 1996).  In this paper we introduce a new simulation code designed to study numerically the formation of \\lya systems. It is based on a combination of Smoothed Particle Hydrodynamics (SPH, Lucy 1977, Gingold \\& Monaghan 1977, see \\eg Monaghan 1992 for a review) and an adaptive P3M (particle-particle--particle-mesh) gravity solver (Couchman 1991). Its efficient gravity solver and SPH implementation lead to a fast and accurate code which has the potential to extend considerably the dynamic range of the simulations. We discuss tests of the new code and perform extensive comparisons against two other simulation codes: \\hydra and \\tree. Both of these are also based on SPH but their gravity solvers differ: \\hydra \\cite{Couchman95} uses the same gravity solver as \\apm but \\tree \\cite{KatzWeinbergHernquist96} uses a tree structure. We discuss in detail the differences between the \\apm and \\hydra codes. We also discuss the changes we have made to the publically available \\hydra code to study the \\lya cloud problem. The overall agreement between the three codes is excellent which shows that hydrodynamic simulations that include cooling are reproducible with different simulation codes. The good agreement also shows that \\hydra can be used to study the \\lya problem and we are currently analysing several large \\hydra simulations performed on the T3D computer to understand in more detail how resolution affects \\lya statistics (the VIRGO consortium, in preparation).  We then use \\apm to perform simulations at increased resolution and establish the extent to which published results are influenced by lack of numerical resolution and other numerical artifacts. Wadsley \\& Bond (1996; see also Bond \\& Wadsley 1997) recently warned simulators of the importance of long-wavelength perturbations on the occurrence of filamentary structures in simulations. This is illustrated explicitly in the work of Miralda-Escud\\'e \\etal (1996) who compare simulations with the same resolution but different box sizes. Unfortunately, current numerical codes do not possess the required dynamic range to resolve the Jeans length in a very large simulation box. We try to gauge the effects of missing waves and of failing to resolve the Jeans length by performing simulations with various box sizes. This paper is organised as follows: Section~\\ref{sect:simulation} discusses the physical model and gives details of the simulation codes, Section~\\ref{sect:comparison} presents the comparisons between codes, Section~\\ref{sect:resolution} addresses the importance of numerical resolution, Section~\\ref{sect:observations} does a comparison of simulations against observations and finally Section~\\ref{sect:summary} summarises. Technical details are relegated to Appendices. ",
        "conclusions": "\\label{sect:summary} We have presented a new simulation tool, \\apm, designed to study numerically the formation of structures responsible for the \\lya-forest. This code is very fast and treats the low density IGM relatively accurately, allowing increased resolution at little extra simulation time. The IGM is allowed to interact with a time-dependent but uniform background of ionising photons assumed to come from quasars, using the rates suggested by Haardt \\& Madau (1996). This background heats the low density gas and changes the form of the cooling function at higher densities. The distribution of the gas in the density-temperature plane can be understood from the relative importance of cooling and heating processes, and from the comparison of the appropriate cooling time scale with the Hubble time.  We performed extensive comparisons of the new code with the \\hydra code of Couchman \\etal (1995), which was adapted to study this problem as well. The agreement between the two codes is excellent for a wide variety of statistics. The distribution of gas in the $(\\rho,T)$ plane is very similar and various statistics on the distribution of halos agree very well. The amount of gas which is able to cool in collapsed halos is similar in the two codes, showing that coding details are not very important in determining this fraction. We are currently analysing several large \\hydra simulations performed on the T3D computer to understand in more detail how resolution affects \\lya statistics (the VIRGO consortium, in preparation). Both \\apm and \\hydra are based on the Lagrangian SPH method, which has high resolution in high density regions. However, since many lines form in {\\em low} density regions where SPH suffers from low resolution, it would still be very valuable to compare in detail with some of the Eulerian codes used by other groups.  We also compared our new code with published results from \\tree (Hernquist \\etal 1996) for simulations started from their initial conditions and confirm their findings. We have also analysed independently our simulated spectra from these runs using a different implementation of automated Voigt profile fitting (VPFIT, Carswell \\etal 1987). The deduced line statistics in terms of column density and $b$-parameter distributions agree well with their published values, showing that Voigt profile fitting gives reproducible results.  We have then used \\apm to study the effects of lack of numerical resolution on quantities deduced from simulated spectra based on Voigt profile fitting. The mean effective hydrogen optical depth is converged in our medium resolution simulation and so are the derived column density distributions (DDFs). The latter also are in good agreement with DDFs deduced from observations, for our assumed background flux and baryon fraction. However, the relative amounts of cool gas are rather different when comparing the A-22-64 with the A-11-64 run, which has eight times better mass resolution, and there are still noticeable differences with our highest A-5-64 run (which has another factor of eight better mass resolution), due to lack of numerical resolution. With increasing resolution, we find that the optical depth decreases, especially for \\Hep, and that the number of lines with small $b$-parameter increases. However, from a comparison of the A-5-64 run with an even higher resolution simulation, A-2.5-64, we find that the A-5-64 box is already very close to convergence and we are relatively confident that we can draw reliable conclusions from this simulation. We found that the deduced $b$-parameter distributions are sensitive to the assumed continuum level, a problem which should also influence observations to some extent. The DDFs, on the other hand, are not very sensitive to the exact VP-fitting procedure.  Some previously published results on the \\Hep forest are unreliable due to lack of numerical resolution. For example at $z=4$, the mean effective \\Hep optical depths are 4.54, 3.52, 2.78 and 2.63 for runs A-22-64, A-11-64, A-5-64 and A-2.5-64, respectively. This shows that the required resolution to get the mean optical depth correct is very high. We interpreted the dependence on resolution as being due to the formation of progressively smaller halos being resolved with better resolution. Low density gas falls into these halos and hence the optical depth decreases. The good agreement between the 5.5Mpc and the 2.5Mpc box increases our confidence that these higher resolution runs have effectively converged.  Turning to a comparison of our highest resolution simulation with observations, we come to the following conclusions.  \\begin{itemize} \\item There is excellent agreement between the observed and simulated \\lya column density distributions at $z=2$ and $3$, provided we divide the ionising background intensity advocated by Haardt \\& Madau (1996) by two, for our assumed baryon fraction of $\\Omega_B h^2=0.0125$. Alternatively, for the intensity of the ionising background as computed by Haardt \\& Madau, we require a higher baryon fraction (Rauch \\etal 1997) \\begin{equation} \\Omega_B h^2 \\gtsima 0.017\\,\\,\\,\\,{\\mbox{\\rm (from DDFs)}}\\,. \\end{equation}  \\item The simulated $b$-parameter distributions peak at lower $b$-values than the observed ones for $z=4$, 3 and 2, suggesting that the simulated IGM temperature in our simulations is too low. We argued that uncertainties in reionization history, combined with non-equilibrium effects and feedback from star formation, might be sufficient to increase the temperature by a factor $\\sim 2$, which would bring the simulated distributions into excellent agreement with the observed ones. However, this would increase the required $\\Omega_B$ even more, since increasing the temperature would decrease the amount of absorption, giving the higher $\\Omega_B$ limit  \\begin{equation} \\Omega_B h^2 \\gtsima 0.028\\,\\,\\,\\,{\\mbox{\\rm (from $b$-parameter distribution)}}\\,. \\end{equation}  \\item The \\Hep optical depth corresponding to our best fit \\H optical depth is lower than observed values, suggesting that the Haardt \\& Madau ionization spectrum may be too hard. The more recent analysis by Zheng \\etal (1997) of observed quasar spectra lead to a similar conclusion. Fitting both \\H and \\Hep optical depths requires a spectral break \\begin{equation} {J_\\h\\over J_\\hep} \\approx 14\\,. \\end{equation}  \\end{itemize}  Overall we find that the level of agreement between simulations of the \\lya forest in a scale-invariant, CDM universe and observations, is still impressive. More detailed comparisons between simulations and observations will allow us to study the thermal history of the universe at even higher redshifts."
    },
    "9805/astro-ph9805269_arXiv.txt": {
        "abstract": "We present a new geometrical approach to the study of accretion flows onto rotating black holes. Instead of Boyer-Lindquist coordinates, the standard choice in all existing numerical simulations in the literature, we employ the simplest example of a {\\em horizon adapted coordinate system}, the Kerr-Schild coordinates. This choice eliminates unphysical divergent behavior at the event horizon. Computations of Bondi-Hoyle accretion onto extreme Kerr black holes, performed here for the first time, demonstrate the key advantages of this procedure. We argue it offers the best approach to the numerical study of the, observationally, increasingly more accesible relativistic inner region around black holes. ",
        "introduction": "Advances in satellite instrumentation, e.g., the Rossi X-Ray Timing Explorer (RXTE), and the Advanced Satellite for Cosmology and Astrophysics (ASCA), are greatly stimulating and guiding theoretical research on accretion physics. The recent discovery of kHz quasi-periodic oscillations (QPOs), extends the frequency range over which these oscillations occur into timescales associated with the innermost regions of the accretion process (for a review see van der Klis 1996). Stella and Vietri (1997) have proposed that observed low frequency QPOs in neutron star X-ray binaries correspond to the precession of the accretion disk, i.e., the Lense-Thirring effect. This could be the first evidence for a genuinely general relativistic effect, i.e., the dragging of inertial frames, in such systems. Morgan, Remillard and Greiner (1997) identified a 67 Hz QPO in the black hole candidate GRS 1915+105 which may be associated with relativistic trapped inner disk oscillations (Nowak et al. 1997). Recently, Cui, Zhang and Chen (1998) have extended the interpretation of Stella and Vietri to black hole binaries.  Within this model, GRO J1655-40 and GRS 1915+105 are found to spin at a rate close to the maximum theoretical limit. Moreover, in extragalactic sources, spectroscopic evidence (broad iron emission lines) increasingly points to (rotating) black holes being the accreting central objects (Tanaka et al. 1995; Kormendy and Richstone 1995). Recently, Bromley, Miller and Pariev (1998) placed a limit on the inner edge of the accretor giving rise to the iron K$\\alpha$ emission in MCG-6-30-15 at about 2.6 Schwarzschild radii. Their estimate is that the black hole is rotating at a rate which is about $(23 \\pm 17)$ \\% of the allowed maximum.  Early theoretical studies indicated that a rotating black hole in the presence of an accretion disk must be spinning at nearly the maximal rate (Thorne 1974). Rotation increases the available energy in the near-horizon region: the binding energy per unit mass of a test particle can reach up to $0.42c^2$ and the innermost stable circular orbit approaches the horizon (Bardeen, Press and Teukolsky 1972) and coincides with it (at least in areal coordinates) in the extreme case $a=M$ ($M$ is the mass of the hole and $a$ its specific angular momentum). In the rotating case, motions away from the equatorial plane are affected by the dragging of inertial frames, while motions in the ergo-region may (more speculatively) extract rotational energy from the hole.  Accretion theory is primarily based on the study of stationary flows and linearized perturbations thereof. Establishing the nature of flow instabilities, though, will almost certainly require highly resolved and accurate, time-dependent, non-linear numerical investigations. Especially intriguing is the possibility of establishing features of the accretion process that are reflecting the nature of the spacetime. Such numerical probes rely crucially on adequate and consistent approximations of the geometry. For a wide range of accretion problems, a Newtonian theory of gravity {\\em is} adequate for the description of the background gravitational forces. The extensive experience with Newtonian astrophysics suggests that the first explorations of the relativistic regime could benefit from the use of {\\em model potentials} (Paczy\\'nski and Wiita 1980). This constitutes the {\\em Newtonian paradigm}, which is still being developed (e.g., Nowak and Wagoner 1991; Artemova, Bj\\\"ornsson and Novikov 1996).  For comprehensive numerical work, a full (i.e., three-dimensional) formalism is required, able to cover also the maximally rotating hole. In rotating spacetimes the gravitational forces cannot be captured fully with scalar potential formalisms. A vivid example is provided by the wave systems examined in Chandrashekhar (1983), in which rotation introduces {\\em frequency} dependent potentials. Additionally, geometric regions such as the ergo-sphere would be very hard to model without a metric description. Whereas the bulk of emission occurs in regions with almost Newtonian fields, only the observable features attributed to the inner region may crucially depend on the nature of the spacetime.  Pioneering numerical efforts in the study of accretion onto black holes (Wilson 1972; Hawley, Smarr and Wilson 1984; Hawley 1991), established the relativistic framework, the so-called {\\it frozen star paradigm} of a black hole. In this, the time ``slicing\" of the spacetime is synchronized with that of asymptotic observers far from the hole. This is a powerful approach, leading to a very economical description of the geometry and can, {\\em in principle}, be used to capture all the interesting effects of the spacetime curvature. We focus here on the limitations of this approach which further motivate this {\\em Letter}. The shortcomings are due to the poor choice of coordinates near the black hole horizon and, hence, manifest themselves only in its neighborhood.  We have argued, though, that this is precisely the region of most interest. A set of consistency problems arises from the need for correct boundary conditions at the horizon. Such conditions are easily imposed on simple supersonic inflows, but become murkier for co-rotating accretion disks on rapidly-rotating Kerr black holes. In addition, imposing boundary conditions on magnetic fields is problematic. Addressing this issue has led to the development of the so-called {\\it membrane paradigm}. Starting with the description of the black hole processes in a non-singular coordinate system (Damour 1978), this approach endows an {\\em approximate} horizon (a timelike worldtube) with special electric and magnetic properties and then reverts back to the frozen star picture for the description of the rest of the spacetime (Thorne and MacDonald 1986). However, the computation is still performed in the original singular system. Imposing boundary conditions near the horizon becomes a demanding practical task, as the singular coordinate coverage of the horizon leads to an unphysical blowup of coordinate dependent quantities. ",
        "conclusions": "We have shown the feasibility of a new geometrical approach to the numerical study of accretion flows onto rotating black holes. Our procedure relies on the use of {\\it horizon adapted coordinate systems} in which all fields, i.e., metric, fluid and electromagnetic fields, are free of coordinate singularities at the event horizon. Among the large family of those systems we propose the use of the Kerr-Schild coordinate system -- the simplest example of the class -- as the natural framework to perform accurate numerical studies of relativistic accretion flows.  We have discussed our approach in the context of the various paradigms in black hole astrophysics. Our proposal shares with the frozen star picture the exact representation of the relativistic geometry. It departs from it in significant ways in the crucial horizon region, in which a choice of different time coordinate allows the explicit use of the ``one way membrane'' picture in the computations.  We conclude that smooth coordinate systems at the horizon become an invaluable tool for the numerical study of accretion phenomena around extreme Kerr black holes. Coupled with high resolution numerical methods these systems may help clarify the basic dynamical processes around accreting black holes."
    },
    "9805/astro-ph9805004_arXiv.txt": {
        "abstract": "The IUE contribution to the understanding of the blazar phenomenon has been of fundamental importance. Here I review the progress obtained with the latest multifrequency campaigns performed with IUE on two prototype objects, the BL Lac PKS 2155 - 304 and the  highly polarized, superluminal quasar 3C 279. \\vspace {5pt} \\\\    Key~words: ultraviolet; blazar continuum; blazar emission mechanisms; jets. ",
        "introduction": "The unique capabilities of IUE in spectrophotometric accuracy and UV sensitivity made it an essential tool in the study of variable X-ray sources. In particular for blazars it allowed to determine the shape of the non-thermal continuum in a region free of the possible contribution from a host galaxy and, together  with coordinated optical and X-ray observations, made possible by the flexible scheduling, over a very wide spectral range. It is worthwhile to recall here that the first evidence of a sytematic difference in spectral shape between X-ray selected BL Lacs and other blazars and \"normal\" quasars was obtained with IUE (Ghisellini et al. 1986). Systematic analyses of IUE data concerning blazar variability are presented by Treves and Girardi (1990),  Edelson (1992), Pian and Treves (1993). Early results from blazar observations with IUE are reviewed in Bregman, Maraschi and Urry (1987).  It became clear after the results of the first pioneering studies that at least some of the sources were varying extremely rapidly also in the UV and that quasi - simultaneous snapshots of the UV to X-ray energy distribution could describe some average \"state\" but were insufficient to probe the correlation between the two wavelength ranges. Simultaneous light curves in X-rays and UV and possibly other wavelengths were  and are needed to address physical models of variability. In fact, despite the long lifetime of IUE well sampled multiwavelength data were obtained in a limited number of cases (see for reviews Wagner \\& Witzel 1995; Ulrich, Maraschi, \\& Urry 1997, UMU97 hereafter). Here I will discuss two sources for which many data have been obtained which are helping us to make progress in the understanding of the blazar phenomenon. The first is the BL Lac object PKS 2155-304, one of the brightest blazars in the UV and soft X-ray sky (Section 3). The second is the superluminal quasar  3C 279, the first and one of the brightest blazars detected in $\\gamma$-rays (Section 4). Before  discussing two apparently different cases I will briefly present a scheme for a unitary phenomenological and theoretical understanding of the non thermal emission of blazars. ",
        "conclusions": "Multifrequency studies of blazars seem to lead to a unified view of the broad band continuum of the whole class. Two spectral components, each exhibiting a broad peak in the $\\nu F_{\\nu}$ representation are present in all objects. The two peak frequencies are different in different objects but their ratio is approximately constant. Irrespective of subclassifications, the blazar continua may be described by a single spectral sequence whereby the values of the peak frequencies are fixed by the radio luminosity: higher luminosity objects peak at lower frequencies.  The overall regularity of the SEDs suggests that the same mechanisms (synchrotron and inverse Compton radiation) operate in all blazars, albeit under gradually different physical conditions. The magnetic field, the critical particle energy (corresponding to the maxima in the SED), the importance of ambient vs. synchrotron photons in the inverse Compton process may change gradually along the sequence (Ghisellini et al. 1997).  Multifrequency studies of objects at the \"blue\" end of the sequence, like PKS 2155-304, revealed the evolution of synchrotron flares from X-rays to UV  frequencies, indicating impulsive injection  of high energy particles in the emission region. It is regrettable that IUE did not survive long enough to contribute to multifrequency campaigns including TeV observations, which only recently yielded positive detections of the brightest \"blue\" blazars.  For the \"red\" blazar 3C 279 multifrequency observations have shown a clear long term correlation of the UV emission (due to the synchrotron process) with the $\\gamma$-ray emission (due to the inverse Compton mechanism) thus providing strong support to the idea that the two spectral components derive from the same population of relativistic electrons. The short time-scale variability requires either a highly non linear variation of the seed photons during a flare or dilution of the synchrotron flare by more stationary synchrotron emission from adjacent regions in the jet."
    },
    "9805/astro-ph9805232_arXiv.txt": {
        "abstract": "We search for a relation between the published distributions of different elements and the calculated magnetic field structure, following from a dipole-quadrupole configuration, of the CP2 star CU Vir. The highest concentration of individual chemical elements on the stellar surface coincides obviously with the regions of the highest values of the magnetic field strength. ",
        "introduction": "The B9pSi star CU Vir (HD 124224, $P_{\\rm rot}$~=~0\\fd52, e.g. Weiss et al. 1976), has an amplitude of the effective magnetic field curve of about 1000 gauss and shows very large spectral variations of helium and silicon. Relations between the structure of the magnetic field and the distribution of some chemical elements on the stellar surface should exist according to, e.g., Michaud (1970), Glagolevskij (1994), and Hatzes (1997). For our investigation we use the values of the longitudinal magnetic field, $B_{\\rm eff}$, measured photoelectrically by Borra and Landstreet (1980) and the distribution of He, Si, and Mg over the stellar surface derived by Goncharskij et al. (1983), Hiesberger et al. (1995), Kuschnig et al. (1997), and Hatzes (1997). ",
        "conclusions": "Our dipole-quadrupole model yields two maxima at the positive halfwave of the magnetic field and a very satisfactory agreement between the calculated and the observed behaviour of $B_{\\rm eff}$. We are not able to fit the strong anharmonic variation of $B_{\\rm eff}$~by a decentred dipole field. The two observed He spots seem to support the assumption of a dipole-quadrupole configuration.\\\\ The maps proposed for Si allow to explain the element distribution over the surface by a concentration of the element at the magnetic poles as well as in the regions where the magnetic field lines have a horizontal direction. To overcome the existing discrepancies, more accurate observations are necessary.\\\\"
    },
    "9805/astro-ph9805138_arXiv.txt": {
        "abstract": "We consider an ultra-relativistic wind consisting of electron-positron pairs and photons with the principal goal of finding the asymptotic Lorentz factor $\\gamma_{\\infty}$ for zero baryon number. The wind is assumed to originate at radius $r_i$ where it has a Lorentz factor $\\gamma_i$ and a temperature $T_i$ sufficiently high to maintain pair equilibrium. As $r$ increases, $T$ decreases and becomes less than the temperature corresponding to the electron mass $m_e$, after which non-equilibrium effects become important. Further out in the flow the optical depth $\\tau$ drops below one, but the pairs may still be accelerated by the photons until $\\tau$ falls below $\\sim 2\\times10^{-5}\\gamma_{i}^{3/4}$. Radiative transfer calculations show that only at this point do the radiation flux and pressure start to deviate significantly from their blackbody values. The acceleration of the pairs increases $\\gamma$ by a factor $\\sim 45$ as compared to its value at the photosphere; it is shown to approach $\\gamma_{\\infty} \\sim 1.4\\times 10^3 (r_i/10^6\\mbox{cm})^{1/4} \\gamma_{i}^{3/4} T_i/m_e$.  The limit of zero baryon number is a good approximation when the mass injection rate $\\mdot$ in the flow is below a critical value corresponding to $(\\edot/\\mdot)_{\\rmn{c,0}}\\sim 5\\times10^7(r_i/10^6\\mbox{cm})T_i/m_e$ for fixed energy injection rate $\\edot$. For large baryon loading, $\\edot/\\mdot\\simless(\\edot/\\mdot)_{\\rmn{c,M}}\\sim 350(r_i/10^6\\mbox{cm})^{1/4}\\gamma_{i}^{3/4}T_i/m_e$, the asymptotic Lorentz factor is $\\gamma_\\infty\\sim\\edot/\\mdot$. Surprisingly, increasing $\\edot/\\mdot$ from $(\\edot/\\mdot)_{\\rmn{c,M}}$ to $\\infty$ only increases $\\gamma_\\infty$ by a factor $\\sim (m_p/m_e)^{1/4}\\approx 6.5$, less than an order of magnitude. ",
        "introduction": "\\label{intro}  The release of a large amount of radiative energy into a small volume can lead to the formation of a fireball, a dense fluid of radiation and particles that expands under its own pressure. Fireball models have become the accepted framework for understanding gamma--ray bursts at cosmological distances and their afterglows (Paczy\\'{n}ski 1986, 1990; Shemi \\& Piran 1990; M\\'{e}sz\\'{a}ros \\& Rees 1993, Piran 1997). Paczy{\\'n}ski (1986) and Goodman (1986) originally considered the possibility that fireballs could originate in the collision of a pair of neutron stars in a binary star system coalescing as a result of gravitational radiation reaction. (See also Naryan, Paczy{\\'n}ski \\& Piran 1992.) In this picture, the thermal energy released in the collision, $\\sim 10^{53}$ ergs, is radiated as a neutrino--anti-neutrino burst. A fraction of that energy may be transformed into electron-positron pairs above the surface of the neutron star (Goodman, Dar \\& Nussinov 1987). There are now additional proposals for the origin of fireballs (e.g. Paczy{\\'n}ski 1997; Fuller \\& Shi 1997; Pen, Loeb \\& Turok 1997).  Close to the radius at which energy is injected, the resulting wind is opaque. The radiation energy that is initially trapped can escape in two different ways further out in the flow: When the plasma becomes optically thin, radiation streams freely to the observer (Paczy{\\'n}ski 1990). Alternatively, if there is a significant baryon contamination in the fireball, it can become matter dominated before radiation escapes. The matter will increase the opacity and, more importantly, convert part of the radiation energy into bulk kinetic energy (Shemi \\& Piran 1990). Interactions between the expanding atmosphere and the surrounding matter provide a way to convert the kinetic energy in the baryons back to radiation at the resulting shock front (M\\'{e}sz\\'{a}ros \\& Rees 1993). Internal shocks due to variations in the velocity of matter is also proposed as a way to dissipate kinetic energy (Rees \\& M\\'{e}sz\\'{a}ros 1994). Fenimore (1997) recently pointed out that the observed temporal structure of gamma ray bursts severely constrains the proposed models of energy conversion by relativistic shocks; conceivably internal structure and shocks can account for some of the observed variability (e.g. Kobayashi, Piran \\& Sari 1997).  In the fireball models invoked to explain the afterglows of GRB 970228 and GRB 970508, the bulk Lorentz factor of the outflow is $\\gamma \\sim 100-1000$ before the outgoing shell is slowed significantly by sweeping up matter from ambient gas (Wijers, Rees \\& M\\'{e}sz\\'{a}ros 1997; Waxman 1997; Waxman, Kulkarni \\& Frail 1998). From a theoretical point of view, values of $\\gamma$ in this range yield acceptable estimates of burst duration and (with additional assumptions) characteristic photon energies. It is presumed that most of the energy originally in the fireball converts to kinetic energy long before deceleration begins, so the (baryon) rest mass of the flow must be nonzero: $\\mdot=\\edot/\\gamma$, where $\\mdot$ and $\\edot$ are the rest mass and total energy injection rate of the flow. Precisely how such small but nonzero $\\mdot/\\edot$ arises is not clear yet; nor is it obvious whether $\\gamma=\\edot/\\mdot$ can be much larger or smaller than 100-1000, the values that seem necessary for modeling gamma ray bursts.  In this paper we reconsider the original steady wind problem first solved by Paczy{\\'n}ski (1986) but for $\\mdot\\equiv 0$. At first sight, one might think that the result would be $\\gamma \\to \\infty$. However, the failure of equilibrium at low temperatures (once the pair density falls sufficiently so that annihilation becomes slow) leads ultimately to a finite $\\gamma$. As we shall see, in regions where the temperature is greater than the electron mass, the fireball is very optically thick because of the large number of electron--positron pairs. As the radius increases and the local temperature decreases, the deviation from equilibrium in the number density of pairs becomes significant. A little further out in the flow the optical depth falls below unity. However, the remaining pairs are still heated and accelerated considerably via their interactions with the radiation field; in the rest frame of the pairs, the radiation field itself remains close to the blackbody form long after the fireball becomes optically thin. The radiation spectrum detected by a stationary observer would not be blackbody, however, but also differs from the power-law spectra of {\\sc grb}s. Although our calculations pertain to a steady, spherical wind, the general result that $\\gamma$ is finite even at zero baryon mass may be true as well for thin shells emitted from impulsive energy release. We consider non-spherical perturbations around our wind solutions in Section {\\ref{pert}}; as we shall see, some memory of surface `hot-spots' may persist out to the photosphere.  The dynamics for winds with sufficiently small baryon number are similar to $\\mdot=0$ outflow. A larger baryon number will increase the inertia of the flow, and so we expect the radiative acceleration and therefore the asymptotic Lorentz factor $\\gamma_\\infty$ to be reduced. However, as the baryon number increases, the photospheric radius tends to increase as well, with a corresponding increase in the Lorentz factor at the photosphere. These two factors combined result in a surprisingly small variation in $\\gamma_\\infty$: For fixed $\\edot$, as $\\edot/\\mdot$ decreases from $\\infty$ to $(\\edot/\\mdot)_{\\rmn{c,M}}\\sim 350(r_i/10^6\\mbox{cm})^{1/4}\\gamma_{i}^{3/4}T_i/m_e$, $\\gamma_\\infty$ is only reduced by a factor of $\\sim 10$. For even smaller $\\edot/\\mdot$, the baryons will dominate the energy of the flow even inside the photosphere, which results in a final Lorentz factor $\\gamma_\\infty\\sim\\edot/\\mdot$.  We present a detailed analytical model of the dynamics of the fireball in Section \\ref{secan}. This includes approximate results for the asymptotic value of the Lorentz factor and of the energy content in the pairs relative to that of the radiation, based on the initial temperature and initial velocity of the flow. Furthermore, we estimate where the pairs go out of equilibrium, the position of the photosphere, and the radius and optical depth at which the radiation fields start to deviate from their blackbody values. In Section \\ref{results} we show results from a numerical calculation to which the analytical model is compared. Next, based on the equation of radiative transfer, the comoving frame photon distribution function is shown to be very close to blackbody even out to quite small optical depths in these ultra-relativistic flows. The dynamical importance of baryons is described in Section \\ref{baryon}. There we obtain approximate results in different regimes characterized by the amount of baryon loading, and we integrate the dynamical equations with baryons included in order to show how they affect the asymptotic Lorentz factor. Finally, we discuss qualitatively several possible extensions of our model, including the effects of (1) muon pairs and even nucleon pairs which could be present at sufficiently high temperatures; (2) a temperature anisotropy at the inner boundary, and (3) magnetic fields. ",
        "conclusions": "\\label{conclusions}  Radiation energy can escape from a fireball in two different ways: If there is significant baryon contamination present, much of the energy will be converted into bulk kinetic energy (Shemi \\& Piran 1990). However, when the expanding atmosphere has swept up a significant amount of surrounding matter, kinetic energy can be converted into escaping radiation at the resulting shock front (M\\'{e}sz\\'{a}ros \\& Rees 1993). In a similar way, internal shocks due to a non-uniform velocity can convert kinetic energy into radiation (Rees \\& M\\'{e}sz\\'{a}ros 1994). The other mechanism for radiation escape is more direct: If the particle content is small, the fireball can become optically thin before being matter dominated.  In this paper we have considered an extreme case of the latter possibility, in which there are no baryons present. The opacity is then due to electron--positron pairs, resulting in a very large optical depth for temperatures greater than the electron mass. Further out in the flow the temperature decreases, pair creation is suppressed and annihilations freeze in; this results in a small but non-negligible amount of surviving pairs. The radiation force acting on the particles accelerates the pairs considerably, even after the flow has become optically thin. We found that the Lorentz factor of the flow approaches the constant value $\\gamma_{\\infty} \\sim 1.4\\times 10^{3} \\gamma_{i}^{3/4} (T_i/m_e)[(\\siga/\\sigt)+(\\sigs/\\sigt)]^{1/4}$ only when the optical depth falls below $\\tau_{\\gamma} \\sim 1.7 \\times 10^{-5} \\gamma_{i}^{3/4}$. This increases the asymptotic energy content of the pairs by a large factor, their fraction of the total energy approaching $\\sim 8.5 \\times 10^{-6} \\gamma_{i}^{3/4} [(\\siga/\\sigt)+(\\sigs/\\sigt)]^{1/4}$. The flow is always radiation dominated for reasonable values of the input parameters.  Even if the initial temperature were much higher than the electron mass, the resulting flow would not deviate significantly from an $e^\\pm\\gamma$ wind. If there are muons present, they will decay before the electron--positron annihilations freeze out, unless $\\gamma_i$ is very large. And even for $\\gamma_i$ high enough for the muons to survive until far outside the photosphere, their added inertia will only reduce the asymptotic Lorentz factor by at most 20 per cent. For even higher $T_i$, one may have nucleon--anti-nucleon pairs present in the flow. However, nucleon--anti-nucleon annihilations freeze out at a relatively low temperature, thus causing the baryon contamination in the flow to be negligible.  The photon distribution function in the comoving frame is very close to that of blackbody radiation. This is because $\\gamma \\propto r$ and $\\gamma T \\approx$ constant are excellent approximations in the flow until $r=r_{\\gamma}$ where the optical depth is $\\tau_{\\gamma} \\sim 1.7\\times 10^{-5} \\gamma_{i}^{3/4}$. Practically all the observed radiation therefore originates from a region where these two approximations hold. As was discussed in Section \\ref{spec}, the conditions $\\gamma \\propto r$ and $\\gamma T =$ constant imply that the equation of radiative transfer is solved by a blackbody distribution function in the comoving frame of the flow. The spectrum seen by an observer in the lab frame will deviate somewhat from a blackbody in that it has a broader peak and a shallower slope at low photon energies. Such spectra are {\\em not} typical of observed $\\gamma-$ray bursts, which are characterized by flat fluxes for logarithmic energy intervals (e.g. Schaefer et al. 1992, 1994; Kouveliotou 1994). A superposition of quasi-thermal spectra from numerous source regions radiating independently (but with different physical parameters) might produce flat spectra. The non-radial perturbation calculations discussed in Section \\ref{pert} lend partial support to this idea.  Ordered magnetic fields (whose energy content is small compared to that of the flow) will make the temperature and the velocity of the flow anisotropic and may enhance the bulk Lorentz factor of $e^\\pm$ pairs beyond the photosphere, but do not affect the spectrum seen by a distant observer significantly.   The results obtained for zero baryon number also apply when the baryon loading is sufficiently small. For very large baryon loading, the flow becomes matter dominated at optical depths larger than one, and in this case the asymptotic Lorentz factor $\\gamma_\\infty\\sim\\edot/\\mdot$. As $\\edot/\\mdot$ increases above $(\\edot/\\mdot)_{\\rmn{c,M}}\\sim 350(Z/A)^{1/4}(r_i/10^6\\mbox{cm})^{1/4}\\gamma_{i}^{3/4}T_i/m_e$, the asymptotic Lorentz factor at first levels off at $\\gamma_\\infty\\sim(\\edot/\\mdot)_{\\rmn{c,M}}$. For still larger $\\edot/\\mdot\\simgreat(\\edot/\\mdot)_{\\rmn{c,P}}\\sim7\\times10^4(Z/A) (r_i/10^6\\mbox{cm})(T_i/m_e)$, the asymptotic Lorentz factor rises $\\sim(\\edot/\\mdot)^{1/4}$, until $\\edot/\\mdot\\sim(\\edot/\\mdot)_{\\rmn{c,0}} \\sim(Am_p/2Zm_e)(\\edot/\\mdot)_{\\rmn{c,P}}$. The $\\mdot\\to 0$ limit applies for $\\edot/\\mdot>(\\edot/\\mdot)_{\\rmn{c,0}}\\sim 5\\times10^7(r_i/10^6\\mbox{cm})(T_i/m_e)$; in this regime, $\\gamma_\\infty\\sim(Am_p/2Zm_e)^{1/4}(\\edot/\\mdot)_{\\rmn{c,M}}$. The fraction of the total wind luminosity that emerges in the form of bulk kinetic energy falls below one at $\\edot/\\mdot\\sim (\\edot/\\mdot)_{\\rmn{c,M}}$, and decreases monotonically until asymptoting to a finite value $\\sim10^{-5}\\gamma_{i}^{3/4}$ as $\\mdot\\to0$ (see equation [\\ref{am_eratio}]). Thus, for all $\\edot/\\mdot>(\\edot/\\mdot)_{\\rmn{c,M}}$, the asymptotic Lorentz factor varies by a factor of only $\\sim(Am_p/2Zm_e)^{1/4}\\sim6$. The maximum possible $\\gamma_\\infty$ is the value found for $\\mdot=0$ and is finite. Although we have only considered steady winds here, it seems likely that similar results would hold for fireballs originated impulsively."
    },
    "9805/astro-ph9805281_arXiv.txt": {
        "abstract": "We derive $I$ band luminosity functions for galaxies in Abell 426 (Perseus and Abell 539, two rich, low galactic latitude clusters at moderate redshift. Cluster members are selected via the color-magnitude relation for bright galaxies. We find $\\alpha=-1.56 \\pm 0.07$ for Perseus over the range $-19.4 < M_I < -13.4$ ($15 < I < 21$) and $\\alpha=-1.42 \\pm 0.14$ for $-18.5 < M_I < -14$ ($17 < I < 21.5$) for A539. These LF's are similar to those derived in Virgo and Fornax, weakly supporting claims for the existence of a universal luminosity function for galaxies in clusters. ",
        "introduction": " ",
        "conclusions": "We derive $I$ band luminosity functions for galaxies in Abell 426 and Abell 539. For these clusters we find a good fit to a power-law with slope $\\alpha=-1.56 \\pm 0.08$ for Perseus and $-1.42 \\pm 0.14$ for A539. Since A426 is the more evolved of the two clusters, our findings run counter to the idea that cluster evolution destroys dwarf galaxies: on the other hand, star formation may be responsible for boosting the LF slope of dwarf galaxies in this system."
    },
    "9805/hep-ph9805419_arXiv.txt": {
        "abstract": "Decaying topological defects, in particular cosmic strings, can produce a significant flux of high energy neutrinos, photons and cosmic rays. According to the prevailing understanding of cosmic string dynamics in an expanding Universe, the network of long strings loses its energy first into string loops, which in turn give off most of their energy as gravitational radiation. However, it has been suggested by Vincent \\textit{et al.} (VHS) that particle emission may be the dominant energy loss channel for the long string network. In this case, the predicted flux of high energy particles would be much larger. Here we calculate the predicted flux of high energy gamma rays, neutrinos and cosmic ray antiprotons and protons as a function of the scale of symmetry breaking $\\eta$ at which the strings are produced and as a function of the initial energy $m_J$ of the particle jets which result from the string decay. Assuming the validity of the VHS scenario, we find that due to the interactions with the cosmic radiation backgrounds all fluxes but the neutrino flux are suppressed at the highest energies. This indicates that the observed events above the GZK cutoff can only be accounted for in this scenario if the primary particle is a neutrino and $\\eta$ is somewhat less than the GUT scale, i.e. $\\eta \\lesssim 10^{23}$ eV. The domain of parameter space corresponding to GUT-scale symmetry breaking is excluded also by the current observations below the GZK cutoff. A new aspect of this work is the calculation of the spectrum of the tau neutrinos directly produced in the decay of the X-particles. This significantly increases the tau neutrino signal at high energies in all cosmic string scenarios. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805295_arXiv.txt": {
        "abstract": "We present the results of near--infrared H band (1.65 $\\mu$m) imaging of 11 BL Lac objects with redshifts ranging from z = 0.05 to 0.9. We are able to clearly detect the host galaxy in seven low redshift (z$\\leq$0.24) BL Lacs, while the four unresolved BL Lacs have either high or unknown redshift. The galaxies hosting the low redshift BL Lacs are large (average bulge scale length R(e) = 8.8$\\pm$9.9 kpc) and luminous (average M(H) = --25.8$\\pm$0.5), \\ie slightly brighter than the typical galaxy luminosity L* (M*(H) = --25.0$\\pm$0.2), and of similar luminosity to or slightly fainter than brightest cluster galaxies (M(H) = --26.3$\\pm$0.3). The average optical/near--infrared colour and colour gradient of the BL Lac hosts (R--H = 2.2$\\pm$0.5; $\\Delta$(R--H)/$\\Delta$(log r) = --0.09$\\pm$0.04) are consistent with the hosts being normal ellipticals, indicating that the nuclear activity has only a marginal effect on the star formation history and other properties of the hosts. The BL Lac hosts appear slightly less luminous than those of higher redshift flat spectrum radio quasars. The nucleus--to--galaxy luminosity ratio of the BL Lacs is similar to that of low redshift radio galaxies and consistent with what found in previous optical studies of BL Lacs. However, it is smaller that that found for flat spectrum radio quasars, suggesting there is a difference in the intrinsic brightness of the nuclear source or in the Doppler beaming factor between the two types of blazars. ",
        "introduction": "BL Lac objects are active galactic nuclei (AGN) characterized by strong and rapidly variable continuum emission and polarization across the electromagnetic spectrum, strong compact flat spectrum radio emission and superluminal motion (see \\eg Kollgaard \\etal 1992 for general references). They share many properties with flat spectrum radio quasars (FSRQ) and are often grouped together as blazars. The clearest difference between them is that the latter have strong broad emission lines, while these are very weak or absent in BL Lacs.  Blazar properties are usually explained by the beaming model (Blandford \\& Rees 1978), where the observed emission is dominated by a synchrotron emitting relativistically boosted jet oriented close to our line--of--sight. This model is supported by the fact that almost all blazars are strong and rapidly variable $\\gamma$--ray sources (\\eg von Montigny \\etal 1995). The beaming model implies the existence of a more numerous parent population of intrinsically identical objects but with their jet oriented at larger angles to our line--of--sight. In the current unified models for radio--loud AGN (\\eg Urry \\& Padovani 1995), BL Lac objects are unified with low luminosity core--dominated (F--R I) radio galaxies (RG) seen nearly along the jet axis, while the high luminosity lobe--dominated (F--R II) RGs represent the parents of FSRQs (Padovani \\& Urry 1990; Urry, Padovani \\& Stickel 1991). However, for potential problems in this simple unification, see Urry \\& Padovani (1995). For a direct test of the unification model, we need to compare orientation--independent properties of BL Lac objects with those of the parents, \\eg extended radio emission, host galaxies and environments.  Considerable amount of optical imaging exists for relatively nearby (z$\\leq$0.5) BL Lac hosts (\\eg Abraham, McHardy \\& Crawford 1991; Stickel, Fried \\& K\\\"{u}hr 1993; Falomo 1996; Wurtz \\etal 1996; Falomo \\etal 1997b; Jannuzi, Yanny \\& Impey 1997). The host galaxies of nearby BL Lacs have turned out to be predominantly giant ellipticals with similar magnitude to F--R I RGs, although there appear to be some cases of disk dominated host galaxies (\\eg McHardy \\etal 1991; Abraham \\etal 1991; Stocke, Wurtz \\& Perlman 1995, but see contradicting views by \\eg Romanishin 1992; Stickel \\etal 1993; Falomo \\etal 1997a). The extended radio power and optical environments of BL Lacs are also consistent with those of F--R I RGs, but suggest a contribution to the parent population from F--R II RGs (Kollgaard \\etal 1992; Pesce, Falomo \\& Treves 1995; Wurtz \\etal 1997).  Very little near--infrared (NIR) imaging exists on BL Lac objects. However, NIR wavelengths may offer some advantages. Optical emission from BL Lacs is often dominated by the nuclear source, while the luminosity of the massive old stellar population peaks in the NIR. This leads to a better contrast of the nebulosity with respect to the nuclear source at these wavelengths. One also needs to apply much lower K--correction in the NIR than in the optical. In this paper we present NIR H--band (1.65 $\\mu$m) images of 11 BL Lac objects and compare the NIR host properties with those of RGs and FSRQs. The BL Lacs were observed during our project to study the host galaxies of a complete sample of FSRQs (Kotilainen, Falomo \\& Scarpa 1998; hereafter KFS98) and thus they do not satisfy any criteria of completeness. However, all the low redshift BL Lacs in this sample have previously been imaged in the optical by us. The same procedure of analysis was performed on the NIR and optical datasets, thus allowing us to investigate the R--H colour of the host galaxies in a homogeneous manner. Properties of the observed objects are given in Table 1. In section 2, we briefly describe the observations, data reduction and the method of the analysis and refer the reader to a more thorough discussion given in KFS98. Our results are presented in section 3 and conclusions in section 4. Throughout this paper, H$_{0}$ = 50 \\kmpspMpc and q$_{0}$ = 0 are used.  \\begin{table*} \\begin{center} \\begin{tabular}{llrllrrr} \\multicolumn{8}{c}{{\\bf Table 1.} Journal of observations.}\\\\ \\hline\\\\ \\multicolumn{1}{c}{Name} & Other name & z & V & M(B) & Date & Exp. time & FWHM \\\\ \\medskip & \t          &   &   &      &      & (min)    & (arcsec) \\\\ \\hline\\\\ PKS 0048--097 & OB-080   & $\\geq$0.5(?) & 16.3 & -- & 21/8/95 & 40 & 1.0 \\\\ PKS 0118--272 & OC-230.4 & $\\geq$0.557 & 15.9 & $\\leq$-26.8 & 18/8/95 & 37 & 1.0 \\\\ PKS 0521--365 &          & 0.055   & 14.6 & -22.3 & 13/1/96 & 21 & 1.2 \\\\ PKS 0537--441 &          & 0.896 & 15.0 & (-27.0) & 13/1/96 & 36 & 1.1 \\\\ PKS 0548--322 &          & 0.069   & 15.5 & -22.0 & 13/1/96 & 28 & 1.0 \\\\ PKS 1514--241 & AP Lib   & 0.049   & 14.9 & -21.7 & 19/8/95 & 15 & 0.9 \\\\ PKS 1538$+$149  & 4C 14.60 & 0.605   & 17.8 & -25.2 & 21/8/95 & 40 & 1.7 \\\\ PKS 2005--489 &          & 0.071   & 14.4 & -24.8 & 19/8/95 & 30 & 1.0 \\\\ MS 2143.4$+$070 &          & 0.237   & 18.0 & -22.8 & 18/8/95 & 40  & 0.9\\\\ PKS 2155--305 &          & 0.116 & 13.5 & -25.9 & 19/8/95 & 22 & 1.0 \\\\ PKS 2254$+$074  & OY 091   & 0.190   & 16.4 & -23.3 & 18/8/95 & 10 & 1.2 \\\\ \\hline\\\\ \\end{tabular} \\end{center} \\end{table*} ",
        "conclusions": "In this paper we have presented the results of a near--infrared imaging study of a sample of 11 BL Lac objects, for most of which the host galaxy is clearly resolved. Consistently with what is found in optical studies, we find that the host galaxies of low redshift BL Lacs are large (average bulge scale length R(e) = 8.8$\\pm$9.9 kpc) and luminous (average M(H) = --25.8$\\pm$0.5); they are more luminous than L$^*$ galaxies (by $\\sim$1 mag) but of similar luminosity to or slightly fainter than the brightest cluster galaxies. Our NIR study was able for the first time to address the issue of the optical--NIR colour of BL Lac host galaxies. The average R--H colour and colour gradient of the BL Lac hosts are consistent with those of non-active early-type galaxies, suggesting that the nuclear activity does not have much effect on the star formation history of the host galaxies. The nucleus--to--galaxy ratio of BL Lacs is similar to that found in low redshift RGs and consistent with what found in previous optical studies of BL Lacs. However, it is smaller that that found for the higher redshift FSRQs (KFS98), suggesting there is a difference in the intrinsic brightness of the nuclear source or in the Doppler beaming factor between the two types of blazars. We finally encourage a systematic NIR multiwavelength study of a large, well defined sample of BL Lac objects and their immediate environments with the new generation large NIR arrays."
    },
    "9805/astro-ph9805360_arXiv.txt": {
        "abstract": "Due to dramatic improvements in the precision of astrometric measurements, the observation of light centroid shifts in observed stars due to intervening massive compact objects (`astrometric microlensing') will become possible in the near future. Upcoming space missions, such as SIM and GAIA, will provide measurements with an accuracy of 4--60 $\\mu\\mbox{as}$ depending on the magnitude of the observed stars, and an accuracy of $\\sim 1~\\mu\\mbox{as}$ is expected to be achieved in the more distant future. There are two different ways in which astrometric microlensing signals can be used to infer information: one possibility is to perform astrometric follow-up observations on photometrically detected microlensing events, and the other is to perform a survey based on astrometric observations alone. After the predictable effects of the Sun and the planets, stars in the Galactic disk play the dominant role in astrometric microlensing. The probability that the disk stars introduce a centroid shift larger than the threshold $\\delta_{\\rm T}$ at a given time for a given source in the Galactic bulge towards Baade's window reaches 100\\% for a threshold of $\\delta_{\\rm T} = 0.7~\\mbox{$\\mu$as}$, while this probability is $\\sim 2\\%$ for $\\delta_{\\rm T} = 5~\\mbox{$\\mu$as}$. However, this centroid shift does not {\\em vary} much during the time in which a typical photometric microlensing event differs from baseline. So astrometric follow-ups (e.g.\\ with SIM) are not expected to be disturbed by the statistical astrometric microlensing due to disk stars, so that it is possible to infer additional information about the nature of the lens that caused the photometric event, as suggested. The probability to observe astrometric microlensing events within the Galaxy turns out to be large compared to photometric microlensing events. The probability to see a variation by more than $5~\\mbox{$\\mu$as}$ within one year and to reach the closest angular approach between lens and source is $\\sim 10^{-4}$ for a bulge star towards Baade's window, while this reduces to $\\sim 6\\cdot 10^{-6}$ for a direction perpendicular to the Galactic plane. For the upcoming mission GAIA, we expect $\\sim 1000$ of the observed stars to show a detectable astrometric microlensing signal within its 5 year lifetime.  These events can be used to determine accurate masses of the lenses, and to derive the mass and the scale parameters (length and height) of the Galactic disk. ",
        "introduction": "It is known for more than one decade (Paczy{\\'n}ski~\\cite{Pac1}) that the nature of matter between the observer and observed source stars can be studied by observing brightenings of a large number of these stars caused by the deflection of light by the intervening material. In addition to this magnification effect, there is also a shift in the light centroid of the observed star introduced by the lens object (H{\\o}g, Novikov, \\& Polnarev~\\cite{Hog}; Miyamoto \\& Yoshii~\\cite{miyamoto}; Walker~\\cite{walker}). Upcoming space missions will enable us to observe this centroid shift (Paczy{\\'n}ski~\\cite{Pac2}; Boden, Shao, \\& Van Buren~\\cite{BSV}). In particular, the Space Interferometry Mission (SIM, Allen et al.~\\cite{Allen1})\\footnote{for information about SIM see also {\\tt http://sim.jpl.nasa.gov}} will allow observations of selected targets with a positional accuracy of $\\sim 4~\\mbox{$\\mu$as}$ for sources brighter than $V = 20$. Moreover, the Global Astrometric Interferometer for Astrophysics mission (GAIA, Lindegren \\& Perryman~\\cite{lind})\\footnote{for information about GAIA see also {\\tt http://astro.estec.esa.nl/SA-general/Projects/GAIA/gaia.html}} will perform an astrometric survey aimed at all-sky coverage (Gilmore et al. 1998) with an accuracy of $20~\\mbox{$\\mu$as}$ ($60~\\mbox{$\\mu$as}$) for sources with $V < 12$ ($V < 15$).\\footnote{ Throughout the paper, we are talking about the accuracy of single astrometric measurements, not the accuracy of parallax measurements obtained from the mission within its lifetime.} These two missions are somewhat complementary: While SIM has the ability to point the instrument to selected targets, it will not perform a large survey program; on the other hand, GAIA will perform an all-sky survey, but will not have the ability to point the instrument to a selected target.  It has been mentioned (Paczy{\\'n}ski~\\cite{Pac2}; Boden et al.~\\cite{BSV}; H{\\o}g et al.~\\cite{Hog}; Miyamoto \\& Yoshii~\\cite{miyamoto}; Walker~\\cite{walker}) that the observation of the centroid shift during a (photometrically discovered) microlensing event will yield additional information about the lens, so that its mass, distance, and velocity can be determined unambiguously.  Most of the discussions in the literature so far have been confined to the centroid shifts of photometrically detected microlensing events which can be detected by an instrument like SIM (e.g.\\ Paczy{\\'n}ski~\\cite{Pac2}; Boden et al.~\\cite{BSV}).  It has been pointed out, however, that the microlensing cross-section for centroid shift measurements is much larger than the cross-section for light amplification (Paczy{\\'n}ski~\\cite{Pac3}; Miralda-Escud\\'e~\\cite{miralda}).  In this paper, we investigate the effects of disk stars on the astrometric microlensing signal (centroid shift). The disk stars can affect this signal in two ways.  First, for a microlensing event that has been detected by its photometric signal, the intervening matter can lead to additional centroid shifts and variations of these shifts with time, which disturb the signal of the centroid shift caused by the lens responsible for the photometrically detected microlensing event.  Second, the disk stars form a population producing microlensing events that can be detected by their astrometric microlensing signal alone in an astrometric survey such as GAIA.  This paper is organized as follows. We discuss the signals of photometric and astrometric microlensing in Sect.~2.  In Sect.~3, the optical depths due to photometric and astrometric microlensing and the differences are discussed. The characteristics of astrometric microlensing events and the prospects for disk stars as lenses are discussed in Sect.~4. In Sect.~5, we show that by observing astrometric microlensing events towards several directions, one can measure the scale parameters of the mass distribution of the Galactic disk. In Sect.~6, the effect of a luminous lens is discussed, while the implications for upcoming space missions are discussed in Sect.~7. Finally, in Sect.~8, the results of the previous sections are summarized. ",
        "conclusions": "Astrometric and photometric microlensing differ in two main points: First, the observed centroid shift is a function of both the dimensionless impact parameter $u$ and the angular Einstein ring radius $\\theta_{\\rm E}$ such that for a given $u$, the observed centroid shift is directly proportional to $\\theta_{\\rm E}$. On the other hand, the observed magnification is a dimensionless quantity which depends only on $u$ and not on any other scale. Second, for large angular separations between the lens and the source, the centroid shift, being proportional to $1/u$, falls off much more slowly than the photometric magnitude shift which is proportional to $1/u^4$.  Due to the dependence of the centroid shift on the angular Einstein radius, astrometric microlensing favors lenses close to the observer, while photometric microlensing favors lenses around half-way between observer and source. Therefore, one gets the largest centroid shifts from nearby objects, which are the Sun and the planets first, whose effect has to be corrected for, and then the disk stars. Because of the slower fall-off with the dimensionless separation $u$ in the astrometric case, detectable signatures occur for much larger angular separations, so that the average duration of an event $<\\!\\!t_{\\rm e}\\!\\!>$ can become much larger than the observation time $T_{\\rm obs}$. For the effect of luminous lenses this means that one can expect the lens to be resolved from the source star in some of the cases that show observable signatures. We have shown that the probability that a disk star introduces a centroid shift larger than a given amount $\\delta_{\\rm T}$ at a given time reaches unity for $\\delta_{\\rm T} \\sim 0.7~\\mbox{$\\mu$as}$ for sources towards the Galactic bulge at a latitude where the mass density of the disk stars is constant, which is a good approximation for Baade's window, while this probability is about $2\\%$ for $\\delta_{\\rm T} = 5~\\mbox{$\\mu$as}$ (see Table 1).  Though there is some chance that the centroid shift of a photometrically observed microlensing event, as observed e.g.\\ by SIM, is disturbed by disk star lensing (a 2nd lens), this additional centroid shift is not expected to vary much during the observation time ($\\sim$ several months), so that the effect expected is a slightly shifted position and the variation of the centroid shift is determined only by the primary lens.  Only if one extends the observations to $\\sim 10~\\mbox{yr}$ after the peak, one has to take the contamination by disk stars into account.  It is also interesting to examine the expected results from a microlensing survey looking for centroid shifts rather than the magnification of stars. As stated earlier, the largest centroid shifts come from nearby objects, which gives an opportunity to infer information about the disk stars. For $\\delta_{\\rm T} \\lesssim 10~\\mbox{$\\mu$as}$ and $T_{\\rm obs} \\lesssim 10~\\mbox{yr}$, $<\\!\\!t_{\\rm e}\\!\\!> \\gg T_{\\rm obs}$. Since one can only measure the variation in the centroid shift, not its actual value, and since the astrometric signal does not drop to zero within $T_{\\rm obs}$, the condition that the centroid shift exceeds the threshold $\\delta_{\\rm T}$ cannot be taken as criterion for an event. Instead, one has to rely strictly on the criterion that the centroid shift varies by more than the threshold $\\delta_{\\rm T}$. For $<\\!\\!t_{\\rm e}\\!\\!> \\ll T_{\\rm obs}$, as for most photometric microlensing events, these two criteria become equivalent. The probability that a source  star in the Galactic bulge towards Baade's window shows a centroid shift variation larger than $5~\\mbox{$\\mu$as}$ within one year is  $\\sim 10^{-3}$, which is about 3 orders of magnitude larger than the probabilities for photometric microlensing (see Table~\\ref{table:gamvar}). Among the events that show significant variations, only a fraction (10~\\% for $\\delta_{\\rm T} = 5~\\mbox{$\\mu$as}$) will have the closest angular separation between the lens and the source within the observing time, which will result in a clear `peak' signature, namely an observed change of sign of the component of the centroid shift parallel to the relative proper motion between lens and source, and a maximum of the centroid-shift component transverse to it. Since every event `peaks' once, the number of events that reach the peak within $T_{\\rm obs}$ is related to the event rate, while events that show significant variations only can show this variation in subsequent time intervals.  For an exponential decrease of the mass density along the line of sight (as it would be the case for lines-of-sight at high Galactic latitudes), the probabilities for events are proportional to the scale parameter in that direction if the source stars are at a distance of a few times the scale parameter or more. For sources perpendicular to the Galactic plane, the probability for a variation by more than $5~\\mbox{$\\mu$as}$ and a peak within $T_{\\rm obs} = 1~\\mbox{yr}$ is $\\sim 6\\cdot 10^{-6}$ (Table~\\ref{table:gamvarpeak}). By observing astrometric microlensing events in different directions, one can not only infer information about the total mass and the mass spectrum but also determine the scale length and scale height of the Galactic disk.  An advantage of astrometric over photometric observations is that the lens mass, distance, and velocity can be extracted individually from the observations (H{\\o}g et al.~\\cite{Hog}; Miyamoto \\& Yoshii~\\cite{miyamoto}; Walker~\\cite{walker}; Paczy{\\'n}ski~\\cite{Pac2}; Boden et al.~\\cite{BSV}).  We expect $\\sim 1000$ astrometric microlensing events to be detected by the GAIA mission during its lifetime of 5 years."
    },
    "9805/astro-ph9805226_arXiv.txt": {
        "abstract": "High-resolution and high S/N CCD spectra were analyzed to determine accurate LTE abundances in four $\\lambda$ Boo stars: $\\pi^{1}$ Ori, 29 Cyg, HR 8203 and 15 And. In general, 14 chemical elements were investigated. The main results are the following: all stars have a strong deficiency of the majority of investigated metals. Oxygen exhibits a moderate deficiency. The carbon abundance is close to the solar one.  The results obtained support an accretion/diffusion model, which is currently adopted for the explanation of the $\\lambda$ Boo phenomenon. ",
        "introduction": "Among the unresolved problems of stellar astrophysics, there is one linked with the $\\lambda$ Boo phenomenon. A comprehensive review of the $\\lambda$ Boo phenomenon was recently provided by St\\\"{u}renburg (1993) and by Paunzen et al. (1997).  To explain the $\\lambda$ Boo phenomenon, Venn and Lambert (1990) adopted an accretion hypothesis. According to that hypothesis, the chemical peculiarity of $\\lambda$ Boo stars originates due to the presence of a circumstellar shell. The circumstellar shell consists of two phases: gas and dust grains. The dust grains accumulate metals having a high condensation temperature (e.g. Si, Fe), but elements with lower condensation temperature (C, N) remain in the gaseous phase. Depleted gas from the circumstellar envelope is accreted by the star, while dust grains drift out of the shell due to radiative pressure.  Further studies of the proposed accretion scenario were made by Charbonneau (1991, 1993), who combined it with the theory of diffusion. Attempts to derive accurate elemental abundances in the atmospheres of $\\lambda$ Boo stars were undertaken in several works (Venn and Lambert, 1990; St\\\"{u}renburg, 1993, etc). ",
        "conclusions": ""
    },
    "9805/astro-ph9805010_arXiv.txt": {
        "abstract": "s{We discuss statistical and physical properties of cosmic microwave background polarization, both in Fourier and in real space. The latter allows for a more intuitive understanding of some of the geometric signatures. We present expressions that relate electric and magnetic type of polarization to the measured Stokes parameters in real space and discuss how can be constructed locally. We discuss necessary conditions to create polarization and present maps and correlation functions for some typical models.} ",
        "introduction": "Cosmic microwave background (CMB) anisotropies offer one of the best probes of early universe, which could potentially lead to a precise determination of a large number of cosmological parameters \\cite{jungman,1.bet,2.zss}. The main advantage of CMB versus more local probes of large-scale structure is that the fluctuations were created at an epoch when the universe was still in a linear regime. While this fact has long been emphasized for temperature anisotropies ($T$), the same holds also for polarization in CMB and as such it offers the same advantages as the temperature anisotropies in the determination of cosmological parameters. The main limitation of polarization is that it is predicted to be small: theoretical calculations show that CMB will be polarized at 5-10\\% level on small angular scales and much less than that on large angular scales. Future CMB missions (MAP, Planck) will have sufficient sensitivity that even such low signals will be measurable. This will allow one to exploit the wealth of information present in the polarization.  Recent work has emphasized the rich geometrical structure present in polarization \\cite{2.uros,2.kks,2.spinlong}. In particular, linear polarization has been decomposed into electric ($E$) and magnetic ($B$) types, which transform as scalars and pseudoscalars, respectively. With polarization there are three additional power spectra that can be measured, in addition to $E$ and $B$ autocorrelation there is also $E$ and $T$ cross-correlation. Each of these can provide unique information about our universe. Most of this work has developed the formalism by using multipole expansion on a sphere or on a plane. Here we will develop some of the properties of polarization fields $E$ and $B$ directly in real space, which allows for a more intuitive understanding of their geometrical properties. ",
        "conclusions": "Polarization has a rich geometrical structure, which can be simply understood using a real space construction of scalar and pseudoscalar fields. These can be constructed as integrals over Stokes $Q$ and $U$ parameters. Finite extent versions exist which allow one to search for $B$ polarization without measuring the whole sky.  Generation of polarization requires both Thomson scattering and significant quadrupole moment of photon distribution in electron rest frame. We give simple arguments why this is so and discuss predictions for some realistic models. Realistic attempts to extract polarization information have to include complications such as instrument noise and foregrounds and are given elsewhere in these proceedings \\cite{bouchet}."
    },
    "9805/astro-ph9805195_arXiv.txt": {
        "abstract": "The ESO Slice Project (ESP) is a galaxy redshift survey extending over about 23 square degrees, in a region near the South Galactic Pole. The survey is $\\sim 85\\%$ complete to the limiting magnitude $b_J=19.4$ and consists of 3342 galaxies with redshift determination. \\\\ The ESP survey is intermediate between shallow, wide angle samples and very deep, one--dimensional pencil beams; the spanned volume is $\\sim 5 \\times 10^4$ \\htre at the sensitivity peak ($z \\sim 0.1$). \\\\ In this paper we present the description of the observations and of the data reduction, the ESP redshift catalogue and the analysis of the quality of the velocity determinations. ",
        "introduction": "The ESO Slice Project (ESP) galaxy redshift survey, which is described in Vettolani et al. (1997, hereafter Paper I), extends over a strip of $\\alpha \\times \\delta = 22^o \\times 1^o$, plus a nearby area of $5^o \\times 1^o$, five degrees west of the main strip, in the South Galactic Pole region. The right ascension limits are $ 22^{h} 30^m$ and $ 01^{h} 20^m $, at a mean declination of $ -40^o 15'$ (1950). We have covered this region with a regular grid of adjacent circular fields, with a diameter of 32 arcmin each, corresponding to the field of view of the multifiber spectrograph OPTOPUS (Lund 1986, Avila et al. 1989) at the 3.6m ESO telescope. The total solid angle of the spectroscopic survey is 23.2 square degrees.  This paper presents the survey data (photometry, spectroscopy, completeness, etc.) which are necessary for a comprehensive study of the sample. It is organized as follows: in Section 2 we describe the photometric sample, in Section 3 the observations and data reduction and in Section 4 the redshift determination. In Section 5 we present the catalogue, in Section 6 we discuss the possible biases in the sample and the velocity errors, and finally Section 7 provides a summary. ",
        "conclusions": "We have described in detail the data of the ESP galaxy redshift survey, which extends over about 23 square degrees, in a region near the South Galactic Pole. The survey is $\\sim 85\\%$ complete to the limiting magnitude $b_J=19.4$ and consists of 3342 galaxies with redshift determination. Although not all galaxies have been observed and not all spectra have produced a measurable redshift, we have shown that these facts do not introduce any bias in the final spectroscopic sample. The only significant bias still remaining in the sample is due to the fact that close pairs of galaxies could not be observed in a single OPTOPUS (or MEFOS) observation. For this reason the fraction of not--observed objects is significantly higher than average for objects which have a companion in the photometric catalogue at a distance smaller than about 50 arcsec. \\\\ For all galaxies we have determined, when possible, both absorption and emission velocities. The median formal errors on the velocities are 64 and 31 km/s for the absorption and emission velocities, respectively. Analysis of the velocity measurements of the galaxies which have been observed more than once shows, however, that these formal errors are significant underestimates of the ``true'' errors. In first approximation the true errors can be obtained by multiplying the formal ones by factors of the order of 1.5 and 2.1 for $v_{abs}$ and $v_{emiss}$, respectively. \\\\ The data of the catalogue, available in electronic form at the the CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or via http://cdsweb.u-strasbg.fr/Abstract.html, provide all the information which is needed (i.e. positions, magnitudes, velocities, completeness) for statistical analyses of this sample, as for example the estimate of the luminosity function and mean galaxy density (Zucca et al. 1997)."
    },
    "9805/astro-ph9805126_arXiv.txt": {
        "abstract": "In a recent analysis of number counts in the ESP survey  Scaramella et al. (1998) claim to find evidence  for a cross-over to homogeneity at large scales, and against a fractal behaviour with dimension $D \\approx 2$. In this comment we note firstly that, if such a cross-over exists as described by the authors, the scale characterizing it is $\\sim 100 \\div 300 \\hmp$. This  invalidates the ``standard'' analysis of the same catalogue given elsewhere by the authors  which results in a ``correlation length'' of only $r_0 = 4 \\hmp$. Furthermore we show that the evidences for a cross-over to homogeneity rely on the choice of cosmological model, and most crucially on the so called K corrections. We show that the $D \\approx 3$ behaviour seen in the K-corrected data of Scaramella et al. is in fact unstable, increasing systematically towards $D=4$ as a function of the absolute magnitude limit. This behaviour can be quantitatively explained as the effect of K-correction in the relevant range of red-shift ($z \\sim 0.1 \\div 0.3$). A more consistent interpretation of the number counts is that $D$ is in the range $2 \\div 2.5$, depending on the cosmological model, consistent with the continuation of the fractal $D\\approx 2$ behaviour observed at scales up to $\\sim 100 \\hmp$. This implies a smaller K-correction. Given, however, the uncertainty in the effect of intrinsic fluctuations on the number counts statistic, and its sensitivity on these large scales to the uncertain K corrections, we conclude that it is premature to put a definitive constraint on the galaxy distribution using the ESP data alone. ",
        "introduction": "In a recent paper Scaramella \\etal (1998; S\\&C ) have applied the same statistical analysis to two deep surveys of large scale structure -the ESP galaxy survey, and the Abell and ACO survey of clusters - as performed by three of us and reported in Sylos Labini \\etal (1998; Paper 1). Despite the adoption of the same method, they",
        "conclusions": "More generally we emphasize that the degree of uncertainty in the results here is related to the fact that we are using only the radial counts from the origin, which are extremely sensitive to these effects which produce systematic distortions relative to this   origin. These effects will be much less  with the full correlation function, which averages over points. In a forthcoming work we will discuss in more detail the effect of different corrections on the various statistics. To draw strong statistical conclusions about the distribution at large scales like those which have been possible at moderate scales (up to $\\sim$ 100$\\hmp$) we will probably have to await the completion of the much larger surveys in progress.  Finally a brief comment on the cluster distributions, of which a detailed discussion is given in Paper 1. An analysis with the conditional average density $\\Gamma(r)$ is possible up to $\\sim 80 \\hmp$, and shows clearly defined fractal properties with $D \\approx 2$. The number counts from the origin show a quite fluctuating behaviour up to about $\\sim 100 \\hmp$, and clear incompleteness at scales considerably larger than this (evidenced by a steep decrease in the number counts). The identification of a range of scale in which the catalogue is `complete'  is itself strongly dependent on the assumption of homogeneity (as can be seen, for example, from Figure 4 showing the raw cluster data and the discussion of it in Scaramella \\etal, 1991 ). In our view the very good fit to a $D=3$ behaviour presented again in S\\&C (based on the analysis in Scaramella \\etal, 1991) is an artefact of this assumption rather than the identification of any real property of the spatial distribution of clusters."
    },
    "9805/astro-ph9805310_arXiv.txt": {
        "abstract": "Recent $N$-body simulations have shown that there is a serious discrepancy between the results of the $N$-body simulations and the results of Fokker-Planck simulations for the evolution of globular and rich open clusters under the influence of the galactic tidal field. In some cases, the lifetime obtained by Fokker-Planck calculations is more than an order of magnitude smaller than those by $N$-body simulations.  In this letter we show that the principal cause for this discrepancy is an over-simplified treatment of the tidal field used in previous Fokker-Planck simulations.  We performed new Fokker-Planck calculations using a more appropriate implementation of the boundary condition of the tidal field.  The implementation is only possible with {\\it anisotropic} Fokker-Planck models, while all previous Fokker-Planck calculations rely on the assumption of isotropy.  Our new Fokker-Planck results agree well with $N$-body results. Comparison of the two types of simulations gives a better understanding of the cluster evolution. ",
        "introduction": "Star clusters range in mass from a few hundred to several million solar-masses.  In order to understand their formation and dynamical evolution, detailed numerical modeling is required. There are, however, many effects which complicate their evolution and numerical models of star clusters are just beginning to incorporate deviations from the ideal star cluster (see Vesperini \\& Heggie 1997; Portegies Zwart et al.\\ 1998a).  \\nocite{vh97}\\nocite{pztl98}  Collisional $N$-body simulations are very expensive in terms of computer time. Even with supercomputers or special-purpose machines, it is impossible to do a simulation with the number of particles comparable to that of a real globular cluster. Therefore we are forced to rely on either $N$-body simulations with smaller number of particles or more approximate methods such as Fokker-Planck techniques. In theory, these two approaches should give identical results.  In order to check the reliability of the Fokker-Planck models with other models ($N$-body, gaseous, Monte-Carlo, etc.), some authors compared the results of various types of numerical simulations (Aarseth et al.\\ 1974; Giersz and Heggie 1994a, 1994b; Giersz and Spurzem 1994; Spurzem and Takahashi 1995). These comparisons demonstrate that for isolated clusters made of point masses the results of Fokker-Planck simulations are in good agreement with $N$-body computations.  Recent comparison between the same techniques for clusters in the galactic tidal field, however, gave a completely different view (Fukushige \\& Heggie 1995; Heggie et al.\\@ 1998); the result of the $N$-body simulations did not seem to converge to that of the Fokker-Planck simulations in the limit for $N \\rightarrow \\infty$, contrary to what was expected.  The disagreement between Fokker-Planck models and $N$-body models was even more clearly shown by Portegies Zwart et al.\\@ (1998b, PZHMM). They performed a series of $N$-body simulations with up to 32768 stars with identical initial conditions as one of the Fokker-Planck simulations of Chernoff and Weinberg (1990, CW).  The results of the computations of PZHMM can be summarized as follows: 1) The $N$-body model with the largest number of particles has a lifetime more than an order of magnitude longer than that of the comparable model of CW.  2) The lifetime of the cluster depends on the number of stars in a rather complex way. Since the fundamental assumption of Fokker-Planck calculations is that the evolution is independent of the number of stars, the results of PZHMM might imply that the results of Fokker-Planck calculations for clusters in a tidal field and with stellar evolution are of questionable validity.  The purpose of this letter is to explore what caused this discrepancy between the $N$-body models of PZHMM and the Fokker-Planck models of CW. ",
        "conclusions": "We have found the reason why the Fokker-Planck calculations of CW and the $N$-body calculations of Fukushige \\& Heggie (1995) and PZHMM gave very different results. The assumption of velocity isotropy and the over-simplified escape criterion (the energy condition and removing stars instantaneously) caused an enormous overestimate of the escape rate. By using an anisotropic Fokker-Planck model with an improved escape criterion, we have succeeded to achieve excellent agreement between Fokker-Planck and $N$-body results.  The dependence of the dissipation time on the number of particles is also understood. Stars need some time to travel away from the cluster in order to be gobbled up by the galaxy.  This timescale is of the order of a crossing time at the tidal radius.  Therefore the escape rate depends on the ratio of the relaxation time to the dynamical time, i.e.; on the number of stars."
    },
    "9805/astro-ph9805060_arXiv.txt": {
        "abstract": "We have discovered that the white dwarf PG 2329+267 is magnetic, and assuming a centered dipole structure, has a dipole magnetic field strength of approximately 2.3MG. This makes it one of only approximately 4\\% of isolated white dwarfs with a detectable magnetic field. Linear Zeeman splitting as well as quadratic Zeeman shifts are evident in the hydrogen Balmer sequence and circular spectropolarimetry reveals $\\sim$10\\% circular polarisation in the two displaced $\\sigma$ components of H$_{\\alpha}$. We suggest from comparison with spectra of white dwarfs of known mass that PG 2329+267 is more massive than typical isolated white dwarfs, in agreement with the hypothesis that magnetic white dwarfs evolve from magnetic chemically peculiar Ap and Bp type main sequence stars. ",
        "introduction": "The possibility that white dwarfs may possess large magnetic fields was first suggested in 1947 (Blackett 1947), however it was not until 1970 that the first detection was made (Kemp et al 1970). Since then 43 magnetic white dwarfs have been found, with field strengths ranging from $\\sim$0.1 to $\\sim$ 1000 MG. The vast majority (96\\%) of white dwarfs have as yet shown no sign of magnetic fields (Schmidt \\& Smith 1995) though this percentage may drop as surveys for magnetic fields are extended to lower field strengths. Serendipitous discoveries made during spectroscopic studies may also add to the number of known magnetic systems. What the actual percentage of magnetic white dwarfs is amongst the complete population, and what the distribution of field strengths amongst this set is, remains unclear.  A more complete knowledge of the magnetic properties of these stellar remnants, particularly at low (sub MG) magnetic field strengths, may allow us to deduce the role played by magnetic fields throughout the life time of the progenitor stars, as the magnetic fields found in white dwarfs are believed to be fossil fields preserved from earlier stages of stellar evolution. There is no known process for the generation of large scale magnetic fields during the degenerate phase of stellar evolution and so they are likely to be amplified versions of the fields which permeated their parent stars. White dwarfs with magnetic field strengths $>$1MG may be explained by evolution from chemically peculiar, magnetic Ap and Bp stars (Angel et al 1981), which have detectable magnetic fields from 100 to 10,000 G. White dwarfs with weaker magnetic fields would require their main sequence progenitors to have fields of only a few Gauss, below the current observational limits. This theory is supported by the similar space density of magnetic degenerates and the expected distribution of the remnants of magnetic main sequence stars (Sion et al 1988), as well as the observed tendency for magnetic white dwarfs to be more massive than non magnetic white dwarfs due to their proposed evolution from more massive progenitors.  The presence of a magnetic field has several detectable effects upon the spectrum of a white dwarf. For magnetic field strengths between 1 and 20 MG the linear Zeeman effect produces a distinctive triplet pattern for each absorption feature. Both the upper and lower atomic levels split into three energetically equidistant sub-levels. This allows transitions between the upper and lower levels to occur at three different energies. The wavelength of the central $\\pi$ component is unaffected by the presence of the magnetic field, however the two $\\sigma$ components are shifted, one to a longer ($\\sigma^{-}$) and one to a shorter wavelength ($\\sigma^{+}$). The degree of this separation ($\\pi - \\sigma$) is determined by the strength of the magnetic field (Landstreet 1994) according to,  \\[ \\Delta\\lambda_{L} \\simeq 4.7 \\times 10^{-7}\\lambda^{2}B_{s} \\hspace{1cm}(1) \\]  \\noindent where $\\lambda$ is measured in Angstroms and the average magnetic field strength over the visible hemisphere of the white dwarf, B$_{s}$ is measured in MG.  Above about 20MG the quadratic Zeeman effect dominates over the linear effect and the spectra become more and more complicated. Even at lower magnetic field strengths the quadratic Zeeman effect is noticeable as a blue shift in the wavelength of all the lines in the spectra. The size of the wavelength shift $\\Delta\\lambda_{Q}$, given by equation 2, is different for each line in the Balmer series, with the higher lines being shifted far more than H$_{\\alpha}$ (Preston 1970),  \\[ \\Delta\\lambda_{Q} \\simeq -5 \\times 10^{-11}\\lambda^{2}n^{4}B_{s}^{2} \\hspace{1cm}(2) \\]  \\noindent where n is the principle quantum number of the upper level of the transition, so for the Balmer series n = 3 for H$_{\\alpha}$ and n = 8 for H$_{\\zeta}$. This simple expression is based on perturbation theory and will break down for high n values, even at quite modest field strengths (Surmelian \\& O'Connell 1974).  The circular polarisation of the light can also be used to measure the magnetic field strength of a white dwarf. Even for weak magnetic fields ($<$1MG), where the Zeeman splitting is not obvious due to the large Stark broadening of white dwarf spectral features, the line profile is still a superposition of the unshifted $\\pi$ component and the two shifted $\\sigma$ components. In a longitudinal magnetic field the two $\\sigma$ components have opposite circular polarisations and hence even though the net circular polarisation of the line is zero, the offset $\\sigma$ components produce a distinctive S shaped feature in the circular polarisation spectrum. The percentage of circularly polarisation (V$_{\\%}$) is proportional to the longitudinal magnetic field strength B$_{e}$ and the normalised flux gradient of the zero field $\\pi$ line, as shown below, where $I_{\\lambda}$ is the flux.  \\[V_{\\%}(\\lambda) = 1.1 B_{e} \\left(\\frac{\\lambda}{4861}\\right)^{2} \\frac{1}{I_{\\lambda}} \\frac{dI_{\\lambda}}{d\\lambda} \\hspace{1cm}(3) \\]  Hence by measuring the degree of circular polarisation we can calculate B$_{e}$, the mean longitudinal magnetic field strength over the visible hemisphere of the white dwarf. ",
        "conclusions": "We have detected a magnetic field from the white dwarf PG 2329+267. The mean surface field strength measured from the degree of linear Zeeman splitting of the Balmer hydrogen lines is B$_{s}$ = 1.58 $\\pm$ 0.08 MG. Similar measurements from the quadratic Zeeman effect yield consistent results only for the Balmer lines up to H$_{\\delta}$, for higher lines the perturbation theory used to calculate the magnetic field strength begins to break down. We have detected approximately 10\\% circular polarisation at H$_{\\alpha}$ and have calculated the mean longitudinal magnetic field strength to be B$_{e}$ = + 462 $\\pm$ 60 KG. The ratio B$_{e}$/B$_{s}$ = 0.29 $\\pm$ 0.04 and the shape of the Zeeman split components suggest we are viewing the white dwarf at an inclination of $i = 60^{\\circ}$ $\\pm$ 5$^{\\circ}$ from the magnetic axis. At this inclination the dipole magnetic field strength will be 2.31 $\\pm$ 0.59 MG making PG 2329+267 the fourth weakest known isolated magnetic white dwarf. We have suggested that PG 2329+267 is more massive than most isolated white dwarfs which supports the hypothesis that magnetic white dwarfs evolve from chemically peculiar main sequence Ap and Bp stars.  \\subsection"
    },
    "9805/astro-ph9805256_arXiv.txt": {
        "abstract": "OSSE observed the transient black hole candidate GRO~J0422+32 (XN~Per~92) between 1992 August 11 and 1992 September 17.  High time resolution data were obtained in several energy bands over the $\\simeq$35--600 keV range with a timing resolutions of 8 ms.  Power spectra at energies below 175 keV show substantial low-frequency red noise with a shoulder at a few $10^{-2}$ Hz, peaked noise with characteristic frequency near 0.2 Hz, and a second shoulder at a few Hz.  The frequencies of the shoulders and the peak are independent of energy and source intensity.  The complex cross spectrum indicates that photons in the 75--175 keV band lag photons in the 35--60 keV band by a time roughly proportional to the inverse of the Fourier frequency.  The maximum lag observed is $\\simeq$300 ms.  The power and lag spectra are consistent with the production of the $\\gamma$ rays through thermal Comptonization in an extended hot corona with a power-law density profile. ",
        "introduction": "The hard X-ray transient GRO~J0422+32 (XN~Per~1992) was discovered by the BATSE instrument on the Compton Gamma Ray Observatory in data from 1992 August 5 (Paciesas et al. 1992), and at its peak reached an intensity in soft $\\gamma$ rays approximately three times brighter than the Crab Nebula and pulsar.  The source was observed by CGRO/OSSE beginning 1992 August 11, approximately at the peak of the outburst.  An optical counterpart was proposed by Castro-Tirado et al. (1992) and confirmed by the soft $\\gamma$-ray observations of SIGMA (Roques et al. 1994). While the mass function of $1.2 \\pm 0.04 \\Msun$ determined by Filippenko, Matheson, \\& Ho (1995) is low enough that the compact object might indeed be a neutron star, the H$\\alpha$ radial velocity curve and the M stellar type of the mass donor imply a mass of 3.6$\\Msun$ for the compact primary.  The photometric measurements of Callanan et al. (1996) support this mass estimate and give a distance estimate of $\\sim$2 kpc.  Broadband energy spectra from TTM, HEXE, and OSSE show that during outburst the source was in the X-ray low, hard state, which coincides with the breaking $\\gamma$-ray state (Grove et al. 1997, 1998, and references therein).  The gamma radiation is thus likely the result of thermal Comptonization in a hot corona near the accretion disk.  The $\\gamma$-ray spectrum hardened ($\\Delta kT / kT \\simeq +20\\%$) as the outburst declined (Grove et al. 1998).  Power spectra above 20 keV show significant red noise and peaked noise components frequently referred to as ``quasi-periodic oscillations'' (QPOs), even though they do not necessarily satisfy the width requirement (FWHM/$f_0 < $ 0.5) for such a label.  BATSE reported ``QPOs'' centered at roughly 0.04 Hz and 0.2 Hz (Kouveliotou et al. 1992), both of which were confirmed by SIGMA (Vikhlinin et al. 1995) and OSSE (Grove et al. 1992, 1994).  The spectral shape, rapid variability, and outburst lightcurve are similar to previous X-ray novae A0620-00 and XN~Mus~1991, both of which have measured mass functions that make them very strong black hole candidates (BHCs).  Based on these similarities, GRO~J0422+32 has been classified as a black hole candidate. ",
        "conclusions": "Generally similar power spectra have been reported from a number of black hole candidates, and beginning with Terrel (1972), they have frequently been modeled as arising from a superposition of randomly occurring bursts, or ``shots''.  If the shots have an instantaneous rise and exponential decay (or vice versa) with time constant $\\tau$, the resulting power spectrum is constant below the characteristic frequency $ 1 / (2 \\pi \\tau )$ and falls as $ 1 / f^2 $ at high frequencies.  This type of model can describe the two breaks and the $ 1/ f^2$ behavior above several Hz in the power spectrum of GRO~J0422+32 if there exist (at least) two independent shot components, with e-folding times $\\tau_s \\simeq 50$ ms and $\\tau_l \\simeq 2.1$ sec.  The PSD of the two-shot model is shown for the 75--175 keV band in Fig. \\ref{power_spec}a. Note that the best-fit values of the long and short e-folding times and the ratio of amplitudes of the two components are independent of energy (Table \\ref{shot_fit}).  Subtracting the PSD of the two-shot model from the observed PSD gives a peaked noise profile that is broad and asymmetric, with a sharp low-frequency edge and a broad high-frequency tail, as shown in Fig. \\ref{power_spec}b.  Plausible alternative descriptions of the continuum between 0.1 and 1.0 Hz, e.g. a simple power law with index $-0.9$, do not significantly alter the shape of the peaked noise, although they may change its amplitude.  The sharp low-frequency edge indicates that the physical process responsible for the peaked noise has a well-defined maximum timescale.  This process may perhaps be thermal-viscous instabilities in the accretion disk (Chen \\& Taam 1994) or oscillations in a Comptonizing corona (Cui et al. 1997).  We attempted to fit the total PSD with simple analytic forms---e.g. in the time domain, multiple exponentially-damped sinusoids; or in the frequency domain, multiple zero-centered Lorentzians to model the continuum and offset Lorentzians to model the peaked noise---but none of these adequately describes the sharp rise and broad fall of the peaked noise, nor do they add significantly to our understanding of the characteristic timescales represented in the PSD. Similarly, the scenario of Vikhlinin, Churazov, \\& Gilfanov (1994), in which shots arise from a common reservoir and are coupled through a weak amplitude or probability interaction that generates QPOs, also fails to describe the observed PSD in detail.  The lag spectrum (Fig. \\ref{lag_spec}) is generally similar to that of several other BHCs in the Ginga or Rossi XTE/PCA band (i.e. below 40 keV).  In the X-ray low, hard state, these include Cyg~X-1, GX339--4, and GS2023+338 (Miyamoto et al. 1992), and 1E1740.7--2942 and GRS~1758--258 (Smith et al. 1997). In the X-ray very high state, BHCs with similar lag spectra are GS1124--683 and GX339--4, subtype ``C+D'' for the latter object (Miyamoto et al. 1993).  Furthermore, the lag spectrum is quite similar to that between 20-50 keV and 50-100 keV from Cyg~X-1, which appeared to be essentially independent of the X-ray or $\\gamma$-ray state (Crary et al. 1998).  The present result is more evidence indicating that the frequency-dependent time lag is a common phenomenon shared by many accreting objects in binaries.  The observed power and lag spectra are at odds with the predictions of accretion models that produce most of the X-ray and $\\gamma$-ray emission from a region whose size is comparable to that of the last stable orbit around a black hole of mass a few M$_{\\odot}$.  The characteristic time scale associated with the dynamics of accretion in such an object is of order $10^{-3}$ sec; hence one would expect most of the associated power in the kHz frequency range.  By contrast there is a remarkable {\\it lack} of power at this range.  Furthermore, under these conditions the time lags, which in these models are indicative of the photon scattering time in the hot electron cloud, should be independent of the Fourier frequency and also of order $10^{-3} $ sec, the photon scattering time in this region.  Miller (1995) has argued that the observed time lags represent lags instrinsic to the soft seed photons, rather than the Comptonizing cloud.  However, Nowak \\& Vaughan (1996) have shown that any intrinsic lag is washed out if the observed photon energies are much greater than the seed photon energies, as is the case here, leaving again a frequency-independent lag due to the difference in scattering times across the cloud.  The discrepancy between observed and predicted power and lag spectra prompted an alternative approach proposed recently by Kazanas, Hua \\& Titarchuk (1997; hereafter KHT) and Hua, Kazanas \\& Titarchuk (1997).  These authors suggested that, while the process  responsible for the formation of the high energy spectra is indeed Comptonization, the hot, scattering electron cloud extends over several decades in radius with a power law profile in density, $n(r) \\propto 1/r^p$. This power-law density profile has a number of properties of interest in interpreting timing and spectral observations.  For a $\\delta-$function injection of soft photons at the center of the cloud, the light curves of the photons  emerging from the cloud at a given energy are power laws extending in time to $\\sim r/c$ ($r$ is the outer edge radius of the atmosphere) followed by an exponential cutoff. For small values of the total Thomson depth $\\tau_0$, the power-law index of the light curve is roughly equal to the power-law index $p$ of the density profile of the scattering cloud, becoming progressively flatter for increasing values of $\\tau_0$ and higher escaping energies (Fig. 1 in KHT). On the other hand, the corresponding light curves for clouds of uniform density are exponentials without power-law portions.  The time dependence of the photon flux can therefore be used to map the radial density profile  of the scattering cloud.   For a uniform cloud, the density profile has index $p=0$, and the light curve has no power law portion, i.e. the resulting PSD is that corresponding to an exponential shot.  For a density profile with index $p=1$ and total Thomson depths in the scattering atmosphere of a few, the PSD is $\\propto 1/f$ (KHT Fig. 1).  One should note that this form of the PSD assumes infinitely sharp turn-on of the shots at $t=0$.  As Kazanas \\& Hua (1997) have shown, a finite turn-on time $t_0$ will introduce an additional steepening of the PSD at frequencies $\\omega \\sim 1/t_0$ extending over a decade in frequency, yielding PSDs in agreement with those of Fig. \\ref{power_spec}a.  The great advantage of the present scheme is therefore the direct physical association of features in the PSD with properties of the source. Modeling of the light curves of GRO~J0422+32 with this type of shot indicates values for $t_0 \\approx 50$ msec.  The model presented in KHT provides constraints on the time lags that can be of great value in probing the structure of the scattering medium.  In the process of Comptonization, photons of energy $E_2$ lag in time behind photons of energy $E_1 < E_2$ simply because more scatterings are required to take a photon from $E_1$ to $E_2$.  The lag in time is proportional to the scattering time, which depends only on the density of the medium. Thus in general, for a uniform medium the lag time is constant (i.e. independent of the Fourier period).  However, in a medium with a power-law density profile, the hard photons sample a range of several orders of magnitude in density, which appears in the corresponding time lags.  In addition, because the probability of scattering at a given density range is constant for a medium with $p=1$, all lags should be present with equal weight, producing a time-lag function $\\propto 1/f$, with a maximum lag at the time scale corresponding to the scattering time at the edge of the power-law atmosphere.  Indeed, Fig. \\ref{lag_spec} is in excellent agreement with the above arguments (see Hua, Kazanas, \\& Cui 1997a for fits to similar lag spectra from Cyg~X-1, and Hua, Kazanas, \\& Cui 1997b for discussion regarding preliminary OSSE data from GRO~J0422+32)."
    },
    "9805/astro-ph9805242_arXiv.txt": {
        "abstract": "It has been suggested that the supermassive black holes, at the centers of galaxies and quasars, may initially form in single collapses of relativistic star clusters or supermassive stars built-up during the evolution of dense star clusters. We show that it may be possible for ICECUBE (a planned 1 km$^3$ neutrino detector in  Antarctica) to detect the neutrino bursts associated with those collapses at redshift $z\\la 0.2$ with a rate of $\\sim$ 0.1 to 1 burst per year. Such detections could give new insights into the formation of structure in the universe, especially when correlated with gravitational wave signatures or even gamma-ray bursts. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805309.txt": {
        "abstract": "We study the stability of trajectories near the disk plane of galaxy models with a triaxial dark matter halo component. We also examine the effect of weak discreteness noise, rapidly rotating bar perturbations and weak dissipation on these trajectories. The latter effect is studied  both by  adding a dissipative component to the force law and by using particle simulations. If the matter distribution is triaxial and has a constant density core, dissipation leads to  inflow of material inside the core radius to the centre (since no non-self-intersecting closed orbits exist in the central areas). This leads to the formation of central masses which in turn destabilise the  trajectories of any stars  formed in these regions. In particular, even if the gas  settles by dissipation into a flat disk, stars formed in that disk will later form a bulge like distribution the extent of which would be related to the core radius of the halo and the original asymmetry in the plane. The process of gas inflow is regulated by the fact that too large and condensed a central mass leads to the creation of stable closed loop orbits in the central area around which the gas can move. This would appear to stop the accumulation of central mass {\\em before} it becomes large enough for rapid loss of halo triaxiality. It was found that weak discreteness noise can increase the fraction of such trajectories significantly and can therefore have important consequences for the modelling of galaxies. The addition of rapidly rotating bar perturbations also increases the degree of instability dramatically. So if bars can form in triaxial haloes they are likely to be quickly destroyed leaving a bulge like structure behind (which may explain the absence of bars in surveys of high redshift galaxies). In they do survive their main effect on the gas dynamics is to create attractors other than the centre around which the gas can move. We discuss some possible consequences of the aforementioned effects. In particular, it is suggested that the halo core radius and initial asymmetry may be important in determining the relative disk-halo contribution to the rotation curve of a galaxy --- and hence its Hubble type. ",
        "introduction": "\\label{tax:wyax}  Numerical simulations of dissipationless collapse starting from cosmological initial conditions consistently predict triaxial final states for cold dark matter (CDM) halos (e.g., Dubinsky \\& Carlberg 1991; Warren et al. 1992; Cole \\& Lacey 1996). Considering the fact that collapsing spherical objects are unstable towards non-spherical perturbations (Lin, Mestel \\& Shu 1965), and that violent relaxation is not completely effective in washing out initial asymmetries (Aarseth \\& Binney 1978), these results are perhaps not that surprising. In fact, given the centrally concentrated  radial density distribution of the resulting halos, the observed asymmetries should be considered as lower limits, since, for such structures, many of the trajectories will be chaotic. As these trajectories explore their phase space, the equilibrium figure will tend to relax towards more symmetric shape. This process  would be expected to be strongly amplified by discreteness noise (Merritt \\& Vallury 1996 and Section~\\ref{tax:noise} of the present paper) which is much larger in numerical simulations than in CDM system.  Using parameters obtained from numerical simulations (of Dubinsky \\& Carlberg 1991), Kuijken \\& Tremaine (1991) estimated that a disk equipotential axis ratio of $\\sim 0.9$ can be inferred. Since then, there has been a number of papers suggesting that such a value is incompatible with observations. These include arguments that the scatter in the Tully-Fisher relation would be too large (Franx \\& de Zeeuw 1992).  As was pointed out  by Rix \\& Zaritsky (1995) however, the fact that the Tully-Fisher relationship is used for distance measurement might make the sample used  particularly biased against galaxies with large-scale distortions. These authors use the shapes of the isodensity contours from near-infrared images of external galaxies to obtain an axis ratio of $\\sim 0.95$ for the disk isopotentials --- this is consistent with a halo potential axis ratio $\\sim 0.9$, if the disk and halo contribute more or less equally to the potential. Another argument for axisymmetric haloes  makes use of the velocity dispersion of gas clouds  in the Milky Way. From such considerations Blitz \\& Spergel (1991) deduce that the potential axis ratio is near 1 (axisymmetric) unless the Sun lies near a symmetry axis of the potential. Kuijken \\& Tremaine (1994), on the other hand, study in detail  both local and global photometric and kinematic properties of our galaxy, and deduce that the Sun lies near the minor axis of the Milky Way (give or take 10 degrees), and that the data is consistent with a disk equipotential axis ratio of $\\sim  0.9$ (meaning a halo potential axis ratio nearer to 0.8). Less controversial are measurements of the flatness of the halo. These can be obtained  from observations of polar ring galaxies. They suggest a density axis ratio of about a half (e.g., Sackett \\& Sparke 1990; Sackett et al. 1994). In general, therefore, one should expect the dark halos of galaxies to be very aspherical --- in contrast to how they are usually modelled in simple treatments.  Whatever the shapes of haloes today, one may like to speculate about their evolution towards the present state and their influence on the early dynamics of disk galaxies. The fact that the flatness of haloes appears to be larger than the asymmetries in the disk plane, and that some studies constrain this asymmetry to be rather small, led some investigators to suggest that the dissipational formation of a disk in the centre of a triaxial halo causes the latter to evolve to more oblate nearly axisymmetric state: dissipation might lead to the formation of central concentration, which in turn might destroy the triaxiality (we consider this in detail later in this paper). However this is expected to be a slow process which should occur over many dynamical times. Indeed, triaxial figures of equilibrium with mild central density cusps have been successfully built in a self consistent manner (Merritt \\& Fridman 1996: see also Schwarschild 1993 for an application to the singular logarithmic potential), and numerical simulations seem to show that there is a threshold for the ratio of the central mass to that of the triaxial figure ($\\sim$ a few percent) below which rapid loss of  triaxiality does not occur (Merritt \\& Quinlan 1997).  Studies of dissipational galaxy formation (e.g., Katz \\& Gunn 1991) appear to contradict  this. It is  found that, in the presence of dissipation, the resulting haloes become nearly axisymmetric, and that this happened in the initial formation stages. However, modelling dissipational galaxy formation is not an exact science since the exact role and relative importance of the various  hydrodynamic and thermodynamic effects is still unclear. Even more difficult to predict is the effect of star formation and the accompanying energy feedback. These effects, if underestimated, can cause over-dissipation leading to the buildup of  central concentrations far greater than what can be realistically expected. In addition, the small number of particles used by Katz \\& Gunn (about 4000), means that the discreteness noise is large. As mentioned in the opening paragraph, this can lead to artificial relaxation. A more controlled study was undertaken by Dubinsky (1994) who considers the slow growth of a Kuzmin Toomre disk initially placed at the centre of a live halo and the motion of which is evolved as a particle in the simulation. His conclusion is that, in the presence of the disk, the halo orbits change adiabatically, in a way as to cause the halo to become  more oblate and less asymmetric in the disk plane. Nevertheless, the halo retains some of its triaxiality.  Indirect  evidence that some disk galaxies might have had a triaxial halo at one point in their history but may have evolved towards axisymmetry relates to how one interprets  the counter-rotating galactic disks that are observed (particularly the case of  the $ {\\rm S0}$ galaxy NGC 4550: Rubin et al. 1992; Rix et al. 1992). According to Evans \\& Collett (1994), these symmetric populations result from stars which were on box orbits when the potential was triaxial, and follow loop orbits when the potential is axisymmetric. The stars on the box orbits in the original triaxial potential are either formed on these eccentric orbits or subsequently ``heated'' into box orbits. In fact, if the inner parts of galaxies are non-axisymmetric and are dominated by a nearly harmonic constant density core, it would be natural for stars confined within that core to move on box orbits. Indeed, the high velocity dispersions found in the inner regions of galactic disks (Lewis \\& Freeman 1989) suggest that a large fraction of stars in these regions either once were or still are on box orbits. Unless the harmonic core is very large,  this population is likely to be confined to the inner few kpc. Moreover, we shall see in this paper, in the presence of a central mass, this group of stars can be transformed into a bulge population. This may explain why large scale counter-rotating disks are not commonly observed (Kuijken et al. 1996).  The work of Dubinsky (1994) and Evans \\& Collett (1994) are examples of the evolutionary effects that disks can induce on galactic halos and vice-versa. The two effects described above however depend on the properties of the regular orbits in the potential: how they can be arranged in different ways.  A richer variety of interesting phenomena  are related to changes in the qualitative structure of orbits --- that is the transition of orbits from regular to chaotic and vice-versa. In this case, a whole set of new phenomena may appear. Strongly chaotic orbits (visiting most of their allowed energy subspace) will usually have a time averaged density in the 3 dimensional configuration space that is much more round and isotropic than the underlying density distribution. The existence of a large number of these in a certain potential therefore implies that the velocity and density fields will evolve towards more isotropic distributions. An example of this effect is now well documented in the case of galactic bars. When a central mass is present, the once regular trajectories trapped around the $x_{1}$ closed periodic orbits (which support the asymmetric structure of the bar) become chaotic as the latter are destabilised --- leading to the dissolution of the bar and the growth of bulge like structures (e.g., Hasan et al. 1993; Pfenniger \\& Friedli 1991; Friedli \\& Benz 1993; Norman et al. 1995). Similar effects are also thought to be important in the evolution of elliptical galaxies, where it is thought that the presence of a central mass or a significant density cusp causes these galaxies to lose their triaxiality (Norman et al. 1985; Gerhard \\& Binney 1985; Merritt \\& Fridman 1996; Merritt \\& Quinlan 1997).  We wish here to study similar effects in models of disk galaxies with triaxial halos. In this situation, both  processes described above may act. One expects that the effect of chaotic behaviour will  modify the disk stellar distribution, while at the same time causing the halo to become more axisymmetric. The   model characteristics that are expected to be important in the determining the extent of chaotic behaviour (the degree of asymmetry and central concentration) are the same in this new context (Section~\\ref{statprop}  and~\\ref{dishac}), though the interpretations may be different. In addition, also interesting here are situations in which both a non-rotating triaxial halo and a rapidly rotating bar are present (Section~\\ref{tax:rapbars}). We will also be examining the effect of weak discreteness noise (Section~\\ref{tax:noise}) and weak dissipation (Section~\\ref{tax:disp}  and~\\ref{tax:disin}). We will find that the latter property has the interesting  effect of triggering significant inflow of material to the central regions of our models.   We will be concerned with  the stability of motion near the disk plane of fully three dimensional models. By symmetry, all orbits have to pass at least once by the disk plane. Trajectories that remain confined near the disk plane however are usually regular. Those exploring the full configuration space bounded by the (approximately ellipsoidal) zero velocity  surface are strongly chaotic (an exception  was found in the case  of models with rotating bars, but here one is mainly interested in the bars' effect on orbits in the plane in any case). Thus, testing for vertical stability near the disk plane, at the same time gives an idea of the fraction of chaotic orbits in a given model, while also determining the fraction of trajectories that would be vertically unstable from an initial distribution  started near the disk plane. These two related effects determine the plausibility of maintaining a triaxial halo, and a flat distribution of stars in the inner hotter areas of the disk. While a search confined to the vicinity of the disk plane is obviously not exhaustive, it should be representative since the three dimensional regular orbits are usually parented by those in the symmetry planes (e.g., Binney \\& Tremaine 1987 (BT)) and strongly chaotic orbits should pass at any given point in the disk plane with a vertical velocity close to zero (Poincar\\'e recurrence theorem). On the other hand, the reduction in the number of phase space  parameters to be searched, can be used in exploring  the extensive set of models which we will now describe. ",
        "conclusions": "\\label{tax:concpos}  \\subsection{Summary of results}  Triaxial structures are expected from the instability of gravitational collapse to non-spherical perturbations and the ineffectiveness of violent (or collisionless) relaxation in washing out completely the initial state. Triaxial dark matter halos  are predicted by hierarchical theories of galaxy formation  and although there is some doubt if they could survive dissipational collapse, it seems that this would bring about at most a rearranging of shape but not complete loss of triaxiality. Here we have mainly concentrated on models with  constant density core halos. These are favoured by observations but are not predicted by cosmological simulations which nevertheless are in agreement with other observed features (Section~\\protect\\ref{tax:halo}).   We have studied the orbital structure near the disk plane of disk galaxy models with triaxial halos.  This was done by examining the vertical stability and calculating the maximal Liapunov exponent of general orbits and studying the stability of periodic orbits that parent them. These effects were, as could be expected, in general correlated. There are three factors which mainly determine the orbital structure of our models: the non-linearity of the force field, the asymmetry in the density distribution (Section~\\protect\\ref{tax:staper}) and the presence of a rotating perturbation (Section~\\protect\\ref{tax:rapbars}). The nonlinearity is determined by the central concentration of the density distribution. In models with core radii of $R_{0}=2$ kpc significant instability is found around the 1:2 resonance of the $x$-axial orbit, which is moved inwards when the asymmetry is greatest. For models with very small core radius, widespread instability occurs even in the axisymmetric case. For models with larger core radius or which were disk dominated in the central areas, the major low order resonances were too far out to be significant even when significant asymmetry is present.  One way of enhancing nonlinearity is by adding a central mass concentration. Then, as is usually the case (e.g., Gerhard \\& Binney 1985), almost all box orbits are destabilised and replaced by either boxlets or chaotic orbits. The effect of the central mass can be understood in terms of the stability properties of the closed periodic orbits. Although the lower order resonances (e.g., the 1:2 resonance) appear at somewhat smaller radii when a central mass is present, the main effect causing the instability seems to be the creation and broadening of higher  order resonances. When the central mass is strong enough, the instability gaps on the $x$ axis orbit merge and, even for  relatively small central masses of about $10^{-5}$ the total galaxy mass at 20 kpc the axis orbits (which parent the boxes) are destabilized and are replaced by higher order (KAM islands known as) ``boxlets''  (Section~\\protect\\ref{tax:staper}). Increasing the central mass destabilizes these and the remaining stable orbits are mainly those of still higher order in the self replicating phase space hierarchy (e.g., LL). This leads to widespread chaotic behaviour in the region of the phase space once occupied by the box orbits. In addition, it was found that the destabilisation of the box orbits takes place out to many core radii. Therefore high energy stars born on eccentric orbits may be also be unstable. For central masses of up to $0.05 \\%$ the total galaxy mass at 20~kpc many trajectories in the region once occupied by the box orbits are either regular or have small Liapunov exponents (and therefore also small diffusion rates) and thus any resulting  evolutionary effects are expected to be slow. For central  masses of $\\sim 1\\%$ the galaxy mass (at 20 kpc) most orbits in the aforementioned region are strongly chaotic and quickly relax to an invariant distribution (Section~\\protect\\ref{dishac}).  Not all unstable periodic orbits are ``centrophylic'' boxlets. In fact many higher order looplets (in particular the 2:2 loops) which have a definite sense of rotation and do not pass near the centre, may be unstable in the presence of a central mass concentration. As might be expected, the fraction of looplet orbits is greater in more axisymmetric potentials, and, in general, the fraction of vertically unstable orbits was found to be larger in flatter potentials. Thus, the presence of a disk may, in some situations, increase the fraction of unstable general orbits. For although there are fewer box orbits to start with, many orbits with a definite sense of rotation (which are parented by higher order looplets) are unstable in the central areas and fewer orbits are parented by stable boxlets.   Even for highly asymmetric systems with strong central mass concentrations, and in the region of phase space once occupied by the box orbits, there remains a significant fraction of stable periodic orbits (albeit parenting  a small fraction of general orbits).  This will mean that even though there is widespread chaotic behaviour, trajectories in non-integrable but smooth galaxy potentials do not approximate strongly chaotic uniformly hyperbolic systems. Such   systems therefore are not structurally stable and thus their qualitative behaviour can be affected by small perturbations (Section~\\protect\\ref{tax:noise}). This was indeed found to be the case. It was found that perturbations that take many Hubble times to change the energy can have a significant effect on the orbital structure. For smaller perturbations, the effect was more or less equally likely to {\\em stabilize} the trajectories as to destabilize them. For stronger perturbations however the latter effect was much more predominant as these destroyed the intricate island structures. The fact that weak discreteness noise can have such an effect on the dynamics is important for both the estimation of the importance of discreteness effects in real systems and the faithfulness to which $N$-body simulations can reproduce the dynamics of galaxies. In particular, the triaxiality of the halos produced by $N$-body simulation should be taken as a conservative  estimate since these simulations employ far fewer particles than present in halos mainly composed of elementary particles.   In the absence of rapidly rotating bar perturbations, the nearly circular closed loops are stable. This is  due to the absence of resonances, which arises from the fact that the angular frequency and the $R$ and $z$ oscillation frequencies have  nearly constant ratios in the type of models chosen (this is  analogous to the situation  when   response is linear, in which case {\\em each} of the frequencies is constant). The addition of a rotating  bar  breaks this symmetry and therefore markedly increases the fraction of chaotic orbits present --- since now widespread chaotic behaviour can occur in {\\em both} regions once occupied by the main orbit families (boxes and loops). The presence of the rotating perturbation within the triaxial halo potential means that the potential is time dependent and that no transformation of coordinates can change that. This will lead to very complicated behaviour near the disk plane, with trajectories suffering large angle scattering. The result is that trajectories are much more irregular in the disk plane where the time dependence is greatest. These trajectories are sometimes not  $z$ unstable even though they may have a very large maximal Liapunov exponent. Thus some additional ``invariant'' may be at least approximately conserved. In this case therefore the maximal exponent does not provide complete information on the phase space transport properties. This was especially true if the bar has a rectangular  shape.   Central masses can be produced if small dissipative perturbations are given to the loop orbits in the inner areas near the halo core radius (where the 1:1 resonance occurs). Trajectories are then observed to spiral inwards in a few rotation times, even though the magnitude of the dissipation would require many Hubble times for this effect to be noticeable in the outer areas of the potential. This behaviour is easy to understand since closed loop orbits (which parent the general loop orbits) become more and more eccentric as one moves nearer to the center of the potential and do not exist at all deep inside the core, at the point where this happens (the ``separatrix'') dissipation rates are greatly enhanced. Beyond this, the trajectories do not find any closed loop orbits to oscillate around and quickly spiral to the centre. However once a strong central mass has formed, closed loop orbits are stabilised in the central area and dissipative trajectories may settle around them in stable limit cycles instead of spiraling to the centre. The build up of the central mass thus saturates. This provides a self regulating mechanism for the formation of central masses in galaxies, and may explain why black holes are observationally found to have masses smaller than those required to destroy the triaxial equilibrium (Merritt \\& Quinlan 1997). Since a central mass of $0.05 \\%$ of the total galactic mass at 20 kpc appears to be enough to essentially stop further gas inflow. And thus from the discussion earlier in this section one expects the evolution caused by the central mass to be slow. The time-scale for the formation of the central mass is about 1 Gyr.  In the presence of time-dependent forcing (that is a rotating bar), dissipative  trajectories were found to settle into long lived chaotic states where their radial coordinates oscillated around definite values. Such strange attractor states may explain some of the observed ring structure in galaxies (Gu et al. 1996)  --- provided of course that a bar can survive long enough in a triaxial halo potential. Trajectories on some of these attractors may also have large $z$ excursions and large radial motion through the galactic disk.    \\subsection{Possible consequences}  The effects described above may have the following  consequences. First there will be a redistribution of the gas accompanied by the fuelling of a central mass. In general therefore, gas  motion in nonrotating  non-axisymmetric potentials may be linked to the fuelling of central black holes or trigger bursts of star formation as the gas moving on self crossing trajectories is shocked and dissipates towards the centre as in the case of barred potentials (Beckman et al. 1991; Pfenniger 1993). This provides an alternative mechanism now that the correlation between the presence of galactic bars and of AGN's is thought to be weak (Mulchaney \\& Regan  1997).  In the absence of a central mass, and within the region delimited by the harmonic core (approximately equal to $R_{0}$ if the halo is dominant but decreases with increasing symmetry), only box orbits can exist and almost all of these are unstable to perturbations out of the disk plane in the presence of a central mass concentration. Therefore, initially any stars born from a settling disk in this region will become unstable when a central mass forms. The vertical instability will lead to the formation of bulge like structures even if this  system of stars, which is dynamically hot, is initially confined near the disk plane. As the central mass builds up, the harmonic nature of the potential is destroyed in its vicinity and closed loop orbits are stabilised at progressively larger radii within the core. Eventually, when the central mass is large enough, the harmonic core is completely destroyed. At that point the gas accretion also stops, since the gas can now settle on the newly created closed loops {\\em anywhere} inside the core region (Fig.~\\protect\\ref{dispc}). One then expects the end result of this process to be the formation of a bulge-like structure, the extent of which is similar to that of the original harmonic core and is formed from layers of unstable trajectories forming from inside out (see the discussion at the end of Section~ref{tax:dispdisk}).      It then follows that haloes with larger cores will be expected to develop centrally condensed ``bulgy'' disks. Once fully evolved these will then be expected to dominate the rotation curve in the inner areas. On the other hand, for halos with smaller core radii, this effect will be less significant and the halo would dominate the rotation curve at all radii. This suggests an explanation as to why in galaxies with widely different disk-halo contribution rotation curves one still gets these to be approximately (but not exactly:  Persic et al. 1996) flat over most of the detected radius of a galaxy. If this is the case then it will have to be that more massive haloes have larger core radii, since it is observed that galaxies in which are disk dominated near the centre have larger terminal rotation velocities.  Since the effective harmonic core radius in the disk plane for a given $R_{0}$ is also a function of the asymmetry in that plane, one may also suggest that larger mass haloes may have a larger asymmetry.  We have seen that, in the presence of bars rotating in triaxial halos, almost all orbits become chaotic. In addition, these orbits do not have a definite sense of rotation. It is not clear if such a bar  could form in the first place because of the large random motion (due to the triaxiality). However, if it is possible to still have a bar unstable disk inside a triaxial halo, it is likely to evolve a random bulge like structure extremely quickly as the bar is destroyed. It may be interesting to note here that  observations from the Hubble deep field survey (van den Bergh et al. 1996) suggest that barred galaxies are rare at high redshifts. It is also found that the fraction of late type galaxies is much lower than in standard catalogues. This may be expected if disks form inside strongly triaxial haloes where star formation may be caused by gas particles moving in the central regions on self-intersecting trajectories. In the outer regions, gas particles may move around (non-self-intersecting) closed loop orbits  so that the star formation rate is much smaller. In galaxies with small halo core radii (which according to our scenario would end up as late type galaxies), the more violent effects are limited to a small central region and the evolution will generally proceed at a much slower rate --- which means less of them are observed at high redshift.    Evidently, as the above processes unfold, the halo gradually loses triaxiality (because of the destruction of the box orbits which are crucial for maintaining the triaxial structure, and the settling of stars onto loop orbits which are alligned perpendicular to the halo). This also can bring an end to the gas inflow and the scattering of stars into a bulge. This process is not expected to be very fast however, since for a weak central mass there is usually a large fraction of regular and weakly chaotic (ie., with small Liapunov exponents and phase space diffusion diffusion rates) orbits remaining in the region once occupied by the box orbits. The threshold central mass needed for the majority of trajectories in that region to be highly chaotic (i.e., to have an exponentiation time-scale of a dynamical time or so) appears to be much larger than the mass needed to stop the accretion. Thus, such central concentrations would probably never be reached because of the self regulating nature of the accretion mechanism, which may explain why central concentrations as high as those apparently required for rapid evolution towards axisymmetry in elliptical galaxies are rarely observed (Merritt \\& Quinlan 1997). Therefore, one may expect there to be a period where the effects described above act on the stellar trajectories before the halo becomes completely axisymmetric (at which point it is then possible to develop a barred disk).    Although no precise determination of the time-scales of the processes described in this section can be obtained without full $N$-body simulation, including hydrodynamic effects, one can, to first order, perhaps give the following order of magnitude estimates. Since the time-scale over which originally planar trajectories 1 are destabilised corresponds to a few exponentiation time-scales, we expect that, after a significant central mass has formed (which takes in $\\sim 500$ Myr), a significant fraction of the orbits within the inner few kpc will be destabilised over a few dynamical times (less than a Gyr). In the outer areas this will correspond to a few Gyr. Complete loss of triaxiality and the evolution towards a final state should take about a Hubble time."
    },
    "9805/astro-ph9805074_arXiv.txt": {
        "abstract": "The positions in the H-R diagram of strongly magnetic Ap and Bp stars are compared with those of normal main sequence stars of types B7 to F2, with a view to investigating possible differences in evolutionary status between magnetic and non-magnetic stars. The normal B7--F2 stars fill the whole width of the main sequence band with some concentration towards the ZAMS, whereas the magnetic stars are only rarely found close to either the zero-age or terminal-age sequences. ",
        "introduction": "\\label{intr} The evolutionary status of Bp and Ap stars was hardly settled in the past. Estimates on their evolution were based on the membership of magnetic stars in open clusters or associations, on their membership in binary systems, or on indirect arguments inferred from the assumption of a rigid rotator geometry. More recently, it was advocated that Ap and Bp stars are distributed uniformly across the width of the main sequence (North 1993; Wade 1997), or alternatively that the magnetic stars are near the end of their main sequence life (Hubrig \\&\\ Schwan 1991; Hubrig \\&\\ Mathys 1994; Wade et al. 1996). One open question was  whether part of the apparent inconsistency between these results might be related to the fact that not all Ap and Bp stars are necessarily strongly magnetic.   Now, with the release of the Hipparcos data, it has become possible to determine the evolutionary state of magnetic stars with more reliability. In order to understand the physical processes taking place in B and A stars, we investigated possible differences of evolutionary state between magnetic and non-magnetic stars. ",
        "conclusions": "In order to test the theories of magnetic field origin, it would be important to probe the evolution of the magnetic field strength across the main sequence. Our sample of magnetic stars is too small to provide a really stringent test. The sample could be only marginally enlarged by incorporating stars with strong longitudinal fields and accurate Hipparcos parallaxes. Another relevant issue that should be considered is the evolutionary state of chemically peculiar B and A stars without detectable or with very weak magnetic fields.  The study of the magnetic field geometry in stars of various ages and rotation rates will provide important clues to test theoretical predictions. Several mechanisms have been proposed by which the angle between magnetic and rotation axes might change during the main-sequence life time. Therefore one goal for observers should be to provide theorists with constraints on the distribution of magnetic field geometries."
    },
    "9805/astro-ph9805304_arXiv.txt": {
        "abstract": "This paper investigates the hypothesis that the lensing objects towards the Large Magellanic Cloud (LMC) are brown dwarfs by analysing the effects of velocity anisotropy on the inferred microlensing masses.  To reduce the masses, the transverse velocity of the lenses with respect to the microlensing tube must be minimised. In the outer halo, radial anisotropy is best for doing this; closer to the solar circle, azimuthal anisotropy is best. By using a constraint on the total kinetic energy of the tracer population from the Jeans equations, the microlensing mass is minimised over orientations of the velocity dispersion tensor. This minimum mass is $\\approx 0.1\\,\\msun$, which lies above the hydrogen burning limit. This demonstrates explicitly that populations of brown dwarfs with smoothly decreasing densities and dynamically mixed velocity distributions cannot be responsible for the microlensing events. Brown dwarfs are no white knights!  There is one caveat.  If there are demons sitting on the microlensing tube, they can drop brown dwarfs so as to reproduce the microlensing data-set exactly. Such a distribution is not smooth and does not give well-mixed velocities in phase space. It is a permissible solution only if the outer halo is dynamically young and lumpy. In such a case, theorists cannot rule out brown dwarfs. Only exorcists can! ",
        "introduction": "The MACHO collaboration has interpreted its observations of microlensing events towards the Large Magellanic Cloud (LMC) as evidence that about one third of the halo of our own halo exists in the form of objects of around $0.5$ solar mass~(\\cite{al97a}). Unfortunately, there are seemingly insuperable objections to all the obvious candidates for the lensing population. Normal stars would be visible~(\\cite{al97a}), white dwarfs are ruled out by current population II abundance ratios~(\\cite{fms,gm}), while the Hubble Deep Field gives stringent restrictions on the contribution of red dwarfs~(\\cite{gf}). The microlensing events would be easier to understand if the characteristic mass of the lensing objects was below the hydrogen-burning limit ($\\approx 0.08 \\msun$). Of course, a lensing population of brown dwarfs would be much too dark to be visible and there is no conflict either with the metallicity data or the Hubble Deep Field star counts. So, it is natural to ask the questions: Can the deflectors be brown dwarfs? Is it possible that the masses of the microlenses have hitherto been overestimated? The aim of this Letter is to answer these questions.   Uncertainties in estimates of the lens candidates arise from two fundamental sources: low number statistics and modelling error. Although the number of microlensing events observed towards the LMC is still low, a determination of the average mass for a {\\em given} halo model can be obtained with perhaps 50\\% accuracy (see eg. Mao \\& Paczy\\'nski 1996; Alcock et al. 1997). This number is expected to improve substantially over the course of the next few years as new events are detected. A much more important source of error comes from our ignorance of the structure of the outer Milky Way halo.  The halo models used by Alcock et al. (1997) are either isotropic, such as the cored isothermal sphere~(\\cite{kim}), or they are very nearly so, such as the power-law models~(\\cite{wyn}).  Alcock et al. (1997; see especially Figures 17 and 24) plot likelihood contours in the plane of the lens mass and baryon fraction of the halo. The striking elongation of the contours along the baryon fraction axis suggests that there is comparatively little uncertainty in the mass estimates of the microlenses for a given model.  One worry is that this propitious state of affairs is a consequence of using halo models that all pretty much look the same!  In a percipient investigation, Markovi\\'c \\& Sommer-Larsen (1997) looked at a wider range of halos, including some with anisotropic velocity distributions. They found that $\\sim 100$ events (an order of magnitude more than presently available) are needed to estimate the average mass. This large error bar includes both the modelling and the statistical uncertainty. The claims of Markovi\\'c \\& Sommer-Larsen (1997) may be somewhat overstated because of the uniform priors used in their Monte Carlo simulations. However, Mao \\& Paczy\\'nski (1996) have also emphasised the difficulty of drawing firm conclusions about the mass distribution of the lenses from the limited sample available. These two papers consider both the statistical and the modelling uncertainties together. The focus of our paper is on the modelling uncertainty alone.  Our aim is to demonstrate unambiguously that the modelling uncertainties cannot be responsible for the high average mass estimates of Alcock et al. (1997).  Low mass lenses such as brown dwarfs are already ruled only for halo models with negligible rotation and isotropic velocity dispersions (e.g., Chabrier, Segretailn \\& M\\'era 1996). To rule out the hypothesis that the lenses are brown dwarfs requires a thorough investigation of halo models with very different kinematics -- in particular with different streaming velocities and different random motions.  Gyuk \\& Gates (1998) have already shown that rotating halos are unable to reduce the microlensing mass estimates below about $0.25\\,\\msun$ (unless all the lensing takes place very close to the Sun). This Letter will examine the effects of anisotropy and show that the associated {\\em modeling uncertainties} cannot cause the high lens mass estimates. ",
        "conclusions": "If the density of the microlenses is smooth and decreasing, then they cannot be brown dwarfs. This holds irrespective of the details of their kinematics. This general result follows because the Jeans equations (or, equivalently, the virial theorem) imply the existence of an irreducible minimum kinetic energy to support the lensing population against gravity. Even in the optimum alignment of the velocity dispersion tensor of the lenses, this must yield sufficient transverse motion so that the minimum mass is $\\approx 0.1\\,\\msun$ for halo models with flat rotation curves. This is above the hydrogen burning limit.  There is a way to save brown dwarfs.  Let us imagine a collection of demons sitting on the microlensing tube. One of the demons at a heliocentric distance of 20 kpc launches a brown dwarf of mass $0.06\\,\\msun$ with a velocity of $106\\, \\kms$ across our line of sight ... and this causes event \\# 4 with a blended timescale of 39.5 days. A second demon sitting on the microlensing tube at 30 kpc lobs a brown dwarf with a velocity of just $75 \\,\\kms$ ... and this gives event \\# 5 with a blended timescale of 55.5 days, and so on. Of course, demons can exactly reproduce the dataset reported by Alcock et al. (1997) by dropping brown dwarfs from the microlensing tube.  The density of brown dwarfs so produced is neither spherical nor axisymmetric nor in a steady-state. The velocity distribution is not dynamically well-mixed and the time averages theorem (Binney \\& Tremaine 1987, p. 171), which is the fundamental result underpinning steady-state stellar dynamics, does not hold. If it did, we could infer the existence of further brown dwarfs at different phases of the same orbits and show that they produce microlensing events that are not seen. Such a model is possible if the halo is very blobby (e.g.~\\cite{donald,donaldruth}). Then, in every direction that one looks (including $\\ell = 280^\\circ, b = -33^\\circ$), there may be garbage heaps of brown dwarfs whose density and velocity distributions are lumpy. This possibility cannot be ruled out from the microlensing data-set alone."
    },
    "9805/astro-ph9805132_arXiv.txt": {
        "abstract": "We investigate self-consistent particle acceleration near a pulsar polar cap (PC) by the electrostatic field due to the effect of inertial frame dragging. Test particles gain energy from the electric field parallel to the open magnetic field lines and lose energy by both curvature radiation (CR) and resonant and non-resonant inverse Compton scattering (ICS) with soft thermal X-rays from the neutron star (NS) surface. Gamma-rays radiated by electrons accelerated from the stellar surface produce pairs in the strong magnetic field, which screen the electric field beyond a pair formation front (PFF).  Some of the created positrons can be accelerated back toward the surface and produce $\\gamma$-rays and pairs that create another PFF above the surface. We find that ICS photons control PFF formation near the surface, but due to the different angles at which the electron and positron scatter the soft photons, positron initiated cascades develop above the surface and screen the accelerating electric field. Stable acceleration from the NS surface is therefore not possible in the presence of dominant ICS energy losses. However, we find that stable acceleration zones may occur at some distance above the surface, where CR dominates the electron and positron energy losses, and there is up-down symmetry between the electron and positron PFFs. We examine the dependence of CR-controlled acceleration zone voltage, width and height above the surface on parameters of the pulsar and its soft X-ray emission.  For most pulsars, we find that acceleration will start at a height of 0.5 - 1 stellar radii above the NS surface. ",
        "introduction": "The theory of particle acceleration in pulsar magnetospheres has been under development for almost three decades.  Although it was well known that rotating magnetic dipoles would induce electric fields in vacuum (Deutsch 1955), it took several years after the discovery of pulsars to realize that the huge vacuum fields could not in practice be available for particle acceleration. The electric field parallel to the magnetic field is at least partly screened by particles supplied from the stellar surface (Goldreich \\& Julian 1969) or by electron-positron avalanches (Sturrock 1971).  The true accelerating voltage of a pulsar must be determined by departures from the corotation, or Goldreich-Julian charge density, that could completely screen the parallel electric field.  Several types of models have studied pulsar acceleration due to charge deficits at different locations in the magnetosphere. Polar cap (PC) models consider the formation of a parallel electric field in the open field region near the magnetic poles, while outer gap models consider accceleration in the outer magnetosphere, near the null charge surface (see Mestel 1998 for the most recent and comprehensive review of pulsar electrodynamics). Ruderman \\& Sutherland (1975; hereafter RS75) introduced a PC model invoking a vacuum gap due to the trapping of ions in the neutron star (NS) crust.  The calculations by Jones (1985, 1986), and Neuhauser et al. (1986, 1987) seem to favor a low value for the work function (at least a factor of 10 less than it was thought earlier) in the NS surface with a strong magnetic field. The important implication of this study is that the possibility of free ejection of charges (actually of both signs) from the NS surface can be now, at least theoretically, justified. In this paper, we concentrate on a space-charge limited flow model (implying very low work function in the NS surface) based originally on the work of Arons \\& Scharlemann (1979; hereafter AS79), who determined the electric field produced by the small deparature from the Goldreich-Julian charge that grows above the surface due to the geometry of the open dipole field. The electric field accelerating electrons in this model developed along only field lines that curved toward the rotation axis (``favorably\" curved field lines), so that acceleration occurred over half of the PC.  The parallel field is shorted-out at a height above the surface where the $\\gamma$-rays from accelerated particles produce sufficient electron-positron pairs in the strong magnetic field.  The accelerating potential is thus limited by such a pair formation front (PFF).  These initial calculations of electron-positron PFFs assumed that the primary electrons began accelerating at the NS surface and that curvature radiation (CR) was the only mechanism for providing pair-producing photons (Arons 1983; hereafter A83). In recent years, it has become clear that inverse-Compton scattering (ICS) of soft thermal X-ray photons from the hot NS surface by the primary electrons is also an important mechanism above the PC. As well as being a significant energy loss (Kardash\\\"ev et al. 1984, Xia et al. 1985, Daugherty \\& Harding 1989, Sturner 1995; hereafter S95) and radiation (Sturner \\& Dermer 1994) mechanism, ICS can also provide photons capable of producing pairs. Pulsed X-rays have been detected from a number of pulsars which are consistent with blackbody spectra at temperatures around $10^5 - 10^6$ K (Ogelman 1995).  Zhang \\& Qiao (1996) and Zhang et al. (1997; hereafter ZQLH97) first explored the effect of the pairs from inverse-Compton photons on the acceleration in a Ruderman-Sutherland type model.  They found that ICS photons may produce a PFF sooner (at a lower altitude) than the CR photons would from the same accelerating electrons.  In fact in this case, the electrons will stop accelerating before they can emit significant CR.  The standard models of PC acceleration thus need substantial revision.  Another effect which has never been included in PC acceleration models is the formation of a lower PFF due to the positrons that are turned around and accelerated downward from the electron PFF. Although the number of positrons which are accelerated downward is small compared to the number of primary electrons and even to the charge deficit near the upper PFF, the multiplicity of the downward cascades is quite large (as we will discuss in Section 3.1).  Thus, the amount of charge produced by only a small number of downward moving positrons may be sufficient to establish a second PFF.  Although downward going cascades have been discussed in previous papers (see e.g. AS79), their effect on the acceleration of primaries has not been investigated. Daugherty \\& Harding (1996, hereafter DH96) qualitatively discussed the effects of pair cascades by returning positrons, their creation of pairs within the acceleration zone and the need for a self-consistent model of PC acceleration.  In this paper, we present a detailed study of the acceleration of primary electrons and secondary (downward-moving) positrons above a pulsar PC, assuming space-charge limited flow (free emission) of particles from the NS surface (see Harding \\& Muslimov 1998, for a review).  We include the general relativistic effect of inertial frame-dragging, which induces a much larger electric field than that expected in the flat spacetime and is not limited to favorably-curved field lines (Section \\ref{sec:Ell}). This is important because, as has been concluded earlier in papers by Fawley, Arons \\& Scharlemann (1977) and A83, the potential drops (derived for flat space-time) are not sufficient to account for oberved pulsar $\\gamma$-rays. Both electrons and positrons suffer energy loss and emit photons from CR and ICS.  The treatment of ICS of both upward and downward moving particles requires revisions from previous studies of only upward moving particles (Dermer 1990; hereafter D90; S95), which are presented in Section \\ref{sec:IC}. We then compute the location of both electron and positron PFFs due to one-photon pair production as a function of magnetic colatitude and height of the lower PFF.  We also discuss in Section \\ref{sec:pfraction} the fraction of positrons that are turned around at the upper PFF.  When the electrons are assumed to accelerate from the NS surface, we find that ICS photons produce the PFFs (Section 3.1), in agreement with the results of ZQLH97. However, we also find that there is substantial difference between the scattering of upward-going electrons and downward-going positrons by the same thermal X-ray photons: the electrons scatter these photons at angles less than $\\pi/2$, while the positrons scatter the photons head-on.  Electrons produce pairs through resonant scattering, while positrons produce pairs by scattering above the cyclotron resonance. The photons scattered by positrons are therefore more energetic and produce pairs in a shorter distance. These pairs may screen the accelerating field up to some distance above the surface.  We find that stable, double PFFs can form only when CR photons produce them, i.e. at a height where CR losses overtake ICS losses.  We compute the height of these stable PFFs as a function of pulsar period and surface field strength (Section 3.1). One interesting result is that the acceleration voltage limited by CR-controlled PFFs is only a function of magnetic colatitude (i.e. geometry of the open field lines), ranging between $\\sim 10^7$ and $3\\times 10^7\\, mc^2$, and is insensitive to pulsar parameters such as period and surface value of the magnetic field strength and even to the height of the acceleration.  The stable location of the lower PFF depends primarily on surface magnetic field, temperature and size of the hot polar cap, ranging between 0.5 and 1.0 stellar radius, but is insensitive to period. Implications of these results for high-energy pulsar emission are discussed in Section \\ref{sec:Dis}. ",
        "conclusions": "\\label{sec:Dis}  In this paper, we have investigated the effect of cascades from downward-accelerated positrons on the electrodynamics of the PC particle acceleration.  We find that when ICS produces pairs in the acceleration zone, the positron cascades may screen the accelerating electric field and disrupt particle acceleration near the NS surface.  Thus, if lower PFFs can develop, the picture of steady particle acceleration from the PC surface must undergo major revision.  We suggest that a stable acceleration zone may exist at an altitude of about one stellar radius above the PC, in which pairs from CR limit the electrostatic acceleration of primaries by screening the electric field near the upper PFF.  The calculations presented here are only a first attempt to describe the physics of what is a very complicated process.  We have made many assumptions and approximations to obtain our results.  While we believe that the gross qualitative results of our study are correct, there are a number of aspects which should be treated more accurately to achieve more solid quantitative results.  We have assumed that the screening of the $E_{\\parallel}$ occurs over the short distance determined by the upper boundary condition of Poisson's equation, not by the dynamics of the pair screening.  The details of the pair screening also determine the fraction of pairs that return to the PC and we have briefly outlined in Section (\\ref{sec:pfraction}) how such a calculation can be done within the electrodynamic framework set up in this paper.  A determination of the returning positron fraction would answer some interesting questions, such as whether PC heating is important relative to cooling in setting the PC temperature and if so, whether there is a feedback loop between the formation of the upper PFF and the PC heating.  The characteristics of the downward cascades need further study and modeling.  While we have done preliminary modeling of these cascades near the PC rim, there will be significant variation in pair yields with magnetic colatitude.  Near the magnetic pole, the pair multiplicity will drop, allowing the lower PFF height to decrease or even disappear.  We have found that the nature of the PC acceleration depends strongly on the characteristics of the radiation from the hot PC. Thus, to solidify the quantitative aspects of our results, it is important to treat the thermal PC radiation and the ICS process as accurately as possible. For example, we have assumed in this paper, that the thermal radiation is uniformly emitted over a PC of size $R_T = 3\\theta_0$ with an isotropic flux distribution.  However, studies of thermal radiation propagating through a strongly magnetized NS atmosphere (Pavlov et al. 1994) will not be isotropic, due to the anisotropy of the magnetized scattering cross section.  The expected radiation pattern consists of a pencil component, beamed along the magnetic field and a fan component perpendicular to the field. Such a beam pattern has been found to be consistent with observed thermal X-rays pulse fractions and pulse profiles for several pulsars (Shibanov et al. 1995, HM98).  In addition, our treatment of ICS uses a combination of the (non-relativistic) magnetized cross section in the Thomson limit for scattering near the fundamental cyclotron resonance, and the (non-magnetic) Klein-Nishina cross section to describe relativistic effects above the resonance. While this hybrid treatment is somewhat inaccurate for $B \\gsim 0.1\\,B_{\\rm cr}$ and therefore not completely satisfactory, the fully relativistic QED magnetic scattering cross section (e.g. DH86) is too complicated for use in this type of calculation.  In particular, the quantization of the electron momentum perpendicular to the field limits the number of Landau states contributing to the cross section for each incident photon energy.  The magnetic QED cross section for scattering just above the fundamental cyclotron resonance, and therefore the loss rate, could therefore be substantially lower than the Klein-Nishina cross section we have used here. Unfortunately, there exists no simplified, approximate expression for the QED scattering cross section, which smoothly bridges, and allows a unified treatment of, the relativistic resonant and non-resonant regimes of ICS.  A physical process which we have neglected in this study, photon splitting, is not expected to be significant for the magnetic fields we have considered, but will be very important for pulsars having surface $B_0 \\gsim 0.5\\,B_{\\rm cr}$. Photon splitting, a third-order QED process in which one photon splits into two, operates only in very high magnetic fields and competes as an attenuation process with one-photon pair production (Harding et al. 1997) because it can occur below pair threshold.  The implications of photon splitting for PC PFFs is profound.  Pulsars having $B_0 \\gsim 0.5\\,B_{\\rm cr}$ will produce fewer pairs, especially near the surface, so that the cascades from downward-moving positrons may not produce a lower PFF.  This is about the same field strength where we found that double PFFs controlled by resonant ICS at the NS surface become possible.  Thus, the stable, CR-controlled double PFF structure we have studied in this paper, which we found to move to higher altitudes with increasing surface field strength, will eventually collapse back to the surface at very high $B_0$.  A single PFF will then form, controlled by ICS, and the acceleration zone will have characteristics similar to that shown in Figure \\ref{fig:pffsx}.  At extremely high surface fields, $B_0 \\gsim 1.0\\,B_{\\rm cr}$ photon splitting will suppress pair creation completely at this single PFF, and the pulsar may be radio quiet (Baring \\& Harding 1998). However, the particle acceleration in these pulsars will operate very efficiently, free of any screening of $E_{\\parallel}$, so they are expected to be observable at high energies.  Bound-state pair production, where photons convert to positronium just below pair threshold rather than converting to free pairs (Usov \\& Melrose 1995), may also come into play at higher field strengths.  Some of the main results that we have presented in this paper have some important implications for pulsar high-energy emission.  One of these results is the insensitivity of the acceleration voltage (maximum particle energy) $\\gamma_{\\rm max}$ to any pulsar parameters such as period, surface magnetic field strength, obliquity (except for nearly orthogonal rotators), and even height. This result showed up many times in the course of our calculations (cf. Figs. \\ref{fig:pcacch}, \\ref{fig:pffxb}, \\ref{fig:pffsx}, \\ref{fig:pffp} and \\ref{fig:pffx}, and equation [\\ref{gmin_CR}]), and seems to be a robust characteristic of this type of PC acceleration model.  The maximum particle acceleration energy varies only within each pulsar, as a function of magnetic colatitude.  This energy, between $5 \\times 10^{12}$ eV and $5 \\times 10^{13}$ eV for CR-controlled acceleration zones, is about two to three times higher than that without frame-dragging (e.g. A83), and is consistent with the primary particle energy required in CR-initiated PC cascade models of $\\gamma$-ray pulsars (e.g. DH96).  The acceleration energy is very high near the magnetic poles, where the radius of curvature goes to infinity, allowing for the possibility of a narrowly beamed, hard component in $\\gamma$-ray pulses.  However, the radiation power emitted by these high-energy particles will be small, because the curvature radiation loss rate is proportional to $\\rho_c^{-2}$. The insensitivity of $\\gamma_{\\rm max}$ to pulsar parameters implies that the primary particles in all pulsars are accelerated to the same energy, and that the luminosity of the high-energy emission should depend only of the flux of primary particles.  This is consistent with trends in the observed $\\gamma$-ray pulsar luminosities (Thompson et al. 1997).  The $\\gamma$-rays will originate within a stellar radius of the upper stable PFF, at height $h_c = h_0 + S^-_c$.  From our results of Figs. \\ref{fig:pffh0} and \\ref{fig:pffp}, and equations (\\ref{h_0}) and (\\ref{Sc_CR}), the height of the $\\gamma$-ray production increases with period, roughly as $P^{1/2}$.  The standard PC half-angle at height $R_{\\gamma } = R + h_c$ will be $\\theta_c \\simeq (\\Omega R_{\\gamma }/c)^{1/2} \\propto P^{-1/4}$.  Thus, the $\\gamma$-ray emission solid angle $\\Omega_{\\gamma}$, which is expected to be, $\\Omega_{\\gamma} \\simeq 2\\pi [1-\\cos(3\\theta_c /2)]$, will be very weakly dependent on period and field strength.  Our conclusion that stable acceleration may occur in most pulsars at some altitude above the surface will have consequences not only for high-energy emission, but for radio emission as well.  If electron-positron pairs are necessary for coherent radio emission, then the dependence of the PFF altitude on pulsar parameters should be taken into account when determining the radio pulsar ``death line\", the line on the period-period derivative diagram beyond which pulsars are incapable of producing pairs.  Our calulcations in this paper suggest that pulsars with long periods do not produce PFFs, and that there will be a ``death line\" at periods below where there are observed radio pulsars.  But this is a long-standing problem of PC acceleration, most recently discussed by Arons (1998), and one that should be addressed in future studies.  Our principal findings can be summarized as follows.  \\begin{enumerate}  \\item Lower PFFs may form by positrons returning to the NS surface from the upper PFF.  \\item Pair creation by the ICS process dominates near the stellar surface, but is not symmetric for upward and downward going particles, so that stable, double PFF formation is very unlikely.  \\item Stable, self-consistent double PFFs can only exist when they are formed by CR. They can only form at a height above the surface where CR becomes dominant.  \\item The maximum particle energy is insensitive to any pulsar parameters such as period, surface magnetic field strength, obliquity, even height, and is sufficient to power $\\gamma$-ray pulsars.  \\end{enumerate}  The main conclusion of this paper is that the cascades from positrons returning to the PC may have a significant effect on the primary particle acceleration in pulsars and should not be neglected.  It is possible that lower PFFs do not form for all pulsars, and may not form over the entire PC.  A detailed study of the screening of the accelerating electric field by the returning positron cascades is beyond the scope of this paper, but will ultimately be necessary to understand PC acceleration. These studies are needed to address the questions of the returning positron fraction and the multiplicity of downward cascades.  If we can show that the returning positrons do not screen $E_{\\parallel}$, then the present assumption of acceleration right from the NS surface is valid.  But if these studies show that screening at a lower PFF is effective, then the possibility of acceleration above the NS surface must be incorporated in PC models.  We thank the referee Bing Zhang for his very careful review and insightful comments. We are also grateful to Joe Daugherty, for help in the downward cascade simulations, to Steve Sturner for discussions on inverse-Compton scattering and Matthew Baring for comments on the manuscript.  \\clearpage"
    },
    "9805/astro-ph9805181_arXiv.txt": {
        "abstract": "We consider constraints on the amplitude of mass fluctuations in the universe, $\\sigma_8$, derived from two simple observations: the present number density of clusters and the amplitude of their correlation function. Allowing for the possibility that the primordial fluctuations are non-gaussian introduces a degeneracy in the value of $\\sigma_8$ preferred by each of these constraints. However, when the constraints are taken together this degeneracy is broken, yielding a precise determination of $\\sigma_8$ and the degree of non-gaussianity for a given background cosmology. For a flat, $\\Omega_m=1$ universe with a power spectrum parameterized by a CDM shape parameter $\\Gamma=0.2$, we find that the perturbations are consistent with a gaussian distribution with $\\sigma_8=0.49^{+0.08}_{-0.07}$ (95\\% limits). For some popular choices of background model, including the favored low matter density models, the hypothesis that the primordial fluctuations are gaussian is ruled out with a high degree of confidence. ",
        "introduction": "Most studies of structure formation in the universe assume that the primordial density perturbations are gaussian. Standard inflationary theories predict gaussian perturbations, and the central limit theorem tells us that any theory involving the superposition of many random processes will give rise to approximately gaussian fluctuations. However, many well motivated theories predict non-gaussian initial conditions, including topological defect theories (Kibble\\markcite{k76} 1976) and certain forms of inflation  (Peebles\\markcite{P83,P97} 1983,1997). So far, no convincing observational evidence has been found to confirm or refute the gaussian hypothesis.  Clusters of galaxies, however, provide us with a unique probe of possible non-gaussianity. Being the most massive collapsed structures, they correspond to rare peaks in the primordial density field, so their statistics respond very sensitively to non-gaussianity in  the initial matter distribution. In addition, an analysis of the formation of clusters from given initial conditions requires primarily gravitational physics, and is largely free of complications from star formation and feedback. Finally, although the gravitational evolution which must be modeled is non-linear, well-tested analytical approximation schemes exist for studying the statistics of the resulting cluster distribution.  In this work we adapt these analytic schemes to the case of non-gaussian fluctuations, and use them to make predictions for two simple observations, the number density of clusters and the amplitude of cluster correlations. We show that the possibility of non-gaussianity introduces a degeneracy in the value of $\\sigma_8$ (where $\\sigma_R$ is the {\\it rms} overdensity in a sphere of radius $R\\;h^{-1}$Mpc) preferred by each of these observations. However, combining the two constraints breaks the degeneracy, allowing a precise determination of $\\sigma_8$ and the degree of non-gaussianity in the universe. ",
        "conclusions": "\\label{conclusions}  We have modified the Press-Schechter and the peak-background split formalisms to compute the cluster number density and cluster correlation amplitude in non-gaussian models. The best-fit value of $\\sigma_8$ and level of non-gaussianity depend on the choice of background cosmology, but we have demonstrated how these two quantities can be constrained simultaneously. For a flat, $\\Omega_\\Lambda=0$, $\\Gamma=0.2$ universe we find that the fluctuations are consistent with gaussian, with $\\sigma_8=0.49^{+0.08}_{-0.07}$ (95\\% limits). For low-$\\Omega_m$ models with $\\Gamma=0.2$, gaussian fluctuations are ruled out with a high degree of confidence, while non-gaussian fluctuations ($T>1$) are strongly preferred. Taken in conjunction with accurate knowledge of the background parameters (which we are rapidly gaining from supernovae and CMB observations), this method can provide a powerful constraint on the nature of density fluctuations in the universe.  Some concerns remain regarding the validity of our modification of the PS and peak-background split formalisms to the case of non-gaussian fluctuations, and some systematic errors may not yet be accounted for. This issue should be settled by an analysis of n-body simulations with non-gaussian initial conditions. However, our formulae have been well tested for the gaussian case, so the result that gaussianity is ruled out for certain choices of background model is not subject to these uncertainties. At the very least, we have demonstrated the power of two simple and well measured datasets, the cluster temperature function and the amplitude of cluster correlations, to simultaneously constrain the amplitude of mass fluctuations and the level of non-gaussianity in the universe.  We would like to thank P.~Ferreira, M.~Davis and M.~Markevitch for helpful discussions. This work has been supported in part by a grant from the NSF, and E.G. acknowledges the support of an NSF Graduate Fellowship.  \\begin{figure*}[t] \\centerline{\\psfig{file=pdf.ps,width=3.5in}} \\caption{ Three PDFs. The heavily shaded region shows the area contributing to peaks of height $3\\sigma$ and above, while the lightly shaded region shows the additional contribution to peaks of height $2.5\\sigma$ and above. } \\label{fig-pdf} \\end{figure*}  \\begin{figure*} \\centerline{\\psfig{file=data.ps,width=3.5in}} \\caption{ The cluster temperature function (black jagged line, top panel), and the cluster correlation length $r_0$ (datapoints, lower panel), with predictions for four models ($\\Omega_m=1.0$, $\\Omega_\\Lambda=0.0$, and $\\Gamma=0.2$, $\\sigma_8$ and $T$ labeled in lower panel). } \\label{fig-data} \\end{figure*}  \\begin{figure*} \\centerline{\\psfig{file=pdf4.ps,width=3.5in}} \\vskip 0.5in \\caption{ Confidence intervals (68\\%---dark band, 95\\%---light band) in the $T$ vs.~$\\sigma_8$ plane for the cluster temperature function (green band, top left to bottom right) and the cluster correlation length (red band, bottom left to top right). The blue dotted line corresponds to gaussianity ($T=1$), and the panels are labeled with the values of $\\Omega_m$, $\\Omega_\\Lambda$, and $\\Gamma$. } \\label{fig-pdf4} \\end{figure*}  \\begin{table*} \\centering \\begin{tabular}{ccccc} \\tableline \\tableline $\\Omega_m$& $\\Omega_\\Lambda$ & $\\Gamma$ & $\\sigma_8$  &      $T$  \\\\ \\tableline 1.0      & 0.0            & 0.1    & $0.42^{+0.08}_{-0.05}$ & $3.8^{+1.7}_{-1.2}$ \\\\ 1.0      & 0.0            & 0.2    & $0.49^{+0.08}_{-0.07}$ & $2.0^{+1.7}_{-1.1}$ \\\\ 1.0      & 0.0            & 0.3    & $0.56^{+0.10}_{-0.08}$ & $0.9^{+1.3}_{-0.7}$ \\\\ 0.3      & 0.0            & 0.1    & $0.59^{+0.15}_{-0.09}$ & $5.8^{+2.8}_{-1.8}$ \\\\ 0.3      & 0.0            & 0.2    & $0.71^{+0.17}_{-0.13}$ & $3.8^{+3.2}_{-2.1}$ \\\\ 0.3      & 0.0            & 0.3    & $0.83^{+0.21}_{-0.16}$ & $2.1^{+3.2}_{-1.6}$ \\\\ 0.3      & 0.7            & 0.1    & $0.61^{+0.15}_{-0.10}$ & $6.3^{+3.1}_{-2.1}$ \\\\ 0.3      & 0.7            & 0.2    & $0.73^{+0.19}_{-0.12}$ & $4.0^{+3.6}_{-2.0}$ \\\\ 0.3      & 0.7            & 0.3    & $0.87^{+0.21}_{-0.17}$ & $2.3^{+3.5}_{-1.7}$ \\\\ \\tableline \\end{tabular} \\vskip 10pt \\caption{ Best fit values of $\\sigma_8$ and $T$, with 95\\% confidence limits. } \\label{tab} \\end{table*}"
    },
    "9805/astro-ph9805238_arXiv.txt": {
        "abstract": "We present an investigation of the scale-dependence of bias described by the linear model: $(\\delta \\rho({\\bf x})/\\bar{\\rho})_{g} = b (\\delta \\rho({\\bf x})/\\bar{\\rho})_{m}$, $b$ being the bias parameter, and $\\rho({\\bf x})_{g}$ and  $\\rho({\\bf x})_{m}$ are the galaxy number density and mass density, respectively. Using a discrete wavelet decomposition, we show that the behavior of bias scale-dependence cannot be described by one parameter $b$. In the linear bias model the scale-dependence should be measured by the $j$-spectra of wavelet-coefficient-represented bias parameters $\\tilde{b}^{(n)}_j$ and $b_j^{(n)}$, $n$ being  positive integers. Because $\\tilde{b}^{(n)}_j$ with different $n$ are independent from each other, a systematic analysis of the $j$-spectra of $\\tilde{b}^{(n)}_j$ and $b_j^{(n)}$ is necessary.  We performed a $j$-spectrum analysis for samples of elliptical and lenticular (EL), and spiral (SP) galaxies listed in the APM bright galaxy catalog. We found that, for statistics of two-point correlation functions or DWT power spectrum, the scale-independence holds within 1 $\\sigma$. However, the bias scale-dependence becomes substantial when phase-sensitive statistics (e.g. $\\tilde{b}^{(n)}_j$ with $n>2$ or $b_j^{(n)}$) are applied. These results indicate that the bias scale-dependence has the same origin as the non-Gaussianity of galaxy distributions. This is generally consistent with the explanation that the bias scale-dependence originated from non-linear and non-local relationship between galaxy formation and their environment. ",
        "introduction": "Bias is introduced to reconcile the amplitude of fluctuations inferred from clustering of galaxies with that derived from mass distributions. Bright galaxies seem to have a stronger clustering  than that of the underlying mass. Therefore, it is generally believed that galaxies are biased tracers of the mass density field, i.e. it follows the clustering of (dark) matter but with an enhanced amplitude.  Generally, bias is phenomenologically modeled by a linear relation as \\begin{equation} \\delta({\\bf x})_{g} = b \\delta({\\bf x})_{m}. \\end{equation} where $\\delta({\\bf x})_{g} = [n({\\bf x})-\\bar{n}]/\\bar{n}$, $\\delta({\\bf x})_{m} = [\\rho({\\bf x})-\\bar{\\rho}]/\\bar{\\rho}$, $\\rho({\\bf x})$ and $n({\\bf x})$ are, respectively, the galaxy number density distribution and  density field of dark matter, and $\\bar{\\rho}$ and $\\bar{n}$ being the average of $\\rho({\\bf x})$ and $n({\\bf x})$. The phenomenological bias parameter $b$ is assumed to be constant, but may be different for different galaxy types, say, $b_{early}$, $b_{late}$ for early and late types of galaxies.  Eq.(1) and its variants are widely used in the determination of cosmological parameters from samples of redshift surveys of galaxies. However, we have really very little idea about which the bias parameter $b$ should be. In fact, both bias ($b>1$) and anti-bias ($b<1$) are employed in current data analysis (e.g. Mo, Jing \\& White 1996.) This prevents unambiguous measures of cosmological parameters, giving only bias-contaminated results.  Theoretically, the physical mechanism responsible for relation (1) is far from clear. The first analytic model of bias, in which objects are identified with high peaks or collapsed halos of the density field, offers a plausible explanation of the bias of galaxy clusters -- the correlation amplitudes of clusters are strong functions of cluster richness (Kaiser 1984). In this case, bias is mainly caused by the mass of collapsed halos. The larger the mass of the halo, the higher the richness of the cluster.  Yet, the formation of galaxies doesn't depend only on the mass, or local mass density, of collapsed halos, but is substantially modulated by various environmental effects, such as the suppression of star formation in neighboring protogalaxies (Rees 1985), stimulating the formation of nearby galaxies (Dekel \\& Rees 1987), dynamical friction effects (Couchman \\& Carlberg 1992), etc. All these environmental effects are beyond local density, and lead to a non-local relation between the number density of galaxies and the background mass field (Bower et al. 1993). Moreover, the rate of star formation is most likely non-linearly dependent on local mass density. A common result of the non-local and/or non-linear relation between $\\delta_{g}$ and $\\delta_{m}$ is the scale-dependence of parameter $b$. Therefore, to have a deep understanding of the mechanism of galaxy bias, searching for the scale-dependence of parameter $b$ is necessary  (Coles 1993, Catelan et al. 1994.)  So far, the results of detecting $b$ scale-dependence are quite scattered. For instance, the values of the bias-contaminated density parameter $\\beta=\\Omega^{0.6}/b$ are found to be in the range of 0.4 - 1, and equal to about $\\sim 0.5-0.6$ at Gaussian smoothing scales of 3-6 h$^{-1}$Mpc, and $\\sim 1$ on scales of $\\sim 12$ h$^{-1}$Mpc (e.g. Dekel, Burstein \\& White 1996). This is, $b$ is probably scale-dependent from 6 to 12 h$^{-1}$ Mpc. On the other hand, some studies conclude that for galaxies of all types and luminosities the scale dependence of the bias parameter is weak (e.g. Kauffmann, Nusser \\& Steinmetz 1997). Why different detections gave different conclusions? This question motivated us to study the physics included in Eq.(1).  In the first part of this paper, we show that even in linear bias model, the behavior of scale-dependence cannot be described by one parameter $b$, but by a series of $j$-spectra of wavelet-coefficient-represented bias parameters $\\tilde{b}^{(n)}_j$ and $b_j^{(n)}$, $n$ being  positive integers. The $j$-spectra of $\\tilde{b}^{(n)}_j$ are all statistically {\\it independent}. {\\it Different} detections may actually measure {\\it different} parameters $\\tilde{b}_j^{(n)}$. Therefore, it should not be surprised that some detections are positive, and some negative. The different behavior of bias scale-dependence given by different detection does not cause confusion, but may greatly be helpful to reveal the physics behind the bias model (1). Therefore, to have a complete picture of bias scale-dependence a systematic analysis of the $j$-spectra of $\\tilde{b}^{(n)}_j$ and $b_j^{(n)}$ is necessary.  In the second part, we performed a systematic detection of the bias scale-dependence with the samples of galaxies listed in the APM bright galaxies catalog (APM-BGC). This analysis shows that for the APM-BGC sample, the scale-independence approximately holds if only the two-point correlation function and power spectrum are involved, while the scale-dependence becomes substantial when higher order or phase-sensitive statistics involved. This result is worth to constrain models of hierarchical clustering of galaxies.  The paper is organized as follows. In \\S 2 we briefly introduce a method of the space-scale decomposition based on discrete wavelet transform (DWT) analysis. With this method, various statistics of measuring the bias scale dependence are developed and presented. \\S 3 describes the sample to be analyzed. In \\S 4 we apply the DWT method to analyze the galaxy sample of the APM bright galaxies. The implications of these results are discussed in \\S 5. ",
        "conclusions": "We showed that, instead of one parameter $b$, Eq.(1) introduces a series of parameters $b^{(n)}_j$ and $\\tilde{b}^{(n)}_j$ to describe the bias scale-dependence. The statistics of $\\tilde{b}^{(n)}_j$ with different $n$ are independent from each others. The DWT analysis provides a simple and effective tool of systematically detecting the scale-dependence of bias parameters on various orders. This method is effective to be employed for analyzing galaxy samples which show morphology- and/or luminosity-segregation.  With this method we detected the $j$-spectra of $\\tilde{b}_j^{(n)}$ and $b_j^{(n)}$ for the distributions of EL and SP galaxies of the APM-BGC samples. The general results indicate that for second order statistics, i.e. two-point correlation function and power spectrum, the bias is approximately scale-independent, but not so for higher order or phase-sensitive (non-Gaussian) statistics. The result is consistent with the following fact: most evidences for weak scale dependence of bias are from statistics of the two-point correlation functions and power spectrum (Kauffmann, Nusser \\& Steinmetz 1997), while the evidence for scale-dependence is from phase-sensitive statistics (Sigad et al. 1998). Therefore, the scale-dependence of galaxy bias may have the same origin as the non-Gaussianity of galaxy distribution. Linear evolution cannot cause non-Gaussianity of mass distributions if the initial perturbations are Gaussian. The scale-dependence of bias of galaxy distribution is most unlikely due to the non-linear evolution of gravitational clustering, and the non-local relationship between galaxy formation and their environment.  With this in mind, the information of bias scale-dependence is worth for developing models of galaxy formation. Indeed, despite the current detection of bias scale dependence is still very preliminary, the result is already able to set useful constraint on models of hierarchical clustering. In these models, galaxy correlations are generally assumed to be described by the hierarchical relation $\\xi_n = Q_n \\xi_2^{n-1}$ where $\\xi_n$ is the $n$-th order correlation function, and $Q_n$ are constants (White 1979). If these hierarchical relations hold exactly, the second order (two-point) correlation function plus all constants $Q_n$ (which may be different for different types of galaxies) completely characterize the clustering of galaxies, including their higher order correlations. This is, all $\\tilde{B}^{(n)}_j$ and $B^{(n)}_j$ can be represented by second order correlation function plus all (scale-independent) constants $Q_n$. Hence, if bias is scale-independent on second order, it will be scale-independent on all orders. Therefore, the non-flatness of $j$-spectra of $\\tilde{B}^{(n)}_j$ and $B^{(n)}_j$ for sample APM-BGC implies that the hierarchical relations may not hold exactly, or the coefficients, $Q_n$, are scale-dependent. Similar conclusion has also been drawn from the detection of the scale-scale correlations of the Ly$\\alpha$ forests of QSO absorption spectrum (Pando et al. 1998.) Thus, the detection of $\\tilde{B}^{(n)}_j$ and $B^{(n)}_j$, joining with other higher order statistics, is effective to reveal the details of the hierarchical clustering scenario.  We thank Drs. J. Pando and Y.P. Jing for many helpful comments. ZGD and XXY were supported by the National Science Foundation of China.  \\newpage"
    },
    "9805/astro-ph9805055_arXiv.txt": {
        "abstract": "A lot of evidence has recently accumulated that physical characteristics of CP stars ($\\delta m_{1}$, Ca/Fe ratio, Li abundance, distribution of orbital periods, $v \\sin i$, $\\Delta (V1-G)$, magnetic field) depend on orbital period or eccentricity. Consequently, the \"tidal mixing + stabilization\" hypothesis has been formulated to account for such features and the \"binarity $\\times$ magnetism\" hypothesis of Abt \\& Snowden (1973) was reanalyzed. We conclude that all mentioned characteristics and properties of CP stars, including even CP star magnetism, may be affected or governed by binarity. ",
        "introduction": "The backbone of the current research of CP stars is their modern classification scheme by Preston (1974). He clearly distinguished a non-magnetic sequence of Am and HgMn stars as well as a magnetic sequence of Ap and some He-weak stars. While, at present, we have at hand some explanation of most of the CP star distinguishing characteristics like the origin of abundance anomalies and slow rotation of these stars, this is not the case of their magnetism and binarity. Both of these phenomena are often referred to as primordial reasons for CP peculiarity in the corresponding area of the HR diagram. Nevertheless, it is slow rotation which is generally accepted as a more direct reason for the CP phenomenon and which is closely related to the primordial causes, namely, being a result of tidal or magnetic breaking mechanisms (Abt 1965, 1979, Wolff 1983). The reason for these primordial causes is however not very clear, and this is just what we aim to touch in these studies. The pioneering works of Abt (1961) and Abt \\& Snowden (1973) completed and reanalyzed e.g. by Abt (1965), Floquet (1983), Abt \\& Levy (1985), Gerbaldi et al.(1985), Lebedev (1987), Seggewiss (1993), North (1994), North at al. (1998) and many others revealed that the frequency of occurrence among binaries is much higher for Am than for Ap stars, short period orbits are much more frequent in Am than in Ap ones and the N(SB2)/N(SB1) ratio is higher in Am's than in Ap's. The reversed appearances of magnetism and binarity in Am and Ap stars suggest a possible relationship between these fundamental CP characteristics and we will handle such an idea as a \"binarity $\\times$ magnetism\" hypothesis. It was first suggested by Abt \\& Snowden (1973). Guided by such motivations we searched (e.g. Budaj 1995, 1996, 1997a, 1997b, Budaj et al. 1997, Iliev et al. 1997) for the various possible imprints of a stellar companion on the CP star. ",
        "conclusions": "It is hardly possible to explain most of the above-mentioned points in the framework of current views on CP stars, i.e. as a result of slow rotation and magnetism in the corresponding area of the HR diagram, reducing the role of binarity effects just to slowing down the rotation. In the following we briefly summarize our interpretation of the above points.  We showed that the lower boundary of about $P_{\\rm orb}\\approx 1.2^{d}$ in the OPD of Am's corresponded to the semidetached systems with the Am stars radii $R=3R_{\\odot}$. One cannot, however, exclude that the Am phenomenon is disturbed slightly before the Roche lobe is filled in what would then result in a little lower radii of Am stars. Consequently, the Am stars' rotation in this area of synchronization could be limited by its orbital period (at some fixed stellar radius) rather than by the generally accepted critical rotation velocity of about 100 km\\,s$^{-1}$.  We argue that a decrease of $\\delta m_{1}$, Ca/Fe, or Li abundance (i.e. increase of peculiarity) with $P_{\\rm orb}$ could be explained by the hypothesis on \"tidal mixing + stabilization\". We propose the mechanism of tidal mixing which acts to smooth the chemical anomalies built by diffusion. It gets more intensive when the components are closer and weakens when (pseudo-)synchronization is approached. The fact that the increase of peculiarity is observed only in a limited range of orbital periods (up to $P_{orb}\\approx 50^{d}$ or $P_{orb}\\approx 200^{d}$) gives rise to the idea of some kind of stabilization mechanism within this range. The latter arises because, without some ``stabilization'', the peculiarity or chemical anomalies should progressively increase with $P_{orb}$ up to infinite $P_{orb}$ and, at the same time, approach the status of ``single'' stars, which is not observed.  The behaviour of the $V_{max}$ curve, which is not a constant but a function of $P_{orb}$, sheds more light on the well known overlap in rotation velocities of normal and Am stars for $40< v < 100$ km\\,s$^{-1}$ (Abt \\& Hudson 1971) or the deficiency of normal stars for $P_{orb}>2.5^{d}$ (Abt \\& Bidelman 1969) because it intersects the $R=2.1\\, R_{\\odot}$ synchronization curve at $v=35$ km\\,s$^{-1}$ and $P_{orb}=3^{d}$. The parallel behaviour of the $M_{const.}$ and $V_{max}$ curves is hardly an accident but it is rather a natural consequence of the idea that each curve corresponds to different intensity of turbulence in the $v$ versus $P_{orb}$ diagram.  The fact that there is a peak in normal stars within the period gap in Am's supports the presence of the mentioned gap. \\footnote{One could estimate its significance to $1/2^{5}$ as there are 5 normal stars. The gap in Ap's is also significant approximately at the level of $2 \\sigma$.} Thus, three independent samples (Am, Ap, normal A4-F1) indicate that there might be some breaking point in the (magneto)hydrodynamics of an AV-type binary at the orbital period of several hundred days. We suggest the \"tidal mixing + stabilization\" hypothesis to account for this gap and propose that there is a peak in the turbulence within the period gap so that the He superficial convection zone  cannot disappear due to He settling. This could result in the observed pattern of all three samples.  Finally, our findings that the degree of peculiarity in Ap's depend on orbital elements strongly supports the \"binarity $\\times$ magnetism\" hypothesis and we can conclude that magnetism, as a necessary condition for Ap phenomena, is (1) affected by binarity or (2) the opposite. In fact, this second possibility cannot be definitely excluded based on the above arguments (including those mentioned in Sect. 1) only. Abt \\& Snowden (1973) preferred just this possibility and suggest e.g. that \"for those Ap stars having strong magnetic field, the formation of binaries with separation $10^{6}-10^{9}$ km is inhibited ...\". Nevertheless, the second possibility seems to be more complicated and thus more unlikely in the light of our recent findings, due to a rather complex behaviour of Ap peculiarity with respect to orbital elements. Thus, we favor the idea that binarity \"governs\" the magnetism. At present it is not very clear what could be the reason of such an interplay but one can speculate that (pseudo-)synchronization might play an important role. Generally, the shorter the orbital period, the higher the degree of (pseudo-)synchronization and the binary components tend to rotate as rigid bodies. One can expect that this will suppress the differential stellar rotation, which is thought to drive the magnetism, at least in the case of the Sun. It could then result in a deficit of short periods or an increase of peculiarity with $P_{orb}$ or low frequency of occurrence or low SB2/SB1 ratio in Ap binaries. Even the decrease of their peculiarity with eccentricity could be understood because highly eccentric orbits have higher velocities at periastra than circular ones, thus also a higher degree of pseudo-synchronization. Also it would not be very surprising to observe just the opposite behaviour, i.e. more pronounced anomalies at larger eccentricities or relative enhancement of short period orbits (except for $P_{orb}<1.2^{d}$) in non-magnetic Am binaries. Nevertheless, the latter facts could also be explained outside the scope of the \"binarity $\\times$ magnetism\" hypothesis as a consequence of the mentioned \"tidal mixing + stabilization\" hypothesis, since tidal mixing should weaken as pseudo-synchronization approached.  To conclude, there is for the first time reliable observational evidence (except for synchronization and circularization) that the hydrodynamics or magneto-hydrodynamics of a ``detached binary'' component is affected by its companion in a surprisingly broad interval of orbital elements .   \\vspace{-1.0mm}"
    },
    "9805/astro-ph9805263_arXiv.txt": {
        "abstract": "Brightness distribution of Gamma-Ray Bursts (GRBs) is studied in detail under the assumption that GRB rate is related to cosmic star formation rate. The two populations of the long- and short-duration bursts in the 4B BATSE catalog are analyzed separately. Taking account of current uncertainties in the observational estimate of star formation rate (SFR), we have tried various models of the cosmic star formation history  and we find that the SFR evolution in $z$ = 0--1 is strongly constrained by the GRB distribution if the standard candle approximation is valid. The strong SFR evolution by a factor of $\\sim$ 15 from $z$ = 0 to 1 inferred from UV observations is too steep to be consistent with the GRB distribution for any distance scale of GRBs. Some possibilities to reconcile this discrepancy are discussed, including the intrinsic luminosity dispersion of GRBs and/or modification of star formation history estimated by UV observations. We argue that SFR increase factor from $z$ = 0 to 1 may be as low as about 4 if we choose different sets of cosmological parameters and/or take account of the evolution of metallicity and dust extinction in the UV data, and this would significantly remedy the discrepancy. ",
        "introduction": "The brightness distribution of gamma-ray bursts (GRBs) observed by the Burst and Transient Source Experiment (BATSE) has been known to be significantly deficient in faint bursts compared with that expected in the Euclidean space (Meegan et al. 1992), and this has been considered as one of the evidences of the cosmological origin of GRBs (Mao \\& Paczy\\'{n}ski 1992; Piran 1992; Dermer 1992). On the other hand, since most of GRB models are associated to death of massive stars whose lifetime is much shorter than the cosmological time scale, the brightness distribution of GRBs reflects not only the cosmological effects but also the cosmic star formation history (Totani 1997, hereafter T97; Sahu et al. 1997; Wijers et al. 1998). The cosmological origin of GRBs is now confirmed by the discovery of metal absorption lines in the optical counterpart of GRB 970508 (Metzger et al. 1997), and hence more detailed analyses of the GRB brightness distribution are required to investigate the origin of GRBs and the cosmic star formation history.  The most important consequence of the possible relation of GRB rate and star formation rate (SFR) is that the distance scale and absolute luminosity of GRBs become larger than those for no-evolution sources, because we need stronger cosmological effect in order to cancel out the rapid increase of the observed star formation rate from $z=0$ to 1 by a factor of more than 10 (Lilly et al. 1996). However, different authors give quantitatively different results: T97 pointed out that the scenario of binary neutron-star mergers (Blinnikov et al. 1984, referred to ``the NS-NS model'', hereafter) results in a better fit than the case that GRB rate is simply proportional to SFR (referred to ``the proportional model'', hereafter), because the time delay during the spiral-in phase of binary neutron stars make the SFR evolution significantly flatter in $z$ = 0--1. On the other hand, Wijers et al. (1998) concluded that the proportional model is in good agreement with the observed GRB brightness distribution when the redshift of the most distant GRBs is about 6. Petrosian \\& Lloyd (1997), however, claimed that neither the NS-NS nor the proportional model is consistent with the observationally determined SFR evolution. This paper investigates the origin of this discrepancy paying special attention to uncertainties in the SFR observations. We found that the GRB rate evolution in $z$ = 0--1 is crucially important in the GRB distribution analysis, and constraints on the cosmic star formation history in this redshift range are obtained, under the assumption of the standard candle approximation.  It is well known that the duration of GRBs shows a bimodal distribution suggesting the existence of two populations of long ($\\gtilde$2 sec) and short ($\\ltilde$ 2 sec) GRBs (Kouveliotou et al. 1993). All previous papers including T97, which investigated the GRB brightness distribution in the context of the cosmic star formation history, however treated the GRBs as a single population. Because the energy spectrum of short GRBs is significantly harder than that of long GRBs (Kouveliotou et al. 1993) and the assumed spectrum affects the GRB brightness distribution analysis, we should analyze the GRB distribution separating the two populations. We define the long GRBs as those with $T_{90} >$ 2 sec, while the short GRBs with $T_{90} <$ 2 sec, where $T_{90}$ is the interval over which 5 \\% to 95 \\% of the burst counts accumulate (Kouveliotou et al. 1993), and analyze them separately in this paper. We also propose a new method to determine the average GRB spectrum in which the curvature of GRB spectrum compared to a pure power-law is taken into account based on the hardness-brightness correlation seen in the BATSE catalog.  The paper is organized as follows: in \\S 2, formulations used in this paper are described. We determine the average spectrum of the short and long GRBs and give the relations between redshift and peak flux of GRBs.  In \\S 3, the models of cosmic star formation history used in this paper are described. The results of fits to the observed GRB distribution of the 4B BATSE catalog (Paciesas et al. 1997) are given in \\S 4. After discussing the result and its implications on the cosmic star formation history in \\S 5, we conclude this paper in \\S 6. Unless otherwise mentioned, we use the Einstein-de Sitter (EdS) universe with $(h, \\Omega_0, \\Omega_\\Lambda)$ = (0.5, 1, 0) as background cosmology. ",
        "conclusions": "In this paper we have presented a detailed study on the possible relation between the brightness distribution of gamma-ray bursts and the cosmic star formation history. The long and short GRBs in the 4B BATSE catalog are analyzed separately. We proposed a new method to determine the average GRB spectrum in which the curvature of GRB spectra compared to a pure power-law is taken into account based on the observed $\\alpha$-$P$ correlation. Various models of the cosmic star formation history are tried considering the present uncertainties in the observational estimate of SFRs, and implications of the GRB distribution on the star formation history and galaxy evolution were discussed.  We have shown that the evolution of SFR in $z$ = 0--1 is crucially important for the fit of GRB brightness distribution, and SFR evolution in this range is strongly constrained if the standard candle approximation is valid. The analysis on the long GRBs suggests that, in the Einstein-de Sitter universe with $(h, \\Omega_0, \\Omega_\\Lambda) = (0.5, 1, 0)$, the SFR increase factor from $z$ = 0 to 1 [$\\xi(1)$] should be smaller than 3.0 (95 \\% C.L.) if the GRB rate is proportional to SFR, and than 3.8 if the GRBs are produced by binary neutron-star mergers. For short GRBs, $\\xi(1)$ is constrained as $\\xi(1) < 4.5$ and $\\xi(1) < 10$ for the proportional and NS-NS models, respectively. These values are significantly smaller than the current estimate ($\\xi(1) \\sim$ 15; Lilly et al. 1996) of SFR evolution based on the UV luminosity density, but marginally consistent with a theoretical estimate ($\\xi(1) \\sim 4$) by galaxy evolution models based on the local properties of galaxies (Totani, Yoshii, \\& Sato 1997). These results are consistent with Totani (1997) in which the star formation history based on the galaxy evolution model gave a good fit while the observational history did not.  We have discussed some possibilities to reconcile this apparent discrepancy between the UV-estimated SFR evolution and GRB distribution, under the condition that GRBs are related to death of massive stars. One possibility is intrinsic luminosity dispersion of GRBs, although neither the width of dispersion nor the shape of distribution is well known. We have also argued that the uncomfortably large $\\xi(1)$ inferred from the UV observation may be an overestimation and the real value could be as low as about 4 if we choose different sets of cosmological parameters and/or take account of the evolution of metallicity and dust extinction. Therefore the UV observation itself can also be consistent with the BATSE data and the galaxy evolution model.  For the case of low values of $\\xi(1) \\sim 4$, we have obtained some constraints on the distance scale of GRBs and production rate of GRBs. We could not find any acceptable fit for the long GRBs with $\\zmax < 2$. Therefore $\\zmax$ is at least larger than 2 and likely in the range of $\\zmax$ = 3--5, if the long GRBs are associated to death of massive stars or NS-NS mergers. The production rate of GRBs from star formation is $\\sim 5 \\times 10^{-8}$ [$M_\\odot^{-1}$] if GRB rate is proportional to SFR, and the beaming factor required for the NS-NS merger scenario is about a few hundreds.  The maximum redshift of the short GRBs seems smaller than that of the long GRBs if the GRB rate evolution is the same for the two populations. This confirms that the total energy emitted by the short bursts is smaller than that by the long bursts by more than one order of magnitudes (see Fig. \\ref{fig:luminosity}), if the beaming factor is the same. On the other hand, the peak luminosity of the two populations is remarkably similar ($\\sim 2 \\times 10^{51} \\ \\rm erg \\ sec^{-1}$ in 50--300 keV). This might suggest that there is a physical mechanism which regulates the energy loss rate of shock heated matter in shocks generated by relativistic motion, in a wide range of total energy liberated as a relativistic fireball.  The author would like to thank an anonymous referee for careful reading of this manuscript and useful comments. He has been supported by the Research Fellowships of the Japan Society for the Promotion of Science for Young Scientists, and the Grant-in-Aid for the Scientific Research Fund (No. 3730) of the Ministry of Education, Science, and Culture of Japan.   \\begin{table} \\begin{center} \\caption{Best-fit Parameters for $\\zmax$, $\\xi(1)$, and $\\xi(4)$} \\label{table:SFR-SFR} \\begin{tabular}{cccccccc} \\hline  \\hline Cosmology\\tablenotemark{\\it a} & Duration & Model\\tablenotemark{\\it b} & $\\zmax$ & $\\xi(1)$ & $\\xi(4)$ & $\\chi^2_{\\min}$ & S.L. (\\%)\\tablenotemark{\\it c} \\\\ \\hline EdS & long & prop. & 3.7 & 1.7 & 4.6 & 2.34 & 80 \\\\ EdS & long & NS-NS & 4.8 & 1.6 & 13.8 & 0.85 & 97 \\\\ EdS & short & prop. & 4.2 & 1.8 & 10.0 & 5.48 & 36 \\\\ EdS & short & NS-NS & 4.8 & 1.0 & 12.6 & 4.87 & 43 \\\\ Open & long & prop. & 2.9 & 1.7 & 2.1 & 2.00 & 85 \\\\ Open & long & NS-NS & 4.7 & 2.5 & 18.2 & 1.22 & 94 \\\\ $\\Lambda$ & long & prop. & 2.5 & 1.0 & 1.0 & 2.05 & 84 \\\\ $\\Lambda$ & long & NS-NS & 4.8 & 1.4 & 7.2 & 1.16 & 95 \\\\ \\hline \\hline \\end{tabular} \\end{center} \\tablenotetext{a}{Cosmological parameters used are $(h, \\Omega_0, \\Omega_\\Lambda)$ = (0.5, 1, 0), (0.6, 0.2, 0), and (0.7, 0.2, 0.8) for EdS, Open, and $\\Lambda$, respectively.} \\tablenotetext{b}{The proportional model in which GRB rate is proportional to SFR and the NS-NS model in which GRBs are produced by neutron star mergers.} \\tablenotetext{c}{The degree of freedom is $9-4 = 5$.} \\end{table}  \\begin{table} \\begin{center} \\caption{Best-fit Parameters for $\\zmax$ and $\\xi(4)$. (Only for the long GRBs and $\\xi(1)$ is fixed.)} \\label{table:zmax-SFR} \\begin{tabular}{ccccccc} \\hline  \\hline $\\xi(1)$ & Model\\tablenotemark{\\it a} & Cosmology\\tablenotemark{\\it b} & $\\zmax$ & $\\xi(4)$ & $\\chi^2_{\\min}$ & S.L. (\\%)\\tablenotemark{\\it c} \\\\ \\hline 4 & prop. & EdS & 5.0 & 15.9 & 19.4 & 0.35 \\\\ 4 & prop. & Open & 5.0 & 14.5 & 6.0 & 42 \\\\ 4 & prop. & $\\Lambda$ & 5.0 & 13.1 & 20.4 & 0.23 \\\\ 4 & NS-NS & EdS & 4.5 & 36.3 & 8.3 & 22 \\\\ 4 & NS-NS & Open & 4.6 & 31.6 & 2.3 & 89 \\\\ 4 & NS-NS & $\\Lambda$ & 4.4 & 24.0 & 9.91 & 13 \\\\ 3 & prop & EdS & 5.0 & 11.0 & 7.9 & 25 \\\\ 3 & prop & Open & 4.9 & 9.6 & 3.0 & 80 \\\\ 3 & prop & $\\Lambda$ & 5.0 & 9.1 & 9.97 & 13 \\\\ 3 & NS-NS & EdS & 4.5 & 25.1 & 4.1 & 67 \\\\ 3 & NS-NS & Open & 4.6 & 21.9 & 1.4 & 97 \\\\ 3 & NS-NS & $\\Lambda$ & 4.5 & 17.4 & 5.6 & 47 \\\\ \\hline \\hline \\end{tabular} \\end{center} \\tablenotetext{a}{The proportional model in which GRB rate is proportional to SFR and the NS-NS model in which GRBs are produced by neutron star mergers.} \\tablenotetext{b}{Cosmological parameters used are $(h, \\Omega_0, \\Omega_\\Lambda)$ = (0.5, 1, 0), (0.6, 0.2, 0), and (0.7, 0.2, 0.8) for EdS, Open, and $\\Lambda$, respectively.} \\tablenotetext{c}{The degree of freedom is $9-3 = 6$.} \\end{table}"
    },
    "9805/astro-ph9805113_arXiv.txt": {
        "abstract": "We investigate the formation of disk-bulge-halo systems by including bulges in the Fall \\& Efstathiou theory of disk formation.  This allows an investigation of bulge dominated disk galaxies, such as S0s and disky ellipticals. These latter systems, which consist of an elliptical spheroid with an embedded disk with a scale-length of typically a few hundred parsecs, seem to form a smooth sequence with spirals and S0s towards lower disk-to-bulge ratio. The aim of this paper is to examine whether spirals, S0s, and disky elllipticals all can be incorporated in one simple galaxy formation scenario.  We investigate an inside-out formation scenario in which subsequent layers of gas cool and form stars inside a virialized dark halo. The inner, low angular momentum material is assumed to form the bulge.  Stability arguments are used to suggest that this bulge formation is a self-regulating process in which the bulge grows until it is massive enough to allow the remaining gas to form a stable disk component.  We assume that the baryons that build the disk do not loose their specific angular momentum, and we search for the parameters and physical processes that determine the disk-to-bulge ratio, and therewith explain to a large extent the origin of the Hubble sequence. The spread in halo angular momenta coupled with a spread in the formation redshifts can explain the observed spread in disk properties and disk-to-bulge ratios from spirals to S0s.  If galaxy formation is efficient, and all available baryons are transformed into the disk-bulge system, cosmologies with $\\Omega_0 \\lesssim 0.3$ can be excluded, since stable spiral disks would not be allowed to form.  However, if we assume that the efficiency with which galaxies form depends on the formation redshift, as suggested by the small amount of scatter in the observed Tully-Fisher relation, and we assume that the probability for a certain baryon to ultimately end up in the disk or bulge is independent of its specific angular momentum, spirals are allowed to form, but only at small formation redshifts ($z \\lesssim 1$).  At higher formation redshifts, stability arguments suggest the formation of systems with smaller disk-to-bulge ratios, such as S0s. Since density perturbations in clusters will generally collapse earlier than those in the field, this scenario naturally predicts a density-morphology relation, the amplitude of which depends on the baryon fraction of the Universe. Disky ellipticals are too compact to be incorporated in this scenario, and they thus do not form a continuous sequence with spirals and S0s, at least not in the sense of the galaxy formation scenario envisioned in this paper.  Alternative formation scenarios for the disky ellipticals, such as gas-rich mergers or an internal mass loss origin for the embedded disks, are much more viable. ",
        "introduction": "One of the most compelling puzzles in present day astronomy is the question how galaxies formed. In particular, we need to understand the wide variety of sizes, masses and morphologies of galaxies observed, as well as their dynamics, ages and metallicities.  In addition, we need to understand the origin of the scaling relations, such as the fundamental plane relations for ellipticals and the Tully-Fisher relation for spirals, as well as the so-called density-morphology relation (Dressler 1980). The latter shows that the more compact galaxies, such as ellipticals and S0s, are preferentially found in overdense regions such as galaxy clusters.  The classical way of depicting the different morphologies of galaxies is by means of the Hubble diagram, which reveals the gross distinction of three sorts of galaxies: spirals, S0s, and ellipticals.  Several studies over the past years have shown that elliptical galaxies can be roughly divided in two subclasses: the rotation supported, low luminosity disky ellipticals and the pressure supported, bright, boxy ellipticals (e.g., Bender 1988; Bender \\etal 1989; Capaccioli, Caon \\& Rampazzo 1990).  This dichotomy has recently been strengthened by properties observed in their central regions (e.g., Nieto, Bender \\& Surma 1991; Jaffe \\etal 1994; Ferrarese \\etal 1994; Gebhardt \\etal 1996; Faber \\etal 1997).  Based on this dichotomy amongst elliptical galaxies, it has been suggested that the classical Hubble diagram should be revised (Kormendy \\& Bender 1996), to use a more physical subdivision of the class of the ellipticals, rather than the apparent flattening.  Kormendy \\& Bender proposed to use the velocity anisotropy which is well measured by the isophote shapes (Bender 1988; Bender \\etal 1989).  The main morphological parameter that sets the classification of galaxies in the (revised) Hubble diagram is the disk-to-bulge ratio $D/B$.  Understanding the origin of the Hubble sequence is thus intimately related to understanding the parameters and processes that determine the ratio between the masses of disk and bulge.  The diskiness of the low luminosity ellipticals is generally interpreted as due to an embedded disk. These disks, which we term `embedded disks' in the following, are smaller and brighter than disks in S0s and spirals (Scorza \\& Bender 1995), and it has been emphasized many times that the disky ellipticals build a continuous sequence with S0 galaxies (Capaccioli 1987; Carter 1987; Bender 1988, 1990; van den Bergh 1990: Capaccioli \\etal 1990; Rix \\& White 1990; Capaccioli \\& Caon 1992; Bender \\etal 1993; Scorza \\& Bender 1995, 1996; Scorza \\& van den Bosch 1998).  This continuity suggests similar formation histories, whereby one or several parameters of the proto-galaxy vary smoothly.  Disk dominated systems such as spirals are believed to have formed by cooling of the baryonic matter inside a virialized dark halo.  As the gas cools, its specific angular momentum is conserved, and the amount of angular momentum of the dark halo thus determines the size of the disk (Fall \\& Efstathiou 1980; see Section~2 for more details on this disk-formation scenario). The formation of bulge dominated disk systems is far less clear.  The aim of this paper is to investigate to what extent the disky ellipticals and S0s can be incorporated in the Fall \\& Efstathiou theory for the formation of galactic disks, by incorporating a simple picture for the formation of the bulge.  We envision an inside-out formation scenario for the bulge.  It is assumed that the bulge forms out of the low-angular momentum material in the halo, which cools and tries to settle into a small, compact disk.  Such disks are however unstable, and we assume here that this instability, coupled with the continuous supply of new layers of baryonic matter that cool and collapse, forms the bulge. This inside-out bulge formation is self-regulated in that the bulge grows until it is massive enough to allow the remaining gas to form a stable disk component.  We do not describe the bulge formation in any detail but merely use empirical relations of the characteristic structural parameters of bulges and ellipticals to describe the end result of the formation process as a realistic galaxy.  We use this simple formation scenario to investigate the predicted disk-to-bulge ratios and disk scale-lengths as a function of the halo angular momentum, and as a function of formation redshift and cosmology. The main focus of this paper is to investigate whether this inside-out formation scenario can account for two orders of magnitude variation in disk-to-bulge ratio, required in order to incorporate a wide variety of disk-bulge systems: from late-type spirals with $D/B \\gta 10$ to S0s ($D/B \\sim 0.1$) to disky ellipticals ($D/B \\sim 0.1$).  This paper is organized as follows. In Section~2 we describe the formation scenario, the galaxy formation efficiency, and the stability criterion for our disk-bulge-halo models. In Section~3 we use these models to investigate the position of different sorts of disks in the parameter space of disk central surface brightness versus disk scale-length.  In Section~4 we discuss to what extent S0s and disky ellipticals can be incorporated in this formation scenario, and what parameters may be responsible for the origin of (the major part of) the Hubble sequence. In Section~5 we briefly discuss alternative formation scenarios for disky ellipticals. Our results are summarized and discussed in Section~6. ",
        "conclusions": "Understanding galaxy formation is intimately linked with understanding the origin of the Hubble sequence. An important clue is provided by comprehending the formation of disky ellipticals and S0s, simply because these systems form the transitional class from the classical ellipticals to the spirals. Disky ellipticals seem to form a continuous sequence with spirals and S0s towards smaller disk-to-bulge ratio, and it is thus tempting to believe that disky ellipticals, S0s and spirals all formed in a similar fashion. The aim of this paper has been to investigate a simple formation scenario for disk-bulge-halo systems, and to search for the main parameters and/or processes that determine the disk-to-bulge ratio and thus explain to a large extent the origin of (the major part of) the Hubble sequence.  We considered the disk formation scenario originally proposed by Fall \\& Efstathiou (1980), in which the size of the disk, which is formed by cooling of the gas in a dark halo, is determined by the amount of angular momentum of the halo. We have extended upon this formation scenario by including a simple picture for the inside-out formation of an additional bulge component out of the inner, low angular momentum material. Stability arguments are used to suggest that the formation of the bulge is a self-regulating process in which the bulge grows until it is massive enough to allow the remaining gas to form a stable disk component.  We do not describe the bulge formation in any detail but merely use empirical relations which allow us to describe the bulge by a single parameter, namely its mass.  Each dark halo contains a fraction $\\Omega_{\\rm bar}/\\Omega_0$ of baryons.  We introduced a galaxy formation efficiency $\\epsilon_{gf}$ which describes the fraction of those baryons that actually build up the disk-bulge system. The theory of spherical collapse coupled with the definition of a virialized halo predicts a Tully-Fisher relation of the form $L \\propto V_c^3$ as observed, with a zero-point that depends on the Hubble constant $H(z)$. Recent observations, however, suggest that the Tully-Fisher zero-point is independent of redshift, implying that the galaxy formation efficiency is proportional to $H(z)$.  A physical explanation for this redshift dependence may be the higher escape velocities and cooling efficiencies at higher redshifts.  For a combined sample of $\\sim 200$ galaxies, varying from spirals to disky ellipticals, we calculated the value of the halo's spin parameter which yields the observed disk properties under the assumption that disk-bulge systems form in the way envisioned here. We compared two cosmologies (SCDM vs. OCDM) and investigated the differences between assuming a galaxy formation efficiency of unity and two (extreme) scenarios in which $\\epsilon_{gf} \\propto H(z)$: a cooling scenario, in which we assume that only the inner fraction $\\epsilon_{gf}$ of the available baryons cools to form the disk and bulge, and a feedback scenario, in which each baryon, independent of its specific angular momentum, has a probability $\\epsilon_{gf}$ of ultimately ending up in the disk or bulge. Our main conclusions are the following:  \\begin{itemize}  \\item Disk-bulge systems do not have bulges that are significantly more massive than required by stability of the disk component.  This suggests a coupling between the formation of disk and bulge, and is consistent with the self-regulating, inside-out bulge formation scenario proposed here.  \\item If we live in a low-density Universe ($\\Omega_0 \\lta 0.3$), the only efficient way to make spiral galaxies is by assuring that only a relatively small fraction of the available baryons makes it into the galaxy, and furthermore that the angular momentum distribution of those baryons is similar to that of the entire system; i.e., the probability that a certain baryon becomes a constituent of the final galaxy has to be independent of its specific angular momentum.  In the cooling scenario, most of the angular momentum of the system remains in the outer layers, and most halos form disk-systems with massive bulges, such as S0s.  If, however, the galaxy formation efficiency is regulated as described by our `feedback' model, a promising scenario unfolds: At formation redshifts $\\gtrsim 3.0$ the galaxy formation efficiency is unity, and systems that form build up a large bulge to support the disk that assembles around them. These galaxies resemble S0s.  Galaxies that form later, at $z \\approx 0$, no longer require a massive bulge, and spirals are preferentially formed.  Coupled with the notion that density perturbations that collapse early are preferentially found in high density environments such as clusters, this scenario automatically predicts a morphology-density relation in which S0s are most likely to be found in clusters. In a SCDM Universe a similar, albeit less restrictive, mechanism is at work, which predicts a morphology-density relation of smaller amplitude.  \\item A reasonable variation in formation redshift and halo angular momentum can yield approximately one order of magnitude variation in disk-to-bulge ratio, and our simple formation scenario can account for both spirals and S0s. However, disky ellipticals have too large bulges and too small disks to be incorporated in this scenario.  Apparently, their formation and/or evolution has seen some processes that caused the baryons to loose a significant amount of their angular momentum.  Merging and galaxy harassment (Moore \\etal 1996) are likely to play a major role for these systems.  \\end{itemize}  Finally we wish to emphasize of few of the major shortcomings of the oversimplified formation scenario discussed here. First of all, we have neglected the fact that the merging of halos is an ongoing process, and that this is very effective in destroying disks (e.g., Toth \\& Ostriker 1992). Gas-dynamical simulations that do not involve the energy and momentum feedback to the gas from supernova explosions, stellar winds, UV radiation etc. produce galactic disks that are some two orders of magnitude smaller than the observed spiral disks (e.g., Navarro \\& Steinmetz 1997): merging (and also harassment) are very effective in transporting angular momentum out into the halo, thus yielding more and more compact galaxies. This problem with forming galactic disks of proper dimensions is often referred to as the angular momentum problem.  Despite our ignorance regarding the effects of merging, harassment, feedback, and (re)-ionization of the Universe, the observed sizes of (spiral) disks clearly suggest however, that the combine effect of all these processes is apparently such that the material that forms galactic disks has not lost much of its original angular momentum acquired from cosmological torques. The use of the Fall \\& Efstathiou disk formation scenario thus seems justified, despite the aforementioned shortcomings.  In the past semi-analytical simulations of galaxy formation have been mainly based on the assumption that all bulges result from the merging of disk galaxies (e.g., Kauffmann, White \\& Guiderdoni 1993; Cole \\etal 1994; Baugh, Cole \\& Frenk 1996; Somerville \\& Primack 1998). In this paper we have examined an inside-out bulge formation scenario, which should be regarded as complimentary to this merging scenario. Our formation scenario can account for both spirals and S0s, but fails to build systems that are even more bulge dominated. Although it has been shown that bar-instabilities can lead to the formation of small (and close to exponential) bulges, a more detailed study is required to investigate whether the inside-out bulge formation scenario discussed here can yield more massive bulges as well. It is at least intriguing that the critical spin parameter is of the same order of magnitude as the typical spin parameter for halos, suggesting that a significant fraction of halos will yield unstable disks, unless part of the baryonic material is transformed into a bulge component as suggested here. So despite the clearly oversimplified nature of the formation scenario envisioned here, it may provide a useful framework for future investigations of galaxy formation."
    },
    "9805/astro-ph9805325_arXiv.txt": {
        "abstract": "We describe the specifications, characteristics, calibration, and analysis of data from the University of New South Wales Infrared Fabry-Perot (\\uw) etalon. \\uw\\ is a near-infrared tunable imaging spectrometer, used primarily in conjunction with IRIS on the AAT, but suitable for use as a visitor instrument at other telescopes. The etalon delivers a resolving power in excess of 4000 (corresponding to a velocity resolution $\\sim75$~\\kms), and allows imaging of fields up to $100''$ in diameter on the AAT at any wavelength between 1.5 and 2.4~$\\mu$m for which suitable blocking filters are available. ",
        "introduction": "The desire for high spectral resolution observations in the near-infrared has been met with three main types of instrument.  The traditional way of mapping an object at high spectral resolution is to use a long-slit cooled grating spectrograph, and step the slit across the sky. Although this technique records data at every spectral point simultaneously, it is highly inefficient on extended objects if only a single wavelength (or a small number of wavelengths) are of interest. The Fourier Transform Spectrometer (FTS) works by Fourier transforming an interferogram produced by a two-beam interferometer, and can perform measurements over a large wavelength range (e.g., 0.9 -- 5.5~$\\mu$m in the case of the CFHT FTS; Bohlender 1994).  The FTS tends, however, to be mechanically large, complex, and expensive, and is also not very efficient for monochromatic applications. Furthermore, every pixel has the noise from the entire continuum in it, and this noise is correlated from pixel to pixel. The Fabry-Perot interferometer, by contrast, is small, has a high throughput (compared to a grating of comparable size and resolving power; Jacquinot 1954), and can deliver consistently high spectral resolution over a wide field, which is imaged directly with an infrared array.  Fabry-Perot etalons have been successfully employed for narrow-band imaging in the optical for many years (e.g., Atherton et al. 1982; Bland \\& Tully 1989; Jones \\& Bland-Hawthorn 1997). However, it is only the recent advent of low-noise, large-area detector arrays that has made their use as tunable narrow-band filters for the near-infrared particularly advantageous. There is clearly much to be gained by observing emission lines in the infrared; for example, aside from the reduction in extinction relative to the optical regime, many rotational and vibrational transitions of molecules (such as H$_{2}$) also become accessible.  Although many of the brighter Galactic sources can be imaged using filters with fixed, narrow ($\\Delta \\lambda / \\lambda \\sim 1$\\%) bandpasses, the use of a Fabry-Perot etalon with resolving power $\\lambda / \\Delta \\lambda \\gtrsim 10^{3}$ confers a number of advantages, including  \\begin{itemize} \\item the ability to resolve closely spaced lines, or resolve the line of interest from adjacent OH airglow or atmospheric absorption lines; \\item the ability to reveal velocity gradients, or even a complete velocity field, when Doppler motions exceed a few tens of \\kms; \\item the reduced sky background and continuum flux passed to the detector. The resultant increase in the line-to-continuum ratio improves the measurement stability, and allows the sky to be sampled less often. \\end{itemize}  In this paper, we describe one such system, named \\uw\\ (University of New South Wales Infrared Fabry-Perot), which is intended to complement the existing near-infrared imaging and spectroscopic capabilities of IRIS\\footnote{IRIS uses a $128\\times128$ HgCdTe array manufactured by Rockwell International Science Centre, CA.} (Allen et~al. 1993) at the Anglo-Australian Telescope (AAT), but which could also function as a visiting instrument at other facilities (e.g., MSSSO 2.3~m, UKIRT). In the next section, we give a brief overview of Fabry-Perot systems, and \\uw\\ in particular. We then describe some of the novel approaches taken to calibrate \\uw\\ and process the resultant data, and give illustrations of some of the early scientific results obtained with \\uw. ",
        "conclusions": "We have successfully developed and commissioned the University of New South Wales Infrared Fabry-Perot (\\uw) on IRIS at the AAT.  Among the notable characteristics of this system, which is based on a Queensgate ET-70WF etalon, are its high Resolving Power ($\\Re>4000$), wide field (up to $100''$), and ability to be tuned across almost the entire $H$ and $K$ bands. A novel method of {\\em in-situ\\/} wavelength calibration has been applied, and a new suite of reduction and analysis software in the {\\sc iraf} environment has been developed. Early results across a wide variety of sources have been presented, and a number of diverse projects are now underway. Further details are available from the \\htmladdnormallinkfoot{\\uw\\ WWW page} {http://www.phys.unsw.edu.au/$\\sim$sdr/unswirf/UNSWIRF.html}."
    },
    "9805/astro-ph9805169_arXiv.txt": {
        "abstract": "Two rival hypotheses have been proposed for the origin of the compact radio flux observed in radio-quiet quasars (RQQs). It has been suggested that the radio emission in these objects, typically some two or three orders of magnitude less powerful than in radio-loud quasars (RLQs), represents either emission from a circumnuclear starburst or is produced by radio jets with bulk kinetic powers $\\sim 10^3$ times lower than those of RLQs with similar luminosity ratios in other wavebands.  We describe the results of high resolution ($\\sim $ parsec-scale) radio-imaging observations of a sample of 12 RQQs using the Very Long Baseline Array (VLBA). We find strong evidence for jet-producing central engines in 8 members of our sample. ",
        "introduction": "The bimodality of radio luminosity in the quasar population (Miller, Peacock \\& Mead 1990) poses many questions about the fundamental nature of the emission mechanisms in Radio-Quiet Quasars (RQQs) and Radio-Loud Quasars (RLQs). The total radio luminosity is typically two or three orders of magnitude lower for a RQQ than for a RLQ (Miller, Rawlings \\& Saunders 1993) with similar luminosity ratios in all other wavebands.  The double, often co-linear, morphology exhibited by many RLQs on arcsecond scales (e.g., Bridle et al.\\ 1994) usually comprises a bright central core with hotspots at the outermost edges of the radio structure. In addition, lobes of {\\em extended emission} are seen which are fed by jets, via backflow out of the hotspots. In contrast, radio images of RQQs often show merely a weak component, coincident with the optical quasar nucleus, which in some cases is resolved (Miller et al.\\ 1993).  It has been proposed that the activity in RQQs is supplied by a `starburst', i.e., strongly radiative supernovae and supernova remnants (SNRs) in a very dense environment (Terlevich et al.\\ 1992). Sopp \\& Alexander (1991) argued this on the basis of the striking continuity between the far-infra red --- radio correlation of RQQs and that of star-forming galaxies, ultra-luminous infra-red galaxies and also Seyfert galaxies. This is offset from the same correlation for RLQs and radio galaxies. Alternatively, if the energy supply arises from accretion onto a massive black hole, the radio emission from RQQs (as in RLQs) is caused by radio jets, but the bulk kinetic powers of these jets are for some reason $\\sim 10^3$ times lower than those of RLQs (Miller et al.\\ 1993).  To address the question of whether the radio emission in radio-quiet quasars is associated with starbursts or with weak-jet producing central engines, we have undertaken a programme of imaging a sample of RQQs using VLBI techniques. This provides a definitive test between the two rival hypotheses for the radio emission in a given RQQ, since one can derive a size scale (from milli-arcsecond resolution measurements) from which the observed luminosity is emitted and compare this with the size scales on which SNRs are found to be distributed. Moreover, a mere detection with VLBI implies the brightness temperature of the emission $T_{\\rm B} \\stackrel{>}{_{\\sim}} 10^6$K, while typical supernovae/supernova remnants have $T_{\\rm B} \\stackrel{<}{_{\\sim}} 10^5$ K (Muxlow et al.\\ 1994).  We previously reported an experiment to image a well-known nearby radio-quiet quasar (E\\,1821+643) with the VLBA (Blundell et al.\\ 1996). In detecting it at 5 and 8 GHz on milli-arcsecond scales, we found a high brightness temperature ($T_{\\rm B} \\sim 10^{9}\\ {\\rm K}$) which precluded the possibility of star-formation as the origin of its radio emission.  It was instead consistent with a mechanism similar to the central engines postulated for radio-loud quasars. In order to establish whether this result was typical of the radio-quiet quasar population we performed detection experiments using the VLBA on a sample of 12 RQQs, the results of which we present here.  We describe in Section 2 of this paper the details of our sample selection and include a discussion of the conventional criteria used to classify whether a particular quasar is radio-loud or radio-quiet. In Section 3 we describe our observing method and we summarise our results. In Section 4 we discuss the physical implications of our results for the models discussed earlier; we also explore when a quasar is appropriately classified as radio-quiet and consider what may be the counterparts of the RQQ population which undergo Doppler boosting. ",
        "conclusions": "\\subsection{Implications}  Our high detection rate from these snapshot observations supports the findings of our initial experiment on the RQQ, E1821+643 (Blundell et al.\\ 1996). The 4 RQQs which we did not detect in these observations are those with the lower VLA flux densities (see Table~\\ref{tab:source}).  \\begin{table*} \\begin{tabular}{lllrccrcl} \\mc{1}{c}{PG name} & \\mc{1}{c}{R.A.}   & \\mc{1}{c}{Dec.}   & \\mc{2}{c}{VLBA}     & \\mc{1}{c}{Synthesised}     & \\mc{1}{c}{Total flux}   & \\mc{2}{c}{Total flux}   \\\\ & \\mc{1}{c}{J2000.0}& \\mc{1}{c}{J2000.0}&\\mc{2}{c}{8.4 GHz peak}& \\mc{1}{c}{beam}          & \\mc{1}{c}{density from} & \\mc{2}{c}{density from} \\\\ &                   &                   &\\mc{2}{c}{(mJy/beam)}& \\mc{1}{c}{({\\sc fwhm}/mas)}& \\mc{1}{c}{VLBA (mJy)}   & \\mc{2}{c}{VLA (mJy)}          \\\\ \\hline \\hline 0003+199 &  \\mc{1}{c}{--} & \\mc{1}{c}{--}  & $<0.8$  &        &                 &                       &\\ph{xxx} &    \\\\ 0007+106 & 00 10 31.00587 & 10 58 29.5037  & 57.7    &        & $1.98 \\ti 0.83$ &  $68.3 \\pm 1.4$ & &            \\\\ 0157+001 &  \\mc{1}{c}{--} & \\mc{1}{c}{--}  & $<1.5$  &        &                 &                 & &            \\\\ 0923+129 &  \\mc{1}{c}{--} & \\mc{1}{c}{--}  & $<0.4$  &        &                 &                 & & 4.1        \\\\ 1116+215 &  \\mc{1}{c}{--} & \\mc{1}{c}{--}  & $<0.3$  &        &                 &                 & & 1.5        \\\\ 1216+069 & 12 19 20.93171 & 06 38 38.4679  &  1.7    &        & $1.87 \\ti 0.84$ &   $2.0 \\pm 0.3$ & &            \\\\ 1222+22  & 12 25 27.40090 & 22 35 13.0522  &  2.1    &        & $1.88 \\ti 0.80$ &   $3.2 \\pm 0.4$ & &            \\\\ 1309+355 & 13 12 17.75278 & 35 15 21.0857  & 14.9    &        & $1.73 \\ti 0.80$ &  $23.8 \\pm 2.1$ & &            \\\\ 1351+640 & 13 53 15.83069 & 63 45 45.6856  &  1.5    &        & $1.66 \\ti 0.76$ &   $2.9 \\pm 0.5$ & &            \\\\ 1407+26  & 14 09 23.90866 & 26 18 21.0557  &  2.3    &        & $2.33 \\ti 0.98$ &   $4.7 \\pm 0.4$ & &            \\\\ 1700+518 & 17 01 24.82093 & 51 49 20.4995  &  0.8    &        & $1.98 \\ti 0.98$ &   $0.8 \\pm 0.2$ & & 4.9        \\\\ 2209+184 & 22 11 53.88876 & 18 41 49.8634  & 69.8    &        & $1.88 \\ti 0.80$ & $179.7 \\pm 17.1$& &            \\\\ \\hline\\hline \\end{tabular} \\parbox{162mm}{\\caption[sourcetable]{\\label{tab:meas_results} Results from the VLBA observations are tabulated: columns 2 and 3 list the (J2000) Right Ascension and Declination respectively and column 4 lists the peak flux densities which were obtained using the Gaussian-fitting task {\\sc imfit} within \\aip\\ in units of mJy/beam. Non-detections are quoted as $< 3\\sigma$. Column 6 shows the total flux density measured by the VLBA in mJy. Column 7 lists the VLA flux densities measured simultaneously with the VLBA observations for those sources where the VLA was used in phased array mode.}} \\end{table*} \\normalsize  \\begin{table*} \\begin{tabular}{llllccrc} \\mc{1}{c}{PG name} & \\mc{1}{c}{$T_{\\rm B}$(K)}& \\mc{1}{c}{Deconvolved} & \\mc{1}{c}{Limit}      &\\mc{1}{c}{Log$_{10}(L_{8.4})$}         & \\mc{3}{c}{Emitting}    \\\\ &                          & \\mc{1}{c}{size (mas)}  & \\mc{1}{c}{$T_{\\rm B}$}&\\mc{1}{c}{(${\\rm W Hz^{-1} sr^{-1}}$)} & \\mc{3}{c}{region (pc)} \\\\ \\hline \\hline 0003+199           &                          &                        &                       &     \\mc{1}{c}{--}                     & & \\mc{1}{c}{--} &          \\\\ 0007+106           &  $7.5\\ti 10^8$           & $0.4 \\ti 0.4$          & $7.7\\ti 10^9$         &   23.22                               & & 2.9  &                   \\\\ 0157+001           &                          &                        &                       &     \\mc{1}{c}{--}                     & & \\mc{1}{c}{--}  &         \\\\ 0923+129           &                          &                        &                       &     \\mc{1}{c}{--}                     & & \\mc{1}{c}{--}  &         \\\\ 1116+215           &                          &                        &                       &     \\mc{1}{c}{--}                     & & \\mc{1}{c}{--}  &         \\\\ 1216+069           &  $2.8\\ti 10^7$           & $0.7 \\ti 0.4$          & $1.6\\ti 10^8$         &   22.92                               & & 8.0  &                   \\\\ 1222+22            &  $8.3\\ti 10^7$           & $0.9 \\ti 0.4$          & $3.5\\ti 10^8$         &   24.99                               & & 19.4 &                   \\\\ 1309+355           &  $2.5\\ti 10^8$           & \\mc{1}{c}{--}          &                       &   23.30                               & & 4.9  &                   \\\\ 1351+640           &  $2.4\\ti 10^7$           & \\mc{1}{c}{--}          &                       &   21.60                               & & 2.4  &                   \\\\ 1407+26            &  $1.1\\ti 10^9$           & $0.7 \\ti 0.4$          & $8.6\\ti 10^9$         &   25.56                               & & 24.3 &                   \\\\ 1700+518           &  $1.0\\ti 10^7$           & \\mc{1}{c}{--}          &                       &   22.45                               & & 11.3 &                   \\\\ 2209+184           &  $9.7\\ti 10^8$           & \\mc{1}{c}{--}          &                       &   23.09                               & & 2.8  &                   \\\\ \\hline\\hline \\end{tabular} \\parbox{162mm}{\\caption[sourcetable]{\\label{tab:der_results} Derived physical parameters from the VLBA observations.  Column 2 shows the brightness temperatures derived according to the formula quoted in the main text, using the sizes of the synthesised beam and total VLBA flux densities listed in Table 2.  Where our snapshot data permitted robust and well-constained deconvolved sizes to the emitting regions to be fitted, we tabulate those in column 3, together with re-derived brightness temperatures using these deconvolved sizes and the peak VLBA flux densities. Also listed are the luminosities (in ${\\rm W\\, Hz^{-1} sr^{-1}}$) at 8.4 GHz of the compact emission of the RQQs we detected together with the characteristic size of the emitting region (this is the geometric mean of the major and minor axes of the synthesised beam converted into a physical distance in parsecs within the assumed cosmology). }} \\end{table*} \\normalsize  Table~\\ref{tab:der_results} shows that the RQQs require central engines which can supply luminosities of $10^{21.6}$ -- $10^{25.6}$ ${\\rm W\\, Hz^{-1} sr^{-1}}$ arising from regions of a few cubic parsec. The peak luminosity at 5 GHz of the most powerful supernova known (1986J) (Rupen et al.\\ 1987) is $\\sim 10^{20} {\\rm W\\, Hz^{-1} sr^{-1}}$ so between 10 and 1000 ($10^5$ for the two high redshift, very luminous objects) of these close to peak luminosity would thus be required to power a single RQQ.  This would require a very sustained supernova rate with an unprecendented supernova space density: the (conservatively derived) values for the size of the emitting region yield volumes between $10^5$ -- $10^7$ times smaller than in the starburst model of Terlevich and Boyle (1993) or observed in the M82 starburst galaxy (Muxlow et al.\\ 1994).  The brightness temperature quantifies this by consideration of the flux emanating from a given solid angle --- thus radio emission powered by a starburst would not be expected to have a high brightness temperature because of the spatial separation expected for the supernovae.  These results strongly suggest that for these radio-quiet quasars which we detected, their radio emission is {\\em not} dominated by starbursts and imply that they have central engines similar to those in RLQs, but producing only weak radio jets.  In a recent study (Falcke, Patnaik \\& Sherwood 1996) high brightness temperatures were found for three RIQs (which are common to our sample) based on fits to the emission region size deconvolved from the synthesised beam of the telescope. Our observations being considerably shorter do not in all cases allow us to similarly obtain robust, well constrained, fits to the sizes of the emission-regions.  \\subsection{When should a quasar be deemed radio-quiet?}  We return to the question of when a quasar should be appropriately classified as radio-loud. Of the three classifications outlined in the Section 2, none considers the contribution to the total radio emission from cores which might be Doppler boosted, if the cores are indeed powered by relativistic jets. The bimodality in radio-loudness would undoubtedly be more pronounced if instead of comparisons based on total radio luminosity, the `total minus core' radio luminosity, (i.e., only the contribution from unbeamed emission), were to be used (see e.g., Kukula et al.\\ 1998). It is imperative to ascertain whether the radio flux densities quoted from the literature for those quasars believed to be non-radio-loud, which we have detected with the VLBA, are representing the extent of the radio-emission from these objects, and thus that our detections are of genuinely radio-quiet quasars. We thus checked the NVSS survey (Condon 1994) (VLA D-array 1.4-GHz maps), the 6C (Hales et al.\\ 1988, Hales et al.\\ 1990) and the 7C (Waldram et al.\\ 1996) surveys at 151 MHz for evidence of diffuse extended lobe emission related to these RQQs\\footnote{1222+22 and 1407+26 are in areas of sky covered by the 7C survey (Waldram et al.\\ 1996). 1309+355 is in an area of sky covered by the 6C-II survey (Hales et al.\\ 1988).  1700+518 and 1351+640 are in an area of sky covered by 6C-III (Hales et al.\\ 1990).}.  For the 6C and 7C surveys the {\\sc rms} background measurements are roughly $\\sim$ 25 -- 50 mJy/beam in the absence of any confusion. For the NVSS survey the {\\sc rms} background is $\\sim 0.2$ mJy/beam.  We found {\\em no} evidence of any related extended lobe emission for our detected RQQs. We therefore believe that all of the objects we detected originally classified as radio-quiet are correctly classifed as such, although their radio-quietness would be more dramatically evident were the criteria to be based on extended radio emission only.  \\subsection{What are the boosted counterparts of RQQs?} \\label{sec:riq} One important question is whether the jets in RQQs are indeed relativistic near the central engine (as in RLQs), albeit with a much lower bulk kinetic power. If so a subset of the RQQ population, namely those whose jet axes are oriented close to our line-of-sight, would be expected to exhibit Doppler boosted emission. Such a scenario was first proposed by Miller et al.\\ (1993) based on a study of the [OIII] luminosity versus radio luminosity plane for a sample of optically selected quasars (those quasars from BQS with $z < 0.5$). They found that radio-loud objects exist only at high [OIII] luminosity and that for RQQs there is a tight correlation between the radio and the [OIII] luminosity. A number of objects in the radio-quiet region of the plot did not lie so tightly on this correlation; Miller et al.\\ suggested that their location on the plane could be explained if their radio emission was Doppler boosted, i.e., they were the beamed counterparts of RQQs (with Lorentz $\\gamma \\sim 5$); they termed such objects radio-intermediate quasars. If the criteria for radio-loudness is based on extended emission alone (as discussed in Section 4.2) then the RIQs are clearly members of the RQQ population.  The cores of RIQs might be expected to have higher $T_{\\rm B}$ than objects whose radio emission is not Doppler boosted.  While there is no such clear correlation from the numbers in this small sample, we note that the comparison of brightness temperatures is an important tool in testing Miller et al.'s hypothesis.  A number of the sources we detected are among those deemed by Miller et al.\\ as RIQs. It is therefore conceivable that the RIQs represent those RQQs with a jet-producing central engine while the true radio-quiet objects lack such a central engine --- but the dichotomy posed at the beginning of the paper is little changed, as it is still necessary to explain why the RIQ population cannot form the powerful radio jets seen in the RLQ population, even though there is now direct evidence that some contain jet-producing central engines.  \\subsection{Conclusions}  Our results have shown that some radio-quiet quasars show evidence for a central engine resembling those in radio-loud quasars; the evidence we present is consistent with the sample objects being boosted examples of a homogeneous population of radio-quiet quasars with relativistic jets.  Our study underlines the need to address the important question of why powerful radio jets are not seen in RQQs even though a significant fraction of their central engines possess the essential characteristics of those in RLQs; we note that there have been various suggested explanations of this, including for example, lack of black-hole spin (Blandford 1993) or the necessary presence of a hot atmosphere around the nucleus (Fabian \\& Rees 1995)."
    },
    "9805/astro-ph9805219_arXiv.txt": {
        "abstract": "\\gdef\\Msun{M_{\\odot}}  Near-infrared imaging and spectroscopy have been obtained  of the gravitationally lensed galaxy at $z=4.92$ discovered in HST images by Franx et al. (1997). Images at 1.2, 1.6 and 2.2$\\mu$m show the same arc morphology as the HST images.  The spectrum with resolution $\\lambda / \\Delta\\lambda\\sim 70$ shows no emission lines with equivalent width stronger than 100\\AA\\ in the rest frame wavelength range 0.34$\\mu$m to 0.40$\\mu$m. In particular, [OII]3727\\AA\\ and [NeIII]3869\\AA\\ are not seen. The energy distribution is quite blue, as expected for a young stellar population with the observed Ly $\\alpha$ flux.  The spectral energy distribution can be fit satisfactorily for such a young stellar population when absorption by dust is included. The models imply a reddening 0.1 mag $< E(B-V) <$0.4 mag.  The  stellar mass of the lensed galaxy lies in the range of 2 to 16 $\\times 10^9$ $\\Msun$. This is significantly higher than estimates based on the $HST$ data alone. Our data imply that absorption by dust is important to redshifts of $\\sim 5$.   \\medskip ",
        "introduction": "\\label{sec:introduction}  The study of galaxies at high redshifts has been revolutionized with the introduction of the Hubble Space Telescope and the new generation of large ground-based telescopes.  Detecting and studying such systems at high redshift is essential to understanding how normal galaxies form and evolve.  Franx et al. (1997) have recently reported the serendipitous discovery of a field galaxy at a redshift $z=4.92$ that has been gravitationally lensed into several arc components, including a highly magnified fold arc,  by  an intervening cluster of galaxies  (CL1358+62) at $z=0.33$.  Franx et al. present HST imaging and Keck spectroscopy of this highly-magnified lensed system.  Because of the redshift, the HST images at $R$ ($F606W$) and $I$ ($F814W$) correspond to wavelengths of 1023\\AA\\ and 1375\\AA\\ in the rest frame of the background objects, and thereby sample only the far UV portion of the light emitted from these galaxies.  In order to study these objects at wavelengths where nearby galaxies have been studied, we observed the brightest (fold) arc of the lensed galaxy at near infrared wavelengths using the W.M. Keck Telescope. In the discussion we adopt the same cosmological parameters as in Franx, et al., i.e. $H_0 =$ 50 Km/s/Mpc and $q_0=$0.5.   \\smallskip ",
        "conclusions": "\\label{sec:discussion}  The discussion of the last section has shown that the lensed galaxy has significant extinction, and we have derived new larger estimates for the stellar mass.  The high stellar masses imply that the galaxy has built up a very dense center. The knot, which has an effective radius of 130 pc, contains roughly half the mass, i.e., 1.2 to $8\\times 10^{9}$ $\\Msun$ (Franx, et al, 1997).  Assuming an isothermal profile, we derive a velocity dispersion of the knot between 100 and 260 km s$^{-1}$.  Our new results have therefore strengthened the case that this young galaxy at $z = 4.92$ has already managed  to build up a very dense core.  The resulting dynamical timescale for the knot is a few times $10^6$ year, comparable to the age derived from a starburst model.  The dynamical timescale for the galaxy as a whole is at least 10 times higher. This exceeds the lifetime of ionizing O stars. Combined with the fact that the emission and absorption line properties observed by Franx et al. (1997) indicate the presence of a strong wind which locally depletes the interstellar medium in this galaxy, it is likely that we are seeing young, short lived condensations of star formation.  Over time, these knots would  brighten  progressively across the galaxy, as suggested earlier by Lowenthal et al (1997).  The infrared photometry has demonstrated that dust is very likely present in the galaxy.  This conclusion clearly depends on the accuracy of the stellar evolutionary models. The reddening is estimated to be between $0.1$ mag$< E(B-V) < 0.4$ mag, depending on the population model and extinction curve.  These values are comparable to those found in nearby, low metallicity, starbursts (e.g., Meurer et al 1997. This implies that the galaxy has been able to produce a significant amount of metals, but the observations to date are inadequate to derive metallicities.   We note that similar results were derived by Ellingson et al. (1996) for MS1512-cB58 and for galaxies in the Hubble Deep Field by Sawicki and Yee (1998). The optical and infrared fluxes of MS1512-cB58 at $z = 2.72$ imply a young age ($10^7$ years) and a significant extinction ($E(B-V) \\sim 0.3$ mag). The constraints on the age of the young population were stronger because the observations extended beyond the Balmer break.  In the case of the  Hubble Deep Field, Sawicki and Yee have found that  spectral energy distributions with reddening of $E(B-V)\\sim 0.3$mag provides better fits to the observations of $z > 2$ Lyman Break galaxies than do dust free models. The best studied high redshift galaxies have significant amounts of dust while evidence for dust exists for other high redshift galaxies based on sparser color information (e.g., Meurer et al. 1997, Pettini et al. 1997). If the lensed galaxies are typical, then a reddening between $E(B-V)\\approx 0.2 - 0.3$ mag should be expected for most high redshift galaxies."
    },
    "9805/astro-ph9805331_arXiv.txt": {
        "abstract": "On the basis of our new simultaneous photometry and spectroscopy (885 $uvby$ differential measurements in 11 nights and 154 spectrograms of the FeII$\\lambda$4508 region in 5 nights), we can detect 12 probable periodicities in the variability pattern of this star, determining the frequencies of 7 without any ambiguity. Through a direct fit of pulsational models to our data, we estimate the inclination of rotational axis to be about 50$^{\\circ}$ and get a reliable identification of 4 modes as well as useful bits of information about the others: no retrograde mode is visible, whereas the star seems to show a certain preference for purely sectorial prograde oscillations. Finally, the attribution of our lowest frequency to the radial fundamental pulsation allows a new calibration of physical parameters. In particular, the gravity can be determined with unusual accuracy and the luminosity evaluation becomes more consistent with the Hipparcos astrometry. ",
        "introduction": "The serendipitous discovery of HD 2724 as a variable star is due to Reipurth (1981), which chose it as one of the comparison objects for his differential photometry of the eclipsing binary AG Phe. The author identified HD 2724 as a probable $\\delta$ Scuti star and guessed a tentative period of 0$^{d}$.174 ({\\em i.e.} a frequency of 5.75 d$^{-1}$). Lampens (1992) met this periodicity again analysing her excellent sequences of absolute measurement obtained at La Silla in 1984-85 by means of the UBVB$_{1}$B$_{2}$V$_{1}$G Geneva photometer. Lampens' analysis shows multiperiodic variations which are typical of the $\\delta$ Scuti light curves: besides the above mentioned frequency, she identified unambiguously another component at $\\sim$7.38 d$^{-1}$ and suggested $\\sim$6.50 d$^{-1}$ and $\\sim$4.34 d$^{-1}$ as two additional candidate frequencies.  HD 2724 is classified as an F2 III star in Hoffleit and Jaschek (1982). Lampens (1992) gets from her photometry T$_{eff}$ = 7180$^{\\circ}$K and M$_{V}$ = 0.93. Physical parameters can be evaluated also by using the $uvby\\beta$ colours published by Hauck \\& Mermilliod (1990). Moon's \\& Dworetsky's (1985) grids lead us to estimate T$_{eff}$ and $\\log g$ at 7280$^{\\circ}$K and 3.56 respectively, while Villa \\& Breger (1998) obtain from their still unpublished calibration, based on Canuto's \\& Mazzitelli's (1991) models and performed using dereddened indices, T$_{eff}$ = 7216$^{\\circ}$K and $\\log g$ = 3.64. As to the absolute magnitude, Crawford's (1979) calibration yields M$_{V}$ = 1.12, E(b-y) = 0.014 and therefore A$_{V}$ = 0.060. Nevertheless, our photometric evaluations of luminosity are now to be revised owing to new astrometric data: in the Hipparcos Satellite General Catalogue (ESA, 1997), this object (HIC 2388) appears with a parallax $\\pi$ = 7.77 $\\pm$.72 mas, which, taking account of the above assessed interstellar extinction, corresponds to an absolute magnitude M$_{V}$ = 0.57 $\\pm$.20. In principle, pulsational masses could help us to adjust these calibrations. It would entail, however, a thorough knowledge of pulsational states, which today might be achieved only by combining photometry and spectroscopy in a synergetic approach (see {\\em e.g.} Bossi {\\em et al.}, 1994, or Mantegazza {\\em et al.}, 1998).  In order to exploit the complementarity between photometry and spectroscopy in studying dynamical processes like stellar pulsations, we are performing for many years simultaneous observational campaigns of $\\delta$ Scuti stars through both these techniques (Mantegazza {\\em et al.}, 1994; Mantegazza \\& Poretti, 1996). The present work on HD 2724 falls within this frame. ",
        "conclusions": ""
    },
    "9805/astro-ph9805107_arXiv.txt": {
        "abstract": "We report the radio continuum structure of the barred galaxy NGC 3367 with an angular resolution of $\\sim4''.5$. The radio structure indicates emission from the disk and from a triple source consisting of the nucleus straddled by two extended sources (the lobes).  The triple source shows an excess of radio continuum emission  compared to the emission expected from the total radio-H$\\alpha$ correlation, suggesting a non-thermal origin probably related to AGN activity and not to star formation processes. The triple source is approximately 12 kpc in extent at a P.A. $\\approx40^{\\circ}$, close (but not aligned) to that of the stellar bar, P.A. $\\approx65^{\\circ}$. Only the southwest lobe is polarized. The polarization asymmetry between the two lobes suggests that the triple source axis is slightly out of the plane. If the origin of the emission is an outflow of plasma from an AGN, similar to weak radio galaxies and NGC 1068, NGC 3367 provides an excellent laboratory object to study a possible interaction of the ejected material and the barred galaxy. ",
        "introduction": "NGC~3367 is a face-on SBc(s) barred spiral galaxy inclined with respect to the plane of the sky at an angle i$\\approx6^{\\circ}$ (\\cite{gro85}). It can be considered an isolated galaxy at a distance of 43.6 Mpc with a far distant neighbor, NGC 3419, behind the Leo Spur group of galaxies (i.e. using H$_0=75$ km s$^{-1}$ Mpc$^{-1}$ and assuming that the Milky Way is moving towards the Virgo Cluster at 300 km s$^{-1}$; \\cite{tul88}). At this distance, an angular diameter of 1$''$ corresponds to $\\approx210$ pc. The stellar bar has an angular diameter of $\\approx32''$ (6.72 kpc) oriented at P.A. $\\approx65^{\\circ}$, and there is a southwest optical structure resembling a ``bow shock'', along which lies a half ring of H$\\alpha$ knots at a radius of about 10 kpc from the nucleus (\\cite{gar96a}). VLA observations at 15$''$ angular resolution by Condon et al., (1990) at 1.46 GHz show emission from the disk as well as from the nucleus and from two sources straddling the nucleus. NGC~3367 is also an X-ray and far infrared emitter (\\cite{gio90,sto91,fab92,soi89}). The X-ray emission extends beyond the disk, but peaks some 21$''$ from the compact nucleus in the southwest direction (Gioia et al. 1990; Stocke et al. 1991; Fabbiano et al. 1992). This is coincident with the southwest radio continuum lobe. Its X-ray luminosity is stronger than other normal spirals, but weaker than Seyfert galaxies. It also has a strong far-IR emission, of about L$_{FIR}\\approx 2\\times 10^{10}$ L$_{\\odot}$ (Soifer et al. 1989), with a relatively high dust temperature of T$_D{\\approx}35~$ K (Garcia-Barreto et al. 1993). The nuclear region also shows a H$\\alpha$ peak intensity of $2.4\\times 10^{-13}$ ergs s$^{-1}$ cm$^{-2}$, with FWHM$\\sim650$ km s$^{-1}$, and [OIII $\\lambda5007$]$\\approx0.5$ H${\\beta}$ (\\cite{ver86}). All these characteristics are similar to those found in Seyfert galaxies, and thus NGC 3367 has been classified as a Seyfert-like galaxy (\\cite{ver86}).  Radio continuum emission from the central regions of disk galaxies has been detected at low resolution from almost every nearby spiral (\\cite{hum80,gio82,con92,gar93,nik97}), and high resolution studies show that this is usually due to circumnuclear emission, with a variety of sizes and morphologies. Double, triple or jet-like radio continuum structures have been found in several barred spirals with well-known Seyfert activity, such as NGC 1068, NGC 4151, and NGC 5728 but only NGC 1068, a Sy 2, displays extended radio continuum disk emission and has a central (0.4 kpc) source with two lobes (\\cite{wil87}), very much like a small-scale radio galaxy (\\cite{van82,wyn85,ulv87,wil87}).  In this paper we present new VLA\\footnote{The VLA is part of the NRAO which is operated by Associated Universities Inc. under contract with the National Science Foundation.} radio continuum observations of NGC~3367 at 1.4 GHz with a beam of $\\sim4''.5$. The radio continuum image is complex, showing emission from the central region, from many unresolved sources associated with star-forming regions, and from two extended regions straddling the central region. For the purpose of analysis in this paper we will refer to the emission from the central region plus the two extended sources as the {\\it triple source} and we will refer to the emission from the disk emission as {\\it star-formation} emission. ",
        "conclusions": ""
    },
    "9805/astro-ph9805277_arXiv.txt": {
        "abstract": "Navarro, Frenk, and White have suggested that the density profiles of simulated dark matter halos have a ``universal'' shape so that a given halo can be characterized by a single free parameter which fixes its mass.  In this paper, we revisit the spherical infall model in the hope of recognizing in detail the existence and origin of any such universality.  A system of particles is followed from linear perturbation, through first shell crossing, then through an accretion or infall phase, and finally to virialization.  During the accretion phase, the system relaxes through a combination of phase mixing, phase space instability, and moderate violent relation.  It is driven quickly, by the flow of mass through its surface, toward self-similar evolution.  The self-similar solution plays its usual r\\^ole of intermediate attractor and can be recognized from a virial-type theorem in scaled variables and from our numerical simulations.  The transition to final equilibrium state once infall has ceased is relatively gentle, an observation which leads to an approximate form for the distribution function of the final system.  The infall phase fixes the density profile in intermediate regions of the halo to be close to $r^{-2}$.  We make contact with the standard hierarchical clustering scenario and explain how modifications of the self-similar infall model might lead to density profiles in agreement with those found in numerical simulations. ",
        "introduction": "Navarro, Frenk and White (1996) have summarized the results of their dissipationless cosmological clustering simulations in terms of a `universal' shape for the density profile of dark halos.  This profile (henceforth the NFW profile),  \\be{nfw_profile} \\rho(r)~=~\\frac{M_s}{r\\left (r+R_s\\right )^2}~, \\ee  \\noindent is characterized by an $r^{-1}$ central cusp and an outer region where the density falls off faster than that of an isothermal sphere.  Good fits are obtained using Eq.\\,\\EC{nfw_profile} for halos that range in mass from $3\\times 10^{11} M_\\odot$ (dwarf galaxies) to $3\\times 10^{15}M_\\odot$ (rich galaxy clusters).  In addition, the results show a strong correlation between the constants $R_s$ and $M_s$ so that there is essentially a single free parameter (Navarro, Frenk, \\& White 1996).  The possible existence of a universal profile suggests that the structure of collapsed objects can be understood from simple physical arguments.  The earliest attempts to understand cosmological structure were based on the spherical radial infall model (e.g., Gunn \\& Gott 1972; Henriksen \\& De Robertis 1980; Fillmore \\& Goldreich 1984, hereafter FG; Bertschinger 1985, hereafter B85; Hoffman \\& Shaham 1985; White \\& Zaritsky 1992) in which a primordial density perturbation, assumed to be spherically symmetric and smooth, slowly accretes matter from the cosmic background.  If this initial perturbation is also scale-free (i.e., $\\delta\\rho_i\\propto r^{-\\epsilon}$, see e.g., FG) with a velocity distribution corresponding to an unperturbed Hubble law, the structure that develops will be self-similar in the sense that the distribution function at one time can be obtained from that at another time by rescaling the phase space coordinates $r$ and $v_r$. One way to determine the scaling law is to note that at each time $t$ there is a single mass shell that is just beginning to break away from the expansion and fall in towards the center.  The radius at which this occurs defines a function of time, $r_{\\rm ta}\\propto t^{\\delta}$ where  \\be{simclass} \\delta= \\frac{2}{3} + \\frac{2}{3\\epsilon}~, \\ee  \\noindent suggesting that the appropriate radial coordinate for the self-similar solution is $X\\equiv r/r_{\\rm ta}$ (FG and B85).  Each FG and B85 solution represents a single trajectory in phase space, albeit one that describes multiple velocity streams in the inner parts of the system.  An alternative approach (Henriksen \\& Widrow 1997, hereafter HW), based directly on the collisionless Boltzmann equation (CBE), treats the distribution function as continuous in the scaled phase-space variables.  In this picture, the single-trajectory solutions of FG and B85 represent a subset of the characteristic curves of the CBE.  Numerical simulations of systems that begin from a ``cold start'' (i.e., single-stream distribution function) show a rapid transition to a distribution function that is continuous, the transition being facilitated by an instability (HW). The form of the scaling in the CBE (dictated by $\\delta$) is equivalent to the similarity `class' in the sense of Carter and Henriksen (1991).  This is determined by the logarithmic derivative of $r_{\\rm ta}$ with respect to future turn-round time {\\it at the epoch of first shell crossing}, and is given by Eq.\\EC{simclass}.  Self-similarity would seem to be an appealing feature for a model that is to explain universal characteristics in cosmological structures of the type described by Navarro, Frenk, \\& White (1996).  This is particularly so given the usual r\\^ole of self-similarity as an attractor.  However, the self-similar infall model (SSIM) has been, by and large, dismissed as a paradigm for structure formation.  First, the SSIM does not include angular momentum which likely plays a key role in determining the density profile, at least in the central regions.  Second, structure formation is thought to develop by way of hierarchical clustering: Small mass objects form first and merge with one another to create systems of ever-increasing size.  This process is neither smooth nor spherically symmetric.  Moreover, the solutions found by FG and B85 describe an eternal infall of matter and say nothing about how a system might enter or exit such a self-similar state.  In the standard scenario for structure formation, gravitational collapse begins only after the Universe has become matter dominated.  Moreover, at late times (e.g., as for galaxies today), the evolution of a system is dominated by infrequent mergers with comparable-sized objects.  Therefore at best, self-similarity will arise as an intermediate phase in the evolution of a system. (This is of course always the case with self-similar behaviour.) Finally, it is generally believed that the SSIM predicts a power-law density profile for ``virialized'' halos, ($\\rho(r)\\sim r^{-\\mu}$ with $2<\\mu<9/4$ where $\\mu$ depends on $\\epsilon$) in contrast with what is found in the simulations (cf. Eq.\\,\\EC{nfw_profile}).  In this paper we argue instead that the SSIM provides a natural framework for understanding cosmological structure formation, A simple shell code is used to treat spherically symmetric collisionless particles (i.e., spherical shells) on radial orbits.  The particles are followed from linear perturbation through first shell crossing, then through the recently recognized self-similar relaxation phase, and into eventual virialization which occurs after the cessation of infall.  Many features of the self-similar solutions discovered by FG and B85 are present in this intermediate phase though much of the fine-grained phase-space structure characteristic of these solutions is washed out.  We maintain that there is a close connection between the SSIM and semi-analytic models of structure formation based more directly on hierarchical clustering (e.g., Lacey \\& Cole 1993).  In addition we will show that simple modifications of the SSIM lead to density profiles in agreement with results from N-body simulations. An advantage of this approach is that it provides a connection, through a simple semi-analytic model, between initial conditions and the distribution function and density profile of the final virialized system.  In a hierarchical clustering universe, the progenitors of present day nonlinear structures are small-amplitude primordial density fluctuations.  Bond et al.\\,(1991) and Lacey \\& Cole (1993) have developed an analytic model for hierarchical structure formation that relates the initial power spectrum to the mass and formation time of dark matter halos.  In their formalism, the linear density field at some early time, $t_i$, is smoothed on various mass scales.  Objects that will collapse by some later time, $t_{\\rm coll}$, are identified at the earlier epoch as regions where the mean density is above a certain threshold.  This threshold is estimated using the spherical top-hat model, i.e., the perturbation within a given radius is modeled as a region of constant density.  Of course in the spherical top-hat model, all parts of the perturbation collapse to the center at the same time and it is therefore necessary to make an {\\it ad hoc} assumption about what happens after collapse.  Lacey \\& Cole (1993) assume that at $t_{\\rm coll}$ the system reaches virial equilibrium with a radius equal to half its maximum or turnaround radius.  Given this assumption, the mean density within a collapsed object at a particular epoch turns out to be roughly $\\sim 200$ times the background density at that time.  It is now common practice to define the `virial radius', $r_{v}$, as the radius of a sphere that contains an overdense region whose mean density is $v$ times the background density.  The claim is that the mass within this sphere has, more or less, reached a final equilibrium state.  $r_v$ is used, for example, to identify virialized halos in N-body simulations (e.g., Cole \\& Lacey 1996, Navarro, Frenk, \\& White 1996).  Let us see how the SSIM might improve this picture. If the initial perturbation spectrum is scale-free ($\\langle |\\delta_k |^2\\rangle\\propto k^n$) the RMS mass fluctuation will be a power-law function of radius, $\\langle \\delta_R({\\bf x})^2\\rangle^{1/2}\\propto R^{-(n+3)/2}$.  In this expression, $R$ is the radius of the window function used in computing the mass fluctuation, ${\\bf x}$ is the position vector about which this window function is centered, and $\\langle\\dots\\rangle$ indicates an average over ${\\bf x}$.  This suggests an initial density perturbation with $\\delta\\rho_i\\propto r^{-\\epsilon}$ ($\\epsilon=(n+3)/2$) as the most appropriate ``toy-model'' for understanding structure formation via hierarchical clustering and allows us to write the similarity class $\\delta$ and final density profile index $\\mu$ in terms of the spectral index $n$:  \\be{deltavsn} \\delta = \\frac{2}{3}\\left (\\frac{n+5}{n+3}\\right )~; \\ee  \\be{muvsn} \\mu=\\left\\{ \\begin{array}{ll} 3\\left (\\frac{n+3}{n+5}\\right )&\\mbox{$n>1$} \\\\ \\\\ 2&\\mbox{$n\\le 1$} \\ \\ \\ . \\end{array} \\right. \\ee  \\noindent Indeed the results obtained from the SSIM are consistent with those found in Lacey \\& Cole (1993). Their {\\it ad hoc} virialization radius $r_{v}$ corresponds to an $X={\\rm constant}$ $\\left (r={\\rm constant}\\times t^\\delta\\right )$ surface and the expression for the mass within this surface, as a function of $t$, leads to their result for the mass-formation time relationship.  While the ``scaled mass'' within $r_{v}$ approaches a constant, the physical mass increases as $t^{2/\\epsilon}=t^{3\\delta-2}$.  A mass flux through the system boundary is required but this is built into the self-similar solution.  The SSIM avoids the singular nature of the collapse in the spherical top-hat model and thereby affords insight into the virialization process and the ultimate fate of the halo once the mass flux has ceased.  But the question remains as to the relevance of the SSIM to halo formation via hierarchical clustering.  Recently, Syer and White (1997) (see, also Nusser and Sheth, 1997) have suggested that the universality seen in the simulations reflects a balance between two opposing processes, dynamical friction and tidal stripping, that act on a small object as it merges with a larger one.  Dynamical friction will bring a satellite to the centre of a parent halo but only if the satellite has not been tidally disrupted.  Whether or not the satellite reaches the centre intact depends on the density profile of the parent.  For a steep density profile, the satellite is disrupted before it reaches the centre so that its material is spread throughout the halo, thus softening the density law.  Conversely, if the density profile of the parent halo is relatively flat, the satellite reaches the centre largely intact and so boosts the density there.  In this way, dynamical friction and tidal dissipation act as negative feedback mechanisms driving the density profile toward a universal shape (Syer \\& White 1997).  (For an alternative discussion in which the universality of dark halo density profiles does not depend crucially on hierarchical merging, see Huss, Jain, \\& Steinmetz 1998.)  A similar situation arises in the SSIM (FG, Moutarde et al.\\,1995). Suppose the initial density perturbation is relatively flat (here flat and steep are with respect to a $\\rho\\propto r^{-2}$ density law). The system evolves toward a universal profile which, in the intermediate regions of the halo, is $\\propto r^{-2}$ and is a self-similar attractor which we discuss in detail below. As in the hierarchical scenario, the central regions are dominated by material that has fallen in recently (FG) since the binding energy, $GM(r)/r$, is an increasing function of $r$.  By contrast, steep initial density profiles evolve stably as one of a 1-parameter continuum of self-similar solutions.  The parameter is the quantity $\\delta$ introduced above (related ultimately to $n$) and the density profile is the same as that of the perturbation at first shell crossing.  The $\\rho\\propto r^{-2}$ attractor is actually the flattest of these self-similar profiles.  We can say then that the universality predicted by the SSIM is `one-sided'.  It is worth mentioning however that the density laws for the stably evolving systems in the SSIM vary only from $r^{-2}$ to $r^{-9/4}$ for $n\\in\\{1,3\\}$.  As discussed above, a major criticism of the SSIM is that the density profile it predicts for collapsed objects does not agree with the results from N-body simulations.  The power-law index in the NFW profile, for example, varies smoothly from $-1$ in the core to $-3$ in the envelope, in contrast with what is found in the similarity solutions discussed above.  There is some controversy over what the true density law within the inner regions of simulated halos is. Moore et al.\\,(1997) find that mass resolution and force softening have a significant effect on the density profiles of collapsed objects in collisionless N-body simulations.  As the mass and force resolution is increased the density profile in the central region becomes steeper.  Even with 3 million particles per halo, the results have not converged to a unique density profile.  Moore et al.\\,(1997) attribute their results to the issue of dynamic range in the clustering hierarchy: With better resolution, smaller halos would collapse earlier causing the density profile to steepen.  A similar situation arises in the SSIM.  As noted above, the solutions derived by FG and B85 correspond to an eternal infall of matter on a single trajectory. If instead, collapse begins at a finite time, there would be relatively fewer particles in the inner regions leading to a shallower density law in the center.  In the real Universe, the size of the first objects to form is set by the horizon size at matter-radiation equality while in the simulations it is set by the force and mass resolution of the experiment.  As noted above, the NFW profile is noticeably steeper than $r^{-2}$ in the outer regions of simulated halos.  We propose that this region forms after the primary accretion phase when the evolution of a system becomes dominated by major mergers.  The system is still accreting mass during this phase, but not at a rate sufficient to maintain self-similar growth.  The infalling material can be treated as test particles and naturally forms an $r^{-3}$ outer halo.  In Section 2 we write the basic equations describing collisionless radial dynamics (CBE, mass conservation and Euler equations, virialization condition) in scaled variables.  We use these variables to follow numerically a collisionless spherical cosmological perturbation from linear perturbation to a self-similar infall phase (steady-state in scaled variables) and finally to a true virialized state which arises once the stream of particles falling into the perturbation is shut off.  The simulations are presented in Section 3. Section 4 presents a more detailed discussion of the transition from infall phase to virialized isolated system.  A summary and conclusions are given in Section 5. ",
        "conclusions": "N-body simulations of collisionless, self-gravitating matter, such as those by Navarro, Frenk, and White (1996) and Moore et al.\\,(1997), have provided some tantalizing results, in particular suggesting that nonlinear structure in the Universe possesses certain scale-invariant or self-similar characteristics.  These results are perhaps not surprising given that most simulations assume an Einstein-de Sitter Universe and an at least approximately scale-free initial spectrum of density perturbations.  With these assumptions, there is only one characteristic scale in the Universe, the expansion rate.  This observation has been exploited by Press and Schecter (1974), Bower (1991), Lacey and Cole (1993), and others to formulate an analytic model for the evolution of nonlinear structure within the framework of the hierarchical clustering scenario.  However, these models are limited in scope.  While they provide predictions for distribution of collapsed objects as a function of mass, they say little about the dynamical process by which a system relaxes to a virialized (or quasi-virialized) state.  The SSIM is, in some respects, complementary to a Press-Schecter type formalism.  The model begins with highly idealized initial conditions (strictly power-law initial density profile; no angular momentum) but allows one to follow in detail the complete evolution of a system. Analytic calculations and numerical simulations provide clues as to the density profile and distribution function of the system both during the infall phase and in the final equilibrium state.  Our results can be summarized as follows:  \\begin{itemize}  \\item  Soon after gravitational collapse begins, the system, driven by infall of mass through its boundary, enters a period of self-similar evolution.  The system quickly virializes once infall has ceased.  It is likely that the final virialized state is one of the stationary self-similar family discussed by Evans (1994) and Henriksen \\& Widrow (1995) with the self-similar `class' $\\delta$ remembered from the dynamic self-similar infall phase.  \\item  During the infall phase, the system relaxes through a combination of phase mixing, phase space instability, and {\\it moderate} violent relaxation.  However, relaxation does {\\it not} completely randomize particle energies, i.e., particles maintain some memory of their initial state.  \\item  As the system relaxes, the single trajectory of the FG and B85-type solution is transformed into a continuous distribution in phase space.  The subsequent cycles of the single trajectory are now regarded as characteristic curves of the smooth distribution function. We have used this approach to study two similarity classes that are important to our arguments; $\\delta=1,2/3$.  \\item  The self-similar phase may be recognized as ``stationary'' in the appropriate scaled variables.  By following the development of a perturbation in these variables, we can achieve a remarkable dynamic range in our simulations.  The size of the system is always a fixed fraction of the turn-around radius.  In addition, the relation for the mass of the system as a function of time is in agreement with that found by Lacey and Cole (1993).  \\item  During the self-similar phase, the system obeys a virial condition $2K/W={\\it constant}$.  The constant differs from the usual value $1$ (though by only about 10\\%) due to mass flux through the system boundary and the time-dependent nature of the infall solution.  \\item  From the observation that the transition from infall phase to isolated state is relatively gentle, follows the conclusion that the distribution function for the final object has the form $F\\propto \\left (-E\\right )^{1/2}$.  In a realistic system, we expect that the distribution function will eventually decrease at large negative energy and so we include an exponential negative temperature factor. This leads to a form for the distribution function that is similar to the ones proposed by Stiavelli \\& Bertin (1985) and Merritt, Tremaine, and Johnstone (1989) and reminiscent of the Fermi-type function proposed by Lynden-Bell (1967).  \\item  The SSIM predicts an effectively universal profile ($\\rho\\propto r^{-\\mu}~{\\rm with}~\\mu\\simeq 2$) for the intermediate region of a dark matter halo.  The system does remember the initial profile for $\\epsilon>2$ (the so-called steep cases) but the effects on the final density profile are relatively minor.  For flat initial profiles, the system is driven towards the limiting self-similar profile ($\\delta=1$) by accretion of spherical shells with ever-increasing binding energy, a process reminiscent of that considered by Syer and White (1997).  In this sense, the density profile is a one-sided attractor.  Because of finite lifetime effects, the inevitable breaking of scale invariance at the centre of the initial perturbation, and the presence of angular momentum, one can expect a physical (although ill-defined) flattening near the centre of the relaxed halo.  In addition, the simulations themselves are sensitive to resolution effects in this region and so the actual flattening may be less pronounced than in the NFW profile.  \\item  The outer parts of the halo, still in a self-similar relaxation phase, can be quite steep either because the particles there are not full relaxed or because they are essentially in Keplerian orbits about the bulk of the halo mass.  In this latter case, the power law is $r^{-3}$ in accordance with the NFW profile and corresponding to the similarity class $\\delta=2/3$.  \\item In our simulations, the outer edge of the initial mass distribution, $r_0$, determines ultimately the radius of the transition region between $r^{-2}$ and $r^{-3}$ behaviour ($R_{\\rm s}$ in the NFW profile). Likewise, the total mass in the simulation fixes $M_{\\rm s}$.  \\item  The evolution of the initially flat systems towards a limiting self-similar state during the infall phase recalls the evolution of an isolated thermodynamic system towards a maximum entropy state.  Here we propose Lyapounov functions that are maximized in the self-similar state and which are monotonic under collisionless evolution of the distribution function during continued infall.  The existence of such functions tends to strengthen our belief in the `universality' of the SSIM.  \\end{itemize}  \\vskip 1in  \\centerline{\\bf Acknowledgements}  This work was supported in part by a grant from the Natural Sciences and Engineering Research Council of Canada.  \\newpage"
    },
    "9805/astro-ph9805041_arXiv.txt": {
        "abstract": "Recent results obtained by various authors on the properties of HgMn stars are reviewed. Substantial progress has been achieved in the study of abundances and isotopic anomalies. The results about the magnetic fields and membership in multiple systems suggest further directions of investigations to be followed in view of answering the question of the development of abundance peculiarities in HgMn stars. ",
        "introduction": "\\label{intr} Recent years have seen a renewed interest in HgMn stars. The main objective of studies of HgMn stars is to provide new, better observational data for the theorists investigating the mechanisms responsible for abundance anomalies in these stars. Much spectroscopic work was devoted to elemental abundance analyses. About 30 papers on abundances in individual stars and in samples of stars appeared in the last three years. Good studies of the correlations between elemental abundances and fundamental parameters are crucial for the understanding of the physical processes taking place in HgMn stars. On the other hand, in the consideration of the physical mechanisms contributing to the development of the wide range of abundance and isotopic anomalies, it is important to know the r\\^ole that magnetic fields play.  Here I wish to discuss some of the observations that deal with the problems of anomalous abundances of the heavy elements Hg and Pt and with the question of the presence of surface magnetic fields. Finally, I shall discuss the results of recent statistics of multiple systems among HgMn stars. ",
        "conclusions": "There has been considerable progress in recent years in our understanding of HgMn stars. However, many gaps still remain in our knowledge of how HgMn stars form and evolve. The mechanisms responsible for abundance anomalies have not been definitely identified yet. Some systematic trends in the abundance data qualitatively support the mechanism of radiatively-driven diffusion. However, it is difficult to determine the mechanisms responsible for abundance anomalies in the absence of accurate observational information about magnetic fields. It is, therefore, essential to settle the issue of the presence of magnetic fields in HgMn stars.  A potentially fruitful area for future research will be the abundance analyses of the components of binary and multiple systems. We have still only relatively rough ideas about the properties of the companions in the systems. On the other hand, such studies are difficult to carry out because most lines of many of the components of a double, triple or quadruple system appear quite weak (2-3\\%\\ deep) in the spectrum, as a result of the dilution by superposition of their continua. To my knowledge, such an abundance analysis has until now been done only for one multiple system, HD~11905 (Zakharova 1997)."
    },
    "9805/astro-ph9805088_arXiv.txt": {
        "abstract": "We use  a  measure of clustering derived   from the  nearest neighbour distribution and the void probability  function to distinguish between regular and clustered structures.  This measure  offers a succinct way to   incorporate    additional  information  beyond    the  two--point correlation  function.   Application to   a supercluster  catalogue by {}\\scite{einasto:supercluster_data}    reveals no   clustering  in the distribution   of  superclusters.   However,    we  show    that  this supercluster catalogue is severely  affected by  construction effects. Taking  these biases into  account we still  find some indications for regularity on the largest scales, but  the significance is only at the one--$\\sigma$ level. ",
        "introduction": "In  a  recent paper  {}\\scite{einasto:120mpc}  report a   peak in  the 3D--power spectrum of a catalogue of clusters on scales of 120\\hMpc. {}\\scite{broadhurst:large-scale} observed  periodicity on  approximately the same scales in an analysis of 1D--data from a pencil--beam redshift survey. As is well  known from the theory  of fluids, the regular distribution (e.g.\\  of molecules  in  a hard--core  fluid)  reveals  itself  in an oscillating   two--point correlation function    and   a peak in   the structure function respectively (see e.g.~\\pcite{hansen:theory}). In accordance with this an oscillating two--point correlation function $\\xi_2(r)$ or at least a first peak was  reported on approximately the same  scale ({}\\pcite{kopylov:possible},  {}\\pcite{mo:typical_scales}, {}\\pcite{fetisova:features}, and {}\\pcite{einasto:supercluster_II}).  In  this  paper    we    analyze  the  supercluster     catalogue   of {}\\scite{einasto:supercluster_data} which     was  constructed from an earlier    version         of     the      cluster     catalogue    by {}\\scite{andernach:current}    using a friend--of--friends  procedure. With methods  based  on  the nearest   neighbour distribution and  the spherical contact  distribution we  can   show that this  supercluster catalogue is regular with 95\\% significance. However,  taking into account  the selection and construction effects, the high significance vanishes and we only find  some indication for a regular distribution on  large scales, showing that  this supercluster catalogue  is seriously  affected by  the  construction method  with a friend--of--friends procedure.  This paper is   organized as follows.   In Sect.~\\ref{sect:methods} we discuss our   methods.  Through   some  examples, we    illustrate the properties of the $J$-function and  show that it  offers a concise way to incorporate information about correlations of arbitrary order. The analysis of the supercluster  distribution and of   a set of  mock supercluster catalogues is presented  in Sect.~\\ref{sect:results}.  We summarize our results in Sect.~\\ref{sect:conclusion}. ",
        "conclusions": "\\label{sect:conclusion}  We have  shown that the   statistical properties of  the  supercluster distribution as   given by  {}\\scite{einasto:supercluster_data}    are seriously  affected  by the  construction  with  a friend--of--friends procedure.    This  is not  astonishing  since   the linking length of 24\\hMpc\\  is already one  fifth  of the   claimed regularity scale  of 120\\hMpc. The distinction between regular  and clustered point patterns with the $J$--function is unambiguous  for  a homogeneous and   isotropic point distribution. In such a case the borderline is given by $J(r)=1$. Our  procedure for  generating  mock supercluster samples,  where  we start with  a Poisson sample,  include the selection effects, and redo the supercluster  identification  with a friend--of--friends algorithm in   the same way   as  for the real  cluster   sample,  results in  a $J(r)\\ne1$ even  though   we  started  from a  Poisson   distribution. Although   it  is  plausible,  that   such  mock--supercluster samples describe  the   borderline between clustered  and  regular structures, there is no proof of this assertion. The apparent regularity  in  the northern  part of  the sample can  be explained as a result    of these construction effects,   whereas  the southern  part still shows  a trend towards regular structures outside the $1\\sigma$--range.  The  results for the   oscillating two--point correlation function  of galaxy clusters and, correspondingly the   peak in the power  spectrum were   obtained with estimators using   weighting schemes and boundary corrections, which rely heavily on the assumption  of homogeneity.  Up to  now  there   is   no  reliable    way to prove    this   from  the three--dimensional distribution of galaxies  and clusters.   There are some  hints  that the   universe  reaches homogeneity on  scales above several  hundreds       of~\\hMpc\\   (see    the      discussions    by {}\\pcite{guzzo:homogeneous}  versus     {}\\pcite{labini:scale}).    We adopted a conservative point of view  and used estimators which do not make any assumptions about the  distribution of superclusters  outside the sample window. In  Sect.~\\ref{sect:methods} we showed that  the  $J$--function can be estimated from one  point set only for scales  smaller than the radius of the  largest   void.   Therefore, we   do not    reach  the claimed regularity  scale at  120\\hMpc. Still, the   measure  $J(r)$ gives  us information about  global properties of the supercluster distribution, in our case a tendency towards regular structures.  We analyzed the  distribution of clusters of  the more recent redshift compilation by {}\\scite{andernach:current} with the $J$--function.  We found   the  expected clumping of   galaxy   clusters, as indicated by $J(r)\\le1$.  Qualitatively, the   $J(r)\\le1$ may be   explained with a $\\xi_2(r)>0$  and the Gaussian approximation in Eq.~(\\ref{eq:Jgauss}). This  clumping  out to scales  of  40\\hMpc\\ is  confined mainly to the interior of the superclusters. Isolated \"field\" clusters were  not included in the  supercluster  sample but may contribute to the correlation seen up to  scales of 50\\hMpc in the cluster samples. One hierarchical level higher, the supercluster  centers  themselves show a tendency  towards regular structures.  Again   this can be explained  qualitatively with the Gaussian approximation (see Fig.~\\ref{fig:gauss}). A theoretical example  illustrating such  a  hierarchical property  is given by Neyman--Scott  processes {}\\cite{neyman:statistical}: In such a process the overall distribution  of points shows correlation (i.e.\\ $\\xi(r)>0$ for small $r$), but the cluster centers of these points are distributed randomly by construction.  Unlike the  two--point correlation function the $J$--function incorporates information  stemming from high  order correlations.  Our example in  Fig.~\\ref{fig:2DxiJ} illustrates, that a regular structure detected unambigously with the $J$--function  may not be visible in an analysis with the two--point correlation function $\\xi_2(r)$ alone.  Another problem is the  fluctuations between the northern and southern parts of the sample. This may be attributed to the different selection effects entering the  Abell and ACO  parts of the sample, probably due to the  different sensitivity of the  photo plates used.   However, in the case of  the IRAS 1.2~Jy  galaxy catalogue such  fluctuations were shown   to        be      real    on     scales        of    200\\hMpc\\ {}\\cite{kerscher:fluctuations}. Also, {}\\scite{zucca:esp-ii} find from  the ESP survey, that  at least in the southern hemisphere the local density is  below the mean sample density out to 140\\hMpc. If we assume that the fluctuations  decrease on scales above 200\\hMpc, the   finding of regular structures  on  such large  scales is a great challenge   to   the standard scenarios   of    structure formation by gravitational  instability,   starting from Gaussian  initial  density fluctuations.    Implications  of these    regular structures  for the standard scenarii  of     structure  formation  are     discussed   in {}\\scite{einasto:supercluster_III} and {}\\scite{szalay:walls}."
    },
    "9805/astro-ph9805367_arXiv.txt": {
        "abstract": "Simulations of structure formation in the Universe predict accretion shock waves at the boundaries of the large-scale structures as sheets, filaments, and clusters of galaxies. If magnetic fields are present at these shocks, particle acceleration should take place, and could contribute to the observed cosmic rays of high energies. When the radio plasma of an old invisible radio lobe is dragged into such a shock wave, the relativistic electron population will be reaccelerated and the plasma becomes radio-luminous again. Such tracers of the accretion shock waves are observed at the boundaries of some clusters of galaxies: the so-called cluster radio relics, which are large regions of diffuse radio emission, without any parent galaxy nearby. The observed properties of the cluster radio relics are naturally explained by accretion shock waves. Radio relics therefore give the first evidence for the existence of accretion shocks of the large-scale structure formation and they allow investigations of the shock properties. ",
        "introduction": "The large-scale structure of the Universe, seen in the structured galaxy distribution, is still forming. Matter is flowing out of the cosmic voids onto sheets and filaments. Within the filaments the matter flows to the density cusps frequently located at the intersection points of filaments: the clusters of galaxies. Whenever the flow passes from one structure into another, its velocity suddenly changes and several Mpc sized shock waves occur. At these shock waves the kinetic energy of the gravitationally accelerated gas is dissipated, mainly thermalized to temperatures of a few $0.1$ keV in filaments and several keV in clusters of galaxies. The shock velocity at filaments is expected to be of the order of several $100$ km s$^{-1}$, and the accretion shocks at clusters $1000-2000$ km s$^{-1}$. From simulation of structure formation rough values of the cluster accretion shock radius and velocity can be given in terms of the cluster temperature as an indicator of the gravitational potential (Kang et al. 1997): \\begin{eqnarray} \\label{eq:rs} r_{\\rm s} &=& 4.24\\, h_{50}^{-1} {\\rm Mpc} \\,( kT_{\\rm obs}/{6.06\\,{\\rm keV}} )^{1/2} (1+z)^{-3/2}\\\\ \\label{eq:Vs} V_{\\rm s, predicted} &=& 1750 \\, {\\rm km\\, s^{-1}}\\, ({kT_{\\rm obs}}/{6.06\\,{\\rm keV}} )^{1/2}\\,\\,. \\end{eqnarray} The dissipated accretion power per shock surface is: \\begin{equation} \\label{eq:Qflow} Q_{\\rm flow} \\approx \\frac{1}{2}\\,n_{\\rm e}\\,m_{\\rm p}\\,V_{\\rm s}^3 \\approx 4\\cdot 10^{44} \\;\\, \\frac{\\rm erg\\,s^{-1}}{\\rm Mpc^2}\\,\\,\\frac{n_{\\rm e}}{10^{-5}\\,{\\rm cm^{-3}}}\\,\\left( \\frac{kT_{\\rm obs}}{6.06\\,{\\rm keV}} \\right)^{3/2}\\,. \\end{equation} ",
        "conclusions": ""
    },
    "9805/astro-ph9805151_arXiv.txt": {
        "abstract": "During the initial data reduction of the Wisconsin H-Alpha Mapper (WHAM) \\ha\\ Sky Survey, we have discovered several very long ($\\sim 30\\arcdeg$--80\\arcdeg) filaments superimposed on the diffuse \\ha\\ background. These features have no clear correspondence to the other phases of the interstellar medium revealed by 21 cm, X-ray, IR, or radio continuum surveys, and they have no readily identifiable origin or source of ionization. In this letter, the data for two of these faint ($I_{H\\alpha} \\approx$ 0.5--1.5 R) structures are presented. The first is an 80\\arcdeg-long, 2\\arcdeg-wide arch that extends nearly perpendicular to the Galactic plane at $\\ell = 225\\arcdeg$ and attains a maximum latitude of $+51\\arcdeg$ near $\\ell = 240\\arcdeg$ before reaching the southern boundary of our survey map at $\\ell = 270\\arcdeg$, $b = +42\\arcdeg$.  The vertical portion of this feature between $b = +10\\arcdeg$ and $+25\\arcdeg$ is associated with a single radial velocity component centered at $\\vlsr = +16$ \\kms\\ with a full width at half maximum of 27 \\kms. A decrease in the velocity is observed from $b = +33\\arcdeg$ through $+48\\arcdeg$ as the feature arches toward higher Galactic longitudes. At this end, the emission component is centered near $\\vlsr = -20$ \\kms.  Where this feature appears to meet the Galactic plane near $\\ell = 225\\arcdeg$, it is directly above the \\hii\\ region surrounding CMa R1/OB1.  A second filament consists of a $\\sim 25\\arcdeg$--30\\arcdeg-long arc spanning $\\ell = 210\\arcdeg$--240\\arcdeg\\ at $b = +30\\arcdeg$ to 40\\arcdeg.  The radial velocity of this feature increases systematically from 0 \\kms\\ at $\\ell = 215\\arcdeg$, $b = +38\\arcdeg$ to +18 \\kms\\ at $\\ell = 236\\arcdeg$, $b = +28\\arcdeg$. Both features have rather constant intensities along their entire lengths, ranging from 0.5--1.5 R (EM = 1--3 cm$^{-6}$ pc) with no obvious trends. ",
        "introduction": "\\label{sec:intro}  The Wisconsin H-Alpha Mapper\\footnote{\\texttt{http://www.astro.wisc.edu/wham/}} (WHAM) survey is providing the first velocity-resolved map of the \\ha\\ emission from our Galaxy's diffuse interstellar medium.  The combination of WHAM's sensitivity ($< 0.1$ R) and velocity resolution (12 \\kms) reveals new details about the large-scale structure and kinematics of the Warm Ionized Medium (WIM).  Studies of faint emission structures in the WIM may help us understand the distribution and morphology of the ionized gas and may even provide direct evidence of ionization sources. In this letter, we present the discovery of two large, faint filaments in our Galaxy. ",
        "conclusions": "\\label{sec:discuss}  The intensities in Table~\\ref{tab:params} are converted to emission measures using the formula: \\begin{equation} \\label{itoem} {\\rm EM\\ (cm}^{-6}\\ {\\rm pc)} = 2.75\\ T_4^{0.9}\\ I_{\\rm H\\alpha}\\ {\\rm (R)}\\ e^{2.2\\,E(B-V)}, \\end{equation} where $T_4$ represents the temperature of the emitting gas in units of $10^4$ K. With typical values for the temperature of the WIM gas, $T_4 = 0.8$ (\\cite{r85}), an \\ha\\ intensity of 1 R corresponds to an EM of 2.25 cm$^{-6}$ pc. Since we are examining mostly high-latitude features, we expect corrections for interstellar extinction to be small. No such correction has been applied to the data presented here.  To derive further physical parameters for the filaments, we must make an assumption about the distances to them. Some evidence suggests that the lower part of Filament 1 is at a distance of about 1 kpc, based on its radial velocity and possible association with CMa R1/OB1. The lower portion has a radial velocity of $+16$ \\kms, which is consistent with a kinematic distance of approximately 1 kpc at this Galactic longitude for a value of Oort's Constant $A = 16$ \\kms\\ kpc$^{-1}$ (\\cite{mb81}). Unfortunately, this argument is weakened by the radial velocities of the filament's high-latitude portion, which are ``forbidden'' in the Galactic rotation model.  However, in Figure~\\ref{fig:images}d ($\\vlsr = +20$ \\kms), the vertical portion of Filament 1 appears to end in the prominent \\hii\\ region surrounding the star-forming association CMa R1/OB1. In fact, the center of the brighter, arc-shaped, \\ha\\ emission features in this \\hii\\ region, as seen on the Palomar Sky Survey red plates, is at $\\ell = 225\\arcdeg$, $b = -1\\arcdeg$ (\\cite{ro78}). Also, the velocity centroid of the \\hii\\ region (Table~\\ref{tab:params}) is similar to that of the first 15\\arcdeg--20\\arcdeg\\ of the filament away from the plane. Unless the agreement between the spatial and velocity information of these two regions is a coincidence, we can use the distance of the \\hii\\ region as an estimate of the distance to the filament. The distance to the CMa OB1 association has been determined photometrically by Clari\\'{a} (1974) to be $1150 \\pm 140$ pc, consistent with a kinematic distance of 800-1100 pc for the \\ha\\ emitting gas in the region (\\cite{ro78}). In this letter, we adopt a distance of 1 kpc for Filament 1 but include $\\dist = d/1$ kpc in the derived parameters.  At a distance of 1 kpc, the vertical extent of Filament 1 ($b \\approx +51\\arcdeg$) translates to a vertical height above the Galactic mid-plane, $Z$, of $1200\\;\\dist$ pc and a width ($\\approx 2\\arcdeg$) of $35\\;\\dist$ pc. If we assume the filament is a cylinder of uniform density gas of width $L$, then the emission measure, \\begin{equation} \\label{em} {\\rm EM\\ (cm}^{-6}\\ {\\rm pc)} = \\int_{0}^{L} n_e^2\\,dl = n_e^2 L, \\end{equation} of the feature can be used to estimate the density. For Filament 1, where $L = 35\\;\\dist$ pc and ${\\rm EM} \\approx 1.1$ cm$^{-6}$ pc, $n_e = 0.18\\;\\dist^{-1/2}$ cm$^{-3}$. The typical column density through the filament is $N_e = 1.9 \\times 10^{19}\\;\\dist^{1/2}$ cm$^{-2}$. The mass of the material in the filament's vertical section can be estimated by $1.4\\;m_H\\;Z\\;L^2\\;n_e = 9.2 \\times 10^{3}\\;\\dist^{5/2}$ M$_\\odot$, where the factor 1.4 is a correction for helium. If the filament is photoionized, our observed EM implies an incident ionizing flux of at least $\\alpha_B\\;{\\rm EM} = 1.0 \\times 10^{6}$ ph cm$^{-2}$ s$^{-1}$. The hydrogen recombination rate within the filament is given by $\\alpha_B\\;n_e^2 = 1.0 \\times 10^{-14}\\;\\dist^{-1}$ cm$^{-3}$ s$^{-1}$, implying that the power required to sustain this rate throughout the volume of the filament is $(\\alpha_B\\;n_e^2\\;Z\\;L^2) \\times 13.6\\ {\\rm eV} = 9.4 \\times 10^{36}\\;{\\dist}^2$ erg s$^{-1}$.  The value of $\\alpha_B$ used in these calculations, $3.10 \\times 10^{-13}$ cm$^3$ sec$^{-1}$, is interpolated from Osterbrock (1989) assuming a gas temperature of 8000 K.  The narrow shape and location of Filament 1, particularly the coincidence in location and radial velocity of one end of the filament with an energetic source in the plane (CMa R1/OB1), suggest one possibility for its origin---a jet-like ejection from the association. However, given the length of the filament at the distance of CMa R1/OB1, the apparently low velocities associated with it (Table~\\ref{tab:params}), and the near constant \\ha\\ intensity profile along its entire length, it is difficult to reconcile a scenario in which ionized hydrogen is being ejected from CMa R1/OB1 with the gas's short recombination times. A parcel of ionized gas at these temperatures and densities will recombine in $t_r = (\\alpha_B\\;n_e)^{-1} = 1.8 \\times 10^{13}\\;\\dist^{1/2}\\ {\\rm s} = 5.7 \\times 10^5\\;\\dist^{1/2}$ yr. To reach the observed height and remain ionized, the gas would need to be ejected at speeds in excess of $Z/t_r = 1400\\;\\dist^{1/2}$ \\kms.  We see no evidence for such speeds in our data.  Furthermore, we would expect a significant gradient in the filament's intensity as a function of height above the Galactic plane, since gas at larger distances from the plane has had more time to recombine.  If another source is responsible for maintaining the ionization of Filament 1, an ejection scenario could still be plausible. If the filament's arc shape is caused by free-fall of ejected gas back to the Galactic plane, an initial velocity of about 70 \\kms\\ is required to reach the observed height above the plane at the distance of CMa R1/OB1. Such speeds agree better with the measured filament component velocities in Table~\\ref{tab:params}, particularly when projection effects are considered. However, the filament's length would then require that the ejecting source be more than $3 \\times 10^7$ yr old. This number can be compared to estimates of $3 \\times 10^6$ yr for the age of CMa OB1 by Clari\\'{a} (1974) and $7.6 \\times 10^5$ yr for the age of a supernova put forth by Herbst \\& Assousa (1977) as being responsible for generating the ring-shaped \\ha\\ nebulosity in the vicinity of CMa OB1/R1 and for being the progenitor of star formation in the R1 association. In this scenario, a diffuse ionization source is probably required to maintain a constant intensity along the filament, since the growing distance from CMa R1/OB1 would produce an intensity gradient if it were the sole source. The efficiency of leaking Lyman continuum radiation from the disk is a longstanding problem for ionizing a thicker layer of the Galaxy.  However, Bland-Hawthorn \\& Maloney (1998), Dove \\& Shull (1994), and Miller \\& Cox (1993) have argued that a substantial ionizing flux can diffuse into the Galaxy's halo from the disk. The required ionizing flux of at least $1.0 \\times 10^6$ ph cm$^{-2}$ s$^{-1}$ is consistent with these models.  Several other phenomena may produce large, faint filaments. Large-scale, kinematic Galactic structures such as superbubbles, chimneys, or worms have been popular explanations for filamentary structures seen in \\ion{H}{1} as well as \\ha\\ (\\emph{e.g.} \\cite{khr92}; \\cite{ro79}). The velocity gradient of Filament 1 is outside the range expected from a simple model of Galactic rotation and therefore may be evidence for a dynamical influence on the filament. However, the observed velocity gradient is inconsistent with this feature being the edge of an expanding shell, since a shell's projected edges should be at a constant radial velocity. Furthermore, for the two cases presented here, we find no obvious correlation between these \\ha\\ filaments and emission at other wavelengths, including \\ion{H}{1} 21 cm, Rosat All-Sky Survey X-Ray, and IRAS bands, making it difficult to relate them to previously identified \\ion{H}{1} ``worms'' and superbubbles. Since the Leiden/Dwingeloo \\ion{H}{1} survey is sensitive to below $5 \\times 10^{18}$ cm$^{-2}$ (\\cite{hb97}), the ionized column density of the filament, $2 \\times 10^{19}$ cm$^{-2}$, suggests that the gas is fully ionized.  Other possibilities include a suggestion by Dupree \\& Raymond (1983) that faint, ionized trails of ionized hydrogen could be produced by high-velocity white dwarfs.  However, the recombination time discussed above suggests unreasonable velocities for the star.  In summary, we present the discovery of two long, faint \\ha\\ filaments at high latitudes from the WHAM \\ha\\ survey. Their origin is not yet identified, but the existence of such features may help to explain processes responsible for the maintenance of the general WIM layer. As additional portions of the \\ha\\ survey are reduced, new clues about the nature of these structures may be revealed. We plan to follow these observations with [\\ion{S}{2}] and [\\ion{N}{2}] observations of this region. Additional observations will provide information about the gas's temperature and ionization state, which are useful in narrowing the kinds of processes that can be producing these filaments.  We thank Kurt Jaehnig and Jeff Percival of the University of Wisconsin's \\emph{Space Astronomy Lab} for their dedicated engineering support of WHAM; Nikki Hausen, Mark Quigley, and Brian Babler for data reduction support; and Trudy Tilleman for essential night-sky condition reports from Kitt Peak, which have made remote observing possible. We acknowledge the use of NASA's \\emph{SkyView} facility (http://skyview.gsfc.nasa.gov), located at NASA Goddard Space Flight Center and the SIMBAD database, operated at CDS, Strasbourg, France. This work is supported by the National Science Foundation through grants AST9619424 and AST9122701.  \\newpage"
    },
    "9805/astro-ph9805198_arXiv.txt": {
        "abstract": "We report on a 50\\,ks observation of the bright Seyfert 1 galaxy MCG$-$6-30-15 with the {\\it Rossi X-ray Timing Explorer}.  The data clearly show the broad fluorescent iron line (equivalent width $\\sim$ 250 eV), and the Compton reflection continuum at higher energies.  A comparison of the iron line and the reflection continuum has enabled us to constrain reflective fraction and the elemental abundances in the accretion disk.  Temporal studies provide evidence that spectral variability is due to changes in both the amount of reflection seen and the properties of the  primary X-ray source itself. ",
        "introduction": "The current paradigm for active galactic nuclei (AGN) is a central engine consisting of an accretion disk surrounding a supermassive black hole (e.g., see review by Rees 1984). The main source of power is the release of gravitational potential energy as matter falls towards the central black hole.  Much of this energy is released in the form of X-rays, some fraction of which are reprocessed by matter in the AGN (Guilbert \\& Rees 1988; Lightman \\& White 1988).   Careful study of the X-ray reprocessing mechanisms can give much information about the immediate environment of the accreting black hole. These effects of reprocessing can often be observed in the form of emission and absorption features in the X-ray spectra of AGNs.  In Seyfert 1 nuclei, approximately half of the X-rays are `reflected' off the inner regions of the accretion disk. Since it is superposed on the direct (power-law) primary X-ray emission, the principle observables of this reflection are a fluorescent iron K$\\alpha$ line, and a Compton backscattered continuum which hardens the observed spectrum above $\\sim 10\\keV$ (see eg. George \\& Fabian 1991). The iron line together with the reflection component are important diagnostics for the geometry and physics of the X-ray continuum source. The strength of the emission line relative to the reflection continuum depends largely on the abundance of iron relative to hydrogen in the disk, as well as the normalization of the reflection spectrum relative to the direct spectrum. (There is also a dependence on the relative oxygen abundance, Reynolds et al. 1995.) The relative normalization of the reflection spectrum probably depends primarily on the geometry (i.e., the solid angle subtended by the reflecting parts of the disk as seen by the X-ray source).  However, it can also be affected by strong light bending effects (e.g., Martocchia \\& Matt 1996) or special-relativistic beaming effects (e.g. Reynolds \\& Fabian 1997). Disentangling the abundance from the absolute normalization of the reflection component is an important first step in constraining these effects and hence the construction of physical models for AGN central regions.   MCG$-$6-30-15 is a Seyfert 1 galaxy that is both bright and nearby (z=0.008).  Since its identification, MCG$-$6-30-15 has been intensively studied by every major X-ray observatory.  An extended {\\it EXOSAT} observation provided the first evidence for fluorescent iron line emission (Nandra et al. 1989) which was attributed to X-ray reflection.  Confirmation of these iron features by {\\it Ginga} as well as the discovery of the associated Compton reflected continuum supported the reflection picture (Nandra, Pounds \\& Stewart 1990; Pounds et al. 1990; Matsuoka et al. 1990).   ASCA data showed the iron line to be broad, skewed, and variable (eg. Tanaka et al 1995; Iwasawa et al. 1997).  In this paper, we present the first data from the {\\it Rossi X-ray Timing Explorer} ({\\it RXTE}) for MCG$-$6-30-15.  Our observation shows clear evidence for a redshifted broad iron line at $\\sim$ 6.1 keV and the reflection continuum above 10 keV.  Due to the larger effective area and longer  exposure of the {\\it RXTE} observation as compared with {\\it Ginga}, we can study the reflection continuum in detail for the first time.   We present preliminary constraints on the abundances of iron and reflective fraction, and investigate the relationship between spectral changes and the reflection component during the different  phases of our data.  Section 2 will detail the data analysis procedure followed by spectral fitting results in Section 3. We present a study of temporal variations on spectral components with particular emphasis on the reflection component in Section 4.  This will follow with a discussion of results and future work in Section 5. ",
        "conclusions": "The purpose of this paper was to show what unanswered questions can be addressed with the large area and wide-band coverage of {\\it RXTE} even with the current uncertainties in spectral calibration. The presence of a broad iron line is clearly evident as shown with a simple power law fit, and is one of the first detections where both features are seen simultaneously.  We add a reflection component to our power law and gaussian fit to find that reflection is necessary to describe our data.  We note also that the steep intrinsic photon index coupled with a narrow $H \\beta$ FWHM implies that MCG$-$6-30-15 can be a possible narrow-line Seyfert 1 galaxy candidate.  While spectral results may change in detail over the course of the next year with further improvements in calibration, we can already begin to place upper bound limits on the relationship between abundance values and reflective fraction.  In Section 4, we study the effects of temporal variability on spectral components and find evidence to support the notion that variability may be due to changes in the amount of reflection seen (e.g. due to gravitational or Doppler beaming of the primary emission towards the disk).  It is however not clear whether this effect may also be coupled with  contributions from changes in the properties of the source itself (e.g. the temperature and optical depth of the coronal plasma).  We expect to be able to resolve these issues better with longer looks and simultaneous ASCA observations.  For the time being, the present results are important observational first  steps in understanding some of the physics of AGN reprocessing mechanisms, and push the limits of our knowledge."
    },
    "9805/astro-ph9805221_arXiv.txt": {
        "abstract": "}[2]{{\\footnotesize\\begin{center}ABSTRACT\\end{center} \\vspace{1mm}\\par#1\\par \\noindent {\\bf Key words:~~}{\\it #2}}}  \\newcommand{\\TabCap}[2]{\\begin{center}\\parbox[t]{#1}{\\begin{center} \\small {\\spaceskip 2pt plus 1pt minus 1pt T a b l e} \\refstepcounter{table}\\thetable \\\\[2mm] \\footnotesize #2 \\end{center}}\\end{center}}  \\newcommand{\\TableSep}[2]{\\begin{table}[p]\\vspace{#1} \\TabCap{#2}\\end{table}}  \\newcommand{\\TableFont}{\\footnotesize} \\newcommand{\\TableFontIt}{\\ttit} \\newcommand{\\SetTableFont}[1]{\\renewcommand{\\TableFont}{#1}}  \\newcommand{\\MakeTable}[4]{\\begin{table}[htb]\\TabCap{#2}{#3} \\begin{center} \\TableFont \\begin{tabular}{#1} #4 \\end{tabular}\\end{center}\\end{table}}  \\newcommand{\\MakeTableSep}[4]{\\begin{table}[p]\\TabCap{#2}{#3} \\begin{center} \\TableFont \\begin{tabular}{#1} #4 \\end{tabular}\\end{center}\\end{table}}  \\newenvironment{references}% { \\footnotesize \\frenchspacing \\renewcommand{\\thesection}{} \\renewcommand{\\in}{{\\rm in }} \\renewcommand{\\AA}{Astron.\\ Astrophys.} \\newcommand{\\AAS}{Astron.~Astrophys.~Suppl.~Ser.} \\newcommand{\\ApJ}{Astrophys.\\ J.} \\newcommand{\\ApJS}{Astrophys.\\ J.~Suppl.~Ser.} \\newcommand{\\ApJL}{Astrophys.\\ J.~Letters} \\newcommand{\\AJ}{Astron.\\ J.} \\newcommand{\\IBVS}{IBVS} \\newcommand{\\PASP}{P.A.S.P.} \\newcommand{\\Acta}{Acta Astron.} \\newcommand{\\MNRAS}{MNRAS} \\renewcommand{\\and}{{\\rm and }} ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805017_arXiv.txt": {
        "abstract": "The position in the HR diagram and the kinematic characteristics of different kinds of CP stars of the upper main sequence are obtained using the LM method (Luri et al., 1996). Most of the CP stars are main sequence stars occupying the whole width of the sequence. From a kinematic point of view, they belong to the young disk population (ages $\\la$\\, 1.5 Gyr). It has also been found that, on kinematic grounds, the behaviour of $\\lambda$\\,Bootis stars is similar to the one observed for normal stars of the same spectral range. On the other hand, roAp and noAp stars show the same kinematic characteristics. The peculiar velocity distribution function has been decomposed into a sum of three dimensional gaussians and the presence of Pleiades, Sirius and Hyades moving groups has been clearly established. Finally, a small number of CP stars are found to be high-velocity objects. ",
        "introduction": "\\label{intr} The release of Hipparcos data (ESA, 1997) allows to reconsider the luminosity of CP stars of the upper main sequence and their kinematic behaviour on sounder bases. In the present paper the following kinds of CP stars have been considered: He-rich, He-weak, HgMn, Si, SrCrEu and the related group of $\\lambda$\\,Bootis stars.  The LM statistical method (Luri et al., 1996) has been applied to the different samples. This method has the advantage that all the available astrometric data (whatever the quality of the parallax is) as well as radial velocity data are used for  each star for the luminosity calibration. It also provides the kinematic characteristics of the samples. The method is able to identify and separate groups of stars with different luminosity, kinematic or spatial properties, allowing the treatement of non homogeneous samples. ",
        "conclusions": "\\vspace{-1mm} Our main results can be summarized as follows:\\\\ - Most CP stars are main sequence objects occupying the whole width of the sequence (about 2 mag). The intrinsic dispersion in absolute magnitude varies from 0.5 to 0.8 mag for all the groups except He-rich stars which spread a large range in luminosities. Some Am stars in the secondary groups are out of the main sequence, but before reaching a definitive conclusion it will be necessary to search for possible misclassifications.\\\\ - From a kinematic point of view, CP stars belong to the disk population younger than 1\\,-\\,1.5 Gyr. The velocity field shows the presence of moving groups as observed for normal stars of the same spectral range. In particular, the presence of Pleiades, Sirius and Hyades moving groups has been clearly established.\\\\ - $\\lambda$\\,Bootis stars are concentrated in the main sequence. The definition of this type of stars is not well established (see Gerbaldi, these proceedings). Their evolutionary status remains controversial, but the kinematic characteristics correspond to those of non-peculiar stars of the same spectral range.\\\\ - roAp and noAp stars show similar kinematic characteristics. \\vspace{-2mm}"
    },
    "9805/astro-ph9805079_arXiv.txt": {
        "abstract": "Soft X--ray Transients (SXRTs) have long been suspected to contain old, weakly magnetic neutron stars that have been spun up by accretion torques. After reviewing their observational properties, we analyse the different regimes that likely characterise the neutron stars in these systems across the very large range of mass inflow rates, from the peak of the outbursts to the quiescent emission. While it is clear that close to the outburst maxima accretion onto the neutron star surface takes place, as the mass inflow rate decreases, accretion might stop at the magnetospheric boundary because of the centrifugal barrier provided by the neutron star. For low enough mass inflow rates (and sufficiently short rotation periods), the radio pulsar mechanism might turn on and sweep the inflowing matter away. The origin of the quiescent emission, observed in a number of SXRTs at a level of $\\sim 10^{32}-10^{33}\\ergs$, plays a crucial role in constraining the neutron star magnetic field and spin period. Accretion onto the neutron star surface is an unlikely mechanism for the quiescent emission of SXRTs, as it requires very low magnetic fields and/or long spin periods. Thermal radiation from a cooling neutron star surface in between the outbursts can be ruled out as the only cause of the quiescent emission.  We find that accretion onto the neutron star magnetosphere and shock emission powered by an enshrouded radio pulsar provide far more plausible models. In the latter case the range of allowed neutron star spin periods and magnetic fields is consistent with the values recently inferred from the properties of kHz quasi-periodic oscillation in low mass X--ray binaries. If quiescent SXRTs contain enshrouded radio pulsars, they provide a missing link between X--ray binaries and millisecond pulsars. ",
        "introduction": "Transient X--ray sources are quiescent and undetected for most of the time and undergo sporadic outbursts, typically lasting for 10--100~d, during which they emit an intense X--ray flux. They were initially classified on the basis of their spectral hardness, owing to the lack of a clear understanding of their nature (Cominsky et al. 1978). The subsequent discovery of a number of phenomena observed also in different classes of X--ray binaries (e.g. X--ray pulsations and bursts; see e.g. White 1989) showed that there is a close relationship between transient and persistent accreting compact sources. White, Kaluzienski \\& Swank (1984) introduced a revised spectral classification that further extends this analogy. The spectra of {\\it hard} X--ray transients (HXRTs) are characterised by equivalent temperatures $\\gsim 15$ keV. These sources often contain a young, pulsating, neutron star orbiting a Be star companion and are clearly associated to persistent X--ray pulsars in high mass binaries (Maraschi, Treves \\& van den Heuvel 1976). The outburst of {\\it soft} X--ray transients (SXRTs), characterised by equivalent temperatures $\\lsim 15$ keV, are often accompanied by a pronounced increase in the luminosity of their (faint) optical counterparts and by the onset of type I burst (thermonuclear flashes on the surface of a neutron star) activity. These properties clearly associate SXRTs with low mass X--ray binaries (LMXRBs) containing an old neutron star. The {\\it ultrasoft} X--ray transients are also associated to LMXRBs, but in this case the similarity with the ``high state\" spectra of persistent black hole candidates (BHCs), together with the absence of bursts and pulsations, suggests that these systems likely harbor a stellar mass black hole (White \\& Marshall 1984). This class has been later extended to include also {\\it hard tail} transient sources, based on the analogy with the spectral characteristic of Cyg~X-1 in its ``low state\". The prediction on the nature of the compact object in different classes of LMXRB transients has been brilliantly confirmed by mass measurements which established A~0620--00, GS~2023+338, GS~1124--68 and GRO J1655--40 as firm BHCs (McClintock \\& Remillard 1986; Casares, Charles \\& Naylor 1992;  Orsoz et al. 1996) and Cen X-4 as a neutron star (Shahbaz, Naylor \\& Charles 1993).  Transients systems are characterised by an X--ray luminosity that varies over many decades (variations between $10^{33}$ and $10^{38}\\ergs$ are not uncommon). Therefore they allow to investigate accretion onto collapsed stars over a much larger range of luminosities, and therefore accretion rates, than persistent sources. This is well illustrated e.g. by the case of the HXRT EXO~2030+375, a 42~s pulsator, that led to a remarkable progress in the understanding of the physics of accretion onto magnetic neutron stars (Parmar et al. 1989;  Parmar, White \\& Stella 1989; Angelini, Stella \\& Parmar 1989).  SXRTs, while still poorly studied, provide a unique opportunity to gain crucial insights in the neutron stars that are hosted in non-pulsating LMXRBs. According to current evolutionary scenarios, the neutron stars in transients as well as persistent non-pulsating LMXRBs are gradually spun-up by accretion torques to limiting periods ranging from milliseconds to tens of milliseconds, depending on the value of their residual magnetic field (Alpar et al. 1982; Bhattacharya \\& van den Heuvel 1991; Phinney \\& Kulkarni 1994). Once accretion from the companion star stops, the neutron stars in these systems are expected to shine as ``recycled\" radio pulsars orbiting a low mass companion. These LMXRBs therefore likely represent the progenitors of the weak magnetic field ($10^8-10^9$~G) millisecond radio pulsars (MSPs) that are discovered in increasing number in the Galaxy and globular clusters (Backer et al. 1982; Manchester et al. 1991; Taylor, Manchester \\& Lyne 1993).  Despite numerous searches, fast coherent X--ray pulsations in the persistent emission of LMXRBs directly arising from the neutron star rotation have proved elusive (see e.g. Vaughan et al. 1994 and references therein). Only after the launch of the Rossi X--Ray Timing Explorer (RossiXTE) kiloHertz (kHz) quasi-periodic oscillations (QPOs) have been discovered in several objects (for a review see van der Klis 1997, 1998). In the great majority of the brightest X--ray binaries two kHz QPOs (between 0.3 and 1.2~kHz) have been discovered. These sources are usually classified according to the track in the colour-colour diagram for different luminosities (e.g. Hasinger \\& van der Klis 1989; van der Klis 1995). Atoll sources (or suspected) usually maintain their frequency difference ($\\sim 300-500$ Hz) constant over large variations of QPOs centroid frequencies and the centroid frequency appears to be positively correlated with the X--ray luminosity (4U 1728--34 Strohmayer et al. 1996; 4U 0614+091 Ford et al. 1997; KS 1731--260 Wijnands \\& van der Klis 1997; 4U 1636--53 Wijnands et al. 1997; 4U 1735--44 Wijnands et al. 1998a; 4U 1820--30 Smale, Zhang \\& White 1997; 4U 1705--44 Ford, van der Klis \\& Kaaret 1998). A different case is presented by 4U 1608--52 for which a varying peak separation has been observed (Mendez et al. 1998).  In the Z source Sco X-1 (van der Klis et al. 1997) the frequency separation between the kHz QPOs decreases with mass accretion rate, but in the other Z sources (GX 5--1 van der Klis et al. 1996; Cyg X-2 Wijnands et al. 1998b; GX 340+0 Jonker et al. 1998; GX 17+2 Wijnands et al. 1998c) the separation remains approximately constant, although a similar decrease in peak separation as found in Sco X-1 can not be excluded.  Nearly constant pulsations at this frequency difference have also been revealed during X--ray bursts in 4U 1728--34 (2.8 ms; Strohmayer et al. 1996) and during a persistent emission interval of 4U 0614+09 (3.1 ms; Ford et al. 1997). In the case of KS 1731--260 and 4U 1636--53 the frequency difference between the two QPO peaks is consistent with half the frequency of the nearly periodic signals at $\\sim 524$ and 581 Hz, respectively, that have been detected during type I bursts from these sources (KS 1731--260: Smith, Morgan \\& Bradt 1997; Wijnands \\& van der Klis 1997; 4U 1636--53: Zhang et al. 1996a; Wijnands et al. 1997). In Aql X-1 a similar coherent modulation during an X--ray burst has been observed at $\\sim 550$ Hz, whereas only one kHz QPO has been revealed at 750--830 Hz (Zhang et al. 1998).  The most straightforward interpretation of these findings is based on magnetospheric beat-frequency models (Alpar \\& Shaham 1985; Lamb et al. 1985; Miller, Lamb \\& Psaltis 1998): the nearly coherent signal during type I bursts corresponds to the spin frequency of the neutron star, whereas the higher frequency kHz QPO arise from the Keplerian motion of matter in the innermost accretion disk region, close to the magnetospheric boundary or at the sonic radius. The lower frequency kHz QPO originates instead from modulated accretion at the beat frequency between the neutron star spin frequency and the Keplerian frequency at the magnetospheric boundary.  These kHz QPOs provide the first evidence that the neutron stars in LMXRBs are spinning at periods of the order of milliseconds and are the likely progenitors of MSPs.  The large accretion rate variations that are characteristic of SXRTs should allow the exploration of a variety of different regimes for the neutron stars in these systems which are unaccessible to persistent LMXRBs. While it is clear that, when in outbursts, SXRTs are powered by accretion, the origin of the low luminosity X--ray emission that has been detected in the quiescent state of several SXRTs is still unclear. An interesting possibility is that a MSP be visible in the quiescent state of SXRTs (Stella et al. 1994;  hereafter Paper I). This would provide a ``missing link\" between persistent LMXRBs and recycled MSPs.  This paper concentrates on various aspects of the physics of the neutron stars in SXRTs. A short account of our main original results has been given in Paper I. After a review of the observations of SXRTs\\footnote{Readers who are familiar with the subject may refer to Table I only.} (Section 2, see also Tanaka \\& Shibazaki 1996), we give a brief description of the models for the outburst mechanism (Section 3). In Section 4 we explore the different regimes that are expected for the neutron stars of SXRTs in the decay phase of their outbursts. In Section 5 we expand on the different emission mechanisms which might be responsible for the quiescent luminosity.  The conditions under which the neutron stars of SXRTs evolve towards and remain within the radio pulsar region of the magnetic field -- spin period ($B-P$) diagram are discussed in Section 6. The main conclusions of the paper are presented in Section 7.  \\section {The properties of Soft X--ray Transients}  SXRTs are a fairly inhomogeneous class\\footnote{Some sources (especially in the vicinity of the Galactic Center) have been classified as transients even if their peak luminosity was only a few times higher than the instrumental detection limit. In the absence of additional evidence for their transient behaviour, these sources should be considered only as variable X--ray sources.}. Often their outbursts consist of a flux increase lasting a few days that reaches X--ray luminosities of $L_X \\sim 10^{37}-10^{38}\\ergs$, followed by a slower, nearly exponential decay with a timescale of weeks to months. Outbursts of this kind have been observed in Cen X-4, Aql X-1, 4U 1730--22 and A 1742--289 (White, Kaluzienski \\& Swank 1984 and references therein) and perhaps in a few other cases (see Table I). Other sources (like EXO~0748--676, 4U~2129+47 and 4U~1608--52), alternate long periods of relatively high (and often variable) X--ray flux with others in which they are detected at a much lower level, if at all. Unfortunately, the transitions between these intensity states have been poorly studied so far.  With the possible exception of Aql X-1 (see below), the intervals between outbursts are irregular, often in the 1--10 yr range, but for many sources only one outburst has been observed so far.  Available X--ray data on SXRTs are still sparse: in most cases the outburst monitoring has been carried out with wide field instruments, characterised by limited effective area, energy range and, especially, sensitivity. A relatively small number of pointed observations close to the outburst maxima have been obtained mainly with large area collimated detectors for a few systems, whereas the quiescent emission, months to years away from the outbursts, has been investigated for $\\sim 10$ SXRTs, mainly with low energy X--ray telescopes. At least two SXRTs, Aql~X-1 and Cen~X-4, have been determined to emit very different spectra in different states. While close to the outburst peak the spectra are relatively soft (equivalent thermal bremsstrahlung temperatures of $k\\,T_{\\rm br} \\sim 5$~keV), at intermediate luminosities ($\\sim 10^{35}-10^{37}\\ergs$) during the rise and decay of the outburst, a high energy tail extending to at least $\\sim 100$~keV is detected (a similar tail is also seen in a few persistent burst sources; e.g. Barret \\& Vedrenne 1994). The evolution at the end of the outburst from luminosities of $\\sim 10^{34}-10^{36}\\ergs$ to quiescence is basically unknown. Only recently, BeppoSAX observations allowed to explore this luminosity range in the case of Aql X-1 (Campana et al. 1998; see Section 2.2.1). The X--ray spectrum of Aql X-1 consists again of a soft component plus a hard energy tail, similar to the one observed at higher luminosities. The spectra in the quiescent state ($L_X \\sim 10^{32}-10^{33}\\ergs$) are characterised by a soft component (equivalent black body temperatures of $k\\,T_{\\rm bb}\\sim 0.1-0.3$~keV). For those SXRTs observed above a few keV, a power-law like high energy tail has been revealed which, in the case of Aql X-1, hardens as the luminosity decreases below $\\sim 10^{33}\\ergs$ (Campana et al. 1998).  Type I X--ray bursts have been observed in the active phase of thirteen SXRTs. These burst of X--ray radiation are most likely due to thermonuclear flashes at the surface of accreting neutron stars (Maraschi \\& Cavaliere 1977). There is no evidence that these bursts differ in any property from those of persistent LMXRBs (e.g. Lewin, van Paradijs \\& Taam 1993, 1995), thus supporting the idea that SXRTs contain weakly magnetic neutron stars accreting from a low mass companion.  Up to now three SXRTs (4U 1608--52 Berger et al. 1996; KS 1731--260 Wijnands \\& van der Klis 1997; Aql X-1 Zhang et al. 1998) have displayed kHz QPOs, when their X--ray luminosity was at a level of $10^{36}-10^{37}\\ergs$. In the case of 4U 1608--52 the centroid frequency of these QPOs (between 600 and 1100 Hz) does not correlate with the observed X--ray luminosity. For Aql X-1 a single kHz QPO in the same frequency range (750--830 Hz) has been observed at two different luminosities ($1.2-1.7\\times 10^{36}\\ergs$). The persistent emission of KS 1731--260 shows twin kHz QPO around 900 and 1160 Hz, respectively. The frequency difference ($\\sim 260$ Hz) between these QPOs is consistent with half the frequency of the nearly periodic 524 Hz signal observed in a Type I burst from this source (Smith, Morgan \\& Bradt 1997; Wijnands \\& van der Klis 1997). When interpreted in terms of beat-frequency models (Alpar \\& Shaham 1985; Lamb et al. 1985; Miller, Lamb \\& Psaltis 1998) these results imply a neutron star spin period of $\\sim 3.8$ ms.  The X--ray outbursts of SXRTs are often accompanied by a considerable enhancement of their optical luminosity (optical novae). Increases up to $\\sim 6$ magnitudes with respect to the quiescent state have been measured, which have greatly helped in the identification of the optical counterparts of seven SXRTs. The optical spectra during outbursts are usually characterised by a rather flat continuum with emission lines (Balmer, He II, N III; van Paradijs \\& McClintock 1995), similar to those of bright persistent LMXRBs. These spectra result mainly from reprocessing of the high energy photons at the accretion disk and companion star. In some cases, the intrinsic spectrum of the companion star becomes detectable in the quiescent state, therefore making detailed photometric and spectroscopic measurements possible. These studies have shown that SXRTs contain late type stars (G or K), and in some cases allowed a determination of the orbital periods. These are known for seven SXRTs and are in the range from 4 to 19 hr, similar to those of LMXRBs (van Paradijs \\& McClintock 1994). Only for one SXRT (Cen X-4) the mass function has been measured (Shahbaz, Naylor \\& Charles 1993).  Radio emission has been observed during the outbursts of A1742--289 (Davis et al. 1976), Aql X-1 (Hjellming, Han \\& Roussel-Dupr\\'e 1990) and Cen X-4 (Hjellming et al. 1988).  SXRTs are located in the galactic plane with a distribution similar to that of LMXRBs (van Paradijs \\& White 1995). The presence of SXRTs in globular clusters is of particular interest, due to their possible evolutionary link with recycled MSPs. It is possible that some of the dim ($L_X\\lsim 10^{34}\\ergs$) X--ray sources in globular clusters are SXRTs in quiescence (e.g. Verbunt et al. 1994a; for an alternative explanation see e.g. Grindlay 1994).  In Table I we summarise the main properties of the presently known or suspected SXRTs. In the following we give a brief outline of the best studied objects. In the compilation of Table I we have included all the transients for which there are indications that they consist of a neutron star with a low mass companion, excluding ultrasoft X--ray transients which likely harbor a black hole. Sources which are sometimes defined as transients in the literature, but for which there is no clear evidence of flux variations greater than a factor of a $\\sim 100$, have not been included. Among these are: KS 1732--273, EU 1737--132, EXS 1737.9--2952, GRS 1741.9--2853, GS 1826--24.  \\setcounter{figure}{0}  \\begin{figure*}[!p] \\psfig{figure=tab.ps} \\end{figure*}  \\subsection{Soft X--ray Transients with fast rise and exponential decay outbursts}  \\subsubsection {Aql X-1}  Aql X-1 (4U~1908+005) is the most active SXRT known: more than 30 X--ray and/or optical outbursts have been detected. This led to several attempts to correlate the properties of different outburst and to look for possible (quasi-)periodicities in the recurrence times. There is evidence that the peak intensity of an outburst correlates with the elapsed time from the previous one (White, Kaluzienski \\& Swank 1984; Kitamoto et al. 1993). A recurrence time of $\\sim 125$ d was quite evident in the 1969--1979 observations from the Ariel V and Vela 5B satellites (Priedhorsky \\& Terrell 1984). However, this periodicity did not extend to the time of more recent Ginga and optical observations (1987--1992), which on the contrary suggest of a $\\sim 310$ d periodicity (Kitamoto et al. 1993). The outbursts of Aql X-1 are generally characterised by a fast rise (5--10 d) followed by a slow exponential decay, with an $e-$folding time of 30--70 d. Type I X--ray bursts were first discovered by Koyama et al. (1981) during the declining phase of an outburst. A periodicity of 132 ms, which persisted for only $\\sim 1$ min, was detected during the peak of a type I burst observed with the Einstein SSS (Schoelkopf \\& Kelley 1991).  For a distance of $d\\sim 2.5$~kpc, the corresponding 1--10~keV luminosity is $L_X \\sim (0.9-4)\\times 10^{37}\\ergs$. Close to the outburst maxima the X--ray spectrum is soft with $k\\,T_{\\rm br} \\sim 4-5$~keV. X--ray observations of Aql X-1 during the decay of an outburst have been collected in the $L_X\\sim 10^{34}-10^{36}\\ergs$ luminosity range. During the 1979 outburst Einstein MPC observed several times Aql X-1 (Czerny, Czerny \\& Grindlay 1987). The 1.2--10 keV spectrum when the source was at a level of a few $10^{36}\\ergs$ is well fit by a thermal bremsstrahlung model with the same temperature as in outburst. At a level of $\\sim 10^{35}\\ergs$ the spectrum cannot be fit with the same model, but is instead consistent with a power-law model with photon index $\\Gamma\\sim 2.3$. The same power-law spectrum was recovered about half a year later when the source was at a level $\\sim 2\\times10^{34}\\ergs$ (Czerny, Czerny \\& Grindlay 1987). The ROSAT PSPC spectra during the 1990 and 1992 outbursts when $L_X\\sim 10^{35}-10^{36}\\ergs$ (0.1--2.4 keV) could not be fit satisfactorily by single component models (Verbunt et al. 1994b). Finally the ASCA spectrum (0.5--10 keV) when the Aql X-1 luminosity was $\\sim 2\\times 10^{35}\\ergs$ was well fit by a single power-law with photon index $\\Gamma\\sim 2$ (Tanaka \\& Shibazaki 1996; Tanaka 1994). Moreover, several episodes of hard X--ray emission have been discovered by BATSE during 1991--1994 (Harmon et al. 1996), when the X--ray luminosity was about $\\sim 4\\times 10^{36}\\ergs$. The X--ray spectra were characterised by a power-law of $\\Gamma\\sim 2-3$ and extending up to 100 keV.  Probably due to its closeness, Aql X-1 is one of a few SXRTs detected in quiescence. The ROSAT HRI and PSPC revealed Aql X-1 on three occasions at a level of $\\sim 10^{33}\\ergs$ in the 0.4--2.4 keV range. During these observations the spectrum was very soft and could be well fit either with a black body model with a temperature of $k\\,T_{\\rm bb}\\sim 0.3$ keV, a thermal bremsstrahlung with a temperature of $k\\,T_{\\rm br}\\sim 0.8$ keV or a power-law with $\\Gamma \\sim 3$. The derived black body temperature implies an emitting radius of $\\sim 10^{5}$ cm (Verbunt et al. 1994b).  An outburst from Aql X-1 reaching a peak luminosity of $\\sim 10^{37}\\ergs$ (2--10 keV) was discovered (Levine et al. 1997) and monitored starting from mid-February, 1997 with the RossiXTE All Sky Monitor (see Fig. 1). Several pointed observations were successively carried out leading to the discovery of a nearly coherent modulation at $\\sim 550$~Hz during a type I X--ray burst and a single QPO peak, with a frequency ranging from $\\nu_{QPO}\\sim 750$ to 830 Hz, at two different luminosities of $1.2-1.7\\times10^{36}\\ergs$ (Zhang et al. 1998).  Observations carried out with the BeppoSAX Narrow Field Instruments (NFIs) starting from March 8$^{th}$, 1997, allowed to study the final stages of the outburst decay  (see Fig. 1; Campana et al. 1998). At the time of the first BeppoSAX observation (which started on March 8$^{th}$, 1997) the source luminosity was decreasing very rapidly, fading by about 30\\% in 11 hr, from a maximum level of $\\sim 10^{35}\\ergs$. The second observation took place on March 12$^{th}$, 1997 when the source, a factor of $\\sim 50$ fainter on average, reduced its flux by about 25\\% in 12 hr. In the subsequent four observations the source luminosity attained its constant value of $\\sim 6\\times10^{32}\\ergs$ (0.5--10 keV). The sharp decrease after March 5$^{th}$ 1997 is well described by an exponential decay with an $e-$folding time $\\sim 1.2$ d (see Fig. 1). The quiescent luminosity is consistent with the value previously measured with other satellites (e.g. Verbunt et al. 1994b).  The X--ray spectra during the fast decay phase, as well as that obtained by summing up all the observations pertaining to quiescence, could be fit with a model consisting of a black body plus a power-law. The soft black body component remained nearly constant in temperature ($kT_{\\rm bb} \\sim 0.3-0.4$ keV), but its radius decreased by a factor of $\\sim 3$ from the decay phase to quiescence. The equivalent radius in quiescence ($R_{\\rm bb}\\sim 10^5$~cm) was consistent with the ROSAT results. The power-law component changed substantially from the decay phase to quiescence: during the decay the photon index was $\\Gamma \\sim 2$, while in quiescence it hardened to $\\Gamma\\sim 1$.  The optical counterpart of Aql X-1 was identified in 1978 with the variable K1IV star V1333 Aql (Thorstensen, Charles \\& Bowyer 1978; Shahbaz et al. 1996, 1997). Its quiescence magnitude is V=19.2 mag. Brightenings of up to $\\sim 5$ mag have been observed during the X--ray outbursts. Chevalier and Ilovaisky (1991, 1997) monitored different optical outbursts of the source and determined a photometric orbital period of 18.9 hr.  Aql X-1 was observed at radio wavelengths with the VLA at 0.4 mJy (at 8.4 GHz) during an outburst (Hjellming et al. 1990). This source was also searched during the quiescent phase for pulsed radio emission at 400 MHz with the Jodrell Bank telescope. This provided an upper limit of $\\sim 10$ mJy during June 1989 (Biggs, Lyne \\& Johnston 1989) and $\\sim 3$ mJy during July--August 1989 (Biggs \\& Lyne 1996).  \\begin{figure*}[!t] \\psfig{figure=ftot.ps,width=7cm} \\caption{ Light curve of the Feb.-Mar. 1997 outburst of Aql X-1. Data before and after MJD 50514 were collected with the RossiXTE ASM (2--10 keV) and the BeppoSAX MECS (1.5--10 keV), respectively. RossiXTE ASM count rates are converted to (unabsorbed) luminosities using a conversion factor of $4\\times 10^{35}\\ergs$ (before MJD 50512) and $2\\times10^{35}\\ergs$ (after MJD 50512) as derived from RossiXTE spectral fits (Zhang, Yu \\& Zhang 1998). BeppoSAX luminosities are derived directly from the spectral data (Campana et al. 1998). The evolution of the flux from MJD 50480 to MJD 50512 is well fit by a Gaussian centered on MJD 50483.2. This fit however does not provide an acceptable description for later times (see the dot-dashed line), not even if the accretion luminosity is calculated in the propeller regime (dashed line). The straight solid line represents the X--ray luminosity corresponding to the closure of the centrifugal barrier $L_{\\rm min}$ (for a magnetic field of $10^8$ G and a spin period of 1.8 ms) and the straight dashed line the luminosity gap due to the action of the centrifugal barrier, $L_{\\rm cor}$. The dotted line marks the minimum luminosity in the propeller regime ($L_{\\rm lc}$). } \\label{fig_aql} \\end{figure*}   \\subsubsection {Cen X-4}  Only two X--ray outbursts have been detected from Cen X-4. The second outburst was observed in 1979, ten years after the first one (Conner, Evans \\& Belian 1969). Their rise times were $\\sim 5-7$~d and $e-$folding decay times of $\\sim 30$~d. During the 1979 outburst Cen X-4 reached a peak flux of $\\sim 5$~Crab, corresponding to $L_X\\sim 4\\times 10^{37}\\ergs$ for $d \\sim 1.2$~kpc (Kaluzienski, Holt \\& Swank 1980). Type I bursts were observed at intermediate luminosity levels (Matsuoka et al. 1980). A variable, hard spectral component extending up to $\\sim 100$ keV and with an equivalent bremsstrahlung temperature $k\\,T_{\\rm br}$ of 30--70~keV was also revealed.  This component appeared slightly before the outburst maximum, with a flux similar to that in the 3--6 keV range, and during the decay phase, with a factor of $\\sim 5$ lower flux than the 3--6 keV flux (around $L_X\\sim 2\\times 10^{37}\\ergs$; Bouchacourt et al. 1984).  The 1979 outburst was particularly well studied and led to the identification of the optical counterpart which had brightened by $\\gsim 6$ magnitudes (Canizares, McClintock \\& Grindlay 1979).  ASCA detected Cen X-4 during quiescence at a level of $2.4\\times 10^{32}\\ergs$ (0.5--10 keV; Asai et al. 1996a).  The X--ray spectrum was well fit by a black body component ($k\\,T_{\\rm bb}=0.16$ keV) plus an additional power-law component with $\\Gamma \\sim 2-3$. The flux from the two spectral components is comparable. The equivalent radius of the black body emission is $\\sim 1.8\\times10^5$ cm, substantially smaller than the radius of a neutron star. A search for X--ray pulsations gave negative results, providing an upper limit to the pulsed fraction of $\\sim 50\\%$ between 8 ms and 8200 s (Asai et al. 1996a).  During quiescence Cen X-4 was also observed with the Einstein IPC (in 1980, $\\sim 440$ d after the 1979 outburst; Petro et al. 1981) and EXOSAT CMA (in 1986, van Paradijs et al. 1987). Assuming the ASCA spectrum, Campana et al. (1997) found that both measurements are consistent with the same value of the X--ray luminosity derived with ASCA. A Ginga observation in 1991 provided an upper limit of $\\sim 5\\times 10^{32} \\ergs$ in the range 2--7 keV (for a thermal brems\\-strahlung spectrum with $T=5$ keV; Kulkarni et al. 1992). A ROSAT HRI observation in 1995 revealed Cen X-4 at a level comparable to that measured by ASCA, but showed a factor of 3 flux variability in a few days (Campana et al. 1997).  Cen X-4 is one of the best studied SXRTs at optical wavelengths. Extensive spectroscopic and photometric measurements of the optical counterpart in quiescence (V=18.7 mag) led to the determination of the orbital period (15.1 hr; Chevalier et al. 1989) and the mass function ($\\sim 0.2 \\msole$, converting to a neutron star mass between $0.5-2.1 \\msole$;  Shahbaz, Naylor \\& Charles 1993). The optical spectrum shows the characteristics of a late K main sequence star, contaminated by lines and continuum emission probably resulting from an accretion disk (see, e.g., Cowley et al. 1988; Chevalier et al. 1989). The exact nature and evolutionary state of the companion star, together with the mechanism responsible for the mass transfer are the subject of a long debate. The most likely scenarios involve either a peculiar subgiant or a stripped giant, filling its Roche lobe, and viewed at low inclination (McClintock \\& Remillard 1990; Shahbaz, Naylor \\& Charles 1993).  Radio emission had been revealed a few days after the 1979 outburst at a level of $\\sim 10$ mJy at both 1.5 and 4.8 GHz (Hjellming et al. 1988).  The quiescent phase of Cen X-4 was observed in the radio band with the VLA, searching for pulsations and/or continuum emission. Radio emission was not detected with an upper limit of 0.4 mJy at 1.4 GHz (Kulkarni et al. 1992).  \\subsubsection {4U 1730--22}  The only known outburst from this source was observed with Uhuru in 1972 (Forman et al. 1978). The decay of the X--ray flux was characterised by an $e-$folding time of $\\sim 30$ d with evidence of a secondary maximum. The thermal bremsstrahlung-like spectrum, with a temperature of $\\sim 4$ keV, clearly indicated that this source belongs to the SXRT class (Cominsky et al. 1978).  \\subsubsection {A 1742--289}  This SXRT, located close to the Galactic Center direction, underwent a strong outburst in 1975, characterised by a fast rise time to a peak flux of $\\sim 2$~Crab (corresponding to a luminosity of $\\sim 4\\times10^{38}\\ergs$ at 8.5 kpc) followed by an exponential decay with an $e-$folding time of 12~d (Branduardi et al. 1976). Though the source lies in a very crowded region, the concomitant radio outburst yielded an accurate position (Davis et al. 1976). ASCA has recently detected A~1742--289 at a 3--10 keV luminosity ranging between $10^{35}-10^{36}\\ergs$ (Maeda et al. 1996). The source exhibited X--ray bursts and eclipses, yielding an orbital period of 8.4 hr. The X--ray spectrum is highly absorbed ($N_H\\sim 10^{23}$ cm$^{-2}$) and could be well fit by a power-law with index 2.4 or a thermal bremsstrahlung model with $k\\,T_{\\rm br}=7.5$ keV.  However, a reanalysis of the Ariel V data from the 1975 outburst provided no evidence for eclipses suggesting that the source detected by ASCA is perhaps a new one, rather than the quiescent counterpart of A~1742--289 (Kennea \\& Skinner 1996).  \\subsection {Soft X--ray Transients with extended on/off periods}  The sources of this sample are characterised by on and off states lasting several months to years. Three out of four systems show partial X--ray eclipses and/or dips, indicating that they are viewed from a high inclination. In some instances there is strong evidence that the central X--ray source is hidden by the accretion disk and the observed X--rays are scattered along our line of sight by an extended photo-ionised corona above the disk (i.e. they are accretion disk corona sources; White \\& Holt 1982). A different case is presented by 4U 1608--52 which shows intermediate properties between this class of SXRTs and those displaying outbursts with fast rise and nearly exponential decay.  \\subsubsection{EXO 0748--676}  EXO 0748--676 was discovered during a slew with the EXOSAT satellite in 1985. The intensity decayed in the first 2 months after the discovery, as expected for classical transients. However, in June--July 1985 and January 1986 it was again in a bright state with $L_X\\sim 10^{37}\\ergs$ (Parmar et al. 1986). The source was still active when reobserved with Ginga in 1989 (Parmar et al. 1991) and more recently with ASCA, even if at a lower level (Corbet et al. 1994; Thomas et al. 1997). Type I bursts (with photospheric radius expansion), dips and partial eclipses at the 3.8 hr orbital period were observed with EXOSAT. Parmar et al. (1986) found a quiescent X--ray luminosity of $10^{34}\\ergs$ (0.5--10 keV), while Garcia \\& Callanan (1998) infer from the same data a black body temperature of $k\\,T_{\\rm bb}\\sim 0.2$ keV. Based on ASCA data Thomas et al. (1997) find evidence of a previously unreported soft excess. Optical observations, which detected EXO 0748--676 in the range V$\\,\\sim 17.5-16.8$ mag also testify that the source remained active during the period 1985--1993. Optical variability as well as optical bursts resulting from reprocessed X--ray bursts have also been observed.  \\subsubsection{4U 2129+47}  4U 2129+47 shows partial X--ray eclipses at the orbital period of 5.2 hours, as well as type I bursts. Before 1984 it was considered a persistent LMXRB, since it had been detected and studied with all the major X--ray satellites. Its relatively low luminosity and smooth orbital modulation at 5.2 hr in the active state ($\\sim 10^{35}-10^{36}\\ergs$ for $d\\sim 6$ kpc) suggests that 4U 2129+47 is an accretion disk corona source (White \\& Holt 1982). During this active state the optical counterpart was identified (Thorstensen et al. 1979). EXOSAT CMA observations in September 1983 provided only upper limits corresponding to a luminosity of $\\lsim 10^{34}\\ergs$ (Pietsch et al.  1986). Subsequent X--ray and optical observations showed that the source entered a long period of quiescence (Garcia et al. 1989; Molnar \\& Neely 1992) and led to its inclusion in the SXRT group. A ROSAT HRI observation detected this source at a level of $\\sim 3\\times 10^{33}\\ergs$ (0.3--2.4 keV; Garcia 1994). Garcia \\& Callanan (1998) derived a quiescent luminosity of $6\\times 10^{32}\\ergs$ (0.5--10 keV) for a black body spectrum with a temperature $k\\,T_{\\rm bb}=0.2$ keV. The quiescent X--ray light curve, obtained with the ROSAT HRI, does not show strong evidence for the partial eclipses characteristic of the active state, indicating that either the vertical extent of the disk is drastically reduced or that the disk is not present during quiescence (Garcia 1994). Note that the optical spectrum of 4U 2129+47 during quiescence does not display any characteristic feature of an accretion disk (Garcia 1994).  The companion was classified as a F9 subgiant, but no evidence was found of the expected ellipsoidal and radial velocity variations at the orbital period (Garcia et al. 1989).  An upper limit of $\\sim 8$ mJy at 610 MHz for pulsed radio emission has been set for 4U 2129+47 during quiescence (Biggs, Lyne \\& Johnston 1989) and of 2 mJy at 400 MHz (Biggs \\& Lyne 1996); at 1.4 GHz the $4\\,\\sigma$ upper limit is of 0.25 mJy (Kulkarni et al. 1992).  \\subsubsection{4U 1658--298}  4U 1658--298, was discovered in 1976 when an isolated Type I burst was detected (Lewin, Hoffman \\& Doty 1976a). The corresponding SXRT was discovered only two years later during a strong outburst reaching a flux of $\\sim 10^{-9}\\ergs$ cm$^{-2}$ (Lewin et al. 1978; Share et al. 1978). During this outburst, X--ray eclipses and dips were also detected (Cominsky, Ossmann \\& Lewin 1983). Hard X--rays up to $\\sim 80$ keV from this source were revealed with the A4 experiment on board HEAO-1 (Levine et al. 1984).  \\subsubsection{4U 1608--52}  The persistent source 4U 1608--52 detected by Uhuru and OSO-7 in 1971--1973, and the transient observed with Ariel V in November 1975 and with Ariel V, SAS-3 and HEAO 1 in July--September 1977 were recognised to be the same object (Fabbiano et al. 1978 and references therein). The type I bursts (with photospheric radius expansion indicating a 3.6 kpc distance) observed from this region in the Norma constellation were also attributed to this source, thus providing the first case of transient-burster association. The bursts from 4U 1608--52 were later confirmed with HAKUCHO (Murakami et al. 1980) and EXOSAT (Penninx et al.  1989).  The Vela 5B data detected 8 outbursts during 1969--1979 as well as a persistent emission at a level of $\\sim 10^{36}\\ergs$ (Lochner \\& Roussel-Dupr\\`e 1994). The outbursts are characterised by either a sharp rise and an exponential decay or a much more symmetric evolution. Like other SXRTs, 4U 1608--52 exhibited a spectral transition from a thermal bremsstrahlung to a power-law like spectrum when the luminosity decreased below $\\sim 10^{37}\\ergs$ (Mitsuda et al. 1989). A hard power-law like spectrum has also been detected with BATSE during the 1991 outburst with a power-law slope of $\\Gamma\\sim 1.8$ and a steepening above $\\sim 65$ keV (Zhang et al. 1996b). The high energy spectrum could be equally well fit by either a Sunyaev-Titarchuk Comptonisation model or a broken power-law.  In 1993 ASCA revealed 4U 1608--52 at a much lower X--ray luminosity of $\\sim 2\\times10^{33}\\ergs$ in the 0.5--10 keV energy range (Asai et al. 1996a).  The spectrum could be well fit by a black body with $k\\,T_{\\rm bb}=0.30$ keV or a bremsstrahlung with $k\\,T_{\\rm br}=0.32$ keV. The black body emission radius is $\\sim 1.5\\times10^5$ cm, substantially smaller than the radius of a neutron star.  A periodicity search based on the ASCA light curves provided an upper limit of 50\\% rms in the range 8 ms--8200 s, which is valid only if the orbital period is longer than 2 d (Asai et al. 1996a).  This source was active again in 1996 and was observed by RossiXTE at a level of $\\sim 2\\times10^{37}\\ergs$ with a $k\\,T_{\\rm br}\\sim 5$ keV bremsstrahlung spectrum (Marshall \\& Angelini 1996). RossiXTE observations revealed also a variable QPO feature at 850--890 Hz, the frequency variations of which did not correlate with intensity changes (Berger et al. 1996). A second kHz QPO has been recently revealed at about 1100 Hz simultaneous with the 600--900 Hz peak (Yu et al. 1997) previously known (Mendez et al. 1998). There is evidence that the frequency separation varied between 230--290 Hz, perhaps providing the first example of a variable kHz peak separation in an atoll source.  The optical counterpart, identified during the 1977 outburst, is a reddened faint star ($I\\sim 18.2$ mag) which becomes fainter in quiescence ($I>20$ mag, $B>22$ mag; Grindlay \\& Liller 1978). This was re-discovered during the 1996 outburst about 150~d after the peak at a level of $R=20.2$ mag and $J=17.2$ mag, when the source was still active in the X--rays (Wachter 1997). One year before the outburst the source was detected at $J=18.0$ mag and $R>22$ mag (Wachter 1997).   \\subsubsection{4U 1730--335: the Rapid Burster}  The Rapid Burster alternates periods of activity lasting several weeks, to period of quiescence, during which the X--ray luminosity decreases by more than three orders of magnitude. During the active periods the Rapid Burster emits a variety of combinations of type I and II bursts, making it unique among LMXRBs (Lewin et al. 1976b). Here we do not describe in detail the very complex phenomenology of this source and we refer to the review by Lewin, van Paradijs \\& Taam (1995) and references therein.  In the state in which the Rapid Burster shows the closest resemblance to other SXRTs, namely persistent X--ray emission and type I X--ray bursts, the average luminosity is about $10^{37}\\ergs$ (for the $\\sim 10$ kpc distance of the globular cluster Liller 1) and the spectrum can be described by a thermal bremsstrahlung with $k\\,T_{\\rm br}\\sim 10$ keV (Barr et al. 1987).  During quiescence an upper limit of $\\sim 10^{34}\\ergs$ was obtained with Einstein (Grindlay 1981). Recently, the Rapid Burster has been detected in quiescence by ASCA at a level of $3\\times 10^{33}\\ergs$ (Asai et al. 1996b). However, the ASCA point spread function ($\\sim 3'$) is comparable to angular radius of the Liller 1 cluster ($3.3'$) and it cannot be ruled out that the measured X--ray flux is due to other sources within the cluster.  RossiXTE observations of X--ray bursts from the Rapid Burster did not lead to the discovery of kHz QPOs (Guerriero, Lewin \\& Kommers 1997). A radio transient with flux density correlated with the RossiXTE ASM X--ray flux of the Rapid Burster has been recently observed (Rutledge et al. 1998).  \\subsection{Soft X--ray Transients with poorly sampled outburst light curves}  \\subsubsection{MX 0836--42}  This source was discovered at the end of 1971 with OSO-7 and Uhuru (Markert et al. 1977; Cominsky et al. 1978) at a level of $8\\times 10^{-9}\\ergs$ cm$^{-2}$. The spectrum was soft, but owing to poor coverage the light curve of the outburst could not be determined accurately. MX 0836--42 remained undetected until a new period of activity occurred in 1990--1991. Ginga detected a variable X--ray flux from this source from the end of November 1990 until February 1991.  ROSAT observed an active state at a level of $\\sim 10^{-10}\\ergs$ cm$^{-2}$ (1.0--2.4 keV) around the middle of November 1990 during the all sky survey and re-observed it at a level $\\sim 15$ times lower in May 1991. These observations led to the discovery of Type I bursts (Aoki et al. 1992) and a substantial reduction of its error box (Belloni et al. 1993), confirming the LMXRB nature of this source.  \\subsubsection{MX 1746--20}  Little is known about this transient which has been unambiguously observed only in January 1972. It was discovered with the OSO-7 satellite (Markert et al. 1975) and tentatively identified with the globular cluster NGC 6440. This association was also supported by the more precise source location obtained with the Uhuru data (Forman, Jones \\& Tananbaum 1976). The peak X--ray luminosity was about $10^{37}\\ergs$ for the distance of this cluster ($d=7$ kpc). Cominsky et al. (1978) classified MX 1746--20 as a soft transient, on the basis of a spectral fit yielding a thermal bremsstrahlung temperature of $\\sim 4$ keV.  It is generally assumed that the dim source detected with the Einstein HRI at a level of $\\sim10^{33}\\ergs$ in the globular cluster NGC~6440 (Hertz \\& Grindlay 1983) is the quiescent counterpart of MX 1746--20. ROSAT HRI observations confirmed this detection at a similar level (Johnston, Verbunt \\& Hasinger 1995).  \\subsection{Galactic Center transients}  Over the last few years several X--ray sources in the Galactic Center region have been discovered with Ginga, TTM, ART-P and SIGMA (e.g. in't Zand et al. 1989; Sunyaev et al. 1991). Little is known about these sources, owing the lack of optical identifications, poor temporal coverage and spectral information.  The maximum X--ray luminosity observed from KS 1730--312, 4U 1735--28, EU 1737--132, GS 1741.1--2859 are of the order of $L_X\\sim 10^{38} \\ergs$. On the other hand the highest detected luminosities of KS 1732--273, EXS 1737.9--2952, GRS 1741.9--2853 are considerably lower ($L_X\\sim 10^{37} \\ergs$). Some other sources were only a factor of a few above the instruments' detection thresholds: among these are KS 1632--477, KS 1724--356, KS 1731--260, KS 1739--304, KS 1741--293. It is therefore unclear whether these sources are to be considered transient sources and, in particular, SXRTs.  \\subsubsection{KS 1731--260}  The best studied of the latter sources, KS 1731--260, shows only modest X--ray luminosity variations, casting doubts on its inclusion in the SXRT class.  The source was discovered by TTM on board MIR-KVANT (Sunyaev 1989) at a level of $\\sim 10^{37}\\ergs$ (2--27 keV) for a distance of 8.5 kpc.  The spectrum was well fit by a bremsstrahlung spectrum with $k\\,T_{\\rm br} \\sim 6$ keV; X--ray bursts were also observed (Sunyaev et al. 1990).  KS 1731--260 was also detected by ART-P, SIGMA (revealing a hard energy tail extending up to energies of $\\sim 40$ keV; Barret et al. 1992) and during the ROSAT all-sky survey (see e.g. Smith, Morgan \\& Bradt 1997). RossiXTE revealed a fairly coherent signal during a $\\sim 2$ s interval starting at the end of the contraction phase of an X--ray burst. These oscillations likely indicate the presence of neutron star spinning at 1.9 ms (Smith, Morgan \\& Bradt 1997). Moreover, two simultaneous QPO peaks around 900 and 1160 Hz have also been discovered with RossiXTE (Wijnands \\& van der Klis 1997), the difference frequency of which is consistent with half the frequency observed by Smith, Morgan \\& Bradt  (1997), suggesting that the neutron star spin period is instead $\\sim 3.8$ ms. ",
        "conclusions": "The properties of SXRTs in outburst are clearly linked to those of persistent LMXRBs, as testified by e.g. their X--ray spectra, the occurrence of type I bursts, optical brightenings and kHz QPOs, indicating that SXRTs host weakly magnetic neutron stars.  For X--ray luminosities much below the outburst maxima, SXRTs show similarities with transient BHCs. In particular, as the luminosity decreases at a level of $\\sim 10^{36}-10^{37}\\ergs$ in a few SXRTs (as well as in persistent LMXRBs) the X--ray spectrum shows a transition from a relatively soft thermal spectrum to a power-law like spectrum extending up to 100 keV (Barret \\& Vedrenne 1994; Mitsuda et al. 1989; Harmon et al. 1996). A similar spectral transition occurs also across the ``high'' and ``low states'' of several BHCs (e.g. Tanaka \\& Shibazaki 1996). The spectral hardening in BHCs has been interpreted in terms of a change in the disk to an advection-dominated regime (Narayan \\& Yi 1995). This regime has also been invoked to explain the low luminosity emission of the BHCs A~0620--00 (Narayan, McClintock \\& Yi 1997) and V404 Cyg (Narayan, Barret \\& McClintock 1997). We note however that in the case of SXRTs the presence of a ``hard surface'' (either the star surface or the magnetosphere) makes the application of simple advection-dominated models questionable.  Also in quiescence, SXRTs and transient BHCs show similar properties, displaying a soft spectral component (with equivalent black body temperatures of $\\sim 0.2-0.3$ keV; e.g. Tanaka \\& Shibazaki 1996). However ASCA and BeppoSAX observations of Cen X-4 (Asai et al. 1996a) and Aql X-1 (Campana et al. 1998) during quiescence demonstrated the presence of hard tails which have not been detected in BHCs.  We note also that the Aql X-1 outburst decay resembles closely the evolution of dwarf novae outbursts, with a drastic turn off of the luminosity (e.g. Osaki 1996). Models of low mass X--ray transient outbursts hosting an old neutron star or a black hole are largely built in analogy with dwarf novae outbursts. In particular, van Paradijs (1996) showed that the different range of time-averaged mass accretion rates over which the dwarf nova and low mass X--ray transient outbursts were observed to take place is well explained by the higher level of disk irradiation caused by the higher accretion efficiency of neutron stars and black holes. However, the outburst evolution of low mass X--ray transients presents important differences. In particular, the steepening in the X--ray flux decrease of Aql X-1 has no clear parallel in low mass X--ray transients containing BHCs. The best sampled light curves of these sources show an exponential-like decay (sometimes with a superposed secondary outburst) with an $e-$folding time of $\\sim 30$ d and extending up to four decades in flux, with no indication of a sudden steepening (Chen, Shrader \\& Livio 1997). In addition, BHC transients display a larger luminosity range between outburst peak and quiescence than neutron star SXRTs (Garcia et al. 1998 and references therein). Being the mass donor stars and the binary parameters quite similar in the two cases, it appears natural to attribute these differences to the different nature of the underlying object. In principle after the decay of a neutron star SXRT outburst the X--ray emission might be dominated by different mechanisms, notably accretion onto the magnetosphere, emission from an enshrouded MSP or cooling from the neutron star surface. None of these has an equivalent in the case of BHCs.  In the case of Aql X-1 for which a spin period frequency of 2--4 ms has been inferred, accretion onto the neutron star surface during the quiescent state of SXRTs would imply a neutron star magnetic field $\\lsim 5\\times 10^6$ G.  Extrapolating this result, one would conclude that SXRTs could not be among the progenitors of MSPs. Cooling of the neutron star surface can easily account for the softness of X--ray spectra. However, the observation of a flux variation on a timescale of a few days in the quiescent flux of Cen X-4 (Campana et al. 1997) can hardly be reconciled with this mechanism alone. Moreover, the observation of a harded spectral component in Aql X-1 (Campana et al. 1998) and Cen X-4 (Asai et al. 1996a) poses problems to accretion onto the surface and cooling mechanisms.  Accretion onto the neutron star magnetosphere is one interesting possibility to explain the quiescent emission of SXRTs. If this regime applies, the allowed region in the $B-P$ diagram, even if far from MSPs near the Eddington spin-up line, contains a relevant number of MSPs and straddles the region where neutron stars of LMXRBs are expected to lie. The spin period evolution in this case is mainly dictated by the ratio of the mean luminosity in outburst and during quiescence, so that either spin-up and spin-down are possible, even if a secular spin-down is more likely (cf. Eq. \\ref{pp}).  Shock emission powered by an underlying MSP is the other plausible mechanism for the quiescent emission of SXRTs. In this regime, SXRTs would occupy a region in the $B-P$ diagram containing a large number of MSPs and likely including kHz QPO sources. The equilibrium line between spin-up during outburst and spin-down due to radiative losses in quiescence, lies just in the middle of the expected neutron star spin-down luminosities. BeppoSAX observations of Aql X-1 provide the best evidence so far for this mechanism (Campana et al. 1998). The predicted spectrum show a substantial flux at energies beyond the soft (e.g. ROSAT) energy range. However, a soft component is present in all SXRTs detected in quiescence, so that other components have to be invoked, like a more complex shock emission mechanism or the contribution from the cooling neutron star.  For the pulsar shock emission regime to apply, the closure of the centrifugal barrier must take place at a level of $\\sim 10^{36}\\ergs$ (cf. Eq. \\ref{lminsp}), with a small jump in luminosity (cf. Eq. \\ref{corbet}), as observed in Aql X-1. This is also the range of luminosities characterising the spectral hardening observed in persistent LMXRBs and SXRTs. This spectral change has been indeed related to the transition to the propeller regime in the case of Aql X-1 (Zhang, Yu \\& Zhang 1998; Campana et al. 1998). In this interpretation SXRTs represent the immediate progenitors of MSPs.  Estimating the current number of SXRTs in our Galaxy is not an easy task, due to the presence of strong selection effects (e.g. large column densities along the galactic plane; lower peak luminosities with respect to BHCs; limited temporal coverage, etc.). A likely number is a few hundreds for mean recurrence time of $\\sim 10$ yr and about a thousand for a recurrence time of $\\sim 50$ yr (Tanaka \\& Shibazaki 1996).  Despite the recent discoveries on Aql X-1 by Ros\\-siXTE and BeppoSAX, deeper studies are necessary to better assess the nature of the neutron stars in SXRTs. The case of Aql X-1 emphasises the importance of combining the nearly continuous monitoring of the outburst evolution that can be obtained by large field X--ray instruments, with deeper and more detailed pointed observations by narrow field X--ray telescopes during the crucial phases of the outbursts.  \\medskip {\\bf Note added in proof.}  After this paper was accepted for publication, we became aware that in April 1998 RXTE revealed a transient X--ray source at a position consistent with SAX J1808.4--3658 (in't Zand et al. 1998), a variable X--ray burster in the direction of the galactic bulge. During pointed RXTE observations highly significant coherent pulsations at $\\sim 401$ Hz were detected. These allowed also to measure an orbital period of $\\sim 2$~hr and a mass function of $\\sim 4\\times 10^{-5}~\\msole$ (Wijnands \\& van der Klis 1998; Chakrabarty \\& Morgan 1998).  These results show that SAX J1808.4--3658 is a LMXRB, the first to show coherent millisecond pulsations in its persistent emission. A radio pulsar (perhaps a partially eclipsing one) might turn on after the X--ray outburst ends. In any case SAX J1808.4--3658 further strengthens the link between SXRTs and MSPs.   \\medskip {\\small {\\it Acknowledgement.} We thank an anonymous referee for providing useful comments. This work was partially supported through ASI grants.}"
    },
    "9805/astro-ph9805235_arXiv.txt": {
        "abstract": "{We recall the current status of the long-standing \\3he problem, and its possible connection with chemical anomalies on the red giant branch. In this context, we collect in the literature all the available observations of the carbon isotopic ratio in field and cluster giant stars. Using the HIPPARCOS parallaxes, we get constraints on the evolutionary status of the field stars of the sample. This allows us to identify the stars that have passed the luminosity function bump and present \\12sur13 ratios in disagreement with the standard predictions of stellar evolutionary models. We determine statistically what fraction of low mass stars experience an extra-mixing process on the red giant branch, and are then expected to destroy their \\3he at this evolutionary phase. The high number we get satisfies the galactic requirements for the evolution of the \\3he abundance. } ",
        "introduction": "The evolution of \\3he in the Galaxy has first been considered to be straightforward, dominated by the net production of this light element by low mass stars (i.e., with masses lower than 2 \\Msun). In these objects, initial D is processed to \\3he during the pre-main sequence phase. Then, as described by Iben (1967), an \\3he peak builds up due to pp-reactions on the main sequence, and is engulfed in the stellar convective envelope during the first dredge-up on the lower red giant branch (RGB). Standard theory predicts that, once in the convective layers of the evolved star, \\3he can not be destroyed because of the too cool temperature in these regions. It is finally ejected in the interstellar medium (ISM) in the latest stages of stellar evolution. In this standard view, the abundance of \\3he must increase in the Galaxy as soon as low mass stars begin to polute the ISM  (Rood et al. 1976). One expects then to have constraints on the cosmological abundance of \\3he (Yang et al. 1984).  Recent observations of a few planetary nebulae (PN; Rood et al. 1992, Balser et al. 1997) led to the measurement of \\3he in one object, NGC 3242 {\\footnote {Balser et al. (1997) got no definitive detection of \\3he in any of the other five PN they observed; however detection of \\3he is probable in two of their objects}}. This PN, which estimated initial mass is 1.2$\\pm$0.2 \\Msun (Galli et al. 1997, hereafter GSTP97), presents a value of \\3he/H$=(7.3 \\pm 1.4) \\times 10^{-4}$, in very good agreement with standard predictions (Vassiliadis \\& Wood 1993, Charbonnel 1995, Dearborn et al. 1996, Weiss et al. 1996 for the most recent computations). This value however differs by almost two orders of magnitude with the \\3he abundance in the proto-solar nebula, \\3he/H=$(1.5 \\pm 0.3) \\times 10^{-5}$ (Geiss 1993), in the local interstellar cloud, \\3he/H=$(2.2 \\pm 0.2) \\times 10^{-5}$ (Gloeckler \\& Geiss 1996), and in galactic HII regions, \\3he/H=$(1~ {\\rm to} ~5) \\times 10^{-5}$ (Balser et al. 1994).  These low values are in clear contradiction with the conventional scenario for galactic evolution of the \\3he abundance, and can not be explained if all low mass stars, such as NGC 3242, happen to return all their \\3he to the ISM. GSTP97 showed that, in order to fit the galactic constraints, \\3he should be destroyed in at least 70$\\%$ of low-mass stars before they become PN. ",
        "conclusions": "We have assembled all the observations of the \\12sur13 ratio in field and cluster giants available in the literature. Using the HIPPARCOS parallaxes, we get constraints on the evolutionary status of our sample stars. We determine that 96$\\%$ of low-mass stars do experience an extra-mixing process on the RGB and are then expected to destroy their \\3he. While consistent ``non-standard\" stellar models are needed to explain the various chemical anomalies in low-mass RGB stars in order to obtain reliable \\3he yields, we can already conclude that the very high percentage we get satisfies the galactic requirements for the evolution of the \\3he abundance."
    },
    "9805/astro-ph9805145_arXiv.txt": {
        "abstract": "We study the vertical heating and thickening of galaxy disks due to accretion of small satellites.  Our simulations are restricted to axial symmetry, which largely eliminates numerical evolution of the target galaxy but requires the trajectory of the satellite to be along the symmetry axis of the target.  We find that direct heating of disk stars by the satellite is not important because the satellite's gravitational perturbation has little power at frequencies resonant with the vertical stellar orbits. The satellite does little damage to the disk until its decaying orbit resonantly excites large-scale disk bending waves.  Bending waves can damp through dynamical friction from the halo or internal wave-particle resonances; we find that wave-particle resonances dominate the damping. The principal vertical heating mechanism is therefore dissipation of bending waves at resonances with stellar orbits in the disk.  Energy can thus be deposited some distance from the point of impact of the satellite. The net heating from a tightly bound satellite can be substantial, but satellites that are tidally disrupted before they are able to excite bending waves do not thicken the disk. ",
        "introduction": "Disk galaxies are observed to be cold and thin, with typical scale heights only 10\\% of their radial scale lengths.  The accretion of satellite galaxies should strongly heat and thicken disks, so this observation limits the satellite infall rate.  T\\'oth \\& Ostriker (1992; hereafter TO) pointed out that thin galactic disks may therefore set important cosmological constraints.  TO estimate the energy deposited in the disk during an accretion event in a simplified manner.  They assume that the satellite galaxy spirals into the parent galaxy on a near-circular orbit as it loses energy through dynamical friction to both the dark matter halo and the disk.  They determine the rate of energy loss to both components using Chandrasekhar's dynamical friction formula (Binney \\& Tremaine 1987, \\S7.1), and deposit the energy locally in the halo and disk, sharing the disk energy between vertical and horizontal motions in a fixed ratio.  TO recognize that their treatment omits all collective effects in the response, but argue that had they ``treated the problem as one of exciting modes in the disk$\\ldots$[they] would have found the same overall energy change in the disk.''  It is not obvious that this claim is correct; there are good reasons to believe that in-plane heating will be increased while vertical heating could be {\\it reduced\\/} by collective effects.  A swing-amplified spiral response (see Binney \\& Tremaine 1987 for a review) will always extract energy from the potential well of the target galaxy, thereby adding to the energy deposited into in-plane random motion.  On the other hand, the following simple thought experiment suggests that vertical heating could be small.  Let us imagine that the disk is very stiff so that the eigenfrequencies of its collective bending modes are very high.  In this case, the time-dependent perturbation of a passing satellite will not contain power at frequencies which are resonant with any collective modes, and there is little internal heating; the disk acts like a rigid plate and the only effect of interaction with the satellite is to tilt or translate the disk. We will develop these ideas further for more realistic disks in \\S 2.  Many of TO's simplifying approximations are avoided in fully self-consistent $N$-body simulations, which have been widely used to study the merger of small satellites with a large disk galaxy (Quinn \\& Goodman 1986; Pfenniger 1991; Quinn, Hernquist \\& Fullagar 1993; Walker, Mihos \\& Hernquist 1996; Athanassoula 1996; Huang \\& Carlberg 1997).  These simulations have mostly confirmed that disks are strongly heated by the accretion of a $\\sim10\\%$ mass satellite and have elucidated other effects, such as the stripping and tidal disruption of the satellite as it approaches the target galaxy, and the tilting of the disk plane due to the gravitational torque from the satellite.  Many of these calculations have begun with models in only approximate equilibrium, and suffered from relaxation and other numerical noise.  Numerical relaxation is reduced by particle softening, but excessive softening impairs the ability of the disk to support collective modes.  Of these simulations, that of Walker et al.\\ was most successful at suppressing relaxation while maintaining a small softening parameter, but their single simulation did not enable them to reach a firm conclusion on many of the issues raised by TO.  Huang \\& Carlberg argue that TO overstate the disk heating associated with mergers; they find that the disk absorbs some of the orbital angular momentum of the satellite simply by tilting, which reduces the energy of vertical oscillation available to thicken the disk.  These simulations have been valuable, but a better appreciation of the physics of the heating process is crucial if we are to understand how to generalize the simulation results.  Since it is the thinness of disks that is hardest to preserve, we focus here on how vertical heating is affected by collective effects, and say little about in-plane heating.  Moreover, vertical heating is a cleaner problem than radial heating, since there is no radial redistribution of the disk matter.  Vertical heating is largely unaffected by the energy deposited into in-plane motion in the short run; it should be noted however, that molecular clouds can gradually scatter horizontal motion into vertical motion, thereby thickening the disk over timescales comparable to a Hubble time (Carlberg 1987).  We first discuss heuristically, in \\S 2, how energy might be deposited by a satellite into vertical heat in the disk.  We test these ideas using numerical simulations which are designed to avoid the complications caused by numerical relaxation and rearrangement of angular momentum. Our simulations utilize an axisymmetric grid and scarcely evolve when unperturbed.  They avoid both internal relaxation and softening while also being much faster than the direct $N$-body methods adopted in previous studies.  We are therefore able to quantify the heating and to explore more parameter space.  The assumption of axisymmetry restricts us to mergers that occur along the symmetry axis, however.  There are several mitigating effects (tidal disruption of the satellite, late star formation, infalling cold gas, etc.)\\ that may reduce the thickening caused by satellite mergers.  We do not address these issues here, but focus on the detailed physical process of vertical heating by a rigid massive satellite. ",
        "conclusions": "The goal of this paper has been to gain physical insight into disk heating and thickening caused by the accretion of small satellite galaxies. We have chosen to explore an axisymmetric system in order to isolate the vertical heating phenomenon.  This strategy has the advantage that our simulations are inexpensive, allowing us to sample a broader region of parameter space while avoiding internal evolution and artificial heating from numerical relaxation.  We believe that most of the conclusions below also apply to the general case of off-axis satellite accretion.  We have found that the satellite loses little energy to direct heating of disk stars, because most of the power in the satellite force is at frequencies lower than the natural oscillation frequencies of the disk stars. There is substantial heating at late stages, however, through the intermediate process of exciting disk bending waves.  Once excited, these waves eventually damp efficiently at wave-particle resonances, thereby heating the disk non-locally.  The only significant radial heating also occurs at the time the disk is thickened by the bending waves.  Thus, satellites which are tidally disrupted before they are able to excite bending waves do not thicken the disk.  Bending waves can damp by several mechanisms, including dynamical friction from the halo, nonlinear damping at the disk edge, and internal wave-particle resonances.  In most cases wave-particle resonance (Landau damping) is by far the strongest damping mechanism; since this mechanism depends strongly on the disk thickness and vertical frequencies, numerical simulations must accurately reproduce these quantities in order to represent the behavior of real galaxy disks.  Thus collective effects can significantly reduce the vertical energy deposited by the satellite only through tilting the disk in an off-axis encounter, as shown by Huang \\& Carlberg (1997; see also Athanassoula 1996). Tilting the target galaxy, reduces the vertical energy of the satellite about the new disk plane.  We have not explored many important issues discussed by TO and others; for example, we have not simulated the full range of satellite impacts with arbitrary orbital orientations and satellite masses, and we have not attempted to estimate the average heating rate.  Since off-center encounters can excite disturbances with $m\\ne0$, which also contribute to heating the disk, we cannot address the actual extent to which disk heating is important in real galaxies.  Many effects that we have ignored, however, such as tidal stripping of the satellite, late star formation, and tilting the disk, tend to make the satellite less destructive.  Thus, the heating we obtain is an upper limit that should result from an axial encounter with a low-mass satellite.  This work was supported by NSF grants AST 93/18617 and AST 96/17088 and NASA Theory grant NAG 5-2803 to JAS and was begun during a long visit by JAS to CITA, whose hospitality is gratefully acknowledged.  We thank Alar Toomre for many discussions and much advice over an extended period and Jeremy Goodman for a thoughtful referee report."
    },
    "9805/astro-ph9805091_arXiv.txt": {
        "abstract": "Some of the  rapidly oscillating (CP2) stars, have \\fqs\\ which are larger than the theoretical \\acfq. As the cut-off frequency depends on the $T(\\tau )$ relation in the atmosphere, we have computed \\mds\\ and adiabatic \\fqs\\ for pulsating Ap stars with $T(\\tau)$ laws based on Kurucz model atmospheres and on Hopf's purely radiative relation.  The fre\\-quen\\-cy-de\\-pen\\-dent treat\\-ment of ra\\-dia\\-ti\\-ve trans\\-fer as well as an improved calculation of the radiative pressure in Kurucz model atmospheres increase the theoretical \\acfq\\  by about 200\\,$\\mu$Hz, which is closer to the observations.  For $\\alpha$\\,Cir we find models with Kurucz atmospheres which have indeed a cut-off frequency beyond the largest observed frequency  and which are well within the \\teff\\ -- $L$ error box. For HD\\,24712 only models which are hotter by about 100\\,K and less luminous by nearly 10\\% than what is actually the most probable value would have an \\acfq\\ large enough.  One may thus speculate that the old controversy about a mismatch between observed largest frequencies and theoretical cut-off frequencies of roAp star models is resolved. However, the observational errors for the astrophysical fundamental parameters have to be reduced further and the model atmospheres refined.  Further details can be found in Audard et al. (1997)  \\thesaurus{06(08.01.3, 08.03.2, 08.09.2, 08.15.1, 08.22.3)} ",
        "introduction": "It has been argued by Shibahashi and Saio (1985) that the cut-off \\fq\\ is largely influenced by the $T(\\tau)$ relation which requires a careful modelling of these layers. Frequently, atmospheres in stellar \\mds\\ are based on an Eddington or Hopf law (e.g. Mihalas 1978), where the  radiative transfer is considered to be \\fq\\ independent (grey case), and convection is not included.  We used the LTE Kurucz {\\sc atlas}9 code (Kurucz, 1993) without the ``overshooting option'' (Castelli 1996) to calculate an interpolation table for $T(\\tau, T_{{\\rm eff}})$. Model \\atms\\ with solar composition were computed for $\\log g = 4.2$ and for \\teff\\ ranging from 7400 to 10000\\,K, and no additional contribution to line opacity by microturbulence has been assumed.  The internal structure models  of 1.8\\,$M_{\\odot}$ representative for CP2 stars were computed with the CESAM code  (Morel 1993 and 1997). We do not include effects from a magnetic field.  We shall call ``Hopf'' and  ``Kurucz'' models full stellar models whose atmospheres are derived from  Hopf's law and Kurucz model \\atm s respectively. ",
        "conclusions": "\\vspace{-1mm} We  have shown that along the main sequence, Kurucz model atmospheres increase the cut-off \\fq\\ by about 8.5\\,\\% relative to the value derived from the Hopf $T(\\tau)$ relation.  For  HD\\,24712 and $\\alpha$\\,Cir, we find models with Kurucz atmospheres and with parameters in agreement with the observational error box which have a theoretical cut-off frequency larger than the largest observed \\fq\\ and hence are in agreement with observations. One may thus speculate that the old controversy about a mismatch between observed largest frequencies and theoretical cut-off frequencies of roAp star models is resolved.  However, effects from e.g. an abundance different from solar might affect the cut-off frequency. Abundant rare-earth elements, through blanketing effects, could decrease the surface temperature and thus increase the cut-off \\fq. This \\fq\\ might also be affected by a chemical composition gradient (Vauclair \\& Dolez 1990). Magnetic field (Dziembowski \\& Goode 1996) and NLTE effects should also be taken into account. \\\\"
    },
    "9805/astro-ph9805058_arXiv.txt": {
        "abstract": "We present a discussion on the effects of convection on the $uvby$ colours of A and F stars. The mixing-length theory used in {\\sc ATLAS9} is compared to the turbulent convection theory of Canuto \\& Mazzitelli (1991, 1992). Comparison with fundamental stars reveals that colours calculated using the Canuto \\& Mazzitelli convection theory are generally in better agreement than those obtained using mixing-length theory. ",
        "introduction": "The colours calculated for stars later than mid A-type are affected by treatment of convection. Small systematic errors were found in the colours calculated using ATLAS6 (Kurucz, 1979), which could be due to convection or missing opacity in the models (Relyea \\& Kurucz, 1978). Recent improvements in opacity in ATLAS9 models (Kurucz, 1991) ought to ensure that opacity is now a less dominant source of the discrepancies. This leaves convection as a possible source of the discrepancies.  We have compared mixing length theory (see Castelli et al., 1997) with the turbulent convection model of Canuto \\& Mazzitelli (1991, 1992). The ATLAS9 code was used to calculate $uvby$ colours. The computations were identical, except for treatment of convection. We considered three cases: \\begin{enumerate} \\item mixing-length theory with approximate overshooting (MLT\\_OV), \\item mixing-length theory without approximate overshooting (MLT\\_noOV), \\item the Kupka (1996) implementation of the Canuto \\& Mazzitelli theory (CM). \\end{enumerate}  Here we present a summary of the main findings from the comparisons between the three treatments of convection outlined above. Full details can be found in Smalley \\& Kupka (1997). ",
        "conclusions": "From a comparison with the observed $uvby$ colours, we have found that the CM grid gives results that are generally superior to those with MLT theory without overshooting (MLT\\_noOV). Models with overshooting (MLT\\_OV) are found to be clearly discrepant.  The metallicity index $m_0$ is not in agreement. The reason for this is unclear, but could be linked to microturbulence."
    },
    "9805/astro-ph9805214_arXiv.txt": {
        "abstract": "We report here the results of high-resolution and spectral imaging X-ray observations, with both the \\ros\\ HRI and PSPC, of the field surrounding the nearby ($D=64$\\,Mpc) type~1 Seyfert galaxy IC~4329A and its giant lenticular companion IC~4329. Many point sources are detected, the brightest being associated with IC~4329A itself, having an extremely bright X-ray luminosity of $6\\times 10^{43}$\\,erg s$^{-1}$, and spectral properties compatible with a single power-law model ($\\Gamma=1.73$), with a spectral break at 0.7\\,keV. Two other bright sources are detected associated with the companion galaxy IC~4329, and a likely quasar 14\\arcm\\ to the south-west. We have also established, through optical observations taken at the European Southern Observatory, that three further X-ray point sources, intriguingly positioned with respect to IC~4329A, are in fact nothing to do with the system, and are merely foreground and background objects.  In addition to point source emission, residual, unresolved emission is detected surrounding the IC~4329A/ IC~4329 pair, extending for some 200\\,kpc. This emission appears markedly two-component, comprising of a spectrally hard and smooth component, circularly-distributed about the central galaxy pair, and a spectrally soft, more clumpy component, positioned almost entirely to the south-east of IC~4329A. The hard component of the residual emission itself appears two-component, one component being due to the `wings' of the intensely bright IC~4329A source, the other, apparently due to hot ($\\sim1.5$\\,keV) gas, likely associated with the galaxy group of which IC~4329A and IC~4329 are members. The soft component of the residual emission may be a larger version of the superwinds seen around some ultraluminous far-infrared galaxies, or may even represent a `stripped wake' of intragroup gas. Evidence for shocked gas due to the central IC~4329A/ IC4329 interaction is also found between the two central galaxies. ",
        "introduction": "\\label{sec_intro}  IC~4329A is a nearby ($z=0.01605$), edge-on, S0, type 1 Seyfert galaxy, at a distance of 64\\,Mpc  ($H_{0} = 75$\\,km s$^{-1}$ Mpc$^{-1}$), situated very close to the centre of Abell cluster A3574. This is unusual given that Seyfert galaxies, like most spirals, are rarely found inside clusters (\\eg Osterbrock \\cite{Osterbrock}). This cluster, first described by Shapley (\\cite{Shapley}), is often referred to as the IC~4329 cluster or group, after its brightest member (\\eg de Vaucouleurs \\etal\\ \\cite{RC2}), or as Klemola~27 (Klemola  \\cite{Klemola}), and is the easternmost and most distant member of the chainlike Hydra-Centaurus Supercluster (Chincarini \\& Rood \\cite{Chincarini}).  The heavily reddened type~1 Seyfert nucleus observed within IC~4329A (Winkler \\etal\\ \\cite{Winkler}) is a strong FIR/X-ray source, the 25$\\mu m$ peak in the {\\em IRAS} flux density, suggestive of there being hot gas close to the nucleus. It has the steepest Balmer decrement ($H{\\alpha}/H{\\beta}\\simeq 12$) of any known Seyfert, a very steep optical spectral index ($\\alpha = 4.4$), and it is seen to show \\nai\\ in absorption (Penston \\& Wilson \\cite{Penston}). Although calculated extinction values indicate that the absolute visual magnitude $M_{v}$, is at least -23, and may be as high as -25, IC~4329A appears optically, to be a largely undisturbed edge-on system, with a prominent dust lane.  The brightest galaxy within the cluster, the giant lenticular IC~4329 has a redshift close to that of IC~4329A ($\\Delta v = 460$\\,km s$^{-1}$), and lies at a projected distance of 59\\,kpc to the west. As pointed out by Kollatschny \\& Fricke (\\cite{Kollatschny}), these two systems appear to be part of a loose group of seven galaxies, indicating that a number of the group member's activity may well be linked to past interactions. The fact that IC~4329 is a shell galaxy, together with the fact low surface brightness features are observed around IC~4329A (Wolstencroft \\etal\\ \\cite{Wolstencroft}), suggest that an interaction is taking place between IC~4329 and IC~4329A. Of the seven group members, only IC~4329A shows any indication of significant nuclear activity.   Further evidence for possibly interaction-induced activity is also apparent from emission-line imaging studies. A halo of emission-line gas can be seen around IC~4329A in the $H{\\alpha} +$[\\nii] image of Colbert \\etal\\ (\\cite{Colbert}), extending along the minor axis, $\\sim$10\\arcsec\\ (3\\,kpc) on both sides of the nucleus, with a luminosity of $\\sim2.5\\times10^{39}$erg s$^{-1}$. An almost identical image is seen in Mulchaey \\etal\\ (\\cite{Mulchaey96}). This $H{\\alpha}$ halo is believed by both sets of authors to represent an outflow from the nucleus, perhaps a superwind of the type commonly seen in edge-on infrared-luminous galaxies (\\eg McCarthey \\etal\\ \\cite{McCarthey}; Armus \\etal\\ \\cite{Armus}).  Radio mapping of the IC~4329A system has brought about some very interesting results also. In a Molonglo Synthesis Telescope (MOST) 843\\,MHz survey of radio sources in southern Abell clusters, IC~4329A is seen to possess an apparent radio tail $\\sim6$\\arcm\\ long (Unewisse \\cite{Unewisse}), corresponding to a linear size of 110\\,kpc. This feature implies a radio structure larger than any other known Seyfert, though it must be stressed that the resolution of the MOST observations make it impossible to be certain whether the observed feature is associated with the galaxy or not. Other significant radio observations include those of Unger \\etal\\ (\\cite{Unger}) at 1490\\,MHz (and 4860\\,MHz), where, besides a bright core being visible, an extended region of emission is observed extending $\\sim6$\\arcsec\\ towards IC~4329. This feature may well be associated with the interaction. Furthermore, Blank \\& Norris (\\cite{Blank}), in a 4\\arcsec\\ resolution 2.3\\,GHz observation, observed two 10\\arcsec\\ extensions emanating from the nucleus. The first is roughly in the same direction as the extension seen by Unger \\etal\\ (\\cite{Unger}). The second, extending to the south-east, perpendicular to the dust-lane of the galaxy, may, according to Blank \\& Norris (\\cite{Blank}), be due to a superwind, driven from the nucleus. Baum \\etal\\ (\\cite{Baum}) found that the kpc-scale radio emission seen in several Seyfert galaxies tended to align itself with the systems' minor axes, and they concluded that circumnuclear starbursts are the most likely cause.  \\subsection{Previous X-ray observations}  X-ray emission associated with IC~4329A was first suspected when Ariel~V detected the steady source 2A~1347-300 (Cooke \\etal\\ \\cite{Cooke}). The error box however, contained IC~4329 as well, and it was only later (Delvaille \\cite{Delvaille}) that the source was identified with IC~4329A. Further observations were performed by HEAO~1, both in scanning (Piccinotti \\etal\\ \\cite{Piccinotti}) and pointed modes (Tennant \\& Mushotzky \\cite{Tennant}; Mushotzsky \\cite{Mushotzky84}), and by HEAO~2 (Petre \\etal\\ \\cite{Petre}). IC~4329A and IC~4329 also appear in Fabbiano \\etal's (\\cite{Fabbiano92}) X-ray catalog and atlas.  IC~4329A has also been observed (Miyoshi \\etal\\ \\cite{Miyoshi}) with gas scintillation proportional counters aboard the Japanese satellite Tenma (Tanaka \\cite{Tanaka}). It had been established prior to this that large amplitude variabilities in the X-ray flux are rare within Seyfert galaxies on timescales of seconds to years (Mushotzky \\cite{Mushotzky84}), and indeed, no significant change in flux or spectral shape was seen over the six day Tenma observation. The spectral shape above 15\\,keV however, an almost flat, rather positively-sloped tail (rare for Seyfert~1 galaxies), appeared to be very different when compared to the HEAO~1 data (Mushotzky \\cite{Mushotzky84}). Observations of IC~4329A with {\\em Ginga} (Piro \\etal\\ \\cite{Piro}) showed a hard X-ray bump above 8\\,keV, probably produced by absorption or reflection of the central emission by a very thick cold medium close to the nucleus, an idea supported by the detection of 6.4\\,keV fluorescence lines.  The \\Ros\\ PSPC data from IC~4329A have been published recently in conjunction with the {\\em COMPTON GRO} observations (Madejski \\etal\\ \\cite{Madejski}, hereafter M95), their main result being that IC~4329A's spectrum is compatible with a single power law, of energy spectral index 1, modified by absorption and reflection extending from soft X-rays to $\\gamma$-rays. What evidence there is of a $\\gamma$-ray spectral break is weak, and in any case, the energy of this possible break is both higher than that of NGC~4151 (Zdziarski \\etal\\ \\cite{Zdziarski}), and higher than that of typical Seyfert 1s (Fabian \\etal\\ \\cite{Fabian}). The \\Ros\\ spectrum of IC~4329 was also presented, it being well described by an optically thin thermal plasma with $kT=0.9$\\,keV. As far as the spatial structure obtained with the PSPC is concerned, M95 describe IC~4329A's X-ray source as being consistent with point-like.  Lastly, {\\em ASCA} observations of IC~4329A (Cappi \\etal\\ \\cite{Cappi}) indicate that the 0.4$-$10\\,keV spectrum is best described by a steep power law spectrum passing through a warm absorber, together with a strong reflection component and Fe K line, confirming both the above {\\em ROSAT-GRO} (M95) and separate {\\em ASCA} (Mushotzky \\etal\\ \\cite{Mushotzky95}) results. Furthermore, as concluded by M95, cold absorption in excess of the Galactic value is required by the data, consistent with the edge-on nature of the galactic disc.  The \\Ros\\ X-ray telescope (XRT), with the Position Sensitive Proportional Counter (PSPC) (Pfeffermann \\etal\\ \\cite{Pfeffermann}) at its focal plane, offers three very important improvements over previous X-ray imaging instruments (such as the \\Ein\\ IPC). Firstly, the spatial resolution is very much improved, the 90\\% enclosed energy radius at 1\\,keV being $27''$ (Hasinger \\etal\\ \\cite{Hasinger}). Secondly, the PSPC's spectral resolution is very much better ($\\Delta E/E \\sim 0.4$ FWHM at 1\\,keV) than earlier X-ray imaging instruments, allowing the derivation of characteristic source and diffuse emission temperatures. Lastly, the PSPC internal background is very low ($\\sim3\\times10^{-5}$\\,ct s$^{-1}$ arcmin$^{-2}$; Snowden \\etal\\ \\cite{Snowden}), thus allowing the mapping of low surface brightness emission. The High Resolution Imager (HRI) on the other hand, because of its excellent spatial resolution (more like $5''$) and relative insensitivity to diffuse emission, is an ideal instrument for further investigation into the point source populations (see Tr\\\"{u}mper (\\cite{Trumper}) for a description of the \\Ros\\ satellite and instruments).  Here we report the results of a 15.5\\,ks \\Ros\\ High Resolution Imager (HRI) observation that addresses one essential aspect of the X-ray emission from IC~4329A and its neighbours that has not been possible until now $-$ that of the high resolution spatial properties of the X-ray emission. Although, as mentioned above, some aspects of the PSPC data have been published (M95), the authors concentrate almost entirely on the spectral properties of the individual systems, and no discussion of the spatial properties is given. We have therefore performed a thorough reanaysis of the 8.3\\,ks of \\Ros\\ PSPC data, concentrating on the spatial properties and on the existence of any extended features. These results are also presented here.  The plan of the paper is as follows. Sect.\\,\\ref{sec_obs_data} describes the observation and the preliminary data reduction methods used, Sect.\\,\\ref{sec_disc_sources} discusses the results as regards the point source emission, Sect.\\,\\ref{sec_disc_unres} discusses the results as regards the remaining unresolved emission, and finally a summary is presented in Sect.\\,\\ref{sec_summary}. ",
        "conclusions": "\\label{sec_disc_sources}  \\subsection{The bright point sources - IC~4329A, IC~4329 \\& H1-P3}  The three bright sources visible in the IC~4329A field are all especially interesting. H1-P3 is bright, shows marginal evidence for extension in the PSPC (though the fact that no evidence for extension is seen in the HRI data implies that the source is truly unresolved) and appears associated with a quite bright (B magnitude = 16.5) star-like object, some 2.9\\arcsec\\ distant. H5-P6 is also very bright, appears to be significantly extended (both in the PSPC and HRI data), and is associated with the giant elliptical galaxy IC~4329. Finally, H11-P8 is extremely bright, showing a great deal of structure, and is undoubtedly due to the Seyfert galaxy, IC~4329A.  It is worth noting again here that an analysis of the PSPC data (in conjunction with {\\em COMPTON GRO observations}) has already been published by Madejski \\etal\\ (\\cite{Madejski}) (M95). They deal however, almost exclusively with the spectral properties of IC4329A (plus those of IC~4329 and H1-P3, or as they call it, S3), and so, in the discussion that follows, many of the results we present have not been addressed by M95, and are new, though we do compare the results of our spectral analysis with those of M95.  As in M95, spectra of all three bright objects (IC~4329A, IC~4329 and H1-P3) were analysed. Spectra were extracted for all three objects from within circles of 1.7\\arcm\\ (for IC~4329A and IC~4329) and 3\\arcm\\ (for H1-P3) at the position of each source (we note that an extraction radius of 3\\arcm\\ for IC~4329A, as used in M95, is likely to be too large, given that this is approximately the distance between IC~4329A and IC~4329). Background spectra were extracted as follows: for IC~4329A, from an annulus 6.6\\arcm\\ to 9.1\\arcm\\ from IC~4329A, thus avoiding contamination from any other bright features; for IC~4329, from a 1.7\\arcm\\ radius circle situated equidistant, on the opposite side of IC~4329A, thus ensuring that contamination from the very bright central source could be removed; and for H1-P3, from a 3\\arcm\\ to 5.5\\arcm\\ annulus centred on H1-P3, again avoiding any bright features.  The three background-subtracted spectra, once corrected for exposure and vignetting effects, were fitted with standard spectral models (thermal bremsstrahlung, power law, blackbody and Raymond \\& Smith (\\cite{Raymond}) hot plasma models). A number of extra, more complex models have been attempted as regards the IC~4329A spectrum, as in M95, and the results of all the best fits are given below in Table~\\ref{table_fits_sources} as follows: Source (col.\\,1), spectral model (whether PL - power law plus absorption, BB - blackbody plus absorption, PL/E - power law plus absorption and an edge, RS Raymond \\& Smith hot plasma plus absorption) (col.\\,2), fitted $N_{\\rm H}$ (col.\\,3), fitted spectral index $\\Gamma$, where $F\\propto E^{-\\Gamma}$ (col.\\,4), fitted temperature (kT, in keV) (col.\\,5) (Note here that in the case of the PL/E model, this column gives the edge energy in keV). The next columns give the metallicity  (solar, where an `F' indicates that the value has been frozen) (col.\\,6), the reduced $\\chi^{2}$ (col.\\,7), and three values of the (0.1$-$2.4\\,keV) luminosity (cols.\\,8-10). Two values of luminosity as calculated using the PSPC results are given; one (col.\\,8) gives the `intrinsic' luminosity of the source (\\ie correcting for the total $N_{\\rm H}$), the second (col.\\,9) gives an `emitted' luminosity (\\ie correcting merely for the Galactic $N_{\\rm H}$). The final luminosity column (col.\\,10) gives the intrinsic (0.1$-$2.4\\,keV) luminosity, using the count rate observed with the HRI, and calculating the fluxes, assuming identical spectral models as inferred from the PSPC data. All luminosities are calculated for an assumed distance of 64\\,Mpc (which is almost certainly incorrect in the case of H1-P3, as discussed below).  \\begin{table*} \\caption[]{Results of the best model fits to the IC~4329A, IC~4329 and H1-P3 spectra (see text). Models are: PL (power law plus absorption), BB (blackbody plus absorption), PL/E (power law plus absorption and an edge), RS (Raymond \\& Smith hot plasma plus absorption). In the case of the PL/E fit, the temperature $kT$ refers to the temperature of the edge. Three (0.1$-$2.4\\,keV) luminosities are tabulated. One, the intrinsic PSPC luminosity of the source, two, the Galactic $N_{\\rm H}$-corrected (\\ie emitted) PSPC luminosity (Galactic $N_{\\rm H} = 4.4 \\times 10^{20}$~cm$^{-2}$), and lastly, the intrinsic HRI luminosity, using the HRI count rates in conjunction with the models suggested by the PSPC data.} \\label{table_fits_sources} \\begin{tabular}{llrrrrrrrr} \\hline \\noalign{\\smallskip} Source & Model & $N_{\\rm H}$ & Photon & $kT$  & $Z$     & red.$\\chi^{2}$ & \\multicolumn{3}{c} {$L_{\\rm x}$ (10$^{42}$\\,erg s$^{-1}$)} \\\\ & & 10$^{20}$\\,cm$^{-2}$& Index    &(keV)&(Solar)& & (Intrinsic)& (Emitted) & HRI (Intrinsic)\\\\ (1) & (2) & (3) & (4) & (5) & (6) & (7) & (8) & (9) & (10) \\\\ \\noalign{\\smallskip} \\hline \\noalign{\\smallskip} IC~4329A & PL  & 22.7$\\pm$3.3 & 1.28$\\pm$0.12 & &             & 1.1 & 41.6$\\pm$0.7 & 23.7$\\pm$0.4 & 29.0$\\pm$0.3 \\\\ & BB  &  7.5$\\pm$0.1 & & 0.52$\\pm$0.02 &             & 1.1 & 23.8$\\pm$0.4 & 22.8$\\pm$0.4 & 16.6$\\pm$0.2 \\\\ & PL/E& 27.9$\\pm$5.1 & 1.73$\\pm$0.33&(0.72$\\pm$0.14)&& 0.9 & 63.0$\\pm$1.0 & 23.4$\\pm$0.4 & 43.0$\\pm$0.5 \\\\ IC~4329  & RS  &  2.7$\\pm$1.2 & & 1.08$\\pm$0.06 & 1.0(F)      & 1.3 & 0.64$\\pm$0.02& 0.64$\\pm$0.02& 0.53$\\pm$0.03 \\\\ & RS  &  4.3$\\pm$1.6 & & 1.07$\\pm$1.07 &0.4$\\pm$0.2  & 1.3 & 0.79$\\pm$0.03& 0.79$\\pm$0.03& 0.61$\\pm$0.04 \\\\ H1-P3    & PL  &  3.2$\\pm$1.3 & 2.35$\\pm$0.33 & &             & 1.4 & 1.04$\\pm$0.04& 1.04$\\pm$0.04& 0.67$\\pm$0.05 \\\\ \\noalign{\\smallskip} \\hline \\end{tabular} \\end{table*}  Although no thermal model (whether a Raymond \\& Smith hot plasma model or a thermal bremsstrahlung model) is able to fit the IC~4329A data adequately, a simple power law model does gives quite an acceptable fit, the fitted parameters agreeing well with M95 and with Rush \\etal\\ (\\cite{Rush}). However, as in M95, close inspection of the residuals does suggest an edge-like feature at around 0.7\\,keV. Incorporating this edge into the model does improve the fit (the data and residuals are shown in Fig.\\,7), and we are able to reproduce the best-fit results of M95 very accurately. Firstly, a photon index of 1.73$\\pm$0.33 is suggested, consistent with M95, with the {\\em Ginga} data (Piro \\etal\\ \\cite{Piro}; Fiore \\etal\\ \\cite{Fiore}), and with the {\\em ASCA} data (Cappi \\etal\\ \\cite{Cappi}). Secondly, the edge feature at 0.72$\\pm$0.07\\,keV is found at exactly the same energy as in M95. As M95 suggest, the energy of this edge is inconsistent with that expected if there were a neutral absorber present, and this strongly suggests the presence of an ionized absorber (O{\\sc vi}, O{\\sc vii}). Further modelling, to address the question of the true nature of this absorber, is possible. However, because of the modest spectral resolution of the PSPC, one cannot distinguish between different models, \\ie between an ionized absorber model, a partial covering by neutral material model, and a high column cold absorber model (see M95 for a detailed discussion).  It is worth noting that the inferred $N_{\\rm H}$, 27.9$\\times10^{20}$\\,cm$^{-2}$, is substantially larger than the Galactic $N_{\\rm H}$ in the direction of IC~4329A (4.4$\\times10^{20}$\\,cm$^{-2}$; Dickey \\& Lockman \\cite{Dickey}),  indicating the presence of a large intrinsic absorption. This is not too surprising given the edge-on nature of the galaxy. In their study of the soft X-ray properties of Seyfert galaxies in the \\Ros\\ All-Sky Survey, Rush \\etal\\ (\\cite{Rush}) also find a very significant excess in the best fit \\nh.  The intrinsic (0.1$-$2.4\\,keV) luminosity of IC~4329A, 6.3$\\times10^{43}$\\,erg s$^{-1}$, is very large, within the top 10\\% or so of the Seyferts within the Rush \\etal\\ (\\cite{Rush}) All-Sky Survey sample. It is an extremely luminous galaxy, and an extremely luminous Seyfert galaxy as well.   \\begin{figure} \\unitlength1.0cm \\label{fig_spec_ic4329a} \\vspace{120mm} \\hfill \\parbox[b]{8.7cm} {\\caption{IC~4329A spectrum with the best-fit power-law plus absorption edge model (see Table~\\ref{table_fits_sources}). The pulse height spectrum of the total X-ray emission is indicated by crosses, and the fit, by the solid line. }} \\end{figure}  The IC~4329 spectrum on the other hand, is only fitted adequately well by a thermal (Raymond \\& Smith hot plasma) spectrum. The best fit, while keeping the metallicity frozen at solar, results in a well-constrained, 1.08$\\pm$0.06\\,keV spectrum, absorbed by a column of 2.7$(\\pm1.2)$\\\\ $ \\times 10^{20}$\\,cm$^{-2}$, a column consistent with (though on the low side of) the Galactic value. This result is entirely consistent with M95. Fitting of the spectrum while letting the metallicity optimize, gives a column entirely consistent with the Galactic value and a low (0.4$\\pm$0.2 solar) metallicity, though the fitted temperature is less well constrained than in the frozen-metallicity case. The intrinsic (0.1$-$2.4\\,keV) luminosity of IC~4329, 7.9$\\times10^{41}$\\,erg s$^{-1}$, is somewhat higher than average when compared to optically similar systems (Fabbiano \\etal\\ \\cite{Fabbiano92}). The fitted temperature is entirely consistent with that of ellipticals, {\\em ASCA} observations resulting in temperatures for several early-type galaxies of between 0.7 and 1.2\\,keV (Matsushita \\etal\\ \\cite{Matsushita}; Rangarajan \\etal\\ \\cite{Rangarajan}). Similarly, the low fitted metallicity appears to be consistent with ellipticals, high-resolution studies with {\\em ASCA} revealing abundances $\\leq0.5$\\,solar in several cases (Loewenstein \\etal\\ \\cite{Loewenstein}; Matsushita \\etal\\ \\cite{Matsushita}).  The H1-P3 spectrum is best fit with a power law model of photon index 2.35$\\pm$0.33, absorbed by a column of 3.2$(\\pm$\\\\ $1.3) \\times10^{20}$\\,cm$^{-2}$, consistent with that out of our own Galaxy. It is almost certainly due to a background quasar, given the facts that it is unresolved in the HRI data, it appears coincident with a quite bright (B mag = 16.5) star-like object, and it has a spectrum consistent with that of quasars (power law with photon indices in the range 2.2$\\pm$0.2; Branduardi-Raymont \\etal \\cite{Branduardi}; Roche \\etal\\ \\cite{Roche}).  Finally, note that, in Table~\\ref{table_joint_sources}, it appeared that, in the cases of these bright sources, the inferred HRI and PSPC fluxes did not agree particularly well. This could have been attributable to time-variability or the assumption of the wrong spectral model. We have already seen however (Figs.\\,4$-$6) that none of these sources appear to be particularly time-variable in either the HRI or the PSPC, and usage of the correct spectral model, as has been done here, has only really aided the situation in the case of IC~4329, and then, only slightly. In the case of H1-P3, it is possible that the source has varied between the HRI and PSPC observations. The fact that the object is very likely to be a QSO or background AGN adds some credence to this. Also bare in mind that the quality of fit to the H1-P3 PSPC spectra is not excellent, the reduced $\\chi^{2}$ being only 1.4. In the case of IC~4329A and IC~4329, the situation is rather intriguing. Both sources appear extended however, and this will lead to a reduction in the calculated HRI count rates, compared to the PSPC count rates. Furthermore, IC~4329A is extremely bright, accentuating the above effect. Finally, as may be the case (Fig.\\.2), and is discussed in detail later, if there were a large amount of low-surface brightness, diffuse emission in the vicinity of these two galaxies, this could very well lead to a reduction in the HRI count rates compared to the PSPC count rates, the HRI being relatively far less sensitive to this type of emission than the PSPC.  \\subsection{The secondary point sources}  Moving on to the remaining point sources, many interesting results have been obtained. Feature P4, for instance, appears elongated in the east-west direction in the PSPC image (Fig.\\,3). The HRI is able to resolve this feature into two separate, equally bright sources (H2 and H4), the more western of which (H2) appears coincident with a bright stellar-like object, with a B magnitude of 10.7. The apparent optical counterpart to H4 (not seen in the APM finding charts of Irwin \\etal\\ (\\cite{Irwin})) is much fainter (B=14.7, see Fig.\\.8(left)). What is rather striking though, is that, what appears to be a `twin' of H2/4-P4 can be seen on the opposite side of IC~4329A, at an extremely similar projected distance from the bright central galaxy. This source, H16-P12, appears coincident with a rather faint (B=17.6) stellar-like object, less than 2\\arcsec\\ west of HRI position (Fig.\\.8(right)). The positioning of these two sources with respect to the central bright galaxy, IC~4329A, is both unusual and intriguing.  On July 6, 1997, during an observing campaign at the 2.2m ESO/MPG telescope at La Silla observatory, we obtained spectra of these optical candidates using the EFOSC2 spectrograph with grism \\#4, a 2\\arcsec-wide long slit and the $2048 \\times 2048$ $15\\mu m$ LORAL CCD. This gave a dispersion of 2 \\AA\\ per pixel, a spectral coverage of 4100--7500 \\AA, and a spectral resolution of 12 \\AA\\ FWHM. The seeing was typically 1.5\". The data were reduced according to the procedure given in Pietsch \\etal\\ (\\cite{Pietsch88}).  Through these optical observations, we have established that all three sources have nothing at all to do with IC~4329A, and are in fact Galactic foreground or background objects. Sources H2 and H4 to the south-west, when compared with Jacoby \\etal's (\\cite{Jacoby}) library of stellar spectra, appear to be foreground stars of types G3~{\\rm V} and M5~{\\rm V} respectively. Furthermore, the north-eastern source looks to be a background quasar with a redshift of $0.5430 \\pm 0.0005$. The HRI count rates measured are consistent with the X-ray fluxes expected from these source classes.  \\begin{figure*} \\unitlength1.0cm \\vspace{85mm} \\hfill \\parbox[b]{18.0cm} {\\caption{Contours of HRI emission overlayed onto optical digitized sky survey images for the H2/H4 field (left) and the H16 field (right). The X-ray image has been smoothed with a Gaussian filter of 10\\arcsec\\ FWHM, and the contours correspond to 0.0625 and 0.1875\\,cts arcsec$^{-2}$. }} \\end{figure*}   All of the remaining non-bright sources appear to have very close ($<3.5$\\arcsec) optical counterparts, the brightest of which is that associated with H17-P14 (B=14.7).   \\label{sec_summary}  We have observed both the ROSAT HRI and PSPC data from fields centred on the edge-on, type~1 Seyfert galaxy IC~4329A and its nearby companion, the giant lenticular IC~4329. 17 and 22 sources are detected respectively in the full HRI and PSPC fields of view, the brightest being associated with the two central galaxies and a further source to the south-west. Many coincidences are seen in the two datasets and the nine most significant HRI sources all have equivalent PSPC counterparts. In addition to point source emission, unresolved residual emission is detected surrounding the IC~4329A/IC~4329 pair. This emission appears markedly two-component, with a smooth, circularly-distributed hard component, centred midway between the two central galaxies, and a more irregular, soft component, situated almost entirely to the south-east of IC~4329A. Our findings with regard to the observed point-source and unresolved emission can be summarized as follows:  1. An extremely bright ($L_{X}=6\\times10^{43}$\\,erg s$^{-1}$) source is detected associated with the central Seyfert IC~4329A. Its spectral properties are compatible with a single power-law ($\\Gamma=1.73$), with a spectral break at 0.7\\,keV, in very good agreement with previous authors' work.  2. Two other very bright sources are detected associated with the nearby giant lenticular IC~4329 ($L_{X}=8\\times10^{41}$\\,erg s$^{-1}$), and with a likely quasar to the south-west. Fitting of standard spectral models to these source spectra again result in fits that agree well with previous authors' work.  3. Many other bright sources are detected both in the HRI and PSPC fields of view, including three point-like sources, symmetrically positioned with respect to the disc of IC~4329A. Optical follow-up observations of these sources with the 2.2\\,m ESO/MPG telescope at La Silla, Chile, has established that they are nothing to do with the central Seyfert, being merely foreground and background objects.  4. None of the sources detected show any significant temporal variability.  5. In addition to the point sources, residual emission is detected, both in the HRI and in the PSPC, surrounding the IC~4329A/IC~4329 pair. This emission appears markedly two-component, comprising of a hard, smooth, circularly-distributed component, centred somewhere between IC~4329A and IC~4329, and a soft, irregular component, situated almost entirely to the south-east of the IC~4329A disc.  6. The hard component of the residual emission appears itself to be made up of two components. One of these is purely the `wings' of the extremely bright IC~4329A source, visible out to several arcminutes. The second component appears to be hot ($\\sim$1.5\\,keV) diffuse gas, with a luminosity of $\\approx5\\times10^{41}$\\,erg s$^{-1}$ and a mass of perhaps $2\\times10^{11}$\\,$M_{\\odot}$. The properties of this emission are very suggestive of it being due to hot gas within the galaxy group of which IC~4329A and IC~4329 are members.  7. The soft component of the residual emission, in terms of its temperature and one-sided nature, bears a good deal of resemblance to proposed starburst driven winds seen in some far-infrared ultraluminous systems. It is however much brighter ($L_{X}=9\\times10^{41}$\\,erg s$^{-1}$), and larger. Another possibility discussed briefly, is that the soft emission may be a `wake' of stripped gas from the galaxy group.  8. A `bridge-like' feature is detected with the HRI between the two central galaxies, and is likely, as is seen in other similar systems, to be due to shocks resulting from the strong interaction taking place between the two systems."
    },
    "9805/astro-ph9805022_arXiv.txt": {
        "abstract": "We present the first K$^{\\prime}$-band image of the carbon star IRC\\,+10\\,216 with 76\\,mas resolution.  The diffraction-limited image was reconstructed from 6\\,m telescope speckle data using the speckle masking bispectrum method.  The image shows that the dust shell of IRC\\,+10\\,216 is extremely clumpy.  Five individual clouds within a $0\\farcs21$ radius of the central star have been resolved for the first time. On the basis of consistent theoretical models we argue that these structures are produced by circumstellar dust formation. The fragmentation of the shell structure gives most likely direct evidence for an inhomogeneous mass-loss process which may be interpreted in terms of large-scale surface convection-cells (Schwarzschild \\cite{Schwschil_75}) being a common phenomenon for red giants. ",
        "introduction": "IRC\\,+10\\,216 (CW Leo) is the nearest and best--studied carbon star.  It is a long--period variable star with a period of about 650\\,d and a spectral type of C9,5 (see e.g.\\ Olofsson et al.\\ \\cite{OlofJoHj_82}). Estimates of its distance range from 100\\,pc (Zuckerman et al.\\ \\cite{ZuckDyCl86}) to 290\\,pc (Herbig \\& Zappala \\cite{HerbZa70}). IRC\\,+10\\,216 is surrounded by a dust shell which is expanding at $v_{\\rm exp} \\approx 15\\,$kms${^{-1}}$, thereby carrying a mass loss rate of $\\dot{M} \\approx 2-5\\,10^{-5}M_{\\odot}\\,$yr$^{-1}$ (e.g.\\ Loup et~al.\\ \\cite{LoupFoOm_93}).  The first high-resolution IR observations of the dust shell of IRC\\,+10\\,216 were reported by Toombs et al. (\\cite{ToomBeFr_72}), McCarthy et al.\\ (\\cite{McCaHoLo80}, \\cite{McCaMcBa90}), Mariotti et al. (\\cite{MariChSi_83}), Dyck et al.\\ (\\cite{DyckZuLe_84}, \\cite{DyckHoZu_87}, \\cite{DyckBeHo_91}), Ridgway \\& Keady (\\cite{RidgKe88}), Christou et al.\\ (\\cite{ChriRiBu_90}), Le~Bertre et al. (\\cite{LeBeMaRe88}), Danchi et al. (\\cite{DancBeDe_94}), Osterbart et al. (\\cite{OsteBaWe_97}), and Weigelt et al. (\\cite{WeigBaHo_97}). Detailed radiation transfer calculations for IRC\\,+10\\,216 have been carried out, for instance, by Groenewegen (\\cite{Groe97}) using a large amount of spectroscopic and visibility data. Consistent time--dependent models describing the circumstellar shells of dust forming long--period variables have been presented by Fleischer et~al.\\ (\\cite{FleiGaSe92}).  A general result of these models is the formation of discrete dust layers, causing pronounced time--varying, step--like surface intensity distributions (Winters et~al.\\ \\cite{WintFlGa_95}).  In this {\\it Letter} we present diffraction-limited 76\\,mas speckle masking observations of the clumpy dust shell of IRC\\,+10\\,216.  We speculate about the origin of these structures in the light of consistent time--dependent model calculations and discuss the red giants' large-scale surface convection (Schwarzschild \\cite{Schwschil_75}) as a possible mechanism for inhomogeneous mass loss. ",
        "conclusions": "Our speckle masking reconstruction of IRC\\,+10\\,216 shows that the circumstellar dust shell of IRC\\,+10\\,216 consists of at least five individual dust clouds B to F within a $0\\farcs21$ radius of the central star.  With $d=170$\\,pc and $R_{\\ast} = 5 \\cdot 10^{13}$cm, the tangential separations of the clouds B, C, and D from the central star correspond to $\\sim$\\,34\\,AU $\\sim 10\\,R_{\\ast}$ (B) and 23\\,AU $\\sim 7\\,R_{\\ast}$ (C,D).  Time-dependent calculations for circumstellar envelopes show the formation of multiple-shell structures, and the synthetic intensity profiles agree well with the measured ones. The prominent clumpiness of these very inner shells gives evidence for an already inhomogeneous mass loss which may be intimately linked with large-scale surface inhomogeneities possibly induced by giant surface convection-cells."
    },
    "9805/astro-ph9805352_arXiv.txt": {
        "abstract": " ",
        "introduction": "The main aims of our work are the following: \\begin{itemize} \\item {\\sl Probe the evolution of massive stars in low metallicity systems.} Since environments with massive stars at metallicities $Z < 1/10 $ \\zsun\\ are not available in the Local Group we use star-forming regions and super star clusters in BCDs. It is important to constrain the populations of massive stars in different environments, since their evolution is dictated by mass loss whose metallicity dependence is not well known.  As a consequence all predictions related to massive stars depend on the adopted mass loss prescriptions. E.g.\\ chemical yields are $Z$ dependent (\\cite{m92}), and the P-Cygni lines detected in several high redshift galaxies (e.g.\\ \\cite{st96}) depend on the mass loss properties.  To constrain the evolution we analyse the Wolf-Rayet (WR) star content since these stars represent bare stellar cores revealed by mass loss. From the WR and O star content we can also derive constraints on the upper mass and the slope of the IMF (cf.\\ \\cite{s96}, \\cite{l98}) %  \\item {\\sl Explain the origin of nebular HeII emission frequently observed in low metallicity extra-galactic HII regions.} The nature of this emission remained puzzling until recently (cf.\\ \\cite{g91} and \\cite{s96}, \\cite{s97}) and indicates a harder ionizing spectrum than commonly thought. \\end{itemize} ",
        "conclusions": ""
    },
    "9805/astro-ph9805164_arXiv.txt": {
        "abstract": "We note that the solution of accretion disk dynamics for an initial delta-function mass distribution gives a light curve that fits both the rise and the decay pattern of the outburst light curves of black-hole soft X-ray transients (BSXTs) until the onset of the first mini outburst quite well. The Green's function solution of Lynden-Bell $\\&$ Pringle~(1974)~is employed for two different time-independent viscosity laws to calculate the expected count rates of X-ray photons in the Ginga energy bands as a function of time. For both models basic characteristics of the outburst light curves of two typical sources~GS 2000+25~and ~GS/GRS 1124-68 are reproduced together with plausible values of the thin disk parameter $~\\alpha~$ and the recurrence times. This agreement with the outburst light curves and the source properties during quiescence support the idea of mass accumulation and the sporadic release of accumulated mass at the outer disk.   {\\bf Key words:} Soft X-ray transients, accretion discs - black-holes, X-rays ",
        "introduction": "~~~~About two thirds of the known LMXBs are persistent and the remaining one third are transient sources (van Paradijs 1995). The transient sources exhibit a soft X-ray spectrum during outbursts (soft X-ray transients $-$SXTs). An optical outburst also occurs together with the X-ray outburst. During an outburst the X-ray and the optical properties of black-hole soft X-ray transients (BSXTs) and neutron star soft X-ray transients (NSXTs) are very similar to each other and also to those of persistent sources. X-ray luminosities  of~ SXTs~ increase from below $~10^{33}~$erg~s$^{-1}$ to $~\\sim10^{37}$-$10^{38}$ erg~s$^{-1}$~ during an outburst. The time scales for decline range from several tens of days to more than one hundred days. Although some sources show rather monotonic decline in many cases there is more complex behavior. Some sources remain persistent for more than one year. Most of the light curves show a steepening around $10^{36}~$erg~s$^{-1}$ corresponding to a mass accretion rate $\\dot{M}_{x}~\\sim10^{16}~$g~s$^{-1}$ inferred from the X-ray luminosity. Later they turn back to their quiescent states with $L_{x}\\sim10^{31}-10^{33}~$erg~s$^{-1}$ and $\\dot{M}_{x}\\sim10^{11}-10^{13}$ g~s$^{-1}$~ (Tanaka $~\\&~$ Shibazaki~1996~ and references therein).  There are nine known X-ray binaries with strong black-hole candidate primaries (van Paradijs 1995). Three of these sources are non-transient high mass systems and the remaining six are transient LMXBs~(Table 1). Four BSXTs in this group, A0620-00, GS 2000+25,~GS/GRS 1124-68~ and GRO J0422+32~ show striking similarities. Their outbursts have fast rise times around a few days. The decay phases can be fit with exponentials with time constants around one month. During decline all four sources exhibit secondary (mini) outbursts. A0620-00, GS/GRS~1124-68 show also tertiary outbursts. The tertiary outburst is absent in the light curve of ~GRO J0422+32, but it may also be present in GS~2000+25. Secondary maxima were detected around 80 days after the onset of the main outburst of GS~2000+25 and GS/GRS~1124-68. A0620-00 and GRO J0422+32 exhibited the secondary maxima 50 days and 125 days after the onset respectively. For both GS~0620-00 and GS/GRO~1124-68 the tertiary maxima were observed about 200 days after the onset of the main outburst (Tanaka $\\&$ Shibazaki~1996). The basic features of the mini outbursts are the sharp increase in luminosity by a  factor of ~$\\sim 1-3$~ and decay patterns which mimic the decay after the first outburst. A0620-00~ is a recurrent transient with a recurrence time of ~58~ years ~(Tsunemi et al 1989). The outbursts of the other sources were detected only once putting a lower limit to their recurrence times of around a few ten years~ (see Table 1). These similarities allow the working hypothesis that all of these sources run with a similar mechanism.  Since the conditions for outburst build up in the quiescent state it is important to treat the observations made during quiescence. According to the standard disk model (Shakura $\\&$ Sunyaev 1973) most of the X-rays ($>70\\%$) come from the inner parts of the disk extending to a radius which is around ten times the radius of the last stable orbit, $~R_{0}=3R_{g}~$ $=3(2GM_{\\mbox{x}}/c^{2})$. During the quiescent state it is possible to fit a black-body curve roughly to the observed X-ray spectrum. However, inner disk temperatures of$~\\sim 0.2-0.3~$keV obtained from these fits give very  small values ~($1-10~$km$^{2}$) for the X-ray emitting area of the disk, while realistic areas imply temperatures that are orders of magnitude smaller than the soft X-ray temperatures that fit the spectra. A disk becomes optically thin below a certain accretion rate depending on the viscosity parameter $\\alpha$ (Shakura $\\&$ Sunyaev 1973). An optically thin or gray disk may be one way of explaining this inner disk problem.  The mass accretion rates obtained from the optical observations during quiescence ($\\sim10^{15}-10^{16}~$g~s$^{-1}$) are orders of magnitude larger than those ($\\sim10^{11}-10^{13}~$g~s$^{-1}$) obtained from the X-ray luminosities (see, e.g., McClintock et al 1995, for these questions in  A0620-00).  An advection dominated inner disk has been proposed to account for the properties of BSXTs. According to this model most of the energy released in the inner disk is advected into the compact star instead of being radiated from the disk (Narayan et al 1996). Regarding the difference between mass accretion rates obtained from optical and X-ray luminosities during the quiescent states, the NSXTs are similar to BSXTs. Although this model could explain the observed spectra and also the low X-ray luminosities of BSXTs in quiescence it is not able to explain the properties of NSXTs because of the existence of the solid surfaces of the neutron stars: Whatever the disk structure is matter finally reaches the neutron star surface which should produce a distinct black-body like emission.  Another possibility is mass accumulation in the outer regions of the disk. This accumulated matter may be the source of the outburst. For both BSXTs and NSXTs, the mass accretion rate inferred from optical observations may indicate the mass accretion rate arriving to accumulate in the outer disk, while the much lower $\\dot{M}_{\\mbox{x}}$ inferred from the X-ray luminosity may reflect the trickle of mass that proceeds through a mass-starved inner disk characteristic of the quiescent state. This qualitative picture would persist until the mass accumulated in the outer disk reaches the level critical for an outburst.  Models for SXTs must address the characteristic time scales: (1) Fast rise of the light curves around a few days; (2) The subsequent decay with a time scale of the order of one month; (3) Recurrence times of the order of a few ten years. Further, as viscosity is involved in determining the decay, values of the viscosity parameter must be plausible. For comparison $\\alpha\\sim 0.1-1$ for dwarf novae  and related systems for the reaction time of the disk to be $10^{5}-10^{6}~$s (Bath $\\&$ Pringle 1981).  There are mainly two types of models for SXTs: the disk instability models (DIM) ~(e.g. Meyer $\\&$ Meyer-Hofmeister 1981; Mineshige $\\&$ Wheeler 1989; Cannizzo 1992; Cannizzo et al 1995~(CCL)) and the mass transfer instability models (MTI) (e.g. ~Hameury et al 1986; 1987; 1988; 1990).  The basic idea for MTI is that the accretion from the secondary star is unstable for a range of accretion rates roughly from $\\sim 10^{12}-10^{15}~$g~s$^{-1}$~to~$\\sim~10^{16}-10^{17}~$g~s$^{-1}$. The secondary star expands under the influence of hard X-rays ($E > 7~$keV) from the primary. The accretion rate exceeds the upper limit of the lower stable range and jumps to an accretion rate greater than the minimum of the higher stable range. The subsequent  burst of mass flow produces the outburst. The disk becomes thicker and thicker, finally shielding the region around the $L_{1}$ point. When the X-ray illumination stops the companion shrinks and the system returns to its quiescent state. MTI models produce the basic characteristics of the light curves and the recurrence times. The problem with MTI is that, according to the model, the hard X-ray flux with $E>$7~keV~ must exceed $\\sim 2.5 \\times 10^{34}~$$ (M_{c} / M_\\odot)^{2}~$erg~s$^{-1}$ in the quiescent state (Hameury et al. 1986) while there is no data showing the existence of photons in this energy range with such high luminosities before the outbursts (Tanaka $\\&$ Shibazaki~1996).  DIMs are highly accepted today, although they do not produce a completely self consistent picture explaining all features of SXTs. Initially DIMs were successful in explaining basic characteristics of dwarf novae (Cannizzo 1994). The model was extended to explain the behavior of SXTs which are similar to dwarf novae in some cases. The disk instability follows a limit cycle mechanism based on an \"S\" shaped surface density $\\Sigma$ versus $\\dot M$ curve. Upper and lower branches of the \"S\" correspond to hot and cool states respectively. The middle branch represents an unstable regime. When ~$\\dot{M}>~\\dot{M}_{min}~$ the disk follows the upper branch switching to the lower branch when ~$\\dot{M}<~\\dot{M}_{max}$. Each radius in the disk has its own \"S\" curve. If the mass transfer satisfies the condition $~\\dot{M}_{max}$$(R_{inner})<$ $\\dot{M}_{T} <$ $\\dot{M}_{min}$$(R_{outer})~$ then the limit cycle mechanism operates through the range between $R_{inner}$ and $R_{outer}$. If this range of radii covers the entire disk then the whole disk can jump from one stable branch to the other. Accumulation of matter at some radius R may cause mass per unit area $\\Sigma(R)$ to exceed $\\Sigma_{max}(R)$. Viscous dissipation increases suddenly. Waves of surface density propagate to both smaller and larger radii. As a result, the surface density at each radius in the disk exceeds the maximum critical value of the lower branch, and the disk jumps to the hot branch. At the end of this process the disk finds itself in a high viscosity state. Unlike the quiescent state now the viscous time scale becomes small and matter flows on to the central object. Because of the matter flow the densities in the outer regions decrease to below the critical densities. This causes a cooling front to propagate throughout the disk decreasing the local surface density at each radius to below the local critical values. Propagation of the cooling front means the decay of the outburst (e.g., Cannizzo 1992; Lasota et al 1996).  The main difficulty of this model is that the $\\alpha$ values needed in order to produce recurrence times of around a few ten years are not able to produce the observed amplitude and duration of the outbursts or vice versa~(for a discussion of the problems see, e.g.  Lasota 1996).  In this work  BSXTs will~be studied with  a different approach concentrating on a simple explanation in terms of a disk dynamics model for both rise and the decay pattern of the light curves until the onset of the first mini outburst. We explore the behavior of the disk following a sudden release of mass in the outer radius in terms of the disk dynamics model of Lynden-Bell $\\&$ Pringle (1974) (LP). The observations of GS 2000+25 and GS/GRS 1124-68 will be used to illustrate the model.  LP made a study for the evolution of viscous disks in general. In particular the solution with no central flux is of interest, corresponding to a disk with a black-hole at its center. The Green's function solution of LP take initial density distributions  in the form of delta functions in radius. This is an idealization of an initial mass enhancement in a ring. Convolution of the  elementary solution of this Green's function model with any initial mass distribution will give the general solution. Initial mass release at a thin ring at the outer radius (or at a specific radius in the disk) already corresponds to a solution with a single delta function initial mass distribution. Thus the Green's function solution of LP is taken here as the physical solution for the BSXT outbursts. The motivation leading us to this approach is the similarity between the observed X-ray light curves of the outbursts and the Green's function luminosity calculated by LP. This hypothesis means that a sudden dumping of mass in the outer disk leads to the  observed outburst. The present work applies the model to the data from ~GS 2000+25~and~GS/GRS 1124-68.  The part of the study by LP related to our application is summarized in the Appendix. In Section 2 data from GS~2000+25 and GS/GRS~1124-68 are examined with the model. The aim is to calculate the photon flux from the disk in the observed X-ray band of Ginga ($1-20~$keV) as a function of time using the LP model, and compare this with observed photon flux data taken with the Ginga satellite ( provided by S.Kitamoto, private communication). A good fit to the data until the first mini outburst is obtained (Figs. 1$-$4). From this fit the characteristics of the  outburst, the viscosity parameter $\\alpha$, the total mass released and the recurrence times will be obtained for both sources. Section 3 summarizes the conclusions, discusses the scope of the model and the problems, and tests for future applications, including the secondary and tertiary maxima commonly seen in the decay phases of the outbursts of BSXTs.  \\begin{table*} \\caption[]{BSXTs with $~M_{\\mbox{x}} > 3M_{\\odot}$~ (Ref:~Tanaka $\\&$ Shibazaki 1996 and references therein)} \\begin{center} \\begin{tabular}{|llllll|} \\hline ~Name~~~&~~BH~mass~~~&~~Outburst year~~~&~~$L_{\\mbox{ave}}$~~&~$\\langle\\dot{M}\\rangle$~& \\\\ ~&$~(M_{\\odot})$&&($10^{35}~$erg/s)&$(10^{15}$~g/s)&\\\\ \\hline J0422+32 &$ >3.2$ &1992&0.8 &0.2 &  \\\\ 0620-00~ & $>7.3$ &1917,1975 &2 &2 &  \\\\ 1124-68~ & $\\sim~6$ &1991&7 &7&  \\\\ J1655-40 & 4-5 &1994&? &? &  \\\\ 2000+25~ & 6-13.9 &1988 &3 &3 &  \\\\ 2023+33~ & 8-15.5 &1956,1979~? &2~? &2~? &  \\\\ &&1989 &0.6&0.6&  \\\\ \\noalign{\\smallskip} \\hline \\end{tabular} \\end{center} \\end{table*} ",
        "conclusions": "We have addressed the general characteristics of the outburst light curves of black-hole soft X-ray transients (BSXTs), through a study of the typical BSXTs GS 2000+25 and GS/GRS 1124-68. The Green's function solution of the accretion disk dynamics model of Lynden-Bell $\\&$ Pringle (1974) (LP), with a delta function initial mass distribution, produces a rise and decay pattern similar to those of BSXTs. It is most remarkable that this simple dynamical model gives both the rise and the subsequent decay. We have employed this solution adopting a delta function initial mass distribution, a mass $\\delta m$ deposited in a ring at radius $R_{1}$ in the disk, for two different time independent viscosity laws. The LP solution restricts the viscosity to be a time independent power law function of the radial distance from the center of the disk.  For the first model we chose a constant kinematic viscosity $\\nu$ for simplicity. For the second model we adopted the commonly used viscosity parameter prescription $\\alpha~=~\\alpha_{0}~(z_{0}/R)^{n}$~ and chose ~$n=-2$~ to remove the time varying temperature dependence of ~$\\nu$. This leads to ~$\\nu=\\nu(h)$, where h is the specific angular momentum.  For both models $kT_{0}$(max), the inner disk temperature at the moment that the light curve for photon count rate reaches its maximum, is the free parameter together with $cos~i$. We adjusted $kT_{0}$(max) to match the maxima of the model and the observed photon flux light curves. Next, from $t_{*}=b~t$ where $ t_{*}$ is dimensionless time and $t$ is real time, the best fit to the observed photon flux was obtained by adjusting the constant $b$. For GS~2000+25 this procedure was performed tracing the possible mass range, $6 M_{\\odot} < M_{\\mbox{x}} < 13.9 M_{\\odot}$, corresponding to the possible inclination angle range, $~47^{\\circ}<~i~<75^{\\circ}$ (Harlaftis at al 1996). For GS/GRS 1124-68 the black-hole mass is $M_{\\mbox{x}}\\sim 6 M_{\\odot}$ (McClintock et al 1992). For this source we tried four different $cos~i$ values (0.25, 0.5, 0.75, 1.0) keeping $M_{\\mbox{x}}$ constant and followed the same steps to obtain the best fit. The results are in Tables (2$-$4)  For both sources the fits are applied to the data until the onset of the first mini outburst. The LP model fits the secular decay in the light curve. The large $\\chi^{2}$ values reflect the excursions about the mean secular decay, which the model does not address. Figs. (1$-$4) show that both models reproduce the average characteristics of the outburst light curves quite well. The requirement of the model that the kinematic viscosity is independent of time and therefore also of the temperature leads to an $\\alpha$ parameter which increases with decreasing temperature. The usual disk instability models require the opposite trend of $\\alpha$ increasing with increasing temperature. Our time independent viscosity models are artificial and may be considered as an effective representation of real viscosity behavior during the hot states of the accretion disks of BSXTs. Empirically the LP models can explain the rise and the decay following the outburst. Our approach here shows that once sudden mass release is triggered at some $R_{1}$, for example by a disk instability, the subsequent light curve can be understood as the dynamical evolution of the disk with viscosity. It seems plausible that the viscosity mechanism must have a temperature dependence at least to start the outbursts. The success of the model means that the real disk dynamics is insensitive to the temperature dependence of the viscosity, or that a constant effective viscosity LP model is a close approximation  to the real luminosity evolution with the integrated effects of variable and temperature dependent viscosity. To investigate the relation between the LP models and disks with variable viscosity is beyond the scope of the present paper, and is to be considered in subsequent work.  The mass dumped in the outburst is $\\delta m \\sim 10^{24}$~g in all our model fits, for both sources. This is consistent with accumulation at $\\langle \\dot{M}_{\\mbox{x}}\\rangle \\sim$ $\\langle\\dot{M}_{\\mbox{x}}\\rangle_{0620}$~ for recurrence times $t_{r}\\sim \\delta m/\\langle\\dot{M}_{\\mbox{x}}\\rangle\\sim 50$~yrs. We expect that the mass release is triggered by a disk instability. The large amounts of mass release $\\delta m$ imply surface densities $\\Sigma~\\sim~(\\delta m/R_{1}^{2})\\sim 10^{4}-10^{5}~$ g~cm$^{-2}$ which are far in excess of the critical (maximum) surface density values for disk instability models, \\ba \\Sigma_{max}=11.4~ \\mbox{g~cm}^{-2}~R_{10}^{1.05} \\L( \\frac{M}{M_{\\odot}} \\R)^{-0.35} \\alpha^{-0.86}   \\nonumber \\ea (Shafter at al 1986). The mass accumulation without triggering an instability then requires small $\\alpha$ values $\\sim 10^{-4}-10^{-5}$ in the quiescent disk. At such $\\alpha$ the quiescent disk could be optically thin.  There is no complete picture yet which explains all properties of BSXTs, in particular the triggering mechanism for the outbursts which drives the accretion disk from the quiescent state to the outburst and back to the quiescent state. In a recent study by Cannizzo et al~(1995)~ exponential decay patterns are produced by the disk limit cycle mechanism, but sudden rise of the outbursts and their long recurrence times are not addressed.   The second (and in two sources, third) outbursts have rise and decay patterns which can be fitted with exponentials with time constants similar to those of the main outburst. In addition to this similarity they also pose the problem of understanding the burst repetition time (Augusteijn et al 1993; Chen et al 1993; Mineshige 1994). What is the source of the mass producing the second  outburst? The data following the second mini outburst does not allow a detailed fit with the LP model. Assuming that the amplitude of the burst is proportional to the accumulated mass, $\\delta m$, and scaling with the primary outburst we find $\\delta m$ values are $\\sim 3\\times 10^{23}~$g and $\\sim 2.5\\times 10^{23}~$g for the second mini outbursts in GS/GRS~1124-68 and GS~2000+25 respectively. Using the time interval between the main and the second mini outburst $\\sim 80~$ days, we find the extra $\\dot{M}$ to supply the $\\delta m$ of the second mini outburst is $\\sim 4.5\\times 10^{16} $g~s$^{-1}$ for GS/GRS~1124-68 and $\\sim 3.5\\times 10^{16}~$g~s$^{-1}$ for GS~2000+25. This is nearly an order of magnitude greater than the long term average accretion rate, $\\langle\\dot{M}\\rangle$. Two possibilities may be considered: (1) All the accumulated mass is not released during the main outburst. The mass released in the second outburst is \"leftover mass\" and not accumulated between the main outburst and the second outburst.  (2) The mass flow from the companion is enhanced by the X-rays coming from the inner disk in the outburst; or both possibilities run together. The X-ray heating of the disk may be important to trigger the mini outbursts whatever the source of the mass creating the second and the third instabilities.  The time between the main and mini outbursts ($\\sim$ few months)must be a characteristic time scale of the \"trigger\". The response time scale of the upper atmosphere of the secondary is too short ( $<\\atop{\\sim}$ $10^{3}$ s)~ (Chen et al 1993 (CLG)) while  the viscous time scale is of the order of a week. The shielding of the region around the inner Lagrangian point, $L_{1}$, has been proposed to explain this time interval (CLG). If the geometrical shielding should end gradually, the abrupt rise of the first mini outburst at a particular time remains unexplained. Two different approaches were proposed by Mineshige (1994). An optically thick Compton cloud above and below the disk which becomes optically thin in a very short time scale just before the first mini outburst may explain the first mini outburst by the response of the companion to the X-rays coming from the inner disk. Alternatively a second thermal instability would be triggered at the outer disk when the strong X-ray heating of the disk has raised the temperature and decreased the critical density for the trigger sufficiently (transient recession of the cooling front). In the former scenario one still has to explain the secondary mini outburst by invoking a different mechanism. and also exhibits similar rise and decay time scales. Similar time interval and characteristics of all the outbursts seem to imply a unique mechanism responsible for the triggering of the outbursts.  It is well known that simple exponential decays after the outburst maximum describe the data well. The mini outburst decays can also be fitted with exponential decays with the same time constants as required by the main outburst(Fig.5). However, the fast rise and the exponential decays (FRED) are put together as separate pieces of the fit rather than being part of a single dynamical model as in the present model. Our LP models fit the main outburst decay quite well. There is an interesting difference from FRED models when it comes to trying to fit the data incorporating the mini outbursts and their decays. In FRED models, the data stream can be fitted well by a superposition of FRED models for the main outburst and the mini outbursts. For our LP model fits, by contrast, when the fit to the main outburst data is extended past the mini outburst(s), it is seen that the model gives a count rate that is greater than the observed count rates, the deviation starting from the onset of the mini outburst. If it is assumed that the ongoing relaxation in the disk after the main outburst is stopped, and the onset of the mini outburst involves an instability that resets the disk to conditions similar to sudden mass release from a local accumulation, then the LP model can be applied to the mini outburst(s) and their decays.  Augusteijn et al (1993) drew attention to the similarity of the main outburst decay, and the mini outburst decays. They related this to a feed-back model invoking modulation of the mass transfer from the companion by irradiation from the outburst. The present approach identifies the similarity with repeated conditions of mass accumulation and release in the disk itself. The trigger of the mass release could be a disk instability. The similarity of main outburst and mini outbursts, requires small scale repetitions of the main outburst with similar initial conditions, rather than superposition or convolution with the disk state as evolved from the main outburst's decay. The presently available data do not allow a detailed fit of the mini outbursts with the LP models.  The behavior of  ~GS~2000+25~ and ~GS/GRS~1124-68~ are similar to those of A0620-00,~and GRO J0422+32. The conclusions may be extended to them as well. These ideas will be developed and a similar detailed study for the other sources will also be attempted in future work.   ~~~~  {\\it \\bf Acknowledgements} This work started from discussions with our late colleague Jacob Shaham. We thank S.Kitamoto for providing GINGA outburst data for the sources GS~2000+25 and GS/GRS~1124-68, and the referee S.Mineshige for his helpful comments. We thank the Scientific and Technical Research Council of Turkey, T\\\"UB\\.ITAK, for support through the grant TBAG \\\"U-18. \\\"U.Ertan thanks T\\\"UB\\.ITAK for a doctoral scolarship. M.A.Alpar acknowledges support from the Turkish Academy of Sciences."
    },
    "9805/astro-ph9805036_arXiv.txt": {
        "abstract": " ",
        "introduction": "\\vspace{-0.2cm}   Since rapidly oscillating Ap stars (roAp) were first discovered (Kurtz 1982), the number observed has increased considerably, making today a total of 28. This discovery of new roAp stars, together with better observations of the ones already known, have brought to light many interesting questions, revealing, at the same time, the need for further theoretical studies on the subject.  Among the many observational facts that need to be understood (for a review on the observational facts of roAp stars see Kurtz, 1990, 1995) are the high frequencies of the modes observed, which can be higher than the theoretical critical cutoff frequency for acoustic modes in these stars, their apparent alignment with the magnetic field, and the fact that some modes cannot be described by one single spherical harmonic.  In this paper the theoretical work on the roAp stars will be reviewed, and the implications of this work on the questions mentioned above will be inspected. In section 2 the different mechanisms proposed to excite pulsations in these stars will be described, and related to the high frequencies of the modes, and their alignment with the magnetic field. In section 3 the methods commonly used to infer information about these stars, from the observation of their oscillations, will be reviewed, and the problems associated with these methods, in particular when the magnetic field is taken into account, will be discussed. \\vspace{-0.1cm} ",
        "conclusions": "\\vspace{-0.2cm}  As discussed in the first section of this paper, two main mechanisms have been proposed to excite oscillations in the roAp stars: the $\\kappa$-mechanism and the magnetic overstability. These mechanisms are completely different in nature, and so are the oscillations excited by each of them. While the $\\kappa$-mechanism drives acoustic modes, the magnetic overstability drives magneto-gravity modes, which are transverse. It should be clear, however, that none of these mechanisms solves, at the present time, the problem of excitation of oscillations in roAp stars, since, for the magnetic overstability there are no calculations of the growth rates, and, for the $\\kappa$-mechanism, the growth rates obtained were negative, meaning no excitation. However, these last calculations should be repeated using the most recent opacities.  Moreover, the recently confirmed success (Kurtz 1998) in the determination of luminosities of roAp stars, using the asymptotics for high-order acoustic modes, is a very strong evidence against the magnetic overstability, since, if the modes were excited through this mechanism, they would not be acoustic modes, and, therefore, the asymptotics would not apply.  Finally, in relation to the critical cutoff frequency, it was concluded that the magnetic field cannot be separated from the problem of the reflection of the high frequency modes by the atmosphere, as these modes are in fact magnetoacoustic oscillations.  As for the asteroseismology in roAp stars, it seems clear that some care must be taken. First, the identification of the modes might be a problem, since the magnetic field tends to distort the eigenfunctions. Secondly, the small and large separations are influenced by non-spherically-symmetric surface effects, like the magnetic field and chemical inhomogeneities. In this case, the greatest problem relates to the small separations, from which no reliable information can be obtained."
    },
    "9805/astro-ph9805085_arXiv.txt": {
        "abstract": "Recent observations of the compact source embedded within the supernova remnant RCW 103 rekindle interest in the origin of this object's emission.   We contrast several models in which neutron-star cooling powers RCW 103.  Specifically, either the presence of an accreted envelope or a sufficiently intense magnetic field can account for the X-ray emission from this object. ",
        "introduction": "Soon after the X-ray source 1E~161348-5055 was first detected by the {\\it Einstein} observatory (\\cite{Tuoh80}) near the center of the supernova remnant (SNR) RCW 103, \\jcite{Tuoh83} proposed that this source is an isolated neutron star emitting thermal radiation. Optical and radio observations have failed to identify a counterpart (\\cite{Tuoh80}; \\cite{Tuoh83}; \\cite{Dick96}; \\cite{Kasp96}), bolstering the interpretation of this source as an isolated neutron star.  Subsequent X-ray observations with {\\it Einstein} and {\\it ROSAT} have not all confirmed the initial detection (\\cite{Tuoh80}; \\cite{Beck93}).  Using recent observations of 1E~161348-5055 with the {\\it ASCA} observatory and archival data from {\\it ROSAT}, \\jcite{Gott97} verify the existence of this source and refocus attention on the interpretation of its emission.  After subtracting a model for the emission of the surrounding SNR, \\jcite{Gott97} find that the point source spectrum is well described by a blackbody having a characteristic temperature $kT = 0.6$ keV and a flux of $6 \\times 10^{-12}$ erg s$^{-1}$ cm$^{-2}$. Estimates of the distance to RCW 103 vary from 3.3 kpc (\\cite{Casw75}) to 6.6 kpc (\\cite{Leib83}).  Combining these values yields an estimated luminosity of $8 d_{3.3}^2 \\times 10^{33}$ erg s$^{-1}$ and an effective emitting area of $7 d_{3.3}^2 \\times 10^{10}$ cm$^2$ where $d_{3.3}$ is the ratio of the true distance to the X-ray source to 3.3 kpc.  This is less than a percent of the total surface area of a neutron star.  So, unless the emission originates from a tiny hotspot, a blackbody cannot account for the emergent spectrum.  \\jcite{Gott97} also find no periodic variation in the flux greater than 13 \\% of the mean count rate. Variation of this order or larger would be expected from a rotating neutron star emitting from a small portion of its surface unless the hot spot coincides with the rotation axis, the object's period is outside the range explored, or gravitational defocusing smooths the periodic signal.  Several models for this object have been proposed since its discovery. \\jcite{Gott97} argue that the object is unlikely to be a cooling neutron star, a plerion, or a neutron star with an ordinary companion. The dismissal of these models prompted \\jcite{Popo97} to argue that the emission from 1E~161348-5055 is powered by accretion onto a neutron star in a binary with another compact object.  Unless the magnetic field of the neutron star is exceptionally weak ($B < 10^8$~G), it will channel the accreted material onto the polar caps producing hotspots and variability.  For an apparently young object to have such a weak field, he argues that the neutron star is {\\it not} coeval with the remnant, but that the remnant resulted from the supernova of the binary companion.  In this {\\it Letter}, we revisit models of 1E~161348-5055 which account for its emission through neutron star cooling.  In the first, a neutron star cooling through an accreted envelope naturally results in a spectrum which greatly departs from a blackbody.  The second model, an ultramagnetized cooling neutron star, results in anisotropic emission from a hotspot with a spectrum which qualitatively resembles a blackbody. ",
        "conclusions": "We find that a young neutron star cooling through a strongly magnetized or a partially accreted envelope can account for the observed emission from 1E~161348-5055.  The detailed models of \\jcite{Pote97} support the conclusions that we have found analytically in this {\\it Letter}. Furthermore, the estimates of the emitting area of 1E~161348-5055 support the conclusion that the spectrum from this object is significantly harder than a blackbody and possibly results from emission through a hydrogen atmosphere.  The recent discovery by \\jcite{Tori98} of a 69-ms X-ray pulsar (J161730-505505) in the vicinity of SNR RCW 103 complicates the evaluation of the possible models.  It has a spin-down age of $8.1 \\times 10^3$~yr, several times larger than that of the remnant. \\jcite{Tori98} examine the possibility that the X-ray pulsar is associated with the remnant, and find that a kick velocity of $1300 d_{3.3} t_{8.1}^{-1}$ km s$^{-1}$ is required to explain its current position relative to the center of the supernova remnant.  Such a large supernova kick velocity is uncommon but has been observed for other pulsars (\\cite{Lyne94}).  However, when this object is compared with other similar rotation-powered, plerionic pulsars, it is a factor of ten underluminous.  We would argue with \\jcite{Gott97} that this source is a heavily absorbed background object, possibly a rotation-powered, plerionic pulsar as \\jcite{Tori98} suggest but located at a distance $\\sim 10$~kpc.  A reanalysis of the spectrum studied by \\jcite{Tori98} may be able to determine the interstellar column density to J161730-505505 and verify its status as a background source.  \\jcite{Gott97} failed to find flux variations at a level of 13 \\% over a wide range of periods.  Although the total flux from a magnetized neutron star may not vary at this level because of gravitational defocussing (\\eg \\cite{Heyl97analns}), the magnetic field causes the atmospheric emission to be highly anisotropic (\\cite{Shib95}; \\cite{Raja97}), so its apparent lack of variability may indicate that it is only weakly magnetized ($B \\sim 10^{11}$~G) or simply that the geometry is not conducive to large flux variations.  Observations of this object with AXAF should be able to distinguish between these models by determining the spectral shape of 1E~161348-5055.  We argue that the X-ray source in the supernova remnant RCW 103 is simply the natural end product of stellar evolution through a supernova: an isolated, cooling neutron star."
    },
    "9805/astro-ph9805029_arXiv.txt": {
        "abstract": " ",
        "introduction": "Multiwavelength continuum and line data information for a key number of   active galactic nuclei (AGN) in the closer Universe is rapidly growing up and becoming more and more complete.  The availability of such multiwavelength dataset allows a more detailed and consistent study of the nuclear emitting regions and   is definitively demanding more elaborated spectral modeling approaches.  The Circinus galaxy  (A1409-65) was first reported by Freeman et al. (1977) with coordinates (1950) R.A. $\\rm 14^{h} 09^{m} 17^{s}.5$, Dec. $\\rm -65^{o} 06^{'} 19^{\"}$ ; $l$= $\\rm 311^{o}.3$,  b= $ \\rm -3^{o}.8$. Because of its proximity  (at a distance of about 4 Mpc), the Circinus galaxy shows one of the richest  optical - IR  nebular spectrum among AGN. The large range of ionization levels and line strengths immediately suggests the presence of a variety of clouds in the nuclear region having different physical conditions and excited by different mechanisms.  Circinus shows a spectacular, one side [OIII] 5007 ionization cone with apex at the nucleus of the galaxy (Marconi et al. 1994). This asymmetry is probably due to extinction by the galaxy disk which is inclined by about 65 degrees, hiding the counter ionization cone to the observer. Circinus also shows a prominent dust lane  South-West of its nucleus (Marconi et al.) which might be causing the obscuration of the intrinsic nuclear light, in particular considering   the non  detection of Circinus nucleus in the UV light.   In most known Seyfert galaxies residing in spirals, the AGN light is   contaminated by circumnuclear star forming regions which largely  complicates the interpretation of the  line and continuum spectra  from the active nucleus. Circinus is not an exception: its Seyfert 2 nucleus  resides on  a Sb-d system and  several pieces of  evidence indicate outgoing star formation in the nuclear vicinity. Resolved star formation is seen up to $\\sim$10 $\\rm arcsec$ from the centre, where a  ring shaped   region is detected (Marconi et al. 1994), whereas  CO band absorptions - most probably from red supergiants - are seen   within the inner 0.75 arcsec region (Maiolino et al. 1998). Thus,  lines from  low ionization states  in the nuclear spectrum might  include substantial  contribution from the   stellar activity. On the other hand, lines  from  highly ionized species, essentially  above  ion IV,  are expected to be produced by radiation from the  active center (AC).  Besides the starburst contribution, the  AGN  emission can itself be dominated by two competing mechanisms, namely, the ionizing radiation from the  AC and  the shocks. Line and continuum emissions from Circinus indicate that both mechanisms are active. They also reflect the different physical conditions of the clouds from which they arise, including the effect of dust. Therefore, if a meaningful representation of the   nuclear region  is sought, any modeling of the continuum and of the line emission should be consistent with each other.  Previous modeling of Circinus  by Oliva et al. (1994), Moorwood et al. (1997), and Binette et al. (1997) provides a first notion of the complex structure of Circinus  nuclear region. In all of the three cases the analysis and results are exclusively based on the modeling of the high ionization line spectrum. However, the available data from Circinus span from radio to X-ray, which requires a self consistent model to explain  the complete dataset. Gas emission accounts for most of the continuum and for the lines from high and low ionization levels, whereas dust emission accounts mostly for the IR continuum.  Thus, the emission from gas and dust should be calculated  consistently, and their effects properly weighted in the different regions of the electromagnetic spectrum. In the modeling of Circinus, several models are used: some provide a better fit to the line spectrum, whereas, others fit better the continuum.  Modeling is based on  hypothesis about the symmetry of the region, stellar emission, central source energetics, etc.,  which added to the input parameters (velocity field, density, dust characteristics) lead to a description of the real galaxy complex. It is difficult to establish errors of the theoretical results since they depend on atomic and molecular data, some of them being still roughly estimated. Therefore, the most probable model should be selected among the best fitting to the observational data throughout consistency.   This paper  is devoted to  a self consistent treatment of the continuum and narrow emission line spectra from an AGN. A multiwavelength  modeling of the nuclear and extended emission region in the Seyfert 2   NGC 5252 was presented in a similar way (Contini, Prieto \\&  Viegas, 1998). The modeling approach is   based on previous work by Viegas \\&  Contini (1997 review  and references therein) which  focuses on the  coupled effect of photoionization  and shocks.   The paper is organized in the following way. In \\S 2 the observational dataset for Circinus is given. The general model and the input parameters are presented in \\S 3. A grid of single-cloud models for  the Circinus galaxy and the fit to the observed line spectra by multi-cloud models are discussed in \\S 4. The spectral energy distribution (SED) of the continuum  given by these  models  as well as by other models necessary to explain specific features of the continuum  appear  in \\S 5. The consistent fit of  selected multi-cloud models to the line and continuum spectra are discussed in \\S 6.  Conclusions  follow in \\S 7. ",
        "conclusions": ""
    },
    "9805/astro-ph9805359_arXiv.txt": {
        "abstract": "We present and analyze observations of the quadruple lensed quasar Q2237+0305, obtained with the {\\it HST} WFPC2 camera in the F336W and F300W filters.  Twenty-five exposures were performed within 15 hours real time on 3 November 1995. On a timescale of 3--4 hours, we observe no variation in component A of greater than 0.02 mag. The other components remain constant over a period of 10 hours to within about 0.05 mag. In the final 5 hours there is some evidence (not conclusive) for variation of component D by about 0.1 mag.  The exposures indicate that component A is brighter than component B by about 0.3 mag. Components C and D are fainter than component A by about 1.3 and 1.4 mag, respectively.  Our results place an upper limit on any fifth (central) component of 6.5 mag fainter than component A.  We determine the astrometric properties of the lens system, using only the exposures of the higher resolution Planetary Camera chip.  We measure the relative distances of the four components with high accuracy. Our values are systematically larger than those of other investigators (by 0.1\\% to 2.0\\%). We discuss the reasons why we believe our results are reliable.  The F336W filter had been chosen for the observations because it corresponds to the redshifted Ly-$\\alpha$ line of the quasar.  This filter might have allowed us to see extended Ly-$\\alpha$ emission from the Broad-Line Region (BLR) of the quasar as Ly-$\\alpha$ arcs, and hence to determine the physical size of the BLR.  However, the quasar components in this filter are consistent with a point source.  We conclude that there cannot be any Ly-$\\alpha$ feature in the image plane brighter than about 23.5 mag in F336W and further from the quasar core than 100 mas. According to a lensing model by Rix, Scheider \\& Bahcall (1992), this would preclude any such features in the source plane further than 20 mas ($\\sim 100 h^{-1}$ pc, assuming $q_0 = 0.5$) from the quasar core and brighter than 25 mag before magnification. ",
        "introduction": "The quasar Q2237+0305 at redshift $z = 1.695$ is gravitationally lensed by a nearby galaxy at $z = 0.039$ (Huchra \\etal~1985).  The galaxy core lies nearly perfectly along the line of sight. Such a configuration results in a symmetric, cross-like arrangement of the four quasar component images, with relative separations in this case between 1.2 and 1.8 arcsec.  Several facts make this lens system useful. First, the closeness of the lensing spiral galaxy allows us to study it in great detail. Second, the large leverage between lens plane and source plane that results from this proximity reduces the time scale for microlensing. Third, the symmetric arrangement of the four components leads to a small relative time delay (of order a day or shorter, \\cf Wambsganss \\& Paczy\\'nski 1994).  This last fact helps to distinguish intrinsic fluctuations of the quasar from microlensing-induced changes: the former have to show up in all four components almost simultaneously, whereas the latter are completely independent of each other, and occur on a timescale of months. Finally, the fact that the four quasar components are bright, well separated, and of comparable optical brightness makes Q2237+0305 an easy target for various photometric, spectroscopic and astrometric studies.  Recently, Q2237+0305 was detected at radio wavelengths with the VLA at 3.6cm and 20cm (Falco \\etal~1996). The relative positions of the components measured in radio agree well with the optical positions.  Furthermore, the relative flux ratios of the components in each wavelength regime are similar, with the exception that component D is brighter in the radio, compared with the optical light.  Thus, the relative brightnesses in radio agree much better with the ratios predicted by various models (Rix, Schneider \\& Bahcall 1992), than do those in the optical; the slight discrepancy in the optical is possibly caused by a demagnification due to microlensing (\\cf Wambsganss, Paczy\\'nski \\& Schneider 1990; Witt \\& Mao 1994), or by dust in the lensing galaxy.  The system Q2237+0305 was the first in which microlensing was detected (Irwin \\etal~1989; Corrigan \\etal~1991). There are various photometric monitoring programs underway; these continue to show fluctuations in the component intensities which can be attributed to microlensing.  Recent results of some of these campaigns were published by Lewis, Irwin \\& Hewett~(1995), and Ostensen \\etal~(1996).  Here we present new data obtained with the WFPC2 camera aboard the Hubble Space Telescope. The motivation and the technical details of the observations are explained in Section 2.  In Section 3 we present the results of these observations, including the photometry and the astrometry of the four components.  We discuss our results and their implications in the final Section 4. ",
        "conclusions": "In this paper, we have presented the results of Hubble Space Telescope exposures on the WFPC2 camera of quasar Q2237+0305; specifications of the images are in Table 1 and we display the image configuration in Figure 1. The F336W and F300W filters were chosen because the relatively red galaxy bulge through which we see the quasar components is faint in the UV, and because one of the filters is very close to the redshifted Ly-$\\alpha$ line of the quasar (at about 3270\\AA), so that it is plausible that any spatially extended structures in the quasar environment would emit in this wavelength regime. We have determined: \\begin{enumerate} \\item The photometry for the four components in the F336W and F300W bands, which is listed in Table 2. The relative brightnesses of the components are known to vary with time due to microlensing. At the time of our observations, component A was the brightest, followed by component B which was fainter by 0.3 mag. Components C and D were fainter than component A by about 1.3 and 1.4 mag, respectively.  On the timescale of 3--4 hours, we can state that we observe no variation in component A of greater than 0.02 mag. For the other components, over a period of 10 hours they remain constant to within about 0.05 mag. In the final five hours there is some evidence for variation of component D of about 0.1 mag. \\item The astrometry of the four components and the galaxy core, which is listed in Tables 3 and 4. The (1$\\sigma$-) uncertainty of these astrometric measurements are about 1.5 mas. We found the system to be consistently larger (by between 0.1\\% and 2\\%) than previous studies have found. We discuss above why we think our values are reliable. \\item The existence of a feature near the A and B components which is bright in F336W, as shown in Figure 5 with a subtraction of F300W from F336W. We conclude that this signal is probably an artifact of the differing diffraction patterns of the two filters. However, we use the brightness of the feature to obtain an upper limit on the brightness of any real Ly-$\\alpha$ regions, plotted as the crosses on Figures 6 and 7. \\item An estimate of the upper limits these images yield for the brightness of any extended image near the quasar, as a function of angular distance from the quasar in the source plane. These upper limits depend on the lensing model and the desired signal-to-noise. They are plotted in Figures 6 and 7. \\item An upper limit on the brightness of central fifth component in our band which is 6.5 mag fainter than component A. \\end{enumerate}"
    },
    "9805/astro-ph9805115_arXiv.txt": {
        "abstract": "The persistent increases in spin-down rate ({\\em offsets}) seen to accompany glitches in the Crab and other pulsars suggest increases in the spin-down torque.  We interpret these offsets as due to {\\em starquakes} occurring as the star spins down and the rigid crust becomes less oblate. We study the evolution of strain in the crust, the initiation of starquakes, and possible consequences for magnetic field and torque evolution. Crust cracking occurs as equatorial material shears under the compressive forces arising from the star's decreasing circumference, and matter moves to higher latitudes along a fault inclined to the equator.  A starquake is most likely to originate near one of the two points on the rotational equator farthest from the magnetic poles. The material breaks along a fault approximately aligned with the magnetic poles.  We suggest that the observed offsets come about when a starquake perturbs the star's mass distribution, producing a misalignment of the angular momentum and spin axes.  Subsequently, damped precession to a new rotational state increases the angle $\\alpha$ between the rotation and magnetic axes. The resulting increase in external torque appears as a permanent increase in the spin-down rate. Repeated starquakes would continue to increase $\\alpha$, making the pulsar more of an orthogonal rotator. ",
        "introduction": "The magnetic braking torque acting on an isolated neutron star would be steady in the absence of abrupt changes to the star's magnetic configuration. Most pulsars, however, do not slow in a regular fashion, but suffer variations in their spin rates in the form of glitches and timing noise. Perhaps the most striking aspect of spin evolution is the persistent increases in spin-down rate that accompany glitches in the Crab pulsar (\\cite{LPS}), PSR 0355+54 (\\cite{lyne87}) and PSR 1830-08 (\\cite{sl96}); see Table 1.  In the Crab, these permanent {\\em offsets} involve fractional changes in the spin down rate of $\\sim 10^{-6}-10^{-4}$ (see Fig. 1). PSR 1830-08 has exhibited one offset of $8\\times 10^{-4}$, and a persistent offset might have followed the large 1986 glitch of PSR 0355+54. Similar offsets might also occur in the Vela pulsar (\\cite{le97}), giving rise to the small braking index of $1.4\\pm 0.2$ reported by Lyne \\etal\\ (1996).  It is striking that all observed offsets are of the same sign, and correspond to {\\em increases} in the spin-down rate.  One interpretation of this phenomenon is that glitches are accompanied by sudden and permanent increases in the external torque acting upon the star (\\cite{leb92}; \\cite{le97}). Such torque changes could occur if either the direction or magnitude of the star's magnetic moment changes. {\\em Starquakes}, occurring as the star spins down (\\cite{starquakes}; \\cite{ruderman76}) would affect the external torque if they change the orientation of the magnetic moment with respect to the spin axis. If structural relaxation occurs asymmetrically about the rotation axis, due perhaps to magnetic stresses or asymmetric material properties, the star's spin and angular momentum vectors would become misaligned. As the star precesses and relaxes to a new rotational state, the magnetic moment would assume a new orientation with respect to the rotation axis, leading to a change in the external torque. In this paper, we study how the rigid neutron star crust relaxes its structure, and consider possible consequences for evolution of the magnetic field and torque.  Alpar and Pines (1993; also \\cite{more_capacitors}) have suggested that the Crab's offsets result from a reduction in the moment of inertia on which the external torque acts. Such a change could occur by either a structural change of the star, \\eg, the star becomes less oblate, or through a decoupling of a portion of the star's liquid interior from the external torque.  If the moment of inertia decreases through structural readjustment, to conserve angular momentum, the star would always spin {\\em more} rapidly than had the glitch not occurred (\\cite{leb92}; \\cite{le97}). The large offsets following the 1975 and 1989 glitches in the Crab, however, eventually caused the star to spin {\\em less} rapidly than had the glitch not occurred (\\cite{LPS}); the observed offsets, therefore, cannot be due {\\em solely} to structural readjustments. In principle, the Crab's spin-down rate offsets could be due to decoupling of a portion of the star's superfluid interior from the external torque (\\cite{capacitors}; \\cite{more_capacitors}), though quantitative agreement of this model with the data has yet to be demonstrated. Torque increases associated with the surface field structure appears to be the most straightforward explanation.  A neutron star relaxes its oblateness as it spins down, moving equatorial material toward the rotation axis and reducing the equatorial circumference. If the stellar crust is brittle, stresses lead to {\\em starquakes} as the yield strength of the crustal material is exceeded (\\cite{starquakes}).\\footnote{Superfluid stresses in the crust (\\cite{ruderman76}) could also drive crust cracking, but will not be considered here.} The actual response of solid neutron-star matter at high pressure to shear is not well-understood. Here we assume that the neutron star crust is brittle and explore the consequences of this assumption.\\footnote{ Known materials exhibit ductile rather than brittle behavior under pressures comparable to their material shear moduli (\\cite{duba90}). Nevertheless, deep-focus earthquakes are known to originate from regions of very high pressure (\\cite{gh95}). These faults are thought to be facilitated by densification phase changes; small regions of higher density nucleate as the material is stressed, and act as a lubricant for shearing motion. Analogous processes might occur in the high-pressure material of the neutron star.} ",
        "conclusions": "The persistent increases in spin-down rate ({\\em offsets}) seen to accompany glitches in the Crab and other pulsars suggest sudden increases in the spin-down torque.  Starquakes occurring as a neutron star spins-down and readjusts its structure affect the spin-down torque exerted on the star by changing the magnetic field geometry or orientation. In this paper we have examined the evolution of strain in the crust of a spinning-down neutron star and the initiation of starquakes as the material reaches critical strain.  Crust cracking occurs as equatorial material shears under the compressive forces arising from the star's decreasing circumference. The star decreases its oblateness as matter is moved to higher latitudes along a fault inclined to the equator.  Magnetic stresses suppress shearing near the magnetic poles, especially shearing motions across the field lines.  Starquakes are thus most likely to originate near the two points on the equator farthest from the magnetic poles and propagate toward the magnetic poles.  Starquake-induced misalignment of the star's angular momentum and spin, associated with glitches, is a possible explanation for the spin-down offsets seen in the Crab pulsar. Following the misalignment, damped precession to a state of larger angle $\\alpha$ between the magnetic and rotation axes could increase the external torque, giving a permanent increase in the spin-down rate. Repeated starquakes would continue to increase $\\alpha$, making the pulsar more of an orthogonal rotator."
    },
    "9805/astro-ph9805323_arXiv.txt": {
        "abstract": "The Cosmic Infrared Background (CIB) is hidden behind veils of foreground emission from our own solar system and Galaxy. This paper describes procedures for removing the Galactic IR emission from the 1.25 -- 240 $\\micron$ {\\it COBE} DIRBE maps as steps toward the ultimate goal of detecting the CIB. The Galactic emission models are carefully chosen and constructed so that the isotropic CIB is completely retained in the residual sky maps. We start with DIRBE data from which the scattered light and thermal emission of the interplanetary dust (IPD) cloud have already been removed. Locations affected by the emission from bright compact and stellar sources are excluded from the analysis. The unresolved emission of faint stars at near- and mid-IR wavelengths is represented by a model based on Galactic source counts. The 100 $\\micron$ DIRBE observations are used as the spatial template for the interstellar medium (ISM) emission at high latitudes. Correlation of the 100 $\\micron$ data with H I column density allows us to isolate the component of the observed emission that is associated with the ISM. Limits are established on the far-IR emissivity of the diffuse ionized medium, which indicate a lower emissivity per H nucleus than in the neutral medium. At 240 $\\micron$, we find that adding a second spatial template to the ISM model can greatly improve the accuracy of the model at low latitudes. The crucial product of this analysis is a set of all-sky IR maps from which the Galactic (and IPD) emission has been removed. We discuss systematic uncertainties and potential errors in the foreground subtraction process that may have an impact on studies seeking to detect the CIB in the residual maps. ",
        "introduction": "The primary scientific goal of the Diffuse Infrared Background Experiment (DIRBE) aboard the {\\it Cosmic Background Explorer (COBE)} spacecraft is the measurement of the cosmic infrared background (CIB) at wavelengths from 1.25 to 240 $\\micron$. This radiation is the cumulative emission of pregalactic sources, protogalaxies, and evolving galaxies, as well as emission from more exotic processes not common in the local universe (e.g. Bond, Carr, \\& Hogan 1986). These sources of the CIB may be found anywhere from the earliest epoch after radiation and matter were decoupled to the present day. The contribution to the CIB from galaxies will be composed of stellar emission from distant galaxies that is redshifted by the cosmological expansion from intrinsically shorter wavelengths, as well as the direct IR emission from stars and dust within galaxies at all distances. The constraints that DIRBE places on the CIB are discussed by Hauser et al. (1998, hereafter Paper I) and Dwek et al. (1998, hereafter Paper IV).  However, in order to detect the CIB, we first need to remove the strong contributions of foreground emission arising within our own solar system and Galaxy. The IR foreground from within our solar system originates from the interplanetary dust (IPD) cloud. The modeling and removal of the scattering and emission from the IPD for the entire cold-mission DIRBE data set are reported by Kelsall et al. (1998, hereafter Paper II). Following the removal of this foreground, we attack the next layer of foreground emission by modeling and subtracting the Galactic IR emission.  The near-IR (1.25 -- 4.9 $\\micron$) emission of the Galaxy is dominated by starlight. Some bright stars and other compact sources are resolved as point sources in the DIRBE data. Most stars blend into a diffuse background, showing a disk and bulge very similar in appearance to many edge-on spiral galaxies. Extinction effects are clearly present at the shortest near-IR wavelengths as a visible dark lane in the inner Galaxy. Papers which have previously examined DIRBE observations of the stellar disk and bulge of the Galaxy are: Weiland et al. (1994), Arendt et al. (1994), Freudenreich et al. (1994), Dwek et al. (1995), Calbet et al. (1996), Freudenreich (1996), Binney, Gerhard, \\& Spergel (1997), Bissantz et al. (1997), Porcel, C., Battaner, E., \\& Jim\\'enez-Vicente, J. (1997), and Fux (1997). The mid- and far-IR (12 -- 100 and 140 -- 240 $\\micron$) emission is dominated by thermal emission from dust in the diffuse ISM and in more compact star-forming regions. Previously published studies of the Galactic ISM based on DIRBE data include: Arendt et al. (1994), Freudenreich et al. (1994), Sodroski et al. (1994, 1995), Bernard et al. (1994), Boulanger et al. (1996), Dwek et al. (1997), Sodroski et al. (1997), Davies et al. (1997), and Lagache et al. (1998).  This paper details the development of models of the Galactic IR emission. The primary intended use for the models is to permit an accurate measurement of an extragalactic IR background. The characterization of the sources that give rise to the Galactic emission is an important secondary result of the process. An overview of the procedures used for modeling the Galactic foreground is given in Section 2. Section 3 describes the DIRBE data set and its preparation, particularly the removal of the IPD scattering and emission (Paper II). Section 4 describes the modeling and removal of starlight from the 1.25 to 25 $\\micron$ measurements. The following section (\\S5) describes in detail the modeling and removal of the ISM emission from DIRBE data. Additional investigation into the far-IR emissivity of the diffuse ionized ISM is contained in the Appendix. The accuracy of the removal of the Galactic emission is a major limitation to the detection of the CIB, and consequently Section 6 discusses estimates of the uncertainties of the data and procedures. A brief discussion of the implications of this modeling for Galactic properties is contained in Section 7. A more detailed analysis of the Galactic ISM as revealed by this work has been reported by Dwek et al. (1997). Finally, a brief summary of results is given in Section 8. Companion papers (Papers I \\& IV) contain analyses of the residual emission, and its implications for the detection of the CIB. ",
        "conclusions": "We have modeled and removed the Galactic IR emission in the DIRBE data in preparation for analysis of the CIB. The procedures used were designed to preserve the emission of the CIB in the residual maps, and not remove it inadvertently with the Galactic emission. The procedures concentrated on producing accurate results at high latitudes where Galactic emission is weakest. At low latitudes, deficiencies in the models are clearly visible in the residual maps.  We find that the stellar emission of the Galaxy is reasonably reproduced by our Faint Source Model, which is based on the SKY model (Wainscoat et al. 1992; Cohen 1993a, 1994a, 1995). An offset of the Sun by $\\sim$ 18 pc from the Galactic plane is required to produce equal residual near-IR intensities at north and south Galactic latitudes. However, there is clearly room for improvement in the geometry or calibration of the FSM.  We find that the ISM can be fairly well modeled by a single spatial and spectral component if we constrain our study to high Galactic latitudes. At 240 $\\micron$, a model of the ISM with two spatial components can produce a much more complete subtraction of the ISM, extending to low Galactic latitudes. The two spatial components can combine to produce a range of color temperatures across the sky. This shows that a complete model of the ISM needs to be able to account for a continuous range of dust temperatures. We are unable to detect any IR emission associated with low density ionized gas at high Galactic latitudes."
    },
    "9805/gr-qc9805014_arXiv.txt": {
        "abstract": " ",
        "introduction": "The combination of particle physics models with general relativity provides one with the possibility to construct quantitative scenarios of the very early Universe.  One important scenario, which helps to solve some of the outstanding problems of standard Big Bang cosmology like the homogeneity and the flatness problems, is that the Universe went through a stage of accelerated expansion, an ``inflationary stage\", in the very early part of its evolution \\cite{lin}. The inflationary scenario not only looses the dependence on peculiar initial conditions, it also provides a quantitative way to understand the formation of structure (galaxies and clusters of galaxies). Indeed, in this scenario, the origin of these large-scale structures can be traced back to vacuum {\\em quantum} fluctuations of scalar fields \\cite{haw82} and the resulting scalar (gravitational potential) fluctuations of the metric. These fluctuations then lead eventually to the formation of large-scale structure in the universe and leave also their imprint as anisotropies in the cosmic background radiation. The anisotropies on large angular scales ($l\\leq 20$, where $l$ is the multipole number) were detected by the COBE satellite. Future satellite missions, Planck Surveyor \\cite{pl} and MAP \\cite{ma}, are scheduled for detection and high precision measurement of these anisotropies up to small angular scales (large $l$'s) and will enable us to possibly test the above scenario. (A recent critical review of some of these aspects is \\cite{BG}.) In addition inflation makes the important prediction of a background of relict gravitational waves \\cite{al79} originating from tensor quantum fluctuations of the metric -- this constitutes an effect of linear quantum gravity.  Though research in this field has entered an exciting stage in which concrete models can be confronted with observations of ever increasing accuracy, an important question of principle is whether and to what extent the quantum origin of the primordial fluctuations can be recognised in the observations. This can only be answered after a thorough understanding of the quantum-to-classical transition for the primordial fluctuations has been achieved. Moreover, such an analysis is anyway necessary for an investigation into the possibilities to observe genuine quantum gravitational effects {\\em beyond} the linear approximation. Such effects may arise, for example, from quantum gravitational correction terms to the functional Schr\\\"odinger equation \\cite{KS} and may in principle be observable in the spectrum of the microwave background \\cite{Ro}.  As a result of the dynamics of the fluctuations produced during inflation, one obtains for almost all initial quantum states a quantum state that is both highly squeezed and highly WKB \\cite{AA94}. The peculiarity of this highly WKB state is characterised in the Heisenberg picture by the fact that the information about the initial momentum becomes lost -- a direct consequence of the vanishing of the decaying mode. This was shown for an initial vacuum (Gaussian) state \\cite{PS1} as well as for initial number eigenstates \\cite{PS2}. As a result, the fluctuations cannot be distinguished from a classical stochastic process, up to a tremendous accuracy well beyond the observational capabilities. This property does not require any environment \\cite{PS1,PS2}.  However, interaction with the environment is unavoidable. This was already stressed regarding the problem of the entropy of the fluctuations \\cite{PS1,LPS}. Usually, classical properties for a certain system emerge by interaction of this system with its natural environment, a process referred to as decoherence (see \\cite{dec} for a comprehensive review). It is therefore important to investigate its importance in the early Universe, since the fluctuations of the scalar field and the metric will most likely interact with various other fields. Highly squeezed states are extremely sensitive to even small couplings with other fields (see Sect.~3.3.3 in \\cite{dec}). Since almost all realistic couplings are in field space (as opposed to the field momentum), the distinguished ``pointer variable\" (defining classicality) is the {\\em field amplitude} which also defines a quantum nondemolition variable in the high squeezing limit \\cite{KPS}. Interferences between different field amplitudes are therefore suppressed in the system itself with the same precision with which the non-diagonal elements of the density matrix describing the system can be taken zero. Environment-induced decoherence is effective when this precision is well beyond observational capabilities.  In \\cite{KPS} we have stressed that these two features play a decisive role in the emergence of classicality. In the present article we shall give more quantitative details than in the above mentioned ones about the nature of this quantum-to-classical transition. We shall present at length some aspects of the free quantum particle which, surprisingly, exhibits many features analogous to primordial fluctuations and discuss some physical ``experiments''. Our paper is organised as follows. Section~2 gives a brief review of the dynamics of cosmological perturbations and clarifies the first of the above two ingredients in the quantum-to-classical transition. Section~3 then explains in what precise sense the system is indistinguishable from a classical stochastic process; this takes place up to an accuracy not only well beyond observational capabilities, but even well beyond the level of accuracy which is meaningful (beyond this accuracy, many other corrections should anyway be taken into account, as stressed in ~\\cite{PS1}). In Section~4 we present the analogy with the free quantum particle. We discuss both the similarities to and differences from the case of primordial fluctuations. Section~5 gives a detailed account of how environment induced decoherence works for the primordial fluctuations. In particular, the rate of de-separation as a measure for quantum entanglement is calculated for various initial states. Section~6 gives our conclusions.  \\vskip 4mm ",
        "conclusions": "We have investigated in detail the quantum-to-classical transition of the fluctuations of quantum origin produced during inflation. When no interaction with the environment is taken into account, such a transition takes place up to a precision well beyond observational capabilities. This is directly related to the fact that it is possible to describe the fluctuations nowadays solely with the help of the ``growing'' quasi-isotropic mode. This transition means that the quantum coherence can be expressed in classical terms, namely the system can be described as a {\\it stochastic} classical system. That this is very far away from a classical (deterministic) system is very clear in the example of a free particle: at very large times, one cannot ascribe anymore to it a definite trajectory in phase space, but rather one has a {\\it classical} probability density with stochastic amplitudes (positions) $x$ and fixed momenta $p_{cl}(x)$. The initial quantum state then completely defines the statistics of the fluctuations through the probability distribution $|\\Psi|^2$. Most inflationary models lead to a Gaussian statistics of the fluctuations, a result in good agreement with observations. Clearly this is very far away from a classical free particle! This aspect is somehow hidden in the case of cosmological fluctuations because in the latter case one is willing to accept the stochasticity of the fluctuations, and it would look absurd to even try a deterministic description of these fluctuations, even if one believes the fluctuations are classical from the very beginning. This explains why the description in terms of a classical stochastic process does not look surprising. It is only when one thinks of the quantum origin of the fluctuations that the peculiar quantum nature appears. We stress also that the quantum-to-classical transition is a result of the expansion of the universe and that it depends on the stretching of the fluctuations while they are outside the Hubble radius. Note that this would not apply to scalar fields with too large mass~\\cite{Mi}, in complete accordance with the fact that these fields cannot be described by just a growing mode.  We again emphasise that the environment has to be taken into account, since the highly squeezed states are extremely sensitive to the presence of an environment, as has been discussed in Sect.~5. When it is taken into account, even coherences which are unobservable in practice but still present {\\it in} the system, essentially disappear from the system itself, since they are ``hidden'' in the correlations with the huge number of degrees of freedom of the environment. However, in the peculiar case of inflation these coherences, not expressible in classical stochastic terms, are anyway tremendously tiny. It is not even clear that environment-induced decoherence would be effective enough to reduce them any further. However, interaction with the environment has an irreversible character and is certainly crucial regarding the problem of the entropy of the fluctuations. It is crucial that this interaction does not spoil the standard predictions of inflationary physics which will be possibly tested in the near future by the satellite missions {\\it MAP} and, with even higher accuracy, by {\\it PLANCK Surveyor}. For example, the fact that the fluctuations have stochastic amplitudes but {\\it fixed} phases results in the appearance of (Sakharov or Doppler or acoustic) peaks on small angular scales in the angular power spectrum of the cosmic microwave background anisotropies.  We note that our discussions exhibit a surprising connection between cosmology -- the origin of structure -- and fundamentals of quantum theory \\cite{KPS}. The quantum-to-classical transition by decoherence is a very general process as studied recently in quantum optical experiments \\cite{Ha}.  We emphasised in Sect.~2 that the high squeezing of the quantum state for the primordial fluctuations is ``generic''. One can, of course, start with any ``quantum state\" at the end of inflation (not necessarily highly squeezed) and evolve it back to the beginning of inflation by the Schr\\\"odinger equation, where it yields an acceptable initial state. However, this state should initially (before inflation) be tremendously {\\em narrow} in $y$ and may be rejected as being unnatural (this is our {\\em quantum no hair conjecture}). Assuming that such an initial state is self consistent with inflation, which is certainly {\\it not} the case for most models, then the longer the inflationary phase, the better our conjecture is expected to work; however, the minimum duration required for inflation to be of cosmological interest is certainly effective enough in this respect. The high squeezing of the fluctuations and the ensuing quantum-to-classical transition is a generic feature of the inflationary phase itself."
    },
    "9805/astro-ph9805053_arXiv.txt": {
        "abstract": "High-resolution spectra of $\\lambda$ Bootis stars reveal the presence of circumstellar gas for example in the Ca K line.  The example of the normal A star $\\beta$ Pictoris shows, that the narrow stable absorption component in Ca K can be reproduced using appropriate disk models and a calcium underabundance in the circumstellar gas of a factor of $\\sim 30$.  Similar models are suggested for the group of metal-deficient $\\lambda$ Bootis stars, but the observational material is still very poor. ",
        "introduction": "High-resolution spectrometry and NLTE analysis of $\\lambda$ Bootis stars reveals the metal-deficient nature of this small subgroup of main-sequence A stars (Venn \\& Lambert 1990, Holweger \\& St\\\"{u}renburg 1993, St\\\"{u}renburg 1993). Venn \\& Lambert (1990) suggested that the abundance anomalies of these stars are due to accretion of circumstellar (CS) gas, which is depleted in condensable elements.  Evolutionary tracks for the $\\lambda$ Bootis stars and a recent high-S/N search for circumstellar Ca K lines raise the question of a possible pre-main-sequence evolutionary status for these stars (Gerbaldi et al. 1993, Holweger \\& Rentzsch-Holm 1995, Paunzen et al. 1998). In this case the circumstellar gas may have remained from the star formation phase.  The disk models presented in the following section are in close analogy to those for more massive disks around T-Tauri stars. A short discussion has been given by Rentzsch-Holm, Holweger \\& Bertoldi (1998). ",
        "conclusions": "For $\\beta$ Pictoris the stable component of the CS absorption line in Ca K can be explained by absorption in a gaseous disk in Keplerian rotation around the star. Ca\\,{\\sc ii} is the dominant ionization stage as already pointed out by Lagrange et al. (1995). Since the inclination of the disk is only a few degrees (Smith \\& Terrile 1984), the results suggest that calcium is underabundant by a factor of $\\sim 30$ in the circumstellar disk around $\\beta$ Pictoris.  The model presents the possibility to quantitatively study the circumstellar lines observed in several $\\lambda$ Bootis stars. Nevertheless further information such as ISO observations or radio observations are needed to constrain the free parameters of the model, like disk masses and extensions."
    },
    "9805/astro-ph9805047_arXiv.txt": {
        "abstract": "The comparison of the observed and computed energy distributions of $\\beta$ CrB has shown that a model with the specific chemical composition of the star can account for the visual enery distribution, while it is still unable to reproduce ultraviolet observations shortward of 1700 \\AA. Furthermore, the predicted absorption of strong Fe\\,{\\sc ii} and Mg\\,{\\sc ii} UV lines is much larger than the observed one. ",
        "introduction": "Some magnetic Ap stars, as  $\\beta$ CrB (HD 137909, HR 5747, F0p) and 33 Lib (HD 137949) show an excess of ultraviolet flux shortward of 2000~\\AA~ when compared both  with other peculiar or normal stars of similar spectral type (i.e. 60 Tau = HD 27628, 78 Tau = HD 28319) and with energy distributions computed for solar or solar scaled abundances.  Hack et al. (1997) explained the UV excess of $\\beta$ CrB ($T_{\\rm eff} = 7950$~K, $\\log g = 4.3$ from their paper) as due to a $\\lambda$ Boo companion having [M/H]=-1.0 and $T_{\\rm eff}$ equal to about 8200~K. However, the hypothesis of a silicon deficiency which increases the ultraviolet flux was also suggested.  In this paper, we show that models computed with approximate specific abundances of $\\beta$~CrB explain the observed depressions at 4200~\\AA~ and 5200~\\AA, but are not able to explain the observed UV excess. ",
        "conclusions": "An atmospheric model computed for the specific abundances of $\\beta$~CrB is still unable to reproduce the ultraviolet observations, which show an excess of ultraviolet flux shortward of 1700~\\AA~ and show profiles of the strong Fe\\,{\\sc ii} and Mg\\,{\\sc ii} lines much weaker than the predicted ones.  Therefore, either the ATLAS12 model can still not predict the spectrum of $\\beta$ CrB, owing to the lack, in the code, of treatment of magnetic field effects and inhomogeneous chemical distributions, or we actually observe the combined fluxes of $\\beta$ CrB and of the unknown companion, which could well be a $\\lambda$ Boo star, but also a much cooler star affected by chromospheric emissions.  33~Lib shows a similar behaviour, but we have not been able as yet to extend the sample to more stars, owing to a dramatic lack of IUE low resolution observations of Ap stars.  \\begin{figure}[hbt] \\epsfysize=11cm \\epsfxsize=5.0cm \\hspace{0cm}\\epsfbox{T5f3.ps} \\vspace{0.5cm} \\caption[h]{The ultraviolet energy distributions (in mag, relative to 5556~\\AA) from both ATLAS9 (full line) and ATLAS12 (dotted line) models are compared with the IUE  (dashed line) and TD1 S2/68 (dark points) observations} \\end{figure}"
    },
    "9805/astro-ph9805337_arXiv.txt": {
        "abstract": "I have collected all available ROSAT observations of VY Scl stars including those during the all-sky survey to investigate the presence, strength and spectrum of soft X-ray emission (0.1--2.4 keV) of this group of nova-like variables. A total of 9 out of the 14 VY Scl stars are detected with ROSAT, mostly during optical high states. Interestingly, all detections during the optical high state have very similar X-ray spectral characteristics. I find that a simple blackbody model gives a reasonably good fit in all cases, with temperatures falling in the narrow range between 0.25--0.5 keV. Possible emission mechanisms are discussed. ",
        "introduction": "VY Scl stars are a subclass of nova-like, cataclysmic variables which are bright most of the time, but occasionally drop in brightness at irregular intervals (e.g. Warner 1995). The transitions between the brightness levels occur on a time scale of days to weeks. These variables are typically found at $P >$ 3 hrs and with large mass transfer rate \\mdot\\ (upper right corner of the $P_{\\rm orb}-$\\mdot\\ diagram of Osaki 1996), and thus are thought to be steady accretors (or dwarf novae in a state of continuous eruption as suggested by Kraft 1964) with hot disks. Their disks are thus assumed to be thermally and tidally stable.  Evidence for high \\mdot\\ is based on measures of the absolute magnitude of VY Scl stars during their high state. The accretion disks then are assumed to be optically thick and can be approximated by models of quasi-steady disks (see e.g. Warner 1987 for a summary). Absolute magnitudes can be determined in various ways, and range between $M_{\\rm V}$ = 3--6 mag for VY Scl stars (Warner 1987, see also Tab. \\ref{xlist}), corresponding to mass transfer rates of up to 5$\\times$10$^{-8}$ \\msun/yr.   An interesting, partly overlapping group of high \\mdot\\ cataclysmic systems are the so-called SW Sex stars (three of the five SW Sex stars belong to the VY Scl group) which all are eclipsing systems but show single-peaked emission lines remaining largely unobscured during primary eclipse. This has been explained in terms of a combination of an accretion stream which  is overflowing the initial impact onto the disk with the effect of a strong accretion disk wind (e.g. Hellier 1996). The observations  of these properties in eclipsing systems only is certainly a selection bias, and it remains to be seen whether other VY Scl stars also exhibit some or all of the properties which presently make up the SW Sex classification criteria.  Based on the fact that VY Scl stars have similar low states like AM Her binaries which have no disks, low states are thought to involve drops in the mass transfer rate from the secondary. Livio \\& Pringle (1994) proposed a model for the optical brightness drops of  VY Scl stars in which the reduced mass transfer rate is caused by a magnetic spot covering temporarily the $L_1$ region. This mechanism works predominantly at short orbital periods because the level of magnetic activity increases with the rotation rate of the star (which in turn is coupled to the orbit). The same idea has recently been expanded and applied to detailed disk instability modelling (King \\& Cannizzo 1998). It was shown that a simple reduction in mass transfer rate from the secondary is not sufficient because then, after the transition of the disk to a cool state, the disk should show outbursts like in dwarf novae. Since such outbursts have never been observed from VY Scl stars during optical low-state, King \\& Cannizzo (1998) concluded that all disk mass must be drained away after the transition of the system into the optical low state.  \\begin{table*} \\caption{Compilation of basic properties of the 14 presently known VY Scl stars$^{(1,2)}$} \\begin{tabular}{cccrccccccc} \\hline \\noalign{\\smallskip} Name~ & R.A. & Decl. & b$^{\\rm II}$ & Mag.  & P$_{\\rm orb}$ & d & $\\!\\!$M$_{\\rm V}^{\\rm low}$/M$_{\\rm V}^{\\rm high}\\!\\!$& i$^{\\rm (3)}$ &    M$_{\\rm WD}$ & M$_{\\rm D}$  \\\\ & (2000.0)   & (2000.0) &    & range   & (min) &       (pc) &  (mag)      & (\\D) & (\\msun) & (\\msun)   \\\\ \\noalign{\\smallskip} \\hline \\noalign{\\smallskip} PX And & 00\\H30\\M05\\fss9 & $\\!\\!$+26\\D17\\amin26\\asec & --36  & 14.8--17.0 & 211 & $>$180 & $<$9/$<$7 & $\\approx$74$^{\\rm ecl}$ & 0.2$^{(4)}$ & 0.32    \\\\ TT Ari & 02\\H06\\M53\\fss1 & $\\!\\!$+15\\D17\\amin42\\asec & --43 & ~\\,9.5--14.5 & 198 & $\\!$180--200$\\!$ & 9/4 &30--40 &$\\approx$1 & 0.35  \\\\ KR Aur & 06\\H15\\M44\\fss0 & $\\!\\!$+28\\D35\\amin08\\asec & +6 & 11.3--19.0 &  234 & 180 & 12/5 & $\\lax$40 & 0.7 & 0.48  \\\\ BZ Cam & 06\\H29\\M34\\fss1 & $\\!\\!$+71\\D04\\amin36\\asec & --24 & 12.5--14.1 & 221 & $\\approx$500$^{(5)}$ & 6/4$^{(5)}$  & $\\lax$40 & 0.1--1.0 & 0.3--0.35  \\\\ BH Lyn & 08\\H22\\M36\\fss1 & $\\!\\!$+51\\D05\\amin24\\asec &  +35 & 13.7--17.2 & 224  &  &   & $\\approx$79$^{\\rm ecl}$ & 0.37--1.4 & 0.22--0.5  \\\\ DW UMa  & 10\\H33\\M53\\fss1 & $\\!\\!$+58\\D46\\amin54\\asec & +50 & 13.8--18.1 & 197 & $\\approx$850 & 7.5/3 &$>$72$^{\\rm ecl}$ & 0.15--0.6 & 0.15--0.3 \\\\ LX Ser & 15\\H38\\M00\\fss2 & $\\!\\!$+18\\D52\\amin02\\asec &  +51 & 13.3--17.4 & 228 & $\\!$250--460$\\!$ & 9/5 & 75$^{\\rm ecl}$ & 0.32--0.48 &0.32--0.39\\\\ $\\!$V442 Oph & 17\\H32\\M15\\fss2 & $\\!$--16\\D15\\amin23\\asec & +9 & 12.6--15.5 &  202  & $>$80 & $<$10/$<$7  & $\\lax$65 & 0.35--0.45 & 0.3--0.37  \\\\ MV Lyr & 19\\H07\\M16\\fss4 & $\\!\\!$+44\\D01\\amin07\\asec & +16 & 12.2--18.0 &  193  & 320 & 10.2/6.0 & 9--14 & 0.4--1.4 & 0.17 \\\\ V794  Aql & 20\\H17\\M34\\fss0 & $\\!$--03\\D39\\amin52\\asec & --21  & 13.7--20.2 & $\\!$240-330$\\!$ & 200 & 14/7  & 22--56 & 0.5--1.2 & 0.46--0.6  \\\\ $\\!$V751 Cyg & 20\\H52\\M12\\fss9 & $\\!\\!$+44\\D19\\amin25\\asec & --0 & 13.2--17.8 & $\\approx$360    & 430 & 8.2/3.6  & & &  \\\\ $\\!$V425 Cas & 23\\H03\\M46\\fss7 & $\\!\\!$+53\\D17\\amin14\\asec & --6 & 14.5--18.5 & 216      &  &      & 16--34 & 0.55--1.2 & 0.29--0.33  \\\\ VY    Scl & 23\\H29\\M00\\fss5 & $\\!$--29\\D46\\amin47\\asec & --72 & 12.9--18.5 & 239 & $\\approx$500 &  9/4   & 25--40 & 0.8--1.4 & 0.23--0.42  \\\\ VZ    Scl & 23\\H50\\M09\\fss2 & $\\!$--26\\D22\\amin53\\asec & --76 & 15.6--20.0 & 208 & 530 & 11.4/3  & $>$76$^{\\rm ecl}$ & 0.3--1.0 & 0.32 \\\\ \\hline \\noalign{\\smallskip} \\end{tabular}  \\noindent{\\Ni $^{(1)}$ The coordinates and most of the optical magnitudes are taken from Downes \\& Shara (1993). Note that the Simbad coordinates are sometimes less accurate. \\\\ $^{(2)}$ References for table values: PX And: Thorstensen \\etal\\ 1991, Still \\etal\\ 1995, TT Ari: Cowley \\etal\\ 1975, Shafter \\etal\\ 1985, KR Aur: Shafter 1983a, Antov \\etal\\ 1996, BZ Cam: Lu \\& Hutchings 1985, Krautter \\etal\\ 1987, BH Lyn: Richter 1989, Andronov \\etal\\ 1989, Dhillon \\etal\\ 1992, Hoard \\& Szkody 1997, DW UMa: Hessmann 1990, Honeycutt \\etal\\ 1993, Dhillon \\etal\\ 1994, LX Ser: Young \\etal\\ 1981, Eason \\etal\\ 1984, V442 Oph: Shafter \\& Ulrich 1982, Szkody \\& Shafter 1983, MV Lyr: Schneider \\etal\\ 1981, V794 Aql: Shafter 1983b,c,  Honeycutt \\& Schlegel 1985, V751 Cyg: Robinson \\etal\\ 1974, Bell \\& Walker 1980, Greiner \\etal\\ 1998, V425 Cas: Shafter \\& Ulrich 1982, Shafter 1983c, Wenzel 1987, VY Scl: Hutchings \\& Cowley 1984, VZ Scl: Schaefer \\& Patterson 1982, Sherrington \\etal\\ 1984, O'Donoghue \\etal\\ 1987 \\\\ $^{(3)}$ Eclipsing systems are marked by ``ecl''. \\\\ $^{(4)}$ The formal best-fit result of M$_{\\rm WD} \\approx$0.2 \\msun\\ has been regarded as implausible because it implies mass transfer on a dynamical time scale (Thorstensen \\etal\\ 1991). \\\\ $^{(5)}$ M$_{\\rm V}$=4 has been assumed in deriving the distance (Krautter \\etal\\ 1987). } \\label{xlist} \\end{table*}  Previously (pre-ROSAT) known X-ray emission from VY Scl stars include V794 Aql (Szkody \\etal\\ 1981), TT Ari during the optical high state (Jensen \\etal\\ 1983), KR Aur during the optical high state (Mufson \\etal\\ 1980, Singh \\etal\\ 1993), LX Ser (Szkody 1981). These investigations have consistently concluded that the  high X-ray luminosity expected from the boundary layer ($\\sim$10$^{34}-10^{35}$ erg/s), based on the high M$_{\\rm V}$ and thus accretion rate of the order of 10$^{-8}$ \\msun/yr) was not detectable.  In order to relate the findings of this X-ray survey of VY Scl stars to some possibly underlying physical quantities, I have compiled some important system parameters like the apparent magnitude range, orbital period, distance, inclination and mass of the binary components from the available literature (Tab. \\ref{xlist}). Also, I have collected the brightness estimates of the VY Scl stars over the last seven years from various sources in order to determine the optical state during which the ROSAT observations have been performed (Fig. \\ref{lc}). Many of the VY Scl stars are monitored by various amateur astronomers around the world and thus a substantial amount of monitoring data was available from the AAVSO, AFOEV and VSNET databases.  In the following I present a complete overview of the ROSAT observations of VY Scl stars during the all-sky survey as well as in subsequent pointed observations. Results for some of the ROSAT observations have been reported already earlier: on MV Lyr and KR Aur during their optical high state (Schlegel \\& Singh 1995; Richman 1996), on TT Ari (Baykal \\etal\\ 1995; Richman 1996), on BZ Cam (van Teeseling \\& Verbunt 1994), and on VZ Scl and DW UMa (van Teeseling \\etal\\ 1996). The results of the \\ros\\ all-sky survey detections of V794 Aql and BZ Cam have already been mentioned  in Verbunt \\etal\\ (1997).   \\begin{table*} \\vspace{-0.15cm} \\caption{ROSAT observations of VY Scl stars$^{(1)}$} \\begin{tabular}{cccccccccr} \\hline \\noalign{\\smallskip} Name & Date  &  T$_{\\rm exp}$ & offaxis  &  CR$^{(2)}$ & HR1 & HR2 & log L$_{\\rm X}^{(3)}$     & opt.  & D$^{(4)}$ \\\\ &      & (sec) & angle & (cts/s) & & & (erg/s) &  $\\!\\!$state$\\!\\!$ &  \\\\ \\noalign{\\smallskip} \\hline \\noalign{\\smallskip} PX And & $\\!\\!$Dec. 31, 1990--Jan. 1, 1991 & ~\\,350 & 0--55\\amin & $<$0.022 & -- & -- & --$^{(5)}$ & & -- \\\\ & Jul. 4/5, 1991  & $\\!\\!\\!$26\\,294 & 0\\farcm3 & 0.0052$\\pm$0.0005 & 0.79$\\pm$0.08 & 0.24$\\pm$0.09 & $>$29.7 & & ~\\,5\\asec \\\\ \\noalign{\\smallskip} TT Ari & Jan. 20/21, 1991 & ~~275 & 0--55\\amin & 0.360$\\pm$0.037  & 0.80$\\pm$0.06 & 0.34$\\pm$0.09 & 31.4 & high & ~\\,4\\asec \\\\ & Aug. 1/2, 1991 & $\\!\\!\\!$24\\,465 & 0\\farcm3 & 0.366$\\pm$0.006 & 0.82$\\pm$0.01 & 0.34$\\pm$0.01 & 31.4 & high & 1\\asec \\\\ \\noalign{\\smallskip} KR Aur & Sep. 14/15, 1990 & ~~360 & 0--55\\amin & 0.065$\\pm$0.014  & 1.00$\\pm$0.00 & 0.75$\\pm$0.14 & 30.7 & high & ~\\,7\\asec \\\\ & Sep. 28--Oct. 6, 1992  & $\\!\\!\\!$17\\,255 & 0\\farcm3 & 0.064$\\pm$0.002 & 0.94$\\pm$0.01 & 0.52$\\pm$0.03 & 30.7 & high & ~\\,2\\asec \\\\ \\noalign{\\smallskip} BZ Cam & Sep. 14/16, 1990 &  ~~430 & 0--55\\amin & 0.077$\\pm$0.014  & 1.00$\\pm$0.00 & 0.17$\\pm$0.18 & 31.9 & & 17\\asec \\\\ & Sep. 29, 1992  & 6\\,117 & 0\\farcm2 & 0.074$\\pm$0.004 & 0.90$\\pm$0.02 & 0.30$\\pm$0.05 & 31.8 & high & ~\\,4\\asec \\\\ & Sep. 3, 1993  & 4\\,591 & 0\\farcm2 & 0.062$\\pm$0.004 & 0.89$\\pm$0.03 & 0.37$\\pm$0.06 & 31.6 & & ~\\,4\\asec \\\\ \\noalign{\\smallskip} BH Lyn & Oct. 7--9, 1990 & ~~460 & 0--55\\amin & $<$0.006 & -- & -- & --$^{(6)}$ & & -- \\\\ \\noalign{\\smallskip} DW UMa & Oct. 25--28, 1990 &  ~~370 & 0--55\\amin & $<$0.033 & -- & -- & & high & -- \\\\ & Oct. 15, 1992  & 3\\,395 & 0\\farcm1 & 0.011$\\pm$0.002 & 0.28$\\pm$0.14 & 0.31$\\pm$0.16 & 31.3$^{(7)}$ & rise & ~\\,4\\asec \\\\ & Oct. 15, 1992  & 1\\,710 & 0\\farcm1 & 0.013$\\pm$0.003 & 0.42$\\pm$0.20 & 0.44$\\pm$0.20 & 31.2$^{(7)}$ & rise & ~\\,6\\asec \\\\ \\noalign{\\smallskip} LX Ser & Aug. 8--10, 1991 & ~~345 & 0--55\\amin & $<$0.037 & -- & -- & $<$31.1$^{(7)}$  & high & -- \\\\ \\noalign{\\smallskip} V442 Oph & Sep. 4/5,   1990  & ~~315 & 0--55\\amin & $<$0.014 & -- & -- &  --$^{(5)}$  &  & -- \\\\ & Sep. 22/23, 1992  & $\\!\\!\\!$11\\,385 & 41 & $<$0.0064 & -- & -- & --$^{(5)}$ & & -- \\\\ \\noalign{\\smallskip} MV Lyr & Oct. 11--15, 1990 &  ~~722 & 0--55\\amin & 0.079$\\pm$0.011  & 0.82$\\pm$0.09 & 0.69$\\pm$0.10 & 31.3 & high & ~\\,6\\asec \\\\ & Nov. 4--8, 1992  & $\\!\\!\\!$20\\,250 & 0\\farcm2 & 0.069$\\pm$0.002 & 0.84$\\pm$0.02 & 0.45$\\pm$0.02 & 31.2 & high  & 4\\asec \\\\ & May 28, 1996  & 2\\,218 & 0\\farcm1 & $<$0.0008$^{(1)}$ & --  & -- & $<$29.7 & low & -- \\\\ \\noalign{\\smallskip} V794 Aql & Oct. 18, 1990 &  ~~240 & 0--55\\amin & 0.093$\\pm$0.021  & 0.87$\\pm$0.11 & 0.44$\\pm$0.19 & 30.8$^{(7)}$ & high & 11\\asec \\\\ \\noalign{\\smallskip} V751 Cyg & Nov. 19/20, 1990  & ~~370 & 0--55\\amin & $<$0.019 & -- & -- &$<$30.3$^{(7)}$ & high & -- \\\\ & Nov. 3, 1992   &  3637 & 52     & $<$0.0058 & -- & --  & $<$30.8$^{(7)}$ & high & -- \\\\ & Jun. 3, 1997   &  4663 & 0\\farcm3 & 0.11$\\pm$0.02$^{(1)}$ & -- & -- & --$^{(1)}$ & low & 2\\asec \\\\ & Dec. 2--8, 1997   & 10813 & 0\\farcm2 & 0.08$\\pm$0.02$^{(1)}$ & -- & -- & --$^{(1)}$ & low & 7\\asec \\\\ \\noalign{\\smallskip} V425 Cas & Dec. 30/31, 1990 & ~~380 & 0--55\\amin & $<$0.019 & -- & -- & --$^{(6)}$ & high & -- \\\\ \\noalign{\\smallskip} VY Scl & Nov. 23--25, 1990 &  ~~~85 & 0--55\\amin & $<$0.13  & -- & -- & $<$31.7$^{(7)}$ & & -- \\\\ \\noalign{\\smallskip} VZ Scl & Nov. 29--Dec. 1, 1990 & ~~290 & 0--55\\amin & $<$0.014 & -- & -- & & high & -- \\\\ & Dec. 5, 1992  & 3\\,354 & 0\\farcm2 & 0.004$\\pm$0.001 & 0.35$\\pm$0.25 & 0.03$\\pm$0.27 & 30.3$^{(7)}$ & high & ~\\,7\\asec \\\\ & Dec. 5, 1992 &  2\\,237 & 0\\farcm2 & 0.006$\\pm$0.002 & 0.70$\\pm$0.24 & 0.12$\\pm$0.29 & 30.5$^{(7)}$ & high & ~\\,6\\asec \\\\ \\noalign{\\smallskip} \\hline \\end{tabular}  \\noindent{\\Ni\\small $^{(1)}$  All observations except the May 1996 pointing on MV Lyr and the 1997 pointings on V751 Cyg have been performed with the ROSAT PSPC. For MV Lyr the HRI count rate has been transformed into a PSPC rate by using the spectral fit parameters of the PSPC observation. For V751 Cyg see Greiner \\etal\\ (1998). \\\\ $^{(2)}$  Count rates are calculated for the 0.1--2.4 keV range (= channels 11-240). Upper limits are 3$\\sigma$ confidence level.\\\\ $^{(3)}$  The distances of Tab. \\ref{xlist} have been used, in particular 200 pc for TT Ari and 460 pc for LX Ser. \\\\ $^{(4)}$ Distance between best-fit X-ray position and optical position of presumed counterpart. \\\\ $^{(5)}$ The upper limit of the count rate has not been combined with the lower limit in distance. \\\\ $^{(6)}$ No distance known. \\\\ $^{(7)}$ A temperature of 0.4 keV has been assumed for the conversion of the upper limit count rates or when the number of counts was below 100. } \\label{xsurvlog} \\vspace*{-0.15cm} \\end{table*} ",
        "conclusions": "The X-ray emission of VY Scl stars during optical high states is characterized by 0.25--0.5 keV blackbody emission from a 50--120 m sized region. In the case of MV Lyr I found evidence for a drop in X-ray flux by a factor of $>$5--7 during an optical low state observation suggesting that the emission is related to the change in accretion rate (which is thought to accompany the high/low state transitions) and not to the secondary coronal emission. At this time, I cannot offer a simple explanation for this emission. I cannot exclude the possibility that with the spectral resolution of the ROSAT PSPC the X-ray continuum determination is in error, and one may be fooled by emission which is not optically thick. Future X-ray observations with higher spectral resolution are thus clearly demanded.  It seems worth investigating the properties of other nova-like cataclysmic variables which do not belong to the VY Scl class to check whether the X-ray properties found for VY Scl stars are unique to this sub-class or are related to the high accretion rate in nova-like CVs in general."
    },
    "9805/astro-ph9805101_arXiv.txt": {
        "abstract": "Geometrical and physical properties of dusty tori of Seyfert nuclei probed by the water vapor maser emission at 22 GHz are discussed. We assume that the dusty torus has a simple cylindrical form and the maser emission can be detected only when we observe the torus from almost edge-on views. The observed low frequency of occurrence of the water vapor maser emission (less than 10 percent) suggests that the torus is a vertically thin cylinder whose outer radius between a few pc and $\\sim$ 10 pc. However, the observed masing regions are concentrated in the inner 1 pc regions of the torus. This property can be explained by that only the inner a few pc regions have physical conditions enough to cause the maser emission; the temperature is as high as several hundred  K and the density is as high as $\\sim 10^{10}$ cm$^{-3}$. ",
        "introduction": "The current unified models of active galactic nuclei (AGN) have introduced a dusty torus which surrounds the central engine. Since the torus is considered to be optically very thick, the visibility of the central engine is significantly affected by viewing angles toward AGN (Antonucci \\& Miller 1985; Krolik \\& Begelman 1988;  see for a review Antonucci 1993). The dusty tori emit their energy mostly in mid- and far-infrared wavelengths. Therefore, infrared spectral energy distributions of AGN have been often utilized to study physical properties of the dusty tori [Pier \\& Krolik 1992b, 1993 (hereafter PK92 and PK93, respectively); Heckman, Chambers, \\& Postman 1992; Heckman 1995; Granato, Danese, \\& Franceschini 1997; Taniguchi et al. 1997; Murayama, Mouri, \\& Taniguchi 1998]. However, geometrical properties of the dusty tori have not yet been studied directly because  the dusty tori are too compact to be resolved spatially at mid- and far-infrared wavelengths.  On the other hand, for these several years, we have learned that the water vapor maser emission at 22 GHz can be used to probe dusty tori directly (Nakai, Inoue, \\& Miyoshi 1993; Miyoshi et al. 1995; Greenhill et al. 1995a, 1995b, 1996; Gallimore et al. 1996; Greenhill \\& Gwinn 1997). In particular, the recent VLBA or VLA measurements of the H$_2$O maser emission of the nearby AGNs, NGC 1068 (Gallimore et al. 1996; Greenhill et al. 1996; Greenhill \\& Gwinn 1997), NGC 4258 (Miyoshi et al. 1995; Greenhill et al. 1995a, 1995b), and NGC 4945 (Greenhill, Moran, \\& Herrnstein 1997a), have shown that the masing clouds are located at distances of $\\sim$ 0.1 - 1 pc from the nuclei. Since these distances are almost comparable to those of the dusty tori, it is suggested that the masing clouds reside in the tori themselves (e.g., Greenhill et al. 1996). Therefore, the H$_2$O maser emission provides a useful tool to study physical properties of dusty tori as well as dynamical ones (e.g., Murayama \\& Taniguchi 1997 and references therein). The recent H$_2$O maser searches [Braatz, Wilson, \\& Henkel 1996, 1997 (hereafter BWH97); Greenhill et al. 1997b] also provides important information.  Although the microphysics of the maser emission has been investigated in detail (Krolik \\& Lepp 1989; Neufeld, Maloney, \\& Conger 1994), the H$_2$O maser data have not yet fully utilized to study geometrical properties of the dusty tori. Therefore, in this {\\it Letter}, incorporating all the above observational results on the H$_2$O maser emission, we discuss the geometrical properties of the dusty tori. ",
        "conclusions": ""
    },
    "9805/astro-ph9805066_arXiv.txt": {
        "abstract": "An efficient method has been developed for solving the inverse problem of Doppler--Zeeman mapping of magnetic, chemically peculiar stars. A regularized iteration method is used to simultaneously solve the integral equations for the Stokes I, V, Q and U parameters. The validity of analytical fits to the local profiles of the Stokes parameters is substantiated. The algorithm had been tested on models and makes it possible to obtain simultaneously, from the observed Stokes profiles, a map of the distribution of a chemical element and the parameters of an arbitrarily shifted magnetic dipole. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805299_arXiv.txt": {
        "abstract": "Previously it was shown that gravitation theory allows the existance of supermassive stable compact configurations of  the degenerated electronic gas ( L.V.Verozub, Astr. Nacr. 317 (1996) 107 ) without events horizon. In the present paper the simplest model of such  kind of  objects in gas environment  has been considered. It is shown that at the spherically symmetric accretion onto the object the luminosity is about $10^{37} erg/s$ for the mass accretion rate of the order of $\\stackrel{\\cdot}{M}=10^{-6} M_{\\odot}/year$ . The vawelength of the radiation maximum is  about $ 400\\div 500 \\stackrel{\\circ}{A}$. There is an  ionization zone around the central objects with the radius about  $10^{-3}pc$ . ",
        "introduction": "The analysis of the observation data gives evidence for the existance of a massive ( about $2.5\\cdot10^{6} M_{\\odot}$ ) compact object in the Galactic Center \\cite{Eckart}. The observation data do not allow to make a definite conclusion about the nature of the object. For this reason it is identified , as a rule, with a supermassive black hole. Another possibility is considered in the present paper.  The  gravitation equations whose spherically symmetric solution have no and physical singularity in flat space-time from the viewpoint of a remote observer where proposed in  the paper \\cite{Verozub1} . According to the equations the events horizon is absent in the spherically symmetric solution. The radial component of the gravity force $F$ affecting a test particle with mass $m$  in the spherical coordinate system in flat space-time is given by \\begin{equation} F=-m\\left[ c^{2}B_{00}^{1} + (2 B_{01}^{0} - B_{11}^{1})v^{2}]\\right] . \\label{gravityforce} \\end{equation} Here $B_{00}^{1}$, $B_{01}^{0}$ and $B_{11}^{1}$ are the nonzero components of the strength tensor $B_{\\alpha \\beta }^{\\gamma }$ of gravity field in flat space-time :\\\\ \\begin{eqnarray} B_{00}^{1}=\\frac{1}{2}\\frac{\\alpha f' f^{4} (1-\\alpha /f)}{f^2 r^{4}},\\\\ B_{01}^{0}=2/r,\\\\ B_{11}^{1}=\\frac{1}{2} \\frac{\\alpha  f'}{f^{2} (1-\\alpha /f)},\\\\ f=(\\alpha ^{3}+r^{3}) , \\end{eqnarray} $\\alpha =2GM/c^{2}$ is the Schwarzshild radius and $v$ is the radial component of the particle velocity.  Fig 1 shows the value of $|F|$ for the particle at rest as a function of $\\overline{r}=r/\\alpha $ .  It is shown in the paper \\cite{Verozub2} that in the above theory there can exist equilibrium stable configurations of the degenerated electronic gas with masses up to $10^{9} M_{\\odot}$ or more than that. This  kind of objects there can exist in  the Galactic Center.  Fortunately, there are some consequences available for observations which can help us to identify the objects with one of the proposed hypotheses. ",
        "conclusions": "We have considered the simplest model of the compact object without events horizon in gas medium with properties of the Galactic center. It leads to some available for observations   consequences. . The observation data speak  in favour these consequences rather than  against them. Consequently, we have an alternative to the supermassive Black Hole hypothesis. We hope that later a more detail consideration of the problem  and an analysis of observations will lead to more definite conclusions."
    },
    "9805/astro-ph9805250_arXiv.txt": {
        "abstract": "$\\kappa$~Psc (HD 220825) is a typical Chromium Ap star that happens to have optimal parameters for Doppler imaging (DI). Its short rotational period of less then 2 days, rotational velocity of $\\sim$40\\,km/s, and a moderate inclination of the rotational axis put modest requirements on spectroscopic observations. Anomalies of iron peak elements are clearly present, but small enough to cause significant deviations from model atmospheres with scaled solar abundances. We applied DI to $\\kappa$~Psc once before, determining the distribution of Cr (Ryabchikova et al. \\cite{rpdp}, hereafter Paper~I). However, due to strong blending of Fe, the image was based on two short ($\\sim$2\\,\\AA ) spectral intervals, dominated by Cr lines. Since the first paper we obtained additional spectra and developed a new code that allows to perform multi-element DI and thus to use larger spectral interval(s). We demonstrate the abilities of the new code and present new maps of Cr and Fe. A much larger time base allowed us to improve the rotational period of $\\kappa$~Psc as well. ",
        "introduction": "Doppler imaging of stars (see e.g. Piskunov \\& Rice 1993) has become a standard technique for studying the distribution of chemical elements on the surface of Ap stars. Up to now DI was limited to a single chemical element. In a few attempts to image several elements, maps were obtained sequentially, each time assuming that all elements but one are distributed homogeneously. Recent improvements in computer performance and more efficient algorithms for solving radiative transfer allowed us to replace pre-computed tables of local line profiles with ``on the fly'' calculations of the emerging spectrum. The new code INVERS11, based on this approach, is not limited by the size of interpolation tables and can handle simultaneous imaging of multiple elements. In addition, we are able to use {\\it blends} of different chemical elements for DI, which was not possible with the old codes.  We applied the new code to the SrCr star $\\kappa$~Psc, classified A0p. The selected spectral region around 5300\\,\\AA\\ is dominated by lines of neutral and ionized Cr and Fe lines, which give additional constraints for the effective temperature. The rotational velocity is $\\sim 38$~km\\,s$^{-1}$ resulting in significant line blending. ",
        "conclusions": ""
    },
    "9805/astro-ph9805120_arXiv.txt": {
        "abstract": "We present MOND fits to 15 rotation curves of LSB galaxies.  Good fits are readily found, although for a few galaxies minor adjustments to the inclination are needed. Reasonable values for the stellar mass-to-light ratios are found, as well as an approximately constant value for the total (gas and stars) mass-to-light ratio. We show that the LSB galaxies investigated here lie on the one, unique Tully-Fisher relation, as predicted by MOND.  The scatter on the Tully-Fisher relation can be completely explained by the observed scatter in the total mass-to-light ratio. We address the question of whether MOND can fit any arbitrary rotation curve by constructing a plausible fake model galaxy. While MOND is unable to fit this hypothetical galaxy, a normal dark halo fit is readily found, showing that dark matter fits are much less selective in producing fits.  The good fits to rotation curves of LSB galaxies support MOND, especially as these are galaxies with large mass discrepancies deep in the MOND regime. ",
        "introduction": "The inability of the visible mass components in disk galaxies to account for the observed rotation curves is usually interpreted as evidence for the existence of an additional, invisible mass component. Other theories suggest that this mass discrepancy is an indication of a breakdown of classical Newtonian dynamics.  It is difficult to evaluate these theories, as only a few make specific and testable predictions.  One of the exceptions is the Modified Newtonian Dynamics (MOND), advocated by Milgrom (1983, 1989) and Sanders (1990, 1996).  This theory postulates that Newton's Law of Gravity should be modified for very small accelerations, with the result that any need for dark matter disappears.  Fits to rotation curves of HSB galaxies using MOND are of equal quality as the fits made using a dark matter halo (see Sanders 1996). MOND is however also able to satisfactorily explain observations of the dynamics of e.g.\\ dwarf galaxies and dwarf spheroidals (see the discussion in Milgrom 1995, and also McGaugh \\& de Blok (1998b) [hereafter Paper II]).  For a complete description of MOND, its predictions, and observational results we refer to Milgrom (1983, 1989), Sanders (1990), Begeman, Broeils and Sanders (1991), Bekenstein \\& Milgrom (1984) and Sanders (1996).  An extensive description of MOND results in the context of LSB galaxies is given in Paper II.  MOND assumes that the force law changes from the conventional Newtonian form when the acceleration of a test particle is much smaller than a limiting acceleration $a_0$, where $a_0$ is a universal constant.  Thus, while the normal Newtonian acceleration $g_N = GM/r^2$ which a mass $M$ exerts on a test particle at distance $r$ is identical to the true test-particle acceleration $g$ for accelerations $g \\gg a_0$, in the MOND limit (i.e., $g \\ll a_0$) the implied Newtonian acceleration is related to the true test-particle acceleration $g$ by $g_N = g^2/a_0$.  The acceleration $a_0$ is a fundamental parameter in the MOND theory. From rotation curve fitting to high-quality rotation curves, Begeman et al.\\ (1991) determined a value of $1.2 \\times 10^{-10}$ m s$^{-2}$ (for $H_0 = 75$ km s$^{-1}$ Mpc$^{-1}$, which we adopt throughout this paper).  As described in Paper II, LSB galaxies provide a strong test of MOND. Their low surface densities imply accelerations $g < a_0$, which means that these galaxies should be almost completely in the MOND regime. Milgrom (1983, 1989) made a number of testable predictions on the shapes of rotation curves, and noted that low surface density galaxies should have slowly rising rotation curves.  This expectation of MOND is confirmed by the observed rotation curves.  In Newtonian terms this translates in these galaxies having large mass discrepancies (McGaugh \\& de Blok 1998, hereafter Paper I).  This brings us to one of the more pronounced differences between MOND and classical Newtonian dynamics, which is the explanation of the Tully-Fisher (TF) relation. As is described in detail in Paper I (see also Zwaan et al. 1995), the fact that LSB galaxies are observed to obey the same TF relation as HSB galaxies implies a strong coupling between the central surface brightnesses of the disks of galaxies and their total mass-to-light ratios (which include dark matter). Assuming standard Newtonian dynamics this implies that LSB galaxies have a higher total mass (within the disk radius) than HSB galaxies of the same asymptotic velocity.  It is hard to derive this result in the standard context without a lot of fine-tuning.  MOND {\\it predicts} that all galaxies should fall on one {\\it mass}-velocity relation, which takes the form $V_{\\infty}^4 = MGa_0$, where $V_{\\infty}$ is the asymptotic velocity and $M$ is the total mass of the galaxy (that is, the mass of stars and gas).  Once the value of $a_0$ is fixed, this relation becomes absolute and can be tested and falsified.  We use the rotation curves of 15 LSB galaxies to do a MOND analysis. Section~2 describes the fitting procedure. Section~3 presents the results. In Sect.~4 we discuss whether MOND can fit any rotation curve, and we present our conclusions in Sect.~5. ",
        "conclusions": "The good fits to LSB galaxies rotation curves support MOND, especially so as these are galaxies with large mass discrepancies in their observed disks, where the MOND effects are strongest.  The MOND fits furthermore {\\it test} the MOND theory.  This is not the case with dark matter fits, where the properties of the dark matter discrepancy are {\\it derived} from the fits. These fits {\\it define} the properties of the mass discrepancy. It is therefore not possible to falsify nor confirm the dark matter hypothesis from rotation curves alone, as can be done with the MOND hypothesis. We show this by fitting a hypothetical non-physical galaxy: the MOND theory is unable to produce a good fit (which in this case is a good thing), whereas we can get a good fit with the dark matter theory (showing it is more flexible in fitting non-physical models).  In principle, there are an infinite number of things the rotation curves of galaxies could do given the presence of an invisible mass component.  There are many plausible predictions for what they are expected to do given various ideas about dark matter.  In the case of MOND, there is one and only one thing rotation curves can do. This is precisely what they do.  One last remarkable result is the great efficiency with which MOND can describe the rotation curves of galaxies of very different types. Even if MOND is ultimately falsified, it would still be worth knowing why it works so well.  Until a MOND cosmology can be derived (see Sanders 1998) it remains a recipe for describing rotation curves, but as a recipe it is far superior to the dark halo recipe, and this raises the question of {\\it why} MOND works so well.  MOND is after all only a simple analytic formula which gets $V(R)$ correct based solely on the luminous mass.  If the MOND phenomenology arises as the result of dark matter, then this would imply that the dark matter must ``know'' about both the distribution of light {\\it and} the MOND formula, and arrange itself appropriately, right down to amplifying the bumps and wiggles in the stellar and gas rotation curves by the appropriate amount."
    },
    "9805/astro-ph9805316_arXiv.txt": {
        "abstract": "We quantify the mean asymmetry of 54 face-on, early type disk galaxies (S0 to Sab) using the amplitude of the $m=1$ azimuthal Fourier component of the R-band surface brightness.  We find that the median lopsidedness, $\\langle A_1/A_0\\rangle$, of our sample is 0.11 and that the most lopsided $20\\%$ of our galaxies have $\\langle A_1/A_0\\rangle\\geq0.19$.  Asymmetries in early type disks appear to be of similar frequency and strength as in late type disk galaxies (Zaritsky and Rix 1997.)  We have observed our early type disks in a bandpass (R-Band) in which the light is dominated by stars with ages greater than $10^9$ yrs, and therefore are seeing azimuthal asymmetries in the stellar {\\it mass} distribution.  The similar degree of lopsidedness seen in disks of very different star formation rates indicates that the lopsidedness in all galactic disks is primarily due to azimuthal mass asymmetries.  Hence, $20\\%$ of all disk galaxies (regardless of Hubble Type) have azimuthal asymmetries, $\\langle A_1/A_0\\rangle\\geq0.19$, in their stellar disk mass distribution, confirming lopsidedness as a dynamical phenomenon. ",
        "introduction": "Just as a human face may mirror traumatic events in its past, so may a galaxy's structure reflect its dynamical history.  In this paper we explore one potential probe of the dynamical past of disk galaxies, their {\\it lopsidedness}.  In the context of this paper, lopsidedness is defined as a significant bulk asymmetry in the stellar mass distribution of a galactic disk.  For nearly face-on galaxies, these asymmetries manifest themselves observationally as a shifting of the outer isophotes of the stellar light.  Baldwin et al. (1980) were the first to point out that the HI profiles of spiral galaxies are frequently lopsided, and suggested that this lopsidedness may stem from weak interactions in the galaxy's past, or from lopsided orbits.  Richter and Sancisi (1994) using a much larger galaxy sample to study the frequency of lopsidedness, concluded that over 50$\\%$ of all disk galaxies had significant asymmetries in their HI profiles.  These HI asymmetries are present both in the projected flux distributions and in velocity space.  A systematic attempt to examine the frequency of asymmetries in the stellar light of galaxies has been made by Rix and Zaritsky, 1995 and Zaritsky and Rix, 1997 (hereafter RZ95 and ZR97) and a recent optical study of symmetry in nearby disk galaxies has been carried out by Conselice (1998.)  Using near-IR photometry of face-on galaxies, ZR97 found that about one fifth of all late type spiral galaxies exhibited significant lopsidedness.  They defined lopsidedness via the amplitude of the $m=1$ Fourier component of the azimuthal decomposition of the galaxy's surface brightness.  It is possible to construct self-consistent asymmetric disk models (Syer and Tremaine, 1996) and some N-body work has shown that lopsided instabilities may occur in the presence of retrograde orbits (Zang and Hohl, 1978; Sellwood and Merritt, 1994) or in the absence of a massive halo (Sellwood, 1985).  These authors also find that co-rotating galaxies with massive halos are stable against intrinsic $m=1$ instabilities.  Since intrinsic disk instabilities in spirals appear not to be the dominant cause of lopsidedness, external potential perturbations are the most plausible candidate mechanism.  N-body simulations (Walker et al. 1996; ZR97) have shown that the merger of a small galaxy with a large disk galaxy can produce the type and degree of asymmetries found in RZ95 and ZR97, if their mass ratio is $\\sim 1/10$.  Recent N-body work by Levine and Sparke (1998) shows that lopsided modes may persist in disk galaxies if the disks are off-center with respect to a massive host halo with a large, flat core.  They propose that a galaxy may become lopsided if it accretes enough matter to push it away from the center of the halo.  It stands to reason that these asymmetries may also be produced by weak, non-cataclysmic encounters with other galaxies (Barnes, Hernquist 1992 and references therein.) Lifetimes of these asymmetries can be estimated from phase mixing (Baldwin et al. 1980; RZ95) or through analysis of N-Body simulations (Walker et al. 1996; ZR97); both estimates point to lifetimes of 5-10 rotation periods, or $1\\sim$Gyr.  For interpreting lopsidedness it is crucial to know whether the observable asymmetries in the stellar {\\it light} reflect asymmetries in the stellar {\\it mass}.  Late type disk galaxies (such as those used in RZ95 and ZR97) with relatively high star formation rates contain many young, bright stars.  As we will show later, in Sbc galaxies, up to 25$\\%$ of the light in the light in the R-Band can come from stars younger than 0.1 Gyrs (OB stars.)  An asymmetric distribution of such a luminous, young population could result in a high observed asymmetry, even if the underlying mass structure were in fact close to symmetric.  To avoid this potential pitfall, we assembled a sample of disk galaxies with low star formation rates, those limited to Hubble Types from from S0 to Sab (Kennicutt et al. 1994), with the aim of comparing the incidence of lopsidedness in this new sample to that in RZ95,ZR97.  If significant asymmetries are similarly frequent, we may conclude that they are in fact dynamical in origin and are not merely azimuthal variations in the mass-to-light ratios within the galaxies.  The layout of the paper is as follows.  In \\S2 we will discuss the sample selection, observations and reduction techniques; In \\S3 we will describe the data analysis methods.  The results, including the frequency of lopsidedness, are presented and discussed in \\S4.  In \\S5 we will present our conclusions as well as the directions for future follow-up work. ",
        "conclusions": "We have determined the median $\\langle \\tilde{A}_1\\rangle$ in our sample of early type disk galaxies to be $0.11$.  We have also determined that $20\\%$ of the galaxies in our sample have $\\langle \\tilde{A}_1\\rangle \\geq 0.19$.  By looking in the R-band and by selecting galaxies with typically low star formation rates, we have picked a compromise between efficiency and the minimization of the contributions from young (i.e. non phase-mixed), bright stars and hence we are primarily viewing actual azimuthal variations in the stellar mass distributions of the early type disks in our sample.  By comparing the value of $\\langle \\tilde{A}_1\\rangle$ which is lower than that for $50\\%$ and $20\\%$ of our galaxy sample, to that measured for late type spirals by ZR97 (0.13 and 0.19 respectively), we can address the issue of whether the asymmetries seen in ZR97 are also caused by variations in the stellar mass distribution of late type disk galaxies.  If asymmetric star formation were important in creating asymmetric light distributions, we would expect that late type disk galaxies would have a higher incidence of lopsidedness than early type spirals.  Since this is not the case, we are led to conclude that lopsidedness in disk galaxies of all types reflects primarily variations in the stellar mass distributions.  Our data then show that one-fifth of {\\it all} disk galaxies have $\\langle \\tilde{A}_1\\rangle\\geq0.19$.  The volume correction discussed in section 4.2 affects our measurement of the frequency of lopsidedness and the possibility of systematic brightening of lopsided galaxies via an increased star formation rate needs to be explored in more detail.  Comparison of the recent star formation history of lopsided galaxies to that of symmetric galaxies is needed to study this effect, and is also needed to help better establish a cause of lopsidedness.  A study of this nature is being carried out this spring by the authors.  Further N-Body simulations also need to be carried out to determine if the types and magnitudes of asymmetries seen in our samples can be reproduced by weak interactions as well as by minor mergers.  It is also necessary to better determine if isolated phenomena (e.g. instabilities) result in asymmetries similar in morphology and magnitude to those we have measured."
    },
    "9805/astro-ph9805193_arXiv.txt": {
        "abstract": "We present the first optical polarimetric measure\\-ments of RX~J1141.3-6410 which confirm that star as a polar. The circular polarization varies between 0 and 13\\% with the orbital period. H$\\alpha$ spectroscopy shows that this line is formed by, at least, two components: a broad and a narrow one. The phase of maximum redshift of the broad component is shifted by 0.5 with the phase of maximum circular polarization which is not usual for this class of stars.  We suggest a geometrical configuration for the system which could explain the main features of the polarimetric and spectroscopic data. ",
        "introduction": "Cataclysmic variables (CVs) are binaries consisting of a red main sequence star and a white dwarf (primary). The secondary fills its Roche lobe and mass is transfered to the white dwarf (WD). This process usually forms an accretion disk. However, some CVs have a magnetic field intense enough to prevent the disk formation. In this case, the matter falls onto the WD near the magnetic pole forming an accretion column.  Another important consequence of the high strength of the magnetic field is the synchronization of the white dwarf rotation with the orbital movement.  These stars are denominated polars and their prototype is AM Her. Reviews can be found in Cropper (1990) and Warner (1995).  The polars can be distinguished from intermediate polars (non-synchronized magnetic systems) by two important features: high circular polarization and strong soft X-ray emission. The former is caused by cyclotron emission in the column. The emission at high frequencies is produced by the shock formed close to the WD surface.  The emission lines in polars are thought to be formed along the trajectory of the material from the secondary to the white dwarf (see review by Mukai 1988). This material leaves the secondary at the inner Lagrangian point ($L_1$) and follows its ballistic trajectory in the orbital plane (horizontal stream) down to the coupling region. In this region the magnetic pressure becomes higher than the ram pressure and the material starts to follow the magnetic lines (accretion stream). The radiation produced near the white dwarf can be reprocessed on the secondary surface contributing to the emission lines.  The emission lines of polars can be formed by up to four components (Rosen et al. 1987, e.g.). In general, at least two components are seen: a broad base component and a narrow peak component.  The accretion stream is responsible for the broad component.  The narrow one is formed nearer the secondary. In some systems, this emission seems to be formed on the secondary surface itself (Liebert \\& Stockman 1985). In others, there is evidence that the horizontal stream produces such emission (Mukai 1988).  Recently, Doppler tomography of polars has improved the understanding of the emission lines (Diaz \\& Stei\\-ner 1994; Schwope et al. 1997; \\v{S}imi\\'c et al. 1998).  An important fraction of the emission seems to be produced near the secondary.  The bulk of the broad component is formed near the coupling region.  These maps do not show an important emission from the region of high velocities near the white dwarf.  Many sources identified by the ROSAT satellite have been shown to be CVs and, more specifically, polars.  Motch et al.  (1996) discovered 7 new CVs and suggested that two of them could be synchronized magnetic systems based on the strength of their emission lines. RX~J1141.3-6410 is one of these systems.  It is associated with a $\\approx$16.5~mag star having a strong \\ion{He}{ii} $\\lambda$4686 emission line. Recently, Cieslinski \\& Steiner (1997) have found a photometric period (P = 0.131\\,517~d) and a light curve consistent with the suggestion of RX~J1141.3-6410 being a polar. However, until now, no polarimetric measurement has been made in order to confirm its magnetic nature.  In this work, we present our optical polarimetric measurements and time-resolved spectral data in the region of H$\\alpha$ for RX~J1141.3-6410.  Some modeling of the intensity and polarization has been made. We suggest a possible geometrical configuration for RX~J1141.3-6410 and the main regions of line formation based on the polarization models and the spectroscopic data. ",
        "conclusions": "The high level of circular polarization observed in RX~J1141.3-6410 confirms this star as a polar. Within our data precision, no peak in linear polarization was observed. The ${\\rm H}\\alpha$ line of RX~J1141.3-6410 shows two components. The maximum blueshift of the broader component is locked with the maximum circular polarization, contrary to most polars.  We suggest that the system is seen near face on and that the magnetic field axis lies near the orbital plane. This configuration is able to explain the main features of RX~J1141.3-6410 data. The narrow component seems to be produced in the horizontal stream, while the broad component may have its origin in the coupling region."
    },
    "9805/astro-ph9805008_arXiv.txt": {
        "abstract": "We present a complete  atlas of the Cygnus Loop supernova remnant in the light of \\o3 ($\\lambda 5007$), \\ha, and \\s2 ($\\lambda\\lambda 6717, 6731$).  We include low-resolution ($25\\arcsec$) global maps and smaller fields at $6\\arcsec$  resolution from observations using the Prime Focus Corrector on the 0.8-m telescope at McDonald Observatory. Despite its shell-like appearance, the Cygnus Loop is not a current example of a Sedov-Taylor blast wave. Rather,  the optical emission traces interactions of the supernova blast wave with clumps of gas. The surrounding interstellar medium forms the walls of a cavity through which the blast wave now propagates, including a nearly complete shell in which non-radiative filaments are detected. We identify non-radiative shocks around half the perimeter of the Cygnus Loop, and they trace a circle of radius $R = 1\\fdg 4$ (19 pc) in the spherical cavity walls. The Cygnus Loop blast wave is not breaking out of a dense cloud, but is instead running into confining walls. Modification of the shock velocity and gas temperature due to interaction of the blast wave with the surrounding medium introduces errors in estimates of the age of this supernova remnant. The optical emission of radiative shocks arises only where the blast wave encounters inhomogeneities in the ambient medium; it is not a consequence of gradual evolution to a global radiative phase. Distance measurements that rely on this uniform blast wave evolution are uncertain, but the radiative shocks can be used as distance indicators because of the spherical symmetry of the surrounding medium. The interstellar medium dominates not only the appearance of the Cygnus Loop but also the continued evolution of the blast wave. If this is a typical example of a supernova remnant, then global models of the interstellar medium must account for such significant blast wave deceleration. ",
        "introduction": "} Supernova remnants greatly determine the large-scale structure of the interstellar medium. The energy of supernova remnants heats and ionizes the interstellar medium (ISM), and their blast waves govern mass exchange between various phases of the ISM.  In doing so, supernova remnants (SNRs) influence subsequent star formation and the recycling of heavy elements in galaxies.  Global models of the interstellar medium that include a hot ionized component (\\cite{Cox74}; \\cite{McK77}) are sensitive to the supernova rate, the persistence of their remnants, and the sizes they attain. A simple calculation of the last of these assumes that the blast wave expands adiabatically in  a uniform medium once the blast wave has swept up mass comparable to the mass of the ejecta. During this Sedov-Taylor phase, the radius of the SNR as a function of $E_{51}$, the initial energy in units of $10^{51}$ erg, $n_o$, the ambient number density in units of ${\\rm cm^{-3}}$, and $t_4$, time in units of $10^4$ yr, is $R=13 (E_{51}/n_o)^{1/5} t_4^{2/5} {\\rm \\,pc}$ in a medium where the mean mass per particle is $2.0 \\times 10^{-24} {\\rm \\, g}$. This phase will last until radiative losses become important.  The beginning of this subsequent phase, marked by the initial loss of pressure behind the blast wave, occurs at $t=1.9\\times 10^4 E_{51}^{3/14} n_o^{-4/7} {\\rm \\,yr}$, when the radius is $R=16.2 E_{51}^{2/7} n_o^{-3/7} {\\rm \\,pc}$ (\\cite{Shu87}), although the radiating shell is not fully formed yet.  We approach these large-scale questions with analysis of complete images of a particular supernova remnant, the Cygnus Loop, in three optical emission lines. This supernova remnant appears to be a limb-brightened shell at radio (\\cite{Keen73}), infrared (\\cite{Bra86}), optical (\\cite{Fes82}), and X-ray (\\cite{Ku84}; \\cite{Lev97}) energies, which at first glance suggests that it is presently in the transition to the radiative stage. The Cygnus Loop has the advantages of being nearby, bright, and relatively unobscured by dust.  This allows us to examine in detail the evolution of various portions of the shock front and to determine physical parameters, such as shock velocity and local ambient density, as they vary throughout the remnant. Despite its appearance, the Cygnus Loop is not a current example of blast wave propagation in a uniform medium at any stage.  Instead, its evolution is governed by the inhomogeneous interstellar medium, which we map using the shock as a probe.  Many of the features we discuss have been noted by others.  Oort (1946) \\nocite{Oort46} first suggested that the Cygnus Loop is an expanding supernova shell. Spectroscopy of radiative shocks in selected locations (e.g., \\cite{Mil74}, \\cite{Ray80a}, and Fesen et al. 1982\\nocite{Fes82}) combined with theoretical models of these shocks (e.g., \\cite{Cox72}, \\cite{Dop77}, \\cite{Ray79}, and \\cite{Shu79}) has been used to derive the physical conditions of the observed shocks. We utilize the radial velocity measures of Minkowski (1958)\\nocite{Min58}, Kirshner \\&\\ Taylor (1976)\\nocite{Kir76}, Greidanus \\&\\ Strom (1991)\\nocite{Gre91}, and Shull \\&\\ Hippelein (1991)\\nocite{Shu91} to discern some of the three-dimensional structure that is ambiguous from the data we present. Many non-radiative or Balmer-dominated shocks in the Cygnus Loop have been identified (e.g., \\cite{Kir76}, \\cite{Ray80b}, \\cite{Tref81}, Fesen et al. 1982\\nocite{Fes82}, \\cite{Fes92}, and \\cite{Han92}). Our observations qualitatively match these, and we rely on these works and others (\\cite{Ray83}; \\cite{Long92}; \\cite{Hes94}) for quantitative measures of parameters such as shock velocity and preshock density. McCray \\&\\ Snow (1979)\\nocite{McC79} and Charles, Kahn, \\&\\ McKee (1985)\\nocite{Cha85} have suggested that the Cygnus Loop is the result of a cavity explosion, and we adapt this global model to interpret the surrounding interstellar medium, as well.  This paper is a companion to the soft X-ray survey presently in progress with the {\\it ROSAT} High Resolution Imager (\\cite{Gra96}; \\cite{Lev97}). With these two surveys, we examine the Cygnus Loop as a whole, not restricting our investigation only to those regions that are exceptionally bright or that appear to be particularly interesting. We hope to understand both the global processes and the specific variations that are responsible for the emission we detect.  The X-rays probe hot (temperature $T\\sim 10^6$ K) gas that shocks with velocities $v_s \\sim 400 \\kms$ heat.  The optical emission is expected from slower shocks ($v_s \\lesssim 200 \\kms$) in which  the post-shock region cools to temperatures $T\\sim 10^4$ K, yet the most prominent regions at optical wavelengths are also bright in X-rays. McKee \\&\\ Cowie (1975) \\nocite{McK75} suggested that the broad correlation of X-ray and optical emission is the result of a blast wave propagating in an inhomogeneous medium.  In this scenario, the shock is significantly decelerated in dense clumps of gas, while portions of it proceed unimpeded through the lower-density intercloud medium. We apply the principles of this basic cloud--blast-wave interaction to a range of locations in the Cygnus Loop. In particular, we refine the cavity model introduced in Levenson et al. (1997)\\nocite{Lev97}, using these optical data to constrain the current ISM in the vicinity of the Cygnus Loop and to determine how the stellar progenitor modified it in the past.  We present the observations in \\S 2. We describe them in detail, noting individual regions of interest, and we use these data to measure the physical conditions of the blast wave and the ambient medium in particular locations in \\S 3.  The purpose of the detailed examination is to combine the results in a complete map of the surrounding ISM.  We present this three-dimensional model while providing a coherent explanation of the history that accounts for it in \\S 4. We predict the future of this SNR and relate its fate to more general theories of supernova modification of the interstellar medium in \\S5 and summarize our conclusions in \\S 6. ",
        "conclusions": "This optical emission line atlas of the Cygnus Loop supernova remnant provides the information to render a portrait of the surrounding ISM.  The ambient medium, which the massive stellar progenitor shaped as it evolved, consists of many large, dense regions. These will significantly affect the subsequent development of the blast wave. The blast wave will decelerate, and it will no longer have sufficient velocity to excite the material through which it propagates to X-ray-emitting temperatures.  Although this is a study of a particular object, it has far-reaching consequences for the ISM as a whole.  Supernova remnants provide the energy to heat the ISM, affect the velocity dispersion of interstellar clouds, and set the stage for future generations of star formation.  Thus, any global model of the gas in the Galaxy critically depends on the evolution of SNR blast waves in their inhomogeneous environments, of which this example is typical. While optical emission-line characteristics identify supernova remnants, they may preferentially select  those whose evolution the extant ISM determines.  Other SNRs need to be examined over a broad range of energies in the same careful way to draw more certain conclusions about their net effect on the interstellar medium."
    },
    "9805/astro-ph9805187_arXiv.txt": {
        "abstract": "Since 1992, observations of roAp stars have been carried out using the dual-channel photometer attached to the 0.8m telescope, which is situated in Central Asia, at the Mt. Dushak-Erekdag station of Odessa Astronomical Observatory. Some results of observations of $\\gamma$ Equ and of HD 134214 are presented. 5 stars were investigated as roAp candidates. The Fourier spectra of 4 stars did not show any variability in the high-frequency region. The Fourier spectrum of HD 99563 revealed a peak at a frequency f=128.9 c/d and with a semi-amplitude of 3.98 mmag. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805302_arXiv.txt": {
        "abstract": "Temporal analyses of the prompt gamma-ray and X-ray light curves of gamma-ray bursts reveal a tendency for the burst pulse time scales to increase with decreasing energy.  For an ensemble of BATSE bursts, Fenimore et al.\\ (1995) show that the energy dependence of burst peak durations can be represented by $\\Delta t \\propto E^{-\\gamma}$ with $\\gamma \\simeq 0.4$--$0.45$.  This power-law dependence has led to the suggestion that this effect is due to radiative processes, most notably synchrotron cooling of the non-thermal particles which produce the radiation.  Here we show that a similar power-law dependence occurs, under certain assumptions, in the context of the blast-wave model and is a consequence of the deceleration of the blast-wave. This effect will obtain whether or not synchrotron cooling is important, but different degrees of cooling will cause variations in the energy dependence of the peak durations. ",
        "introduction": "Since their discovery in the late 1960s, gamma-ray bursts (GRBs) have remained enigmatic despite the fact that thousands of bursts have been detected by various instruments over the last $\\sim 30$ years.  The recent X-ray, optical and radio afterglow observations of bursts detected by the BeppoSAX satellite have enabled significant advances in our understanding of these objects.  Redshift measurements associated with the afterglows of bursts GRB~970508 (Metzger et al.\\ 1997), GRB~971214 (Kulkarni et al.\\ 1998), and GRB~980425 (Tinney et al.\\ 1998) provide convincing evidence that GRBs are extragalactic and may be as distant as $z = 3.4$ or as nearby as $z = 0.0085$.  The temporal decay of the afterglow emission is roughly consistent with the simplest fireball/blast-wave interpretations (Wijers, Rees \\& M\\'esz\\'aros 1997; Waxman 1997), and these models provide a specific theoretical context in which to understand the wealth of burst data which existed prior to BeppoSAX and which is still largely unexplained.  It has been argued that the spectral shapes of a substantial number of bursts, particularly those which have detailed spectra from the Burst and Transient Source Experiment (BATSE) and the other {\\it Compton Observatory}\\ instruments, are due to synchrotron emission from a shock accelerated distribution of non-thermal electrons (e.g., Tavani 1996).  Furthermore, the spectra of X-ray, optical and radio afterglow emission also appear to be due to synchrotron radiation and exhibit characteristic signatures such as power-law behavior consistent with optically thin synchrotron emission in the optical and X-ray bands (Djorgovski et al.\\ 1997; Frontera et al.\\ 1998; Galama et al.\\ 1998), self-absorption in the radio band (Katz \\& Piran 1997; Frail et al.\\ 1997), and spectral index changes of $\\Delta \\alpha = 0.5$ in the optical, indicative of synchrotron cooling (Galama et al.\\ 1998).  However, even in the context of a specific dynamical and emission model, the varied complexity of the spectral and temporal properties of prompt GRB light curves remains an outstanding problem.  In this work, we address one aspect of this problem: the tendency, for a given GRB, for longer burst peak durations at lower observed energy bands. We examine this issue in terms of the blast wave model and demonstrate how a similar tendency arises due to blast wave deceleration and how different degrees of synchrotron cooling can modify the basic effect to produce a range of behavior.  In the remainder of this paper, we give a brief summary of the relevant aspects of the observed energy dependence of burst light curves (\\S~2), describe the dynamics and emission properties of the basic blast-wave model and how they relate to this effect (\\S~3), and finally, discuss some of the strengths and weaknesses of this interpretation and suggest some avenues for further investigation (\\S~4). ",
        "conclusions": "Given our assumptions, the effect of pulse broadening at lower energies will obtain for any blast wave deceleration model, whether the prompt burst emission is due to internal shocks or external shocks, as we have described here.  Furthermore, the energy dependence of the pulse widths will vary depending on the degree of synchrotron cooling, and we find that the range of this energy dependence ($E^{-0.4}$--$E^{-0.66}$) is suggestively close to the results found by Fenimore et al.\\ (1995).  This simple picture, however, cannot explain the pre-activity or precursor behavior described by Fenimore (1998).  In addition, especially for strong cooling, the pulse shapes will be less sharply peaked at lower energies, which implies that the autocorrelation function may not have the universal shape described by Fenimore et al.\\ (1995).  A substantially more complex blast wave model than the one we have presented here is certainly required to describe GRBs, and any additional complexities may either mitigate or worsen the above discrepancies with the observations.  At a minimum, the electron and magnetic field equipartition parameters are not constant throughout the blast wave evolution as we have assumed.  As the blast wave evolves, the expansion time scale also grows with increasing radius. There would then be more time for equipartition to obtain, and one might expect the electron and magnetic field components to come closer to true equipartition with the swept-up protons at later times.  Such an effect may account for the softer pre-activity phase since $E_\\peak$ will increase as $\\xi_e$ and $\\xi_B$ increase until the temporal dependence implied by the deceleration and synchrotron losses begin to dominate and the effects we have described will then control the light curve evolution.  The effects of evolving equipartition parameters in more complex blast wave model calculations can certainly be studied, but one may suspect that only rather contrived electron energy and magnetic field evolution will reproduce the behavior described by Fenimore (1998).  In this case, the ultimate answers may lie in doing the microphysics: magnetic field generation and particle acceleration, in other words, turbulent relativistic magneto-hydrodynamics.  On the observational side, at least three additional investigations should be conducted in order to help guide the theoretical modeling. First, a cross-correlation or similar analysis of the prompt burst light curves in the various energy bands should be performed in order to quantify the energy-dependent delay of the pulse peaks.  Second, in addition to computing properties averaged over an ensemble of bursts, the {\\em distributions} of the energy dependences of the pulse widths and peak delays should also be computed.  This information will determine the range of parameter space for which any given model must account.  Finally, insofar as signal-to-noise limitations prevent time-resolved spectra, the spectral properties integrated over the prompt burst phase should be analyzed for each individual burst. Specifically, if there is evidence for synchrotron cooling in the spectrum, which will be signified by a $\\nu^{1/2}$-dependence in the energy bands of interest below the $\\nu F_\\nu$ peak, then synchrotron cooling may well be important, and in the context of the blast wave model, such spectral features should be correlated with the energy dependence of the pulse durations."
    },
    "9805/astro-ph9805134_arXiv.txt": {
        "abstract": "The detection of resonance absorption lines against known objects such as individual galaxies and clusters of galaxies is a powerful approach for studying the gas content of these systems.  We describe an efficient method of identifying background quasars suitable for absorption line studies.  In this finding technique, we identify serendipitous X-ray sources, about 1/8 of which are suitably bright quasars at moderate redshift. We identify 16 new quasars and galaxies with active galactic nuclei (AGNs), and confirm 5 known quasars and AGNs, superimposed behind elliptical galaxies and clusters of galaxies.  We also present 3 QSO/AGN candidates with uncertain redshift identifications. ",
        "introduction": "Our present understanding of the gaseous content of elliptical galaxies and of clusters of galaxies results largely from studies of the emission of photons from ionized plasmas and excited atoms. Studies of emission require that gas of particular temperature have an adequate emission measure, so it is insensitive to even large amounts of material for which the species are in their ground state.  However, weakly excited gas can produce absorption lines against background continuum sources, so this gas can be detected, and with considerable sensitivity.  In order to search for the absorbing gas, we have developed a strategy to identify background continuum sources, mostly quasars, behind elliptical galaxies and clusters of galaxies. The identification of this absorbing material not only improves our census of the interstellar content of these systems, it also tests several astrophysical models for its origin.  A variety of astrophysical events can occur in ellipticals and in clusters of galaxies that are unlikely to have a detectable emission signature but may lead to detectable absorption lines.  For example, cooling and cooled gas is anticipated from the radiative losses (the X-rays) of the hot interstellar gas in these systems (e.\\ g., from cooling flows, Fabian, Nulsen, and Canizares 1991; see also reviews by Fabbiano 1989 and Sarazin 1990). Two other sources for absorbing gas in elliptical galaxies are mass shed from evolving stars and accretion of gas onto the galaxy (Mathews and Baker 1971, Bregman 1978, White and Chevalier 1983). Important sources of absorbing gas in clusters of galaxies are from gas stripped out of galaxies (Gaetz, Salpeter, and Shaviv 1987) and infall of material into the cluster (Metzler 1995).  The properties of the absorbing gas are often different for the various processes, so the detection of absorption lines can potentially identify the important processes in these systems.  Although many quasars are known, galaxies and clusters subtend a small solid angle on the sky, so there are relatively few coincidences with quasars.  (Given the redshift range that we are searching for backgound objects, {z $\\sim$ 0.1--1, we choose to use the terminology ``quasar'' in this paper, rather than active galactic nuclei (AGN) to designate emission line candidates.) Furthermore, most searches for quasars and active galactic nuclei avoided the region near bright galaxies. Therefore, there is a need to identify a reasonable number of new quasars near ellipticals and clusters of galaxies.  These quasars should be bright enough to permit absorption line observations to be made with present telescopes, such as the Hubble Space Telescope (HST), in a reasonable amount of time (most of the quasars identified were scheduled for HST observations).  Here we describe an effective technique for finding background quasars that makes efficient use of telescope time. ",
        "conclusions": "We have identified a successful process for locating quasar candidates behind individual galaxies and clusters of galaxies. Using X-ray images from archival data, we search for QSO candidates which are then identified via optical spectroscopy.  We have presented 16 new quasars and galaxies with active galactic nuclei (AGNs), and confirmed 5 known quasars and AGNs.  We also presented 3 QSO/AGN candidates with uncertain redshift identifications. We find that approximately 1/8 of X-ray identified sources prove to be QSOs with the correct redshift, spectral shape, and magnitude to be used in HST studies of the instellar and intergalactic media of galaxies and clusters of galaxies."
    },
    "9805/astro-ph9805244_arXiv.txt": {
        "abstract": "The surface distribution of five elements: $iron$, $chromium$, $titanium$, $magnesium$ and $manganese$ on the magnetic A0pCr star $\\epsilon$ UMa, have been calculated using the Doppler imaging technique. We found that $iron$, $chromium$ and $manganese$ are correlated with the assumed dipole magnetic field geometry of this star, which is apparently not the case of $magnesium$ and $titanium$. ",
        "introduction": "The scientific goal of applying the Doppler imaging technique to Ap stars, is to provide observational constraints on the diffusion mechanism in the presence of a global magnetic field. $\\epsilon$ UMa (HD 112185, HR 4905), an A0pCr star, is known as the brightest member (V=1.77) of the class of peculiar A type stars. Bohlender and Landstreet (1990) measured a weak, reversing magnetic field for $\\epsilon$ UMa, that appears to be dominated by a dipole component with a polar magnetic field strength in the order of 400 Gauss.  Furthermore, maps of $chromium$, $iron$ (Rice \\& Wehlau, 1997), $oxygen$ and $calcium$ (Babel et al., 1995) have been published, whereby the distribution of each of these elements appears to be correlated with the assumed dipole magnetic field geometry.  \\section {Observations}  Observations of $\\epsilon$ UMa were done in June 1994 and in March 1995 at the Observatoire de Haute-Provence using the spectrograph AUR\\'ELIE (attached to the 1.52-m telescope) in two spectral regions: 4060 - 4260 \\AA \\ and 4440 - 4640 \\AA. The spectral resolution is about 20000 and the Signal-to-Noise ratio above 150. ",
        "conclusions": "The surface abundance distributions of the five elements we treated can be divided into two groups.  The $iron$ and $chromium$ (Figure 1) distributions show a clear depleted band which coincides with the assumed magnetic equator, confirming the results of Rice \\& Wehlau.  The $manganese$ distribution is very similar to that of these elements, which accumulate near the magnetic poles. They are all slightly overabundant compared to solar values: $manganese$ and $chromium$ are about 0.8 dex above solar values, while $iron$ is 1 dex above.  \\begin{figure} \\centerline{\\psfig{figure=P22f1.ps,height=7cm}} \\caption{The chromium abundance distribution of $\\epsilon$ UMa was obtained from the Cr\\,{\\sc ii}, 4558 \\AA\\ line. This element appears to be on average, about 0.8 dex more abundant than in the Sun.} \\end{figure}  However, the $titanium$ (Figure 2) and $magnesium$ surface structures have much less contrast in terms of peak-to-peak abundances and are apparently not significantly correlated with the magnetic dipole geometry. So far, seven different elements have been mapped for $\\epsilon$ UMa.  Together with the $oxygen$ and $calcium$ maps published by Babel et al. (1995), which reveal abundance enhancements located at the magnetic equator, the maps of $iron$, $chromium$, $magnesium$, $manganese$ and $titanium$ provide important constraints for building models of diffusion in the presence of a global magnetic field. This should provide a better understanding of the hydrodynamics in the atmospheres of Ap stars.  \\begin{figure} \\centerline{\\psfig{figure=P22f2.ps,height=7cm}} \\caption{Titanium abundance distribution of $\\epsilon$ UMa. The Ti\\,{\\sc ii}, 4163 \\AA\\ line was used for the inversion procedure. Titanium is on average slightly depleted on the surface of this star compared to solar abundance.} \\end{figure}"
    },
    "9805/astro-ph9805072_arXiv.txt": {
        "abstract": "Some CP stars have recently been discovered by Catalano et al. (1991) to be variable also in the near infrared, although with smaller amplitudes than in the visible. Hence an observational campaign was started in which the infrared light variability of a number of CP2 and CP4 stars has been investigated at the ESO-La Silla Observatory in the bands $J$, $H$, and $K$. As a general result, infrared variations show the same behavior in all three filters but amplitudes are smaller than in the visible. ",
        "introduction": "\\label{intr} Kroll et al. (1987) showed that the near infrared fluxes and colors of Chemically Peculiar stars (or CP stars, according to Preston's (1974) scheme), when compared to a black body, are normal, like that of early main sequence stars. IRAS data could even prove that the normality of IR fluxes is guaranteed to at least 25$\\mu$ (Kroll 1987): only two CP4 stars showed flux excesses longward of 60$\\mu$, showing cold circumstellar material, which is not uncommon among early B stars. Moreover Leone \\& Catalano (1991) have shown that the solar composition Kurucz model atmospheres, which are used to fit the spectra of CP stars from $\\lambda$5500 to $\\lambda$16500~\\AA, give a fair representation of the overall flux distribution, with the exception of the Balmer region, where CP stars appear generally brighter than normal, this excess being just a few percent of the total flux. \\\\ \\begin{table}[t] \\footnotesize \\begin{center} \\caption{The CP stars checked for variability in the near infrared.} \\label{t1} \\begin{tabular}{rrrrrr}  \\hline\\hline SrCrEu & HD~~~3980 & HD~~24712 & HD~~49976 & HD~~72968 & HD~~83368 \\\\ & HD~~96616 & HD~~98088 & HD~101065 & HD~111133 & HD~118022 \\\\ & HD~125248 & HD~126515 & HD~137949 & HD~148898 & HD~153882 \\\\ & HD~164258 & HD~203006 & HD~206088 & HD~220825 & HD~221760 \\\\ Si et al. & HD~~10783 & HD~~12447 & HD~~74521 & HD~~90044 & HD~116458 \\\\ & HD~119419 & HD~125630 & HD~147010 & HD~166469 & HD~170397 \\\\ & HD~187473 & HD~223640 &  &  &  \\\\ Si & HD~~12767 & HD~~19832 & HD~~25267 & HD~~29305 & HD~~37808 \\\\ & HD~~54118 & HD~~56455 & HD~~66255 & HD~~73340 & HD~~92664 \\\\ & HD~114365 & HD~116890 & HD~122532 & HD~124224 & HD~133880 \\\\ & HD~144231 & HD~145102 & HD~203585 & HD~221006 &  \\\\ He weak & HD~~~5737 & HD~~22470 & HD~~28843 & HD~~35456 & HD~~37151  \\\\ & HD~~49333 & HD~~74196 & HD~125823 & HD~137509 & HD~142990 \\\\ & HD~144334 & HD~148199 & HD~168733 & HD~175362 &    \\\\ He rich & HD~~36485 & HD~~37017 & HD~~37479 & HD~~37776 & HD~~59260 \\\\ & HD~~60344 & HD~~64740 &  &  &   \\\\  \\hline\\hline \\end{tabular} \\end{center} \\end{table} \\vspace{-2mm}  However, in spite of this normal infrared behavior, peculiar abundances and/or magnetic fields seem to affect the near infrared too; in fact, Catalano et al. (1991) have shown that, out of the eight CP stars monitored throughout their rotational periods, at least six are variable in the near infrared, although with smaller amplitudes than in the visible. This unexpected result led us to start an observational campaign aimed at searching for infrared variability and also to better understand the origin of the light variability, which is one of the outstanding observational aspects of these stars.  \\vspace{-3mm} ",
        "conclusions": "\\vspace{-2mm} Near infrared variability has been found to be present in the large majority of the CP2 stars studied. The typical trend of CP2 stars to present smaller amplitude light variations at increasing wavelength is confirmed: the amplitudes in the near infrared are smaller than in the visible. In most cases the variations have been found to show very similar behavior and in phase with each other in all filters.  In a previous paper (Catalano et al. 1991) we investigated the effects of high metallicity at the near infrared wavelengths and showed that a Kurucz model atmosphere with a metal content ten times the solar one could explain a three percent variation in the near infrared brightness, which is the typically observed value.  The influence of the magnetic field in the atmosphere structure has been quantitatively discussed by some authors in some particular configurations, however the most general approach has been carried out by Stepien (1978) who showed that, according to the direction of the toroidal electric currents in the outermost layers, the star's shape can be prolate or oblate with respect to the magnetic axis: the differences between the polar and equatorial values of the radius being up to 3\\%. The results obtained by Stepien lend support to a distorted figure of the star up to a few percent and to small variations (2-3\\%) of the effective temperature over the surface, which in some cases, can contribute to the observed light variations. While this explanation is not valid as far as it concerns the visible light variations of many CP stars, because of the different behaviours presented by the $u$, $v$, $b$, and $y$ curves, it cannot be excluded that the non-spherical shape of the star as seen at the infrared wavelengths could contribute to the observed variability, since the magnetic pressure importance increases in the outer layers.  After completing the analysis of our infrared data, we hope to be able to disentangle the relative contributions of these two mechanisms from the study of the phase relation between the magnetic field and infrared variations.  \\vspace{-3mm}"
    },
    "9805/astro-ph9805354_arXiv.txt": {
        "abstract": "We report on the discovery of a new pulsating X--ray source during {\\it Rossi X--ray Timing Explorer} observations of a low galactic latitude field centered at RA (J2000) $= 19^{hr}05^{m}43^{s}$ and Dec (J2000)$=+08^{\\circ}58\\arcmin 48\\arcsec$. Significant pulsations were detected by both the PCA and HEXTE instruments aboard {\\it RXTE} at a fundamental period of $89.17\\pm0.02$ seconds, with higher harmonics also visible in the $2-10$ keV power spectrum. The folded lightcurve from the source is multiply peaked at lower energies, and changes to single peaked morphology above $\\sim 20$ keV. The phase averaged spectrum from the source is well fit by strongly absorbed power law or thermal bremsstrahlung spectral models of photon index $1.9\\pm0.1$ or temperature $19.5\\pm4.6$ keV, respectively. The mean neutral hydrogen column density is $N_{H}\\approx10^{23}$ cm$^{-2}$, suggesting a distance of $>10$ kpc to the source and a minimum $2-10$ keV X--ray luminosity of $2\\times10^{35}$ ergs s$^{-1}$. By comparison with other pulsars with similar periods and luminosities, we suggest that XTE J1906+09 has a supergiant companion with an underfilled Roche lobe. We speculate further that one of the M stars in a peculiar M star binary system may be the companion. ",
        "introduction": "The discovery of a new member of a small class of astrophysical sources is important because each source has the potential to constrain theoretical models of the source physics. Neutron star binary systems are an example of a small class of X--ray sources, with numbers presently totaling total $\\sim200$. Accretion--powered pulsars can be divided into two broad classes based on their X--ray luminosities and spectra (\\cite{white83}). The X--ray spectra of the higher luminosity sources ($L_{X}\\sim10^{36}-10^{37}$ ergs s$^{-1}$) are typically characterized by a hard power law of photon index $1<\\Gamma<2$ out to an energy of $10-20$ keV, above which the spectrum ``breaks'' to a much steeper index ($\\Gamma>3$). The lower luminosity accreting X--ray pulsars tend not to have this broken power law shape, and instead have a softer spectral shape ($\\Gamma>2$) over a broad range of X--ray energies. Furthermore, observations of low luminosity burst sources have revealed weak power law components ($\\Gamma\\sim2$) when the inferred accretion accretion rate drops below a critical value (\\cite{barret94}). In this {\\it Letter} we present the detection of a new source which appears to be a low luminosity accreting X--ray source. ",
        "conclusions": "We have presented the detection of a new $89$ second pulsating X--ray source with {\\it RXTE}. The source was detected during observations of a region of Galactic plane containing the SGR 1900+14 error box. Characteristics of the folded lightcurves and spectrum suggest that the source, XTE J1906+09, is a low luminosity X--ray binary located beyond the Galactic center. The low luminosity and long pulse period of XTE J1906+09 indicate that this source is probably a high mass X--ray binary accreting via a stellar wind. We raise the possibility that the new source is associated with the highly absorbed double M supergiant binary system on the edge of the SGR 1900+14 error box. If this identification is correct, XTE J1906+09 is the first known X--ray triple system, and a candidate counterpart for SGR 1900+14."
    },
    "9805/astro-ph9805162_arXiv.txt": {
        "abstract": "In this paper we address the issue of the origin of LBV bipolar bubbles. Previous studies have explained the shapes of LBV nebulae, such as $\\eta$ Car, by invoking the interaction of an isotropic fast wind with a previously deposited, slow aspherical wind (a ``slow torus\").  In this paper we focus on the opposite scenario where an {\\it aspherical fast wind} expands into a previously deposited {\\it isotropic slow wind}. Using high resolution hydrodynamic simulations, which include the effects of radiative cooling, we have completed a series of numerical experiments to test if and how aspherical fast winds effect wind blown bubble morphologies.  Our experiments explore a variety of models for the latitudinal variations of fast wind flow parameters. The simulations demonstrate that aspherical fast winds can produce strongly bipolar outflows.  In addition the properties of outflows recover some important aspects of LBV bubbles which the previous \"slow torus\" models can not. ",
        "introduction": "In just a few years the HST has transformed our understanding of the massive unstable stars know as Luminous Blue Variables (LBVs).  Recent observations have revealed a number of LBVs or LBV candidates to be surrounded by extended {\\it aspherical} outflows. The most extraordinary of these is the markedly bipolar nebula surrounding $\\eta$ Carinae (``the homunculus'': \\cite{Hesterea91}; \\cite{Ebbetsea93}; \\cite{HumDav94}).  Other LBVs show nebulae with varying degrees of asphericity from elliptical (R127: \\cite{Notaea95}) to strongly bipolar (which we define though the presence of an equatorial waist: HR Carinae: \\cite{Notaea95}; \\cite{Weisea96}).  These morphologies are quite similar to what has been observed in Planetary Nebulae (PNe) which arise from low mass stars (\\cite{Balick87}; \\cite{Machadoea96}). The aspherical shapes of PNe have been successfully explained through a scenario termed the ``Generalized Interacting Stellar Winds'' model (GISW: \\cite{Kwok78}, \\cite{Kwok82}, \\cite{Kahn83}, \\cite{Frankea93}; \\cite{FM94}; \\cite{MF95}).  In the GISW model an isotropic fast wind from the central star (a proto-white dwarf) expands into an aspherical (toroidal) slow wind ejected by the star in its previous incarnation as a Asymptotic Red Giant. High densities in the equatorial plane constrain the expansion of the fast wind.  The expanding shock which forms quickly assumes an elliptical prolate geometry.  If the ratio of mass density between the equator and pole (a parameter we call $q$, $q_\\rho = \\rho_e/\\rho_p$) is high enough, then the elliptical bubble eventually develops a waist and becomes bipolar.  The similarity of PNe and LBV nebulae has led to the suggestion that both families of objects are shaped in similar ways. In \\cite{FBD95} (hereafter: FBD) a GISW model for $\\eta$ Car was presented in which a spherical ``outburst'' wind expelled during the 1840 outburst expanded into a toroidal ``pre-outburst'' wind.  FBD showed that the resulting bipolar outflow could recover both the gross morphology and kinematics of the Homunculus. \\cite{Notaea95} (hereafter NLCS) used a similar model for other LBV nebulae presenting a unified picture of the development of LBV outflows. More recently \\cite{MacLowea96} (hereafter GLM) presented a model which also relied on the GISW scenario but which changed the order of importance of the winds.  The novel aspect of GLM's study was to include the effects of stellar rotation.  Using the Wind Compressed Disk model of \\cite{BjCas92}, GLM showed that a strong equator to pole density contrast would likely form {\\it during the outburst} when the star is close to the Eddington limit and rotation can deflect wind streamlines toward the equator.  In their model it is the ``post-outburst\" mass loss (which was not considered in either FBD or NLCS) that acts as the fast wind.  The post-outburst wind in GLM's model ``inflates'' the bipolar bubble via its interaction with the toroidal outburst wind.  All these models have demonstrated the potential efficacy of the GISW scenario by recovering the basic shapes observed in LBV nebulae. However, by relying on the presence of a slow torus they are are troubling in their mutual inconsistency.  Specifically the question ``Where is the torus?'' must be answered.  Does the torus form during the outburst phase as in GLM or does it form in the pre-outburst wind as in FBD and NLCS?  Without invoking binary interactions or a pre-existing disk left over from the stellar formation process, it may be difficult to get a strong toroidal density contrast in the pre-outburst environment.  Stepping back further one can also ask if a disk is needed at all?  The latter question arises from consideration of new HST images of $\\eta$ Car (\\cite{Morseea97}) which reveal the disk to be so highly fragmented that it may be more reasonable to consider the structure to be a ``skirt'' of individual clumps of ejecta rather than a continuous feature.  This point is crucial since a discontinuous equatorial spray of isolated bullets can not hydrodynamically constrain an isotropic stellar wind to form a bipolar outflow.  Thus we are led to consider an alternative model to the one proposed by FBD, NLCS and GLM which further generalizes the GISW model by turning that scenario on its head.  It what follows we consider the case of an {\\it aspherical fast wind} interacting with an {\\it isotropic slow wind}.  We imagine a fast wind ejected with higher velocity along the poles than along the equator.  The question we wish to answer is can such a wind, expanding into an isotropic environment, account for the shapes of LBV nebulae.  There are a number of reasons for pursuing this line of investigation some of which were cited above as criticisms of the ``classic'' GISW model.  More importantly, however, theoretical models admit the possibility of aspherical fast winds in massive stars. \\cite{PaulPuls90} have shown that a discontinuity (bistability) in mass loss and velocity occurs when the effective gravity of the star drops below a critical value.  \\cite{LamPaul91} used these results to demonstrate that stellar rotation can induce latitudinal changes in $g_{eff}$ and the optical depth of the wind.  The change in optical depth puts the polar and equatorial regions of the star on different sides of the bistabilty limit.  A high velocity, low density wind forms at the poles, and low velocity, high density wind forms along the equator.   It should be noted that the Wind Compressed Disk (WCD) model of \\cite{BjCas92} also produces aspherical winds since the equatorial focusing occurs close to the star.  Thus a wind that has been shaped by the WCD mechanism, if it is expanding into a slower moving environment, should be considered an aspherical fast wind, \\ie the issue is always the velocity (and density) of previously ejected material.  It is worth noting however that recent numerical models of the WCD mechanism (\\cite{Owockiea96}) which include non-radial line forces found inhibition of the wind compression and mass loss in the equator.  Instead a net flow in the pole-ward direction was formed.  This is, therefore, yet another way by which fast winds might become aspherical. Finally, and most importantly, there is direct evidence for asphericity in fast winds. Observations of the wind of AG Carinae (\\cite{Leithererea94}) imply a pattern of densities and velocities from pole to equator much like that described in \\cite{LamPaul91}.  Finally we note that it is worthwhile to pursue this kind of investigation simply because it has not been done before.  The GISW model and its variations has been very successful in accounting for a variety of bipolar outflow phenomena (\\cite{Melal91}; \\cite{BloLund92}; \\cite{FrBaLi96}).  Since the effect of aspherical fast winds has yet to be investigated the potential of finding useful results is high which argues for a detailed study.  We note that this paper represents an initial exploratory study.  We are using an admittedly ad-hoc formalism to control the asphericity of the fast wind and we have not tuned our parameters to the accepted values for any particular LBV.  Our purpose in this paper is to map out the broad consequences of including aspherical fast winds into the GISW formalism with an application to LBVs as a class of outflows. In future papers we will attempt to apply our results to individual LBVs in an attempt to make detailed contact with observational results.  The organization of the paper is as follows: In section II we describe the numerical method and initial conditions used in our simulations.  In Section III we present and discuss the results of our simulations. In section IV we present our conclusions along with a discussion of some issues raised by the simulations. ",
        "conclusions": "The results of our simulations demonstrate that bipolar wind blown bubbles can result purely from the action of an aspherical fast wind.  In previous studies of LBV bubbles (FBD, NLCS, GLM) it has been assumed that a slow moving torus or disk of gas was a necessary precondition for the development a bipolar bubble.  Our results indicate that the properties of LBV bubbles may not require such a torus to form either before (FBD, NLCS) or during (GLM) the outburst.  Our scenario has a number of attractive features.  First it is both observationally and theoretically motivated.  From the observational side there is already evidence that LBV winds can take on aspherical velocity and density distributions.  From the theoretical side the theory of \\cite{LamPaul91} have demonstrated that \"bistable\" winds are possible around massive hot stars.  In addition, the diversity of shapes of LBV bubbles (NLCS) may be difficult to achieve with pre-exiting disk models.   The models presented here can recover the diversity of shapes simply by changing $q_v$ though it is certainly true that this also begs the question of what drives the velocity contrast in the fast wind.  Currently it is unclear what form the mass distribution takes in the lobes of LBV bubbles.  One critical test of the different models for LBV nebula shaping would entail comparison of latitudinal variations of mass.  If the caps of a bipolar lobe have densities that are comparable to that in the lobe's flanks it would present difficulties for the slow torus models. In a bipolar bubble resulting from a spherical fast wind driving into a slow torus the caps of the bubble should be the location of the lowest density.    For aspherical fast winds however high or equal density in the poles poses no significant problem since we are then seeing a signature of the latitudinal dependence of the fast wind density (consider Runs C and F in Fig.~3). We wish to note also that \\cite{Currie96b} and \\cite{Morseea97} have found that the shape of $\\eta$ Car is best matched by by a geometry which can be described as a \"double flask\" rather than an two oscullating spheres. Based on comparison of published results the models presented here seem to do a better job of recovering such a shape than either FBD, NLCS or GLM.  It is noteworthy however that the scenario presented here would not produce the equatorial ``skirt'' seen surrounding the waist of the homunculus in $\\eta$ Car.  The presence of that feature is what motivated the original application of the GISW slow torus models. Within the current formulation of the aspherical fast wind model there is not a likely means of producing such a feature.  A few points are worth noting however.  The equatorial skirt is not a continuous or even axisymmetric feature. Thus whatever its origin it is hard to imagine that it can be the agent which constricts a spherical fast wind and produces the bipolar bubble.  In addition there are a number of ``spikes'' extending beyond, but connecting with the homunculus that have been reported to have velocities in excess of 1000 km/s (\\cite{Meaburnea96}).  They yield dynamical timescales $\\le t_o$ where $t_o$ is the time since the outburst.  Thus it is possible that the equatorial skirt is actually a spray of material which ejected at some point after the outburst of 1849 and which was decelerated by its passage through the dense shell of the outburst wind. The non-axisymmetric distribution of the skirt may then be a consequence of the intrinsic pattern of ejecta or of impulsive instabilities which will occur when the ejecta is driven through the outburst wind.  Regardless of the answer to this issue the results presented here show that there are two very different scenarios for the formation of LBV bubbles.  Either they form via the interaction of a spherical fast wind driving into an aspherical slow wind (a slow torus) or they form from an aspherical fast wind driving into an isotropic pre-existing environment. This embarrassment of riches can be eventually be dealt with by comparing the latitudinal distributions of mass and momentum observed in real bipolar LBVs with what is predicted for the various models.  Such an approach was used successfully by \\cite{CherMas92} in evaluating different models of molecular outflow formation in YSOs.  This project is currently in progress."
    },
    "9805/astro-ph9805212_arXiv.txt": {
        "abstract": "A companion paper presents a worked model for evolution through inflation to initial conditions for an isocurvature model for structure formation. It is shown here that the model is consistent with the available observational constraints that can be applied without the help of numerical simulations. The model gives an acceptable fit to the second moments of the angular fluctuations in the thermal background radiation and the second through fourth moments of the measured large-scale fluctuations in galaxy counts, within the possibly significant uncertainties in these measurements. The cluster mass function requires a rather low but observationally acceptable mass density, $0.1\\lsim\\Omega\\lsim 0.2$ in a cosmologically flat universe. Galaxies would be assembled earlier in this model than in the adiabatic version, an arguably good thing. Aspects of the predicted non-Gaussian character of the anisotropy of the thermal background radiation in this model are discussed. ",
        "introduction": "An accompanying paper (Peebles~1998a; hereafter Paper~I) presents a worked example of the evolution of a cosmological model through inflation to initial conditions for an isocurvature (ICDM) model for structure formation in a universe that now is dominated by cold dark matter. Here I show that the model can be adjusted to fit main observational constraints.  As in Paper~I, I attempt to keep the discussion simple and definite by adopting a specific set of model parameters chosen to give a reasonable approximation to the observations. More detailed parameter studies that seek to minimize $\\chi ^2$ measures of fit to the full suite of constraints would be interesting but perhaps are not yet a pressing need because many important observational constraints still are preliminary and may harbor systematic errors.  The adiabatic cold dark matter (ACDM) model for structure formation has been subject to searching tests from numerical simulations (eg. Governato {\\it et al.} 1998; Springel {\\it et al.} 1998; and references therein). I hope the simpler observational tests presented here show that the considerable effort needed for a meaningful application of numerical simulations of the ICDM model would be worthwhile.  The model parameters are listed in \\S 2. Second moments of the angular distribution of the thermal background radiation (the CBR) and the large-scale space distribution of galaxies are presented in \\S 3. In the ICDM model the primeval CDM mass distribution is proportional to the square of a random Gaussian process with zero mean. In \\S 4 I discuss the nature of the large-scale non-Gaussian fluctuations in the mass distribution and compare them to third and fourth moments of galaxy counts. The mass function of rich clusters of galaxies is discussed in \\S 5. Because the distribution of mass fluctuations is broader than a Gaussian with the same standard deviation, rare mass concentrations form earlier than in an ACDM model. The ICDM model thus requires a lower mean mass density for given normalization of the power spectrum, and the cluster mass function changes significantly less rapidly with redshift than in the ACDM model. In \\S 6 I present the scaling relation between the epochs of assembly of the dark matter concentrations in galaxies and in rich clusters of galaxies.  The relatively early assembly of protogalaxies in the ICDM model is arguably attractive. Finally, \\S 7 presents some considerations of the higher moments of the angular fluctuations of the CBR. As an example I compute the third moments of the quadrupole and octupole components of the CBR anisotropy. Concluding remarks are presented in \\S 8. ",
        "conclusions": "\\subsection{Is the ICDM Model Attractive from a Theoretical Point of View?}  Paper I compares the fields, parameters, and functional forms of the potential energy in the ICDM model and other physical theories for the seeds of structure from inflation. Here I consider some broader issues.  Our experience in particle physics might lead us to suspect that the laws of physics relevant to the early universe will be found to be elegant and simple, albeit in some deeply subtle way, and that once we understand the physics we will see that the universe is an expression of the physics. This world view informs many studies of inflation. The night thought of a physical scientist might be that Newtonian mechanics is expressed in fully developed turbulence, but it is not likely we would know much about turbulence if we had not seen it. A knowledge of the physics of the early universe might not be of much use if its expression were complex.  Two themes could be accepted in either world view (as well as by those, perhaps the majority, with more moderate opinions). First, the construction of a specific internally consistent example of how evolution from very high redshift could have led to the present state of the universe is a valuable demonstration of consistency of the set of ideas on which it is based. We have examples from the adiabatic CDM family of models. I have argued for yet another, an isocurvature CDM model. Second, the models and their parameters will be reconsidered with each significant advance of knowledge of the physics and astronomy, a process that will lead us to abandon some models, adjust others, and maybe introduce new ones. Perhaps this process will back us into that narrow corner of model and parameter space that is a useful approximation to what  really happened. We may have a modest example in the fact that this latest version of the isocurvature model (earlier steps of which may be traced back through Peebles 1997a) has the same dynamical actors as the ACDM model, though it remains to be seen whether this is a lasting situation.  A mature physical theory must be falsifyable; there is good reason for our conditioned dislike of theories that can be adjusted to fit whatever is measured. On the other hand, if the evolution of the early universe were moderately complex we likely would need a flexible model to fit it. The isocurvature ICDM model in Paper I assumes power law inflation because that makes it easy to select the fields and their potential energy functions to produce a power law CDM fluctuation spectrum $P_\\rho\\propto k^{m_\\rho }$ over a wide range of scales. But the evidence may lead us to another functional form. In the power law ICDM models shown as the solid lines in Figure~1 the CBR anisotropy $T_l$ at $l\\sim 100$ may be too high (Netterfield {\\it et al.} 1997). That could be remedied by taking the power law index $m_\\rho$ to be closer to zero, but that would make $T_l$ unacceptably small at $l\\lsim 10$. The dotted line shows one way out: change the power spectrum to \\beq P_\\rho\\propto (k_x/k)^3 + (k_x/k)^{3/2}, \\eeq with \\beq k_x = 0.01h\\hbox{ Mpc}^{-1}, \\eeq and the other parameters the same as for the middle solid line. There has to be another bend to $P\\propto k^{s}$,  $s<-3$, at $k\\sim 1$~pc$^{-1}$ (Paper~I). These bends are quite inelegant, unless Nature has chosen them.  The flexibility of the ICDM model is limited. For example, it is difficult to lower the spectrum at $l\\sim 100$ without significantly lowering the peak at $l\\sim 300$. The advances in observational constraints from work in progress will show whether the ICDM model is a useful approximation. If the observations to come in the next decade are fitted in all detail by one of the simple structure formation models now under discussion it will compel acceptance. If improving observations require increasingly baroque models it may mean we have missed the correct elegant picture, or that the evolution of the universe does not agree with our standards of elegance.  In my reading of the first of the world views mentioned above the ICDM model is quite inelegant because it was constructed {\\it ad hoc} to fit the observations and it is flexible enough to be capable of adjustment to fit some substantial changes in the observational situation. In the second world view, a model that fits significant observational constraints within a sensible reading of the physics may not be all bad.  \\subsection{Is the Model Attractive from a Phenomenological Point of View?}  The ICDM model has some possibly significant successes and problems. Both will have to be reconsidered with each advance of the observations and their interpretation, of course. I hope is is not entirely self-serving to note that a model that is close to reality may encounter apparent problems as we sort out the ambiguities in the evidence. Following are some considerations.  \\noindent i) The model parameters that fit the CBR angular fluctuation spectrum $T_l$ in Figure~1 fit the second moment $P(k)$ of the large-scale galaxy distribution in Figure~2, a significant success. The low measured value of $T_l$ at $l\\sim 100$ may require adjustment of the model, or perhaps will prove be in some part a systematic error in exceedingly difficult measurements.  \\noindent ii) The skewness and excess kurtosis of galaxy counts are not far from that of the model, a not insignificant result. The major open issue is the correction for nonlinear evolution: does the non-Gaussian primeval mass distribution of the model evolve into the galaxy clustering hierarchy? An example of the predicted non-Gaussian higher moments of the multipole expansion components $a_l^m$ of the angular distribution of the CBR is presented in \\S 7. A comparison to the measurements remains to be done.  \\noindent iii) The cluster mass function agrees with a Press-Schechter approximation under the assumption that galaxies trace mass at $\\Omega = 0.2$. Chiu, Ostriker, \\&\\ Strauss (1998) point out that this probes the nature of the primeval mass fluctuations, because the cluster mass function depends on the tail of the distribution and the rms galaxy peculiar velocity field on the standard deviation. The density parameter in the ICDM model discussed here, $\\Omega =0.2$, agrees with the low peculiar velocities indicated by many analyses (e.g. Peebles 1986; Bahcall, Lubin, \\&\\ Dorman 1995; Peebles 1997b; Willick \\&\\ Strauss 1998). The ICDM model thus seems to pass the Chiu et al. test. If further work showed that the mass fraction $f_{\\rm cl}$ in clusters is not near the upper end of the range in equation~(\\ref{eq:fcl}), and the Press-Schechter method is a good approximation, it would require a lower value of $\\Omega$, increasing $T_l$ and tending to spoil the general consistency with the measured $T_l$ and $P_\\rho (k)$.  \\noindent iv) The model predicts relatively early galaxy assembly; it is an open issue whether this is a success or problem. The model may be considered a success from a theoretical point of view because I arrived at it by a search for galaxy formation at high redshift, when the mean mass density would have been considerably closer to the relatively high density characteristic of the luminous parts of normal galaxies. The line of thought originated in Partridge \\&\\ Peebles (1967); a recent version is in Peebles (1998b). The scaling arguments in \\S 6 suggest the ICDM model has some attractive features as a model for galaxy formation. A more detailed examination by numerical simulation remains to be done.  My conclusion, from the second of the world views presented at the beginning of this section, is that the ICDM model is attractive because it fits a significant set of observational constraints within what appears to be an acceptable physical model."
    },
    "9805/astro-ph9805024_arXiv.txt": {
        "abstract": "We have determined the absolute magnitude at maximum light of SN 1992A by using the turn--over magnitude of the Globular Cluster Luminosity Function of its parent galaxy, NGC 1380.  A recalibration of the peak of the turn--over magnitude of the Milky Way clusters using the latest HIPPARCOS results has been made with an assessment of the complete random and systematic error budget.  The following results emerge: a distance to NGC 1380 of 18.6$\\pm 1.4$ Mpc, corresponding to (m--M)=31.35$\\pm 0.16$, and an absolute magnitude of SN 1992A at maximum of M$_B{\\rm (max)}=-18.79\\pm0.16$.  Taken at face value, SN 1992A seems to be more than half a magnitude fainter than the other SNeI-a for which accurate distances exist. We discuss the implications of this result and present possible explanations.  We also discuss the Phillips's (1993) relationship between rate of decline and the absolute magnitude at maximum, on the basis of 9 SNeI-a, whose individual distances have been obtained with Cepheids and the Globular Cluster Luminosity Function.  The new calibration of this relationship, applied to the most distant SNe of the Calan-Tololo survey, yields H$_\\circ= 62\\pm 6$ km s$^{-1}$ Mpc$^{-1}$. ",
        "introduction": "\\bigskip  The study of Supernovae Ia at maximum light is important for two reasons.  On the one hand Supernovae Ia are commonly regarded as reliable standard candles.  In the recent past, mainly on the basis of photographic data, a number of authors [e.g. Leibundgut and Tammann (1990), Miller and Branch (1990), Della Valle and Panagia (1992), Vaughan et al. (1995)] were able to show that the absolute magnitudes at maximum of SNeI-a have a small dispersion, of the order of $\\lsim 0.3$ mag. In principle, used as standard candles, these objects could provide distance measurements with an uncertainty of only $\\lsim \\pm 14\\%$.  However, `high--quality' observations, obtained in the last 6--7 years, appear to complicate the previous `idyllic' scenario. Phillips (1993), on the basis of a sample of SNeI-a whose light-curves were well sampled at maximum light, has considerably strengthened a former suggestion by Pskovskii (1967) concerning the possible existence of a relationship between the Absolute Magnitude at Maximum of type Ia Supernovae and their Rate of Decline (=AMMRD). Branch, Romanishin and Baron (1996) have found that SNeI-a occurring in early type galaxies are {\\underline {on average}} $\\sim 0.3$ magnitudes fainter than Ia in spirals.  Finally, Sandage et al. (1996) were not able to confirm the Phillips's relationship. As a consequence of these uncertainties, the value of {\\sl H$_\\circ$} measured with SNeI-a varies between $\\sim 50 $ and almost $70 $ km s$^{-1}$ Mpc$^{-1}$ (Lanoix 1998, Hamuy et al. 1995).  On the other hand, a number of observational studies have pointed out the existence of significant intrinsic differences between SNeI-a occurring in spirals and early type galaxies: these observations concern their spectroscopic (Branch, Drucker and Jeffery 1988, Filippenko 1989, Branch and van den Bergh 1993, Nugent et al. 1994), and photometric evolution (van den Bergh and Pierce 1992, van den Bergh and Pazder 1992, Suntzeff 1996, Riess et al. 1996), their rate and place of occurrence (e.g. Bartunov et al. 1994, Della Valle and Livio 1994, Cappellaro et al. 1997a, Wang et al. 1997), all of which question the uniqueness of the progenitors for type Ia Supernovae. Since the peak luminosity is proportional to the synthesized nickel mass (e.g. Cappellaro et al. 1997b), one of the most direct observational ways to prove the existence of intrinsic differences between the progenitors is to measure the differences in absolute magnitude at maximum for a sample of SNeI-a.  Such differences might then be attributed to either differences among the progenitors (Branch et al. 1995 for a review) and/or to differences in the mechanism of the explosion (e.g. Canal, Isern and Lopez 1988).  These problems have motivated our study of SN 1992A. In section \\ref{sec.gclf}, we will briefly discuss the data and methods which we have used to determine the turn--over magnitude of the Globular Cluster Luminosity Function (=GCLF) of NGC 1380.  In section \\ref{sec.distance}, we determine the distance to NGC 1380 and in section \\ref{sec.abs_mag} we derive the absolute magnitude at maximum light for SN 1992A.  In section \\ref{sec.abs_mag_general} we determine the AMMRD. In section \\ref{sec.discussion} we discuss the results and the possible implications for the calibration of the extragalactic distance scale and in the final section (\\ref{sec.conclusion}) we summarize our conclusions. \\bigskip ",
        "conclusions": "\\label{sec.conclusion}  In this paper we have determined the distance to NGC 1380, an S0 galaxy host of the type Ia SN 1992A, through the use of the TO magnitude of the GCLF. We find a distance modulus of 31.35$\\pm 0.16$ corresponding to a distance of 18.6$\\pm 1.4$ Mpc.  This is consistent with the distances to other members of the Fornax cluster, utilizing the same method, as well as the Cepheid distance to NGC 1365.  By applying this distance to the apparent magnitude of SN 1992A, we find that at peak brightness SN 1992A reached $M_B= -18.79\\pm0.16$, which is about 0.4~mag fainter than expected for typical SNeI-a in early type galaxies (Branch, Romanishin and Baron 1996), and about 0.7 magnitudes fainter than SNeI-a in spirals, if one accepts as zero point M$_B=-19.53\\pm 0.07$ (Tammann et al. 1996), the absolute magnitude at maximum of SNeI-a in Spirals.  It is worthwhile noting in this respect that recent work by Mazzali et al.  (in preparation) shows a good correlation between the velocity widths of the nebular lines (at around 300 days after the maximum) and the rate of decline (and therefore the absolute magnitude at maximum). The velocity widths for SN 1992A would indicate that it is subluminous relative to other normal type Ia supernovae.  The close similarity with the apparent magnitude at maximum exhibited by SN 1980N and SN 1981D, two other SNeI-a in the Fornax Cluster, could indicate that SN 1992A is a peculiar object only if NGC 1316, the parent galaxy of SN 1980N and 1981D, is placed on the far side of the cluster at d$\\gsim 20.5$ Mpc. In this case the absolute magnitudes of SN 1980N and SN 1981D would be slightly brighter than M$_B\\sim -19$, in agreement, within the errors, with the value reported by Branch, Romanishin and Baron (1996) for SNeI-a occurring in early type galaxies. Finally, we are not certain that all necessary corrections were made to the TO magnitude of the GCLF to guaranty its reliability as standard candle. Indeed, in section 3 we discussed systematic effects which could conspire to make the TO magnitude of NGC 1380 fainter by $\\sim 0.3$ mags.  We have derived the absolute magnitude at maximum of a number of historical SNeI-a which have occurred in early type galaxies, whose distances are known through the GCLF. Together with high quality data, collected from the literature, we have produced a linear fit to the data points in the $\\Delta m(15)$ {\\sl vs.} $M_B$ plane. We find a slope similar to that measured by Phillips (1993). The $\\sim 0.4$ mag difference in the intercept probably reflects the zero point difference existing between the distances obtained via Cepheids/GCLF (adopted in this paper) and Surface Brightness Fluctuation and Tully-Fisher methods (adopted by Phillips 1993).  After excluding SN 1991bg from the fit and adopting the slope derived by Hamuy et al. (1996) we have determined a new zero point for this relationship and H$_\\circ=62\\pm 6$ km s$^{-1}$ Mpc$^{-1}$. It is apparent that any correction for reddening would tend to increase the obtained value of H$_\\circ$. Our plot in Fig. 6 shows that:  1) SNeI-a in Spirals are located in the bright and slow part of the diagram whereas the SNe in early type galaxies fall in the faint and fast part of the plot. Owing to the intrinsic dispersion of the relationship, it is apparent that an analysis based only on SNe discovered in Spirals or in early type galaxies would hardly reveal any correlation between the rates of decline and absolute magnitude at maximum, then indicating that the dependence of the absolute magnitude at maximum on $\\Delta m(15)$ and on the {\\sl Hubble type} of the parent galaxies are almost equivalent. On the basis of data reported in Tab. 5 (but excluding 1991bg), the unweighted mean of the absolute magnitude at maximum of SNe Ia in early type galaxies is M$_B=-19.05\\pm 0.16$. By comparing this figure with M$_B=-19.53\\pm 0.07$ (Tammann et al. 1996) for Ia in Spirals, we obtain $\\Delta M=0.48\\pm 0.17$.  2) Sub-luminous objects such as 1991bg would indicate that the M$_B$ {\\sl vs.} $\\Delta m(15)$ relationship becomes nonlinear for high rate of decline (see also the case of SN 1992K pointed out by Hamuy et al. 1995).  3) We find a difference of $\\Delta m=0.31 \\pm 0.14$ mag between the absolute magnitude at maximum of SNeI-a calibrated with GCLF and PNLF (see Tab. 6 and Tab. 7). To check the significance of this result we have compared the distance moduli of the ten galaxies for which the distances have been determined via PNLF and GCLF (see Tab. 1 of Jacoby (1997) and Whitmore 1997). By using the TO mag reported in our Tab. 3 we find a difference (m-M)$_{GC}$--(m-M)$_{PNe}=0.56 \\pm 0.24$ (and (m-M)$_{GC}$--(m-M)$_{PNe}= 0.61 \\pm 0.19$ after excluding NGC 3379). This result may put doubt on the consistency of the zero points of these distance indicators. However, if we assume as calibrator of the absolute magnitude at maximum of SNeI-a in early type galaxies the data provided by PNLF ( Tab. 7), the price to pay to fit the AMMRD with 0.784 slope (Hamuy et al. 1996) is to make considerably fainter the absolute magnitude at maximum of SNeI-a in spirals, close to M$_B\\sim -19$. We note that Kennicutt et al. (1998) have estimated that metallicity effects on the distance scale derived with HST observations of Cepheids could affect the distance moduli, to the respective parent galaxies, only at $\\sim 0.2$ mag level.  At least four critical observations could significantly improve our present understanding of the problem:  a) determination of the distance of NGC 1316, parent galaxy of SN 1980N and 1981D, through the use of the GCLF (the PNLF distance already being available). This should also enable us to clarify whether or not SN 1992A is a peculiar (sub-luminous) object, despite its `protonormal' spectroscopic evolution.  b) determination of the distance of NGC 1380 via the PNLF to make a sensible comparison with the distance obtained via GCLF in this paper.  c) determination of the distance to NGC 4526, parent galaxy of SN 1994D, through the use of the GCLF and PNLF. This SN may be a conspicuous exception to the AMMRD relationship, unless we assume that NGC 4526 is located on the near side of the Virgo cluster at $\\sim 13$ Mpc (see Tonry 1995).  However, this last possibility appears quite unlikely, because it is known that early type galaxies are normally concentrated towards the core of the clusters. As an alternative, NGC 4526 may not belong to Virgo Cluster, rather being a foreground galaxy.  d) measurement of the distance, with Cepheids, to NGC 3627 and NGC 4527, parent galaxies of SN 1989B and SN 1991T.  Indeed, these objects have been well studied, close to maximum, in the past (Barbon et al.  1990, Wells et al. 1994, Phillips et al. 1992, Ruiz-Lapuente et al. 1992, Filippenko et al. 1992b), and therefore,  once their position in the M$_B$ {\\sl vs.} $\\Delta m(15)$ plane is firmly established, the large error still associated with the intercept of [1] will be considerably reduced.  \\subsection*"
    },
    "9805/astro-ph9805268_arXiv.txt": {
        "abstract": "The unprecedented detail of the WFPC2 colour-magnitude diagrams of the resolved stellar population of Leo~A presented here allows us to determine a new distance and an accurate star formation history for this extremely metal-poor Local Group dwarf irregular galaxy.  From the position of the red clump, the helium-burning blue loops and the tip of the red giant branch, we obtain a distance modulus, m$-$M=24.2$\\pm$0.2, or 690 $\\pm$ 60 kpc, which places Leo~A firmly within the Local Group.  Our interpretation of these features in the WFPC2 CMDs at this new distance based upon extremely low metallicity (Z=0.0004) theoretical stellar evolution models suggests that this galaxy is predominantly young, {\\it i.e.} $<$~2~Gyr old. A major episode of star formation 900$-$1500~Gyr ago can explain the red clump luminosity and also fits in with our interpretation of the number of anomalous Cepheid variable stars seen in this galaxy.  We cannot rule out the presence of an older, underlying globular cluster age stellar population with these data.  However, using the currently available stellar evolution models, it would appear that such an older population is limited to no more than 10\\% of the total star formation to have occured in this galaxy.  Leo~A provides a nearby laboratory for studying young metal poor stars and investigations of metal-poor galaxy evolution, such as is supposed to occur for larger systems at intermediate and high redshifts. ",
        "introduction": "Small galaxies are common and apparently structurally simple.  They may provide an important perspective on how luminous structures have evolved in the Universe.  Despite a variety of theoretical models (e.g., Dekel \\& Silk 1986; Hensler \\& Burkert 1990), we cannot predict how even the simplest galaxies have changed over time.  In particular the internal clocks which set the time interval for major star formation are seen to be highly variable in the Galactic retinue of dSph, ranging from ancient systems where star formation was complete 10 Gyr ago to galaxies that are mainly intermediate age (e.g., van den Bergh 1994).  The presence of numerous small, actively star forming field galaxies at moderate redshifts of 0.3$<$z$<$1 suggests that such asynchronous behavior may be the rule rather than the exception amongst smaller galaxies (e.g., Babul \\& Ferguson 1996).  However, it still remains to be understood exactly what are faint blue galaxies (FBGs) that are found in deep imaging surveys (e.g., Ellis 1997). Their sheer numbers make them a cosmologically significant population and an important tracer of the star formation history (SFH) of the universe. There is considerable evidence that these FBGs are predominantly intermediate redshift ($z<1$, or a look-back time out to roughly half a Hubble time), late type, intrinsically {\\it small} galaxies, undergoing strong bursts of star formation (e.g., Glazebrook {\\it et al.} 1996; Lilly {\\it et al.} 1996; Odewahn {\\it et al.} 1996).  The best way to understand FBGs is to discover their nearby counterparts which we can study in detail, and directly observe the absence or presence of past bursts in colour-magnitude diagrams (CMD).  Irregular galaxies are very strong candidates for what is left of the FBGs in our nearby universe.  They are systems which may have undergone one or more bursts of star formation in the last few Gyr (Matteucci \\& Tosi 1985).  They exist in large enough numbers that, could they be made bright enough for a short time in the past, they could easily account for the population of FBGs (e.g., Babul \\& Ferguson 1996).  It is still an open question - why dwarf galaxies in the Local Group display such a wide range in stellar age distributions.  In our sample (Skillman 1998) we are finding enhanced star formation rates (SFR) over Gyr time scales, during which the bulk of the stellar populations within small galaxies are formed, rather than very short discrete bursts ($\\lesssim$ a few $\\times 10^7$yrs).  The physical mechanisms responsible for this behaviour are uncertain.  Current theoretical models mainly consider cases where SFR is tied to the gas supply.  Epochs of active star formation can then occur either due to delayed cooling of gas associated with a small galaxy (e.g., Kepner, Babul, \\& Spergel 1997, Spaans \\& Norman 1997), or through rejuvenation of a pre-existing small galaxy due to gas capture within a galaxy group (e.g., Silk, Wyse, \\& Shields 1987).  Another possibility is that dwarfs were formed from tidal debris produced by interactions between galaxies in the Local Group, although it is not clear what objects could have spawned the Leo I and considerably more distant Leo~A dwarfs (e.g., Hunsburger, Charlton \\& Zaritsky 1996).  Here we present the results for Leo~A ($\\equiv$ DDO~69, Leo III, UGC~5364), a nearby Magellanic dwarf irregular galaxy.  A variety of studies have given this galaxy a large range of possible distances (e.g., Hoessel {\\it et al.} 1994 [hereafter, H94], and references therein).  There are several faint HII regions distributed along the ridge of highest column density HI (Tolstoy~1996; Hunter, Hawley \\& Gallagher 1993), which provide a very low limit to the current SFR (over the last 10Myr) of $< 10^{-4} \\Msun yr^{-1}$.  The brightest H$\\alpha$ emission in the galaxy comes from a planetary nebula, which yields an extremely low oxygen abundance of $\\sim$~2.4\\% solar (Skillman, Kennicutt \\& Hodge 1989).  Recent HI observations (Young \\& Lo 1996), show that the optical galaxy is surrounded by a large HI halo, extending out $\\sim$3 times the optical diameter at a column density of 4$\\times 10^{19} \\rm cm^{-2}$.  The detected HI flux corresponds to an HI mass of $(8.1\\pm 1.5) \\times 10^7 \\rm ~\\Msun$, of which $\\sim$30\\% is in the halo at column densities below $2\\times 10^{20} \\rm cm^{-2}$.  The observed HI velocity gradient across Leo~A in HI is so small that the changes in the velocity dispersion most likely reflect the conditions in the ISM rather than rotation or velocity crowding effects.  These HI observations show that the physical state of the ISM in Leo A is surprisingly similar to that in other, larger, more metal-rich galaxies (including our own), despite the fact that Leo~A is dominated by internal motions with very little effect coming from the exceedingly low global rotation measures.  The resolved stellar population of Leo~A has been studied before by Tolstoy~(1996) in the Thuan-Gunn filter system.  It was concluded that the SFR in the galaxy must have been higher in past than at the present time by a factor of $\\sim$~3.  Although the interpretation only went back $\\sim$1~Gyr because of the limitations of crowding and sensitivity of the ground based images.  This study adopted the distance as determined by H94. Our present observations favor a much smaller distance (see \\S 3.1), which has significant impact on the interpretation of the CMD.  The crowding in the ground based images resulted in the Red Giant Branch (RGB) population being misidentified as the Red Super Giant (RSG) population.  Here we show how the details of the SFH of the last few Gyr obtained from uncrowded WFPC2 imaging helps us understand the properties of this galaxy, and why we believe the variable stars that H94 identified are in fact W~Virginis (W~Vir) or Anomalous Cepheid (AC) variable stars. ",
        "conclusions": "Our data indicate that the majority of stars in the Leo~A dwarf irregular galaxy have ages of less than a few Gyr.  Is this reasonable? One way to test this conclusion is to see if other galaxies show independent evidence for star formation in the comparatively recent past.  For example, if our model is correct then we might expect to see a considerable range in mean stellar population ages among other extreme dwarf galaxies in the Local Group.  Since Leo~A is currently dim ($M_V = -$11.5 for our adopted distance; see Mateo 1998), we should be comparing it with the smallest dwarfs, which are mainly dwarf spheroidal (dSph) systems.  The stellar populations of Local Group dSph galaxies are being explored in detail through a combination of ground- and space-based observations (see reviews by Gallagher \\& Wyse 1994, Da~Costa 1994, 1997, Mateo 1998).  These investigations reveal that several of the Galactic dSph contain large complements of intermediate age stars, e.g., Carina (Smecker-Hane {\\it et al.} 1994) and Leo I (Lee {\\it et al.}  1993, Caputo, Castellani, \\& Degl'Innocenti 1996), while other dSph, such as the Ursa Minor system, are mainly composed of ancient, globular-cluster stellar populations.  Leo I has an especially prominent intermediate age stellar population component that was probably formed a few Gyr ago.  The dSph therefore display the kind of range in stellar ages that one might expect if major star-forming episodes have occurred sporadically in small galaxies during the past 10~Gyr or so.  That we are seeing Leo~A at a time when most of the stars have formed in the past few Gyr is evidently not a special case, but instead seems to be a relatively common phenomenon.   There remains a distinct possibility that Leo~A is a a purely young system. The ambiguities in interpreting the RGB mean that we still need to find unequivocal proof of an old population (e.g., RR~Lyr variable stars, or old MSTOs).  The new closer distance for Leo~A makes it possible to observe stars as faint as M$_V$=+4, equivalent to globular cluster age MSTOs. Stellar spectroscopy is also possible of RGB stars to try and look for evidence of metallicity evolution.  In any case, Leo~A is an excellent candidate for further study of star formation in a very low metallicity environment.  Even if it is does not contain solely a young population, it is still the nearest by example of what a FBG may look like."
    },
    "9805/astro-ph9805097_arXiv.txt": {
        "abstract": "The growth of structure from scale-free initial conditions is one of the most important tests of cosmological simulation methods, providing a realistically complex problem in which numerical results can be compared to rigorous analytic scaling laws.  Previous studies of this problem have incorporated gravitational dynamics and adiabatic gas dynamics, but radiative cooling, an essential element of the physics of galaxy formation, normally introduces a preferred timescale and therefore violates the conditions necessary for self-similar evolution.  We show that for any specified value of the initial power spectrum index $n$ [where $P(k) \\propto k^n$], there is a family of power-law cooling functions that preserves self-similarity by ensuring that the cooling time in an object of the characteristic mass $M_*$ is a fixed fraction $\\hat{t}_C$ of the Hubble time.  We perform hydrodynamic numerical simulations with an Einstein-de Sitter cosmology, a baryon fraction of $5\\%$, Gaussian initial conditions, two different power spectrum indices, and four values of $\\hat{t}_C$ for each index, ranging from no cooling to strong cooling.  We restrict the numerical simulations to two dimensions in order to allow exploration of a wide parameter space with adequate dynamic range.  In all cases, the simulations are remarkably successful at reproducing the analytically predicted scalings of the mass function of dissipated objects and the gas temperature distributions and cooled gas fractions in collapsed systems. While similar success with 3-D simulations must still be demonstrated, our results have encouraging implications for numerical studies of galaxy formation, indicating that simulations with resolution comparable to that in many current studies can accurately follow the collapse and dissipation of baryons into the dense, cold systems where star formation is likely to occur. ",
        "introduction": "Studies of evolution from scale-free initial conditions provide idealized but illuminating examples of the more general process of hierarchical structure formation.  So long as the background cosmology, input physics, and initial conditions are scale-free, even an inherently complex and highly nonlinear process such as gravitationally driven, hierarchical structure formation must evolve self-similarly in time.  Self-similar scaling offers a powerful analytic guide to the behavior of such complex systems, and investigations of scale-free clustering have yielded important insights concerning the growth of cosmological structure (e.g., Davis \\& Peebles 1977; Kaiser 1986; Efstathiou et al.\\ 1988). The evolution of scale-free initial conditions is also one of the few cosmological problems in which numerical simulations can be tested against rigorous analytic predictions, and the study of self-similar gravitational clustering by Efstathiou et al.\\ (1988) is one of the key pieces of evidence for the accuracy of cosmological N-body methods.  In Owen \\etal\\ (1998b; hereafter Paper I), we extended this approach to adiabatic gas dynamics, presenting a set of smoothed particle hydrodynamics (SPH) simulations designed to study self-similar evolution of structure in a mixed baryon/dark matter universe.  We found that the resulting population of collapsed structures demonstrated the expected self-similar scalings, so long as we were careful to account for the numerical limitations of each experiment.  However, while the models presented in Paper I included gravitational, pressure, and shock processes, they neglected the effects of radiative energy loss from the gas, a critical element in the formation and evolution of galaxies (White \\& Rees 1978).  In this paper we extend our earlier work to include radiative dissipation; for each choice of the scale-free initial power spectrum $P(k) \\propto k^n$, we construct artificial cooling laws that maintain the scale-free nature of the physics.  Our approach in this investigation differs from that of Paper I in a few important respects.  First, in Paper I we considered a set of 3-D simulations, while for this study we restrict ourselves to 2-D simulations. The restriction to 2-D allows us to perform a larger number of high dynamic range experiments than would be practical in 3-D. While 3-D simulations are clearly necessary for realistic studies of galaxy formation, for our present purposes we wish to study the effects of radiative dissipation in a variety of idealized, scale-free models, and 2-D experiments provide an economical starting point. Our investigation represents a first attempt to examine self-similar evolution in models that incorporate the physical processes most essential to galaxy formation: gravitational collapse and merging, shock heating, and radiative cooling.  We will use our results to guide the choice of parameters for more expensive, 3-D models in a future study.  Section 2 discusses the analytic scaling of characteristic group properties in 2-D models.  In Section 3 we derive the cooling functions that preserve self-similarity, for both 2-D and 3-D models.  Section 4 describes the simulations and their numerical limitations, and it presents our first important numerical result, the scaling of the mass functions of dark matter, baryon, and dissipated baryon groups.  Section 5 presents tests of mass, temperature, density, and Bremsstrahlung luminosity scaling, similar to those used in Paper I.  Section 6 examines the central issue of the paper, the scaling of the cooled baryon component in collapsed objects.  In Section 7 we summarize our conclusions and discuss their implications for numerical studies of galaxy formation. ",
        "conclusions": "We analyze a set of 2-D hydrodynamic simulations designed to study the self-similar evolution of hierarchical structure in the presence of radiative cooling.  The radiative cooling law for a primordial H/He plasma would introduce a preferred timescale in these models, so we instead use artificial cooling laws that maintain the scale-free nature of the physics: $\\Lambda(T) \\propto T^\\beta$ where $\\beta$ is a function (eq.~[\\ref{beta2d.eq}]) of the spectral index $n$ of the initial density fluctuations. We simulate eight distinct physical models, with $n=0$ and $n=+1$ initial power spectra and four different amplitudes of the corresponding radiative cooling laws, ranging from no cooling to very strong cooling. For each of these models we evolve a simulation with $128^2$ gas and $128^2$ dark matter particles, using ASPH to model the hydrodynamics and a Particle-Mesh technique to follow the gravitational interactions. We also repeat one model with $256^2$ gas and dark matter particles, to directly assess any numerical resolution dependencies.  Because both the physics and initial conditions of these models are scale-free, their physical properties must evolve self-similarly in time, in accord with analytic scaling laws. These rigorous analytic predictions provide a stringent test of our numerical methods, since numerical parameters like particle mass and force resolution stay fixed as the system evolves and will therefore act to break self-similar behavior if they limit the physical accuracy of the calculation. We test for self-similar evolution by identifying objects as groups of particles within a given overdensity contour at various times, measuring global properties characteristic of these structures, and examining the evolution of these quantities over a range of output times.  In general we find excellent agreement between the analytically predicted and numerically measured scalings for the masses, temperatures, Bremsstrahlung luminosities, and (to a lesser extent) gas densities over the range of expansion factors that we would naively expect to be accessible based on the simple resolution arguments presented in \\S \\ref{Sims.sec}. Our most significant result, demonstrated directly in Figure~\\ref{Mcoolfrac.fig} and elaborated in Figures~\\ref{fM.fig} and~\\ref{TtoTcfM.fig}, is that the fraction of baryonic material that cools in collapsed objects of specified mass follows the expected scaling with remarkable accuracy.  One of the impressive successes of hydrodynamic cosmological simulations with CDM initial conditions and realistic cooling laws is that they produce dense clumps of cold gas with masses, sizes, and overdensities comparable to the luminous regions of observed galaxies (Katz et al. 1992, 1996; Evrard et al.\\ 1994; Summers, Davis, \\& Evrard 1995; Frenk et al.\\ 1996; Weinberg et al.\\ 1997; Pearce 1998). With reasonable prescriptions for galactic scale star formation, these objects are converted into dense, tightly bound clumps of stars and cold gas.  The resulting stellar masses are not sensitive to the parameters of these prescriptions, at least within some range, because the star formation rate is governed mainly by the rate at which gas cools and condenses onto the central object (Katz et al.\\ 1996; Pearce 1998).  If numerical simulations are to provide accurate predictions of quantities like the galaxy luminosity function, star formation rates in high redshift systems, or even the bias between galaxies and mass, then they must accurately follow the gravitational collapse, shock heating, and subsequent dissipation and condensation of baryons into these dense systems.  Our results provide encouraging evidence that cosmological simulation methods can indeed rise to this challenge.  Specifically, they show that 2-D calculations that resolve individual systems with a few dozen to a few hundred particles correctly follow the formation of dissipated objects.  They also suggest that computing the galaxy baryon mass function may be an easier problem than computing the cluster X-ray luminosity function despite the higher densities of the objects in question and the greater complexity of the physics (with radiative cooling playing a larger role).  The $\\rho^2$ dependence of Bremsstrahlung emissivity implies that a simulation must resolve an object's central density in order to compute its X-ray luminosity accurately. However, thermal instability is a threshold phenomenon, and once a simulation resolves an object's density at the cooling radius (where the cooling time equals the dynamical time), it can compute the dissipated baryon mass with reasonable accuracy.  While it is clearly important to repeat the self-similar evolution tests with 3-D calculations, we see no reason that the behavior in three dimensions should be fundamentally different.  For nearly two decades, cosmological N-body simulations have provided an indispensable tool for predicting the large scale distribution of dark matter in different cosmological models.  It appears that the more ambitious goal of predicting the properties and distribution of galaxies with hydrodynamic simulations is now well within reach."
    },
    "9805/astro-ph9805118_arXiv.txt": {
        "abstract": "We report on a detailed analysis of the  correlation between the optical-UV and X-ray luminosities of quasars by means of Monte Carlo simulations, using a realistic luminosity function. We find, for a quasar population with an intrinsically constant,  mean X-ray loudness \\malpoxe, that the simulated \\alpoxe\\,--\\,$L_{\\rm o}$ relation can exhibit various `apparent' properties, including an increasing \\malpox with $L_{\\rm o}$, similar to what has been found from observations. The determining factor for this behavior turns out to be the relative strength of the dispersions of the luminosities, i.e.\\ their deviations from the mean spectral energy distribution at the  optical and X-ray bands, such that a dispersion larger for the optical luminosity than for the X-ray luminosity tends to result in an apparent correlation. We suggest that the observed  \\alpoxe\\,--\\,$L_{\\rm o}$ correlation can be attributed, at least to some extent, to such an effect, and is thus  not an underlying physical property. The consequences of taking into account the luminosity dispersions in an analysis of the observed luminosity correlations is briefly discussed. We note that similar considerations might also apply for the Baldwin effect. ",
        "introduction": "A study of the dependence of the spectral energy distribution (SED) of quasars  on their luminosity and/or on cosmic epoch is particularly important for understanding the quasar phenomenon. In the optical-to-X-ray regime, the SED can be characterized by the broad band spectral index between 2500{\\AA} and 2\\,keV, which is defined as $\\alpha_{\\rm ox} = - 0.384 \\log (L_{\\rm2keV}/L_{\\rm2500{\\AA}})$. Quasars are known to exhibit strong luminosity evolution in the X-ray and the optical wave bands (e.g.\\ Boyle 1994). However, there have been controversial discussions in the past as to whether the evolution law is the same in these two energy bands. A dependence of \\alpox on redshift or optical luminosity would indicate different evolution in the optical and the X-ray regime. Further, if \\alpox depends on optical luminosity, this is equivalent to a non-linear relationship between X-ray and optical luminosity ($L_{\\rm x}\\propto L_{\\rm o}^{\\rm e}$, $e \\not=1$).  While most of the analyses agree on the result that \\alpox is redshift independent, it has been claimed  that \\alpox increases with $L_{\\rm o}$, which implies that the objects with high optical luminosities are under-luminous in X-rays compared to their low luminosity counterparts (Avni \\& Tananbaum 1982, 1986; Kriss \\& Canizares 1985; Wilkes \\eta\\ 1994; Avni \\eta\\ 1995; Green \\eta\\ 1995). Generally, for a functional dependence of the form \\alpoxe\\,$\\sim \\beta \\log L_{\\rm o}$, a canonical slope of $\\beta\\sim 0.1$ was obtained, which is equivalent to a non-linear relation of the form $L_{\\rm x} \\propto L_{\\rm o}^{0.7}$.  Based on Monte Carlo simulations, Chanan (1983) suggested that a non-linear relation might arise even for an intrinsically  linear dependence from observational flux limits and the large intrinsic scatter in the data. Thus, the observed \\alpoxe\\,--\\,$L_{\\rm o}$ correlation should not be considered as an underlying physical reality. The author also claimed that the choice of $L_{\\rm o}$ as the independent variable is not justified. However, Kriss \\& Canizares (1985) criticized these results by pointing out that they depend critically on the assumption of a Gaussian distribution for the luminosity functions.  A study by La~Franca \\eta\\ (1995) reinforced the idea of a linear relationship between the X-ray and the optical luminosity for quasars. They applied a regression algorithm to a large sample of quasars detected with {\\it Einstein}, which accounts for errors in both variables and the intrinsic scatter in the data, and found $L_{\\rm x} \\propto L_{\\rm o}$.  In a recent study of ROSAT detected quasars by Brinkmann \\eta\\ (1997), it has been shown by means of a simple argument that an apparent correlation between \\alpox and $\\log L_{\\rm o}$ can indeed emerge even for intrinsically uncorrelated variables. Motivated by this idea, as well as by recent improvements concerning the shape of the quasar luminosity functions in the optical and the X-ray regime, we carry out a detailed study of this controversial problem by means of a Monte Carlo analysis. We mostly use the logarithms of luminosities  and denote them as $\\mbox{\\lxe}\\,=\\log L_{\\rm x}$ and $\\mbox{\\loe}\\,=\\log L_{\\rm o}$. We use $q_{\\rm 0} = 0.5$, $H_{\\rm0} = 50~{\\rm kms^{-1}Mpc^{-1}}$ throughout this paper. All errors quoted are at the $1 \\sigma$ level unless mentioned otherwise. ",
        "conclusions": "We have performed Monte Carlo simulations to study the luminosity correlation between the optical and X-ray bands for quasars. We have used a generalized model in which the luminosities in the two wave bands are represented in terms of the respective expected luminosities with dispersions ($l_{\\rm o} = \\overline{l}_{\\rm o} + \\delta\\l_{\\rm o}$, $ l_{\\rm x} = \\overline{l}_{\\rm x} + \\delta\\l_{\\rm x}$). We have shown that the increase of \\alpox with $L_{\\rm o}$ (equivalent to $L_{\\rm x} \\propto L_{\\rm o}^e$ with $e<1$), as found in observational data, can emerge in a sample with an intrinsic luminosity independent \\alpox  (or $L_{\\rm x} \\propto L_{\\rm o}$), provided that the dispersion of the optical luminosities deviating from the average SED are similar to or larger than that of the X-ray luminosities. Our simulations  verified the results of  Chanan (1983), which were achieved for special assumption about the luminosity functions. We suggest that the {\\em observed} \\alpoxe\\,--\\,$L_{\\rm o}$ correlation is, at least to a large extent, apparent and not necessarily an intrinsic property of the quasar population.  Our model is more general than  previous considerations. For $\\mbox{\\Re}\\ll 1$ (implying $l_{\\rm o} \\sim \\overline{l}_{\\rm o}$ and $\\sigma_{\\mbox{\\alpoxe}} \\sim 0.384 \\sigma_{\\rm x}$) the model reduces to the commonly used description, in which \\lo is the primary luminosity and the dispersion in the SED is attributed to that in the X-ray luminosity. The same  holds for $\\mbox{\\Re}\\gg 1$, but with interchanged roles of \\lo and \\lxe. We argue that the effect of the relative strength of the individual luminosity dispersions in the two bands should be taken into account in analyses of quasar luminosity correlations. Since the arguments are valid for any other two wave bands, we expect this effect to play a role in luminosity correlations between radio, infrared, optical, and X-ray wave bands as well. The determination of the \\Re-parameter, and thus of the luminosity scatter in the individual wave bands, is important to understand the broad band emission of quasars.  We finally note that a similar effect as presented for the $\\alpha_{\\rm ox} - l_{\\rm o}$ correlation might also apply for the well-known Baldwin effect, i.e.\\ the inverse correlation of optical emission line equivalent width with optical luminosity (Baldwin 1977). Since the equivalent width basically can be regarded as the ratio of two luminosities (emission line and underlying continuum), the structure of the problem is similar to the one presented in this paper. This is particularly interesting, because there still is no accepted physical explanation for the Baldwin effect."
    },
    "9805/astro-ph9805083_arXiv.txt": {
        "abstract": "If the Galaxy contains $\\sim\\!10^{11}\\Msol$ in cold gas clouds of $\\sim$Jovian mass and $\\sim$AU size, these clouds will act as converging lenses for optical light, magnifying background stars at a detectable rate. The resulting light curves can resemble those due to gravitational lensing by a point mass, raising the possibility that some of the events attributed to gravitational microlensing might in fact be due to ``gaseous lensing''. During a lensing event, the lens would impose narrow infrared and far-red $\\HH$ absorption lines on the stellar spectrum. Existing programs to observe gravitational microlensing, supplemented by spectroscopy, can therefore be used to either detect such events or place limits on the number of such gas clouds present in the Galaxy. ",
        "introduction": "A number of authors have proposed that the Galaxy could contain a hitherto-unrecognized population of small, cold, dense self-gravitating gas clouds, in numbers sufficient to contribute an appreciable fraction of the gravitational mass of the Galaxy (Pfenniger, Combes, \\& Martinet 1994; Gerhardt \\& Silk 1996; Combes \\& Pfenniger 1997).  Walker \\& Wardle (1998; hereafter WW98) pointed out that if such clouds existed in the Galactic halo, each would have an ionized envelope which could explain the ``Extreme Scattering Events'', or ``ESEs'', (Fiedler et al. 1987) during which extragalactic point radio sources occasionally undergo substantial frequency-dependent amplification and deamplification, apparently due to refraction by a plasma ``lens'' moving across the line-of-sight. WW98 proposed that the observed frequency of such ESEs could be explained if there was a population of $\\sim\\!10^{14}$ cold self-gravitating gas clouds, with mass $M\\approx 10^{-3}\\Msol$, and radius $R\\approx 3\\AU$.  Gerhardt \\& Silk (1996) and WW98 noted that if the clouds were opaque, their presence would have been revealed by existing stellar monitoring programs studying gravitational ``microlensing'' (Paczynski 1986; see the review by Paczynski 1996, and references therein), as these experiments would have detected occultation events of duration $\\sim\\! R/200\\kms\\approx 40 {\\rm \\,days}$. Such occultation events have not been reported. However, the hypothesized clouds could be essentially transparent at optical wavelengths: they could have formed from primordial gas, or, if formed from gas containing metals, the grains could have sedimented to form a small core.  In this {\\it Letter} we point out that even transparent clouds would have lensing effects which would be detectable by the stellar monitoring studies currently underway to study gravitational lensing by compact objects in our Galaxy. Existing data can thus test the hypothesis that cold gas clouds contribute an appreciable fraction of the mass of the Galaxy. ",
        "conclusions": "It is not obvious how cold self-gravitating gas clouds with the properties suggested by WW98 might have formed, or whether such clouds would be stable for $\\sim\\!10^{10}\\yr$. However, if they do exist, WW98 show that they could solve two longstanding problems: (1) their ionized envelopes could account for some of the ``Extreme Scattering Events''; and (2) they could contain the ``missing'' baryons in the Galaxy. It is notable that these same clouds could ameliorate a third problem: the fact that microlensing searches detect a larger number of amplification events toward the LMC than expected for lensing by stars and stellar remnants. Some of these events could be due to gaseous lensing. Indeed, lensing by the hypothesized clouds would be so frequent that existing programs to observe gravitational microlensing can already place strong limits on the cloud parameters (see Fig.\\ \\ref{fig:forbid}), but clouds with $M\\approx10^{-3}\\Msol$ and $R\\approx10\\AU$ are still allowed.  With a predicted lensing rate $\\dot{P}_{\\rm OL}\\approx 2\\!\\times\\!10^{-4}\\yr^{-1}$, the typical lens must be weak, with $S\\ltsim 0.3S_c$. The distribution of cloud properties and distances could produce occasional strong gaseous lensing events with $S\\gtsim 1$, perhaps accounting for some of the events attributed to gravitational microlensing.  For non-caustic lensing, the dispersive effects of the gas would produce slightly larger amplification in the blue, but this is counteracted by Rayleigh scattering by the $\\HH$. A number of quadrupole lines of $\\HH$ would be detectable in absorption during the lensing event; this would be the most unambiguous signature of ``gaseous lensing''.  Existing programs to observe gravitational lensing, supplemented by spectroscopy during lensing events, can therefore be used to either detect gaseous lensing events or place limits on the number of $\\sim\\!10^{-3}\\Msol$ $\\HH$ clouds in the Galaxy."
    },
    "9805/astro-ph9805030_arXiv.txt": {
        "abstract": "The determination of the ``Fundamental Parameters''  $T_{\\rm eff}$ and  $\\log g$ for a set of dwarf A0-type stars is discussed in terms of consistency when comparing these values determined through different methods. The position of these stars in the HR diagram are discussed, taking into account the HIPPARCOS data. A large number of binary stars with components of similar spectral types has been found from this spectroscopic survey. ",
        "introduction": "\\label{intr} Our purpose is to analyse a set of A0-type, non-giant stars, in order to define a sample to be used as {\\it standard stars} for further studies. These stars have parallaxes mesured from the HIPPARCOS experiment and their position in the HR diagram will be used to derive their evolutionary status. The spectra obtained in the H$_{\\gamma}$ region are used to detect peculiarities as well as to check the fit with theoretical spectra computed with $T_{\\rm eff}$ and $\\log g$ values determined from calibrations of colour indices. ",
        "conclusions": "\\vspace{-2mm} With the subsample of normal stars selected above, we have plotted the HR diagram (Fig. 5). In abscissa, $T_{\\rm eff}$ is that from the MD calibration, and $M_{\\rm V}$ is derived from the HIPPARCOS parallax data. We remark that there is no direct relation between the luminosity class given by the MK classification and the $M_{\\rm V}$ value; stars belonging to luminosity class V cover a broad range of $M_{\\rm V}$. The evolutionary tracks by Claret \\& Gim\\'enez (1992) are overplotted on this diagram.   The spread on Fig. 5 points out the difficulty to calibrate the spectral type A0 and luminosity class V in terms of $M_{\\rm V}$. This spread is so large that it prevents the detection of binaries simply from an anomalous position in the HR diagram of stars for which only the spectral classification is known.  \\vspace{-2mm}"
    },
    "9805/astro-ph9805206_arXiv.txt": {
        "abstract": "We report the first identification of the Eu\\,{\\sc iii} $\\lambda$ 6666.317 line in optical spectra of CP stars. This line is clearly present in the spectra of HR 4816, 73 Dra, HR 7575, and $\\beta$ CrB, while it is marginally present or absent in spectra of the roAp stars $\\alpha$ Cir, $\\gamma$ Equ, BI Mic, 33 Lib, and HD 24712. ",
        "introduction": "Magnetic Chemically Peculiar stars (CP2 stars) are known to have large overabundances of rare-earth elements (REE) in their atmospheres. Among all REE, europium shows the most prominent overabundances of up to +5.0 dex, in many CP2 stars violating the odd-even pattern observed in the solar atmosphere. In the atmospheres of many CP2 stars the dominant europium ion is Eu\\,{\\sc iii}. The strongest lines of Eu\\,{\\sc iii} are located in the UV region. A few relatively intense lines are observed in the optical spectral region (Sugar \\& Spector 1974). This fact justified a careful study of a few CP2 stars in the spectral region 6620-6680 \\AA, where unblended lines of both Eu\\,{\\sc ii} $\\lambda$ 6645.05 and the strongest optical Eu\\,{\\sc iii} $\\lambda$ 6666.35 are located. ",
        "conclusions": ""
    },
    "9805/astro-ph9805176_arXiv.txt": {
        "abstract": "We sing the praises of the central limit theorem.  Having previously removed all other possible causes of significant systematic error in the statistical parallax determination of RR Lyrae absolute magnitudes, we investigate systematic errors from two final sources of input data: apparent magnitudes and extinctions.  We find corrections due to each of about 0.05 mag, i.e., about half the statistical error.  However, these are of opposite sign and so approximately cancel out.  The apparent magnitude system that we previously adopted from Layden et al.\\ was calibrated to the photoelectric photometry of Clube \\& Dawe.  Using Hipparcos photometry and archival modern ground-based photometry, we show that the Clube \\& Dawe system is about 0.05 mag too bright.  Extinctions were previously based on the map of Burstein \\& Heiles which was constructed from HI maps.  We argue that extinctions should rather be estimated using the new map of Schlegel, Finkbeiner \\& Davis based on {\\it COBE} and {\\it IRAS} measurements of dust emission.  This substitution increases the mean estimated extinction by about 0.05 mag, primarily because of a difference in the zero point of the two maps.  Our final estimate for the absolute magnitude is $M_V=0.77\\pm0.13$ at [Fe/H]$=-1.60$ for a pure sample of 147 halo RR Lyrae stars, or $M_V=0.80\\pm 0.11$ at [Fe/H]$=-1.71$ if we incorporate kinematic information from 716 non-kinematically selected non-RR Lyrae stars from Beers \\& Sommer-Larsen.  These are $2\\sigma$ and $3\\,\\sigma$ fainter than recent determinations of $M_V$ based on main-sequence fitting of clusters using Hipparcos measurements of subdwarfs by Reid and Gratton et al.  Since statistical parallax is being cleared of systematic errors and since the probability of a more than $2\\,\\sigma$ statistical fluctuation is less than 1/20, we conclude that these brighter determinations may be in error.  In the course of these three papers, we have corrected 6 systematic errors whose absolute values total 0.20 mag.  Had these, contrary to the expectation of the central limit theorem, all lined up one way, they could have resolved the conflict in favor of the brighter determinations.  In fact, the net change was only 0.06 mag. ",
        "introduction": "Statistical parallax appears to be an extremely robust method for measuring the absolute magnitude of halo RR Lyrae stars.  Nevertheless, the results of this method are in serious conflict with several other determinations.  In Paper II of this series (Popowski \\& Gould 1998b) we found \\be M_V=0.74\\pm 0.12,\\quad {\\rm at}\\ \\left<\\rm [Fe/H]\\right>=-1.60 \\qquad ({\\rm pure}\\ {\\rm RR}\\ {\\rm Lyrae}),\\label{eqn:pureRR} \\ee for a sample of 165 halo RR Lyrae stars with high-quality proper motions from the Hipparcos (ESA 1997) and Lick NPM1 (Klemola, Hanson, \\& Jones 1993) surveys.  We also combined this result with a separate determination based on a non-kinematically selected sample of 103 RR Lyrae stars and 724 non-RR Lyrae stars from Beers \\& Sommer-Larsen (1995) and (taking account of the 0.45 correlation coefficient between the two samples) found \\be M_V=0.77\\pm 0.10,\\quad {\\rm at}\\ \\left<\\rm [Fe/H]\\right>=-1.71 \\qquad ({\\rm combined}).\\label{eqn:combined} \\ee The former value can be compared with measurements based on main-sequence fitting of globular clusters to subdwarfs with Hipparcos parallaxes which yields  $M_V\\sim 0.44\\pm 0.08$ (Reid 1997) or $M_V\\sim 0.49 \\pm 0.04$ (Gratton et al.\\ 1997; Gratton 1998) at the same metallicity. (These comparisons take account of differences in the metallicity scales used by different authors as we discuss more fully in the Appendix.)\\ \\  If equation (\\ref{eqn:pureRR}) is combined with the measurement of the dereddened apparent magnitude of RR Lyrae stars in the Large Magellanic Cloud (LMC) of $V_0 = 18.98\\pm 0.05$ (Hazen \\& Nemec 1992; Popowski \\& Gould 1998a -- Paper I), this yields a distance modulus $\\mu_\\lmc = 18.24\\pm 0.14$. (Here we have assumed an LMC metallicity [Fe/H]$=-1.8$, and a slope $M_V =$const $+0.15$[Fe/H], but the exact value of the slope makes very little difference because the metallicities in eq.\\ \\ref{eqn:pureRR} and of the LMC are so similar.)\\ \\  This result is quite low compared to the ``traditional'' value $\\mu_\\lmc = 18.50$ and is even lower compared to those derived using Hipparcos-based calibrations of RR Lyrae stars and Cepheids: $\\mu_\\lmc = 18.65\\pm 0.1$ (Reid 1997), $\\mu_\\lmc = 18.63\\pm 0.06$ (Gratton et al.\\ 1997), and $\\mu_\\lmc = 18.70\\pm 0.10$ (Feast \\& Catchpole 1997).  In principle, these discrepancies could be due to a greater than $2\\,\\sigma$ statistical fluctuation.  However, for Gaussian statistics, the probability of a $2\\,\\sigma$ fluctuation is $<1/20$. (Moreover, for the statistical parallax determination, we have checked that the distribution of errors has Gaussian tails, even when the input data are not Gaussian distributed.)\\ \\ The usual cause of $>2\\,\\sigma$ discrepancies is not statistical fluctuations but systematic errors, and one is therefore led to suspect that there are unrecognized systematic errors in one or several of these measurements.  Moreover, the conflict with equation (\\ref{eqn:combined}) is even stronger, about $3\\,\\sigma$.  While there are some additional assumptions that go into equation (\\ref{eqn:combined}) that make it overall less robust than equation (\\ref{eqn:pureRR}), the combined determination nevertheless argues against a large statistical fluctuation as the source of the discrepancy.  This is the third and final paper in a series designed to essentially eliminate the possibility of a significant systematic error in the statistical parallax determination.  Statistical parallax works in effect by forcing equality between the velocity ellipsoids as determined from radial velocities, and from proper motions.  That is, one can measure the nine parameters describing the velocity ellipsoid (three components of bulk motion, $w_i$, plus six independent components of the velocity-dispersion tensor, $C_{i j}$) from radial velocities alone.  On the other hand, if one {\\it assumes} some arbitrary absolute magnitude for the RR Lyrae stars, then one can infer their distances from their measured apparent magnitudes and estimated extinctions.  The distances and proper motions yield the transverse velocities, and from these one can again estimate the nine parameters of the velocity ellipsoid.  One could then adjust the assumed absolute magnitude so that the velocity ellipsoid from proper motions matched the velocity ellipsoid from radial velocities as closely as possible.  In practice, one fits for all 10 parameters (nine for the velocities plus the absolute magnitude) simultaneously using maximum likelihood.  Logically, there are three possible ways for systematic errors to enter the determination.  First, the {\\it mathematics} of the method itself could introduce biases.  Second, the RR Lyrae sample could fail to satisfy some of the {\\it physical} properties assumed by the method. Third, one or more of the four major {\\it observational inputs} (proper motions, radial velocities, apparent magnitudes, and extinctions) could be systematically in error.  (A fifth observational input, metallicities, requires a separate discussion.  Different studies may be on systematically different metallicity scales, and care must therefore be exercised when comparing the results from these investigations. See Appendix.)\\ \\  In Paper I, we investigated possible systematic errors arising from the mathematical method and physical assumptions.  An example of a potential mathematical problem is that the likelihood method explicitly assumes that the velocity distribution is Gaussian while, as we showed, the actual distribution is highly non-Gaussian. An example of a potential physical problem is that the method implicitly assumes that the velocity-dispersion tensor does not depend on location despite the fact that the stars are found at distances ($\\la 2\\,$kpc) that are a significant fraction of the Galactocentric distance ($R_0\\sim 8\\,$kpc). We examined a large number of such effects, some by vigorous Monte Carlo simulations and some with the aid of mathematical arguments.  We corrected for all of them although most were smaller than 0.01 mag and tended to mutually cancel one another.  The largest correction (0.03 mag fainter) was due to Malmquist bias which had been previously recognized but not previously incorporated into the analysis.  In Paper II, we investigated systematic errors arising from the first two observational inputs, proper motions and radial velocities. The proper motions are of greater concern because they are intrinsically more difficult to measure and hence have larger fractional errors.  We had already noted in Paper I that if the proper-motion {\\it errors} are misestimated, this can introduce significant systematic errors even if the proper motion themselves are unbiased.  We used the precise Hipparcos proper motions (when available) to test the two large catalogs, Lick and Wan, Mao, \\& Ji (1980, WMJ), that had previously been used and found that indeed the Lick errors had been slightly underestimated and the WMJ errors had been seriously underestimated.  These two corrections moved $M_V$ brighter by 0.04 mag, but this was mostly compensated by random changes induced by substituting the more precise Hipparcos proper motions (when available) for the previous values.  We also tested all three catalogs to search for non-statistical errors and removed five questionable stars.  Radial velocities are in principle much easier to measure than proper motions.  However, for pulsating variables, the measured velocity of (the atmosphere of) the star can differ from its center of mass by $\\sim 50\\,\\kms$ and hence an accurate velocity determination requires many measurements and/or good phasing.  The quality of the radial velocity data varies from star to star, and it was therefore possible that the errors had been either systematically overestimated or underestimated.  In Paper II, we checked the entire system of the radial-velocity measurements by, in effect, determining the radial-velocity ellipsoid from the Beers \\& Sommer-Larsen (1995) non-kinematically selected sample of metal-poor {\\it non-RR Lyrae} halo stars.  The resulting $M_V$ was consistent with the one derived from the pure RR Lyrae sample, indicating that the radial velocities are not a source of significant systematic error.  In brief, Paper I checked for and removed all sources of systematic error coming from the mathematics of the method and the physical assumptions about the sample, down to a level well below the statistical error.  Paper II did the same for two of the observational inputs: proper motions and radial velocities.  Here we turn our attention to the remaining two observational inputs: apparent magnitudes and extinctions.  At first sight, it does not seem that there could be much controversy about the apparent magnitude of $V\\sim 12$ stars.  However, exactly because the stars are bright, many were measured long ago.  Layden et al.\\ (1996) compiled photometric measurements from several sources and attempted to put them on a common system aligned with their large subsample from Clube \\& Dawe (1980) which has photoelectric photometry and which they assumed to be equivalent to the modern (Landolt 1992) system.  In particular, they found the photoelectric photometry of Bookmeyer et al.\\ (1977) to be on average 0.06 mag fainter than that of Clube \\& Dawe (1980) and transformed it accordingly (see Layden et al.\\ 1996, Table 1).  Thus, there are uncertainties in the apparent-magnitude scale of order 0.06 mag which, according to equation (\\ref{eqn:pureRR}), is half the size of the statistical error.  In \\S\\ 3, we test the Layden et al.\\ (1996) system against Hipparcos photometry.  We show that for $V\\la 12$, the {\\it untransformed} Bookmeyer et al.\\ (1977) photometry is in good agreement with Hipparcos.  The Clube \\& Dawe (1980) photoelectric photometry also agrees well with Hipparcos for $V\\la 10.5$ but is systematically brighter than Hipparcos by $\\sim 0.06$ mag for $11\\la V \\la 12$.  The most straight forward interpretation of these results is that the Bookmeyer et al.\\ (1977) photometry is more reliable than the Clube \\& Dawe (1980) photometry and that therefore the Layden et al.\\ (1996) system is too bright by about 0.06 mag. This conclusion is confirmed by the good agreement between Hipparcos and the high quality photometry of Jones et al.\\ (1992), Schmidt (1991), and Schmidt, Chab, \\& Reiswig (1995).  Extinctions pose another set of problems. For stars that are far from the Galactic plane, one can assume that they are above essentially all of the dust along their line of sight. One can therefore adopt the extinctions as measured for extragalactic objects along the same (or very nearby) lines of sight.  Burstein \\& Heiles (1982, BH) have constructed a map of such extinctions over a large fraction of the sky by combining galaxy counts and HI measurements.  Layden et al.'s (1996) extinction estimates are based primarily on this map for the great majority of the sample. However, there are some lines of sight (particularly at low latitudes) for which BH do not give extinctions and others where the star is relatively close to the plane so that some of the dust may lie behind the star.  In the latter cases, the BH map would overestimate the extinctions.  For these stars, Layden et al.\\ (1996) adopted other methods to estimate the extinction, notably the colors of the stars.  Since the intrinsic color of RR Lyrae stars is a function of the period with relatively little scatter, this method should work well at least on average.  However, Sturch (1966) had earlier used colors to estimate the extinctions toward a large sample of RR Lyrae stars, and Layden et al.\\ (1996) found that these estimates were systematically higher than their BH-based values by 0.11 mag.  Layden et al.\\ (1996) argued that the BH-based system was correct, and attempted to put stars without BH extinctions on the same BH system. Nevertheless, it is important to note that if the Sturch (1966) system were correct, the dereddened apparent magnitudes would be systematically brighter by 0.11 mag, and hence the absolute-magnitude estimate would be brighter by the same amount.  This would move the statistical-parallax estimates of $M_V$ and $\\mu_\\lmc$ closer by $1\\,\\sigma$ to the estimates obtained using competing methods, and would thus help significantly to resolve the controversy.  In \\S\\ 4, we therefore re-evaluate the extinctions using a different approach.  First, we base our determinations on the new extinction map of Schlegel, Finkbeiner, \\& Davis (1998, SFD). We argue that the SFD map is superior to the BH map both in its level of detail and in its zero point.  (The zero point of SFD is about 0.06 mag higher in $A_V$ than the BH map.)\\ \\ Second, we restrict attention to stars that are more than 300 pc from the Galactic plane.  These lie beyond most of the dust and therefore the SFD extinctions require only small corrections. Third, we exclude the four stars with SFD extinctions $A_V>0.56$ since comparison with Layden et al.\\ (1996) shows a systematic deviation for these stars and we are unable to determine which system is in error. We find a correction due to revised extinctions which makes $M_V$ about 0.05 mag brighter.  Combining the corrections due to revised apparent magnitudes and extinctions, we find that equations (\\ref{eqn:pureRR}) and (\\ref{eqn:combined}) are each increased (made fainter) by 0.03 mag.  These changes are in fact mainly due to random fluctuations caused by the fact that we are using a slightly different sample (147 vs.\\ 165 stars).  The two systematic effects that we identify here almost precisely cancel out. Our results are all presented in \\S\\ 5, and we discuss the implications of these results in \\S\\ 6.  We begin in \\S\\ 2 by describing our basic sample. ",
        "conclusions": "We take equation (\\ref{eqn:pureRR2}) as the primary result of this paper because the statistical-parallax solution for the pure RR Lyrae sample requires essentially no additional assumptions.  By contrast, the non-kinematic solution of equation (\\ref{eqn:nonkin}), and by implication the combined solution of equation (\\ref{eqn:combined2}), require the additional assumption that metal-poor RR Lyrae stars have the same kinematics as metal-poor non-RR Lyrae stars.  While this assumption is not absolutely secure, there are a number of very strong arguments in its favor.  First, as we discussed above and in Paper II, there is no evidence that RR Lyrae and non-RR Lyrae stars can be distinguished kinematically.  Second, the available evidence suggests that kinematics are independent of metallicity for [Fe/H] $<-1.5$.  Third, the pure RR Lyrae and non-kinematic solutions for $\\eta$ agree within their errors, even taking account of the 0.44 correlation coefficient between them (see Tables 1 and 2).  Fourth, there is no statistically significant difference between the {\\it individual velocity components} of the solution in Table 1 for 147 RR Lyrae stars and the individual velocity components of the solution based {\\it only} on the 716 metal-poor Beers \\& Sommer-Larsen (1995) stars.  The latter differs somewhat from the solution in Table 2, and is given by $w_i = (-2.2\\pm 9.6, 38.3\\pm 11.0,1.1\\pm 5.5)\\,\\kms$ and $C_{i i}^{1/2} = (160.0\\pm 10.1 ,118.7\\pm 13.1,92.6\\pm 6.1)\\,\\kms$.  Taking the difference between these six parameters and those in Table 1 and dividing by the errors yields $(0.5, -1.4, -0.2, 0.8, -1.2, -0.3)$. Of these six, only one ($w_\\theta$) is possibly inconsistent with a normal statistical fluctuation.  However, this is the {\\it one} component that we expect to be different because the pure RR Lyrae sample was selected by removing stars with prograde orbits, while the non-kinematic sample was, of course, selected without kinematic criteria.  Finally, if there were any systematic difference between the RR Lyrae stars and the non-RR Lyrae stars used in the non-kinematic sample, we would expect that it would be in the sense of the non-RR Lyrae stars having more extreme kinematics because they are on average more metal poor.  This would drive the radial-velocity ellipsoid to higher dispersions and faster (relative to the Sun) bulk motion, and hence would cause one to {\\it overestimate} distances (and luminosities) of the RR Lyrae stars when one attempted to match their proper motions to these high, non-RR Lyrae radial velocities.  That is, the only plausible bias of this method is in the {\\it same direction} as would be needed to resolve the discrepancy between statistical parallax and other methods of determining the absolute magnitude of RR Lyrae stars and {\\it opposite in sign} from the actual difference between the non-kinematic and pure RR Lyrae samples. In brief, while equation (\\ref{eqn:combined2}) does not sit on as firm a foundation as equation (\\ref{eqn:pureRR2}), it does argue very strongly against the idea that equation (\\ref{eqn:pureRR2}) is the result of a large statistical fluctuation, particularly a fluctuation in the direction of underestimating the RR Lyrae luminosity.   As we discussed in the introduction, equation (\\ref{eqn:pureRR2}) is in conflict at the $2\\,\\sigma$ level with the values determined from main-sequence fitting of clusters at the same metallicity of $M_V\\sim 0.44\\pm 0.08$ (Reid 1997) or $M_V\\sim 0.49 \\pm 0.04$ (Gratton et al.\\ 1997; Gratton 1998). There are only four possible explanations for such a discrepancy: 1) a rare $(<1/20)$ statistical fluctuation, 2) a substantial difference between cluster stars and field stars in the magnitude of the horizontal branch, 3) a systematic error in the main-sequence fitting distances to clusters, or 4) a systematic error in the statistical parallax measurement.  In this series of three papers, we have eliminated explanation (4). Explanation (1) is of course always possible, but is unlikely. Gratton (1998) has suggested explanation (2), that field and cluster horizontal branches might be different.  However, two lines of evidence weigh against this possibility.  First, as Gratton (1998) notes, comparison of the apparent magnitudes of RR Lyrae stars in LMC clusters with those of neighboring field RR Lyrae stars (for which the reddening should be quite similar) shows a mean offset of only $0.05\\pm 0.02$.  However, this argument strictly applies only to LMC RR Lyrae stars: there still could be a difference between field and cluster RR Lyrae stars in the Galaxy which, unlike the LMC, is a large spiral and probably has had quite a different formation history.  In fact, Sweigert \\& Catelan (1998) have produced models of two clusters with rising blue horizontal branches (NGC 6388 and NGC 6411) that have RR Lyrae stars several tenths of a magnitude brighter than those of canonical horizontal branch scenarios. The models invoke non-standard features, either high helium abundance, high rotation velocity, or helium mixing at the tip of the giant branch, which cause the stars to have longer periods at fixed temperature and metallicity.  Such long periods are actually observed for the two known RR Lyrae stars in NGC 6388, enhancing the plausibility of this explanation.  However, Catelan (1998) has shown that the period-temperature diagrams for RR Lyrae stars in five clusters that have been used for main-sequence fitting are actually quite similar in appearance to those of field stars of similar metallicity.  Hence, while some cluster horizontal branches may be brighter than those of the field, this does not appear to be the case for the clusters with main-sequence fitting distances.  We therefore consider that explanations (1), (2), and (4) are all rather unlikely and that a systematic error in the main-sequence fitting distances is the most plausible explanation for the discrepancy.  One possible cause of a systematic error in the main-sequence fitting distances is that the metallicities of the local subdwarfs might be on a different scale from those of the clusters (determined from giants).  Specifically, if the subdwarf metallicities were too low (or the giant metallicities too high) then intrinsically brighter subdwarfs would be matched to the cluster main sequences, leading to an overestimate of the cluster distance and of the luminosity of its horizontal branch.  Recently, King et al.\\ (1998) have found intriguing evidence of a possible misalignment of this sort.  They measured the metallicities of M92 subgiants (not quite subdwarfs, but with higher gravities than giants) and obtained metallicities up to half a dex lower than those of M92 giants.  While there are a number of possible explanations for this result, one is that the metallicities of giants are being systematically overestimated or those of subdwarfs are being systematically underestimated."
    },
    "9805/astro-ph9805340_arXiv.txt": {
        "abstract": "In this paper we critically examine predictions of the Ly$\\alpha$ forest within the standard cold dark matter (SCDM) model, paying particular attention to the low end of the column-density distribution.  We show in particular that the width of these lines, typically measured by the $b$-parameter of a Voigt profile, is sensitive to spatial resolution in numerical simulations and has previously been overestimated.  The new result, which predicts a distribution with a median $b$ of around 20-22 km/s at $z=3$, is substantially below that observed. We examine a number of possible causes of this discrepancy and argue that it is unlikely to be rectified by an increase in the thermal broadening of the absorbing gas, but is instead telling us something about the distribution of matter on these scales.  Although the median differs, the shape of the $b$-parameter distribution agrees quite well with that observed, and the high-end tail is naturally produced by the filamentary nature of gravitational collapse in these models.  In particular, we demonstrate that lines of sight which obliquely intersect a filament or sheet tend to produce absorption lines with larger $b$ parameters.  We also examine the physical nature of the gas which is responsible for the forest, showing that for lines with neutral column densities below $N_{HI} \\sim 10^{14}$ cm$^{-2}$ (for this model at $z=3$), the peculiar infall velocity is actually slower than the Hubble flow, while larger lines have, on average, turned around and are collapsing. ",
        "introduction": "A physical picture of the Ly$\\alpha$ forest in Cold Dark Matter (CDM) dominated cosmologies has recently emerged from numerical simulations (e.g. \\cite{cen94}; \\cite{zha95}; \\cite{her96}), and other approximation techniques (\\cite{bi93}; \\cite{bi97}; \\cite{hui97}; \\cite{gnehui97}).  In this context, the absorbers that give rise to low column density lines ($N_{HI} < 10^{15}$ cm$^{-2}$) at $z \\sim 3$ are large, unvirialized objects with sizes of $\\sim$ 100 kpc, and low densities, comparable to the cosmic mean (\\cite{zha97b}).  The width of the lines, as measured by the $b$ parameter of a Voigt profile is set not only by thermal broadening or peculiar velocities, but also by the Hubble expansion across their width (\\cite{wei97}).  In such a situation, there is a monotonic relationship between baryonic density and optical depth which can be exploited to investigate the density distribution along the quasar lines of sight (e.g. \\cite{cro97}).  This explanation of the forest (for previous ideas along this direction, see also \\cite{bon88}, \\cite{mcg90}, \\cite{bi93}, and \\cite{mei94}), arising from gravitationally amplified primordial fluctuations, allows us to test various models of cosmological structure formation.  This can be done, for example, by comparing the distribution of column densities (\\cite{gne97}; \\cite{bon97}; \\cite{mac98}).  These studies have found that the overall normalization of the distribution depends approximately on the parameter $\\Omega_b^2 h^3/\\Gamma$ (there is also some dependence on gas temperature).  Here $\\Gamma$ is the HI ionization rate, $h$ is the Hubble constant in units of 100 km/s/Mpc, and $\\Omega_b$ is the ratio of the baryon density to the critical density required to close the universe.  For a given model, this parameter is often set by requiring that the mean optical depth match that observed (\\cite{pre93}; \\cite{zuo93}; \\cite{rau97}).  The distribution of column densities is close to a power law and at least approximate agreement seems to be found for a number of popular cosmological models.  Another diagnostic is the distribution of $b$ parameters found by fitting Voigt profiles to the spectra.  This has been suggested as a probe of the reionization history (\\cite{hae97}).   In this paper, we undertake a systematic evaluation of numerical uncertainties in simulations of the Ly$\\alpha$ forest, and in doing so, explore the physical structure of the lines.  We will show that the line structure can be rather simply understood, but does require relatively high spatial resolution to model accurately, particularly for the fluctuations at or below the cosmic mean which give rise to the low end of the column density distribution.  We cannot address either Lyman-limit or damped Ly$\\alpha$ systems as they require more spatial resolution and physical processes than these simulations provide.  The paper is structured in the following way: in section~\\ref{sec:resolution_intro}, we examine the effect of resolution on various statistical measures of the forest and the gas that gives rise to it, looking first at distributions of density and temperature (section \\ref{sec:physical}), and then at the properties of the lines fit to simulated spectra, in section~\\ref{sec:lines}.  We analyze the physical nature of the lines in the next two sections (\\ref{sec:nature} and \\ref{sec:jeans}), other non-parametric measures of the spectra (section~\\ref{sec:non_param}), and the effect of our limited computational volume (section~\\ref{sec:power}).  Then we discuss these results and the corresponding observations in section~\\ref{sec:conclusions}.  While this paper was in the final stages of preparation, a preprint (\\cite{the98}) was circulated which examined some of the same issues, although with a different numerical method.  Where there is overlap, our results are in agreement.  In particular, they also find that the b-parameter distribution requires very high resolution. ",
        "conclusions": ""
    },
    "9805/astro-ph9805294_arXiv.txt": {
        "abstract": "Gravitational lensing, caused by matter perturbations along the line-of-sight to the last scattering surface, can modify the shape of the cosmic microwave background (CMB) anisotropy power spectrum. We discuss the detectability of lensing distortions to the temperature, polarisation and temperature-polarisation cross-correlation power spectra and we analyse how lensing might affect the estimation of cosmological parameters. For cold dark matter-like models with present-day matter power spectra normalised to match the abundances of rich clusters of galaxies, gravitational lensing causes detectable distortions to  cosmic variance limited CMB experiments sampling high multipoles ($\\ell \\simgt 1000$).  Gravitational lensing of the CMB, although a small effect, allows independent determinations of the curvature of the universe  and the cosmological constant, \\ie  breaking the  so-called {\\it geometrical degeneracy} in CMB parameter estimation discussed by Bond, Efstathiou \\&\\ Tegmark (1997) and Zaldarriaga, Spergel \\&\\ Seljak (1997). Gravitational lensing of the CMB temperature and polarisation patterns should be detectable by the Planck Surveyor satellite leading to useful independent constraints on the cosmological constant and spatial curvature. ",
        "introduction": "Since the early papers on the cosmic microwave background anisotropies (CMB) by   Peebles \\&\\ Yu (1968), Do\\-ro\\-shke\\-vich, Zel'dovich, \\&\\ Sunyaev (1978),  Wilson \\&\\ Silk (1981) and others, it has been evident that the  CMB anisotropies are  sensitive to fundamental cosmological parameters. These include parameters that define the background cosmology (such as the geometry and matter content) and parameters that define the nature of irregularities in the early Universe (such as the amplitude and shape of the fluctuation spectrum). Early attempts to constrain the parameters of cold dark matter (CDM) models were made by  Bond \\&\\ Efstathiou (1984) and Vittorio \\&\\ Silk (1984). More recently, the parameters of CDM-type models have been constrained using the $COBE$--DMR data alone (e.g. Bennett \\etal 1996, Bunn, Scott \\&\\ White 1995, Stompor, G\\'orski \\&\\ Banday 1995, G\\'orski \\etal 1998), and $COBE$ combined with degree-scale measurements of CMB anisotropies (e.g. Hancock \\etal 1997, Lineweaver \\etal 1997, Bond \\& Jaffe 1997).  In the near future, long-duration balloon flights and satellite experiments promise to provide a wealth of high quality data on the CMB anisotropies. This has stimulated a number of theoretical investigations on the determination of cosmological parameters from observations of the CMB anisotropies (\\eg Jungman \\etal 1996, Bersanelli \\etal 1996, Bond, \\etal 1997, Zaldarriaga \\etal 1997, Efstathiou \\&\\ Bond 1998, Eisenstein, Hu \\&\\ Tegmark 1998). These studies have confirmed that many cosmological parameters, or combinations of parameters, can be determined by future satellite missions to unprecedented precisions of a few percent or better. However, these studies have identified some degeneracies between sets of cosmological parameters\\footnote {\\ie parameter sets that lead to almost indistinguishable CMB power spectra.} estimated from the linear CMB power spectra alone.  Since the entire statistical information on the CMB anisotropies in Gaussian theories is contained in the power spectrum, such parameter degeneracies impose serious limitations on the ability of CMB experiments to constrain cosmological parameters without invoking additional external constraints.   In particular, Bond \\etal (1997) and Zaldarriaga \\etal (1997) have emphasized that cosmological models with identical fluctuation spectra, matter content and angular diameter distance to the scattering surface (see Section 2.1 below) will produce statistically almost indistinguishable power spectra of CMB fluctuations.  This property (which we call the {\\it geometrical degeneracy} hereafter) means that in the limit of validity of linear perturbation theory, CMB measurements cannot set strong independent bounds on the spatial curvature and cosmological constant and hence cannot unambiguously constrain the spatial geometry of the Universe.   In fact there are many additional observational constraints that can be used to break the geometrical degeneracy. Examples include accurate measurements of the Hubble constant, the age of the Universe and the geometrical constraints imposed by Type Ia supernovae light curves [see Figure \\ref{fig1} and the more detailed discussions by White (1998), Tegmark, Eisenstein \\&\\ Hu (1998) and Efstathiou \\&\\ Bond (1998)]. However, before invoking  more conventional astronomical observations, it is worthwhile analysing whether there are non-linear contributions to the CMB anisotropies that can break the geometrical degeneracy. If such effects are present, then it may be possible to break the geometrical degeneracy using measurements of the CMB alone. In this paper, we analyse the effect of gravitational lensing on the CMB anisotropies.  Although acknowledged to be small (Blanchard \\&\\ Schneider 1987, Cole \\&\\ Efstathiou 1989, Sasaki 1989, Seljak 1996), the gravitational lensing effect may be detectable by the high precision observations of the CMB anisotropies expected from future satellite experiments. The possibility of utilising gravitational lensing to break the geometrical degeneracy has been noticed independently by Metcalf \\&\\ Silk (1998). In this paper, we analyse the effects of gravitational lensing on the temperature, polarisation and temperature-polarisation cross-correlation power spectra and assess whether it is possible to observe these effects with the  MAP (Bennett \\etal 1997) and Planck (Bersanelli \\etal 1996) satellites. ",
        "conclusions": "Observations of the CMB anisotropies promise a dramatic improvement in our knowledge of the formation of cosmic structure and  of the values of fundamental cosmological parameters that define our Universe.  According to linear perturbation theory, however, there exists a near exact geometrical degeneracy that makes it nearly impossible to disentangle the values of $\\Omega_K$ and $\\Omega_\\Lambda$ from observations of the CMB anisotropies alone. In reality, the CMB temperature and polarisation anisotropies will be modified by gravitational lensing caused by the irregular distribution of matter between us and the last scattering surface. The effects of gravitational lensing, although small, might be detectable by the Planck satellite for reasonable values of the amplitude of the present day mass fluctuations (\\ie values that reproduce the present day abundance of rich clusters of galaxies).  In this paper, we have computed the effects of gravitational lensing on both the temperature and polarisation pattern and demonstrated that lensing can break the geometrical degeneracy inherent in the linear CMB power spectra. We have performed a Fisher matrix analysis to show how gravitational lensing affects estimates of cosmological parameters. The Fisher matrix requires derivatives of the CMB power spectra with respect to the cosmological parameters. Since numerical errors in these derivatives can artificially break real parameter degeneracies, we have made a detailed analysis of numerical errors in our computations and shown that they are small.  The results of our Fisher matrix analysis are summarized in Tables 3 \\&\\ 4 for an idealized two dimensional space of $\\omega_\\Lambda$ and $\\omega_K$ and for a more realistic space of six cosmological parameters.  These show that gravitational lensing is detectable by a Planck-type experiment and must be taken into account when estimating the values of cosmological parameters. The effects of gravitational lensing are detectable in both the temperature and polarisation anisotropies. For some experimental parameters, the effects of lensing are more easily detectable in the polarisation signal (because of the sharpness of the peaks and minima in the polarisation power spectrum) than in the temperature power spectrum, even though the anisotropies are polarised at only the few percent level.  Gravitational lensing of the CMB anisotropies breaks the geometrical degeneracy and so it should be possible to set limits on the values of $\\omega_\\Lambda$ and $\\omega_K$ from observations of the CMB anisotropies alone. For example, from the 6 parameter analysis in Table 4 for model 1a (a spatially flat $\\Lambda$-dominated universe) it should be possible to set $1\\sigma$ limits of $\\delta \\omega_\\Lambda \\approx 0.03$ and $\\delta \\omega_K \\approx 0.003$ using temperature and polarisation measurements and limits of $\\delta \\omega_\\Lambda \\approx 0.04$ and $\\delta \\omega_K \\approx 0.004$  from observations of temperature anisotropies alone.  This shows that for certain target models a Planck-type experiment is capable of setting tight limits on the geometry of the Universe. Furthermore, the possibility of detecting gravitational lensing adds to the  scientific case for measuring CMB polarisation at high sensitivity and angular resolution.   The lensing constraints on $\\omega_\\Lambda$ and $\\omega_K$ are sensitive to the normalisation of the present day mass fluctuations and the growth rate of the matter fluctuations hence we find less stringent limits for a standard CDM model normalised to $\\sigma_8(t_0) = 0.52$ (Tables 3 and 4). Nevertheless, even in this case, a Planck-like experiment can set $1 \\sigma$ errors of $\\omega_\\Lambda \\approx 0.06$ and $\\omega_K = 0.01$.  The geometrical degeneracy can be broken by applying constraints derived from more conventional astronomical techniques. For example, accurate measurements of the Hubble constant, age of the Universe, the luminosity distances of Type 1a supernovae, measurements of large-scale galaxy clustering can be used, with various assumptions, to break the geometrical degeneracy (see Efstathiou \\&\\ Bond 1998). However, as described in this paper, gravitational lensing breaks the geometrical degeneracy and so one can disentangle the values of $\\Omega_K$ and $\\Omega_\\Lambda$ from accurate observations of the CMB anisotropies. Comparison of results of CMB-based experiments with those obtained with more conventional techniques can provide consistency checks and tests of possible systematic errors.    \\vskip 0.1truein  \\noindent {\\bf Acknowledgements} GPE would like to thank PPARC for the award of a Senior Research Fellowship. RS is supported by UK PPARC grant and acknowledges help of Polish Scientific Committee (KBN) grant No. 2 P03D 008 13.X2. We thank the referee Matias Zaldarriaga for many useful comments and for encouraging us to include the temperature-polarisation power spectrum in the Fisher matrix analysis of Section 4."
    },
    "9805/hep-ph9805238_arXiv.txt": {
        "abstract": "The presence of neutrinos from Boron decay in the flux observed on Earth is attested by the observation of their energy spectrum. Possible distortions of the spectrum investigated in current detectors are often interpreted in terms of evidence in favour or against various schemes of neutrino oscillations. We stress here that a distortion of the spectrum at high energies could also result from an increase in the ratio of neutrinos originating from ($^3$He+p) and $^8$B reactions. While a $^8$B neutrino depletion would contribute to this effect, an increase in the Hep contribution seems also needed to reproduce the preliminary data. ",
        "introduction": "We want to study the effects of possible changes in the ratio between the fluxes of solar neutrinos produced respectively by the ($^3$He+p) reaction and by Boron decay. For simplicity, we will refer to this ratio by the shorthand $Hep/B$.  In particular, an increase in $Hep/B$ could account for the increase in the number of neutrinos observed in the high-energy part of the spectrum, as suggested by the preliminary data of Super-Kamiokande. This is a crucial point to investigate, as such an increase, thus far interpreted as a distortion of the Boron neutrinos spectrum, is the only direct evidence (by this, we mean largely independent of the solar models) for solar neutrino oscillations.  Boron abundance in the Sun has been considerably discussed since its energetic decay neutrinos play a leading role in most experiments, far out of proportion to their sheer numbers. Furthermore, the $^{8}$B reduction mechanism depends on the poorly known $^{7}$Be$(p,\\gamma )^{8}$B production cross section\\footnote{ For a recent reevaluation of this important quantity see Ref.~\\cite{HAM}, and references therein.}. Although the close relation between Boron and Beryllium abundances makes it unlikely to account for all observations by a reduction of the $^{8}$B abundance alone \\cite{HAT}, the impact of a shift in $^{8}$B abundance on the spectrum distortion, and the importance of the latter in discriminating among oscillation schemes makes it an essential element of a complete analysis.  We began with the question: assuming that the apparent depletion of Boron-produced neutrinos is genuine (i.e.~not due to oscillations) would the corresponding change in spectra  effectively mimic the Super-Kamiokande signal? This is indeed largely the case, as we see in the first figure below, if the comparison is made directly between the inferred electron recoil curves and the preliminary data. It turns out, however, that a severe smoothing occurs, due to the limited energy resolution of the experiment, and this must be included in the comparison. This is done in the second figure, which shows clearly that a much larger increase in $Hep/B$ is needed to reproduce the data. Such a large increase cannot stem from a reduction in the Boron contribution alone, as such a suppression would contradict the data. Instead the possibility of an enhancement of the Hep contribution, either for nuclear or astrophysical reasons, must be called into play.  Even apart from possible astrophysical effects, it turns out indeed  that the ($^3$He+p) reaction is in fact poorly known, and could strongly increase $Hep/B$.  After discussing the effect of varying $Hep/B$, we take the opportunity to review in simple terms how it would interfere with the expected spectrum distortions in various oscillation schemes ",
        "conclusions": "Future experiments and extra statistics from Super-Kamiokande will improve understanding of the solar neutrino problem. In comparing with models, we insist, however, that both the ratio $Hep/B$ of Hep to $^{8}$B neutrinos (as well as its implications for other experiments) and various oscillation schemes must be considered simultaneously. Preliminary values favour a large ratio of Hep to $^{8}$B neutrinos. Alternatively, a cut in the energy spectrum might allow a nearly independent study of both effects, since $^{8}$B depletion and/or Hep enhancement show up significantly only in the higher parts of the energy spectrum."
    },
    "9805/astro-ph9805227_arXiv.txt": {
        "abstract": "We have completed a high-resolution, high-signal-to-noise, spectroscopic survey of the northern members of the peculiar $\\lambda$ Boo stars in order to investigate the frequency of the incidence of nonradial pulsation (NRP) in these metal-deficient stars.  Of 18 objects observed, 9 show conclusive evidence of NRP, which suggests that pulsation instability is a common phenomenon in the $\\lambda$ Boo class. ",
        "introduction": "The peculiar HgMn, Am, and magnetic Ap stars are now thought to be reasonably well understood, but the same can not be said about the $\\lambda$ Boo stars.  These objects have Ca\\,{\\sc ii} K and metallic-line types near A0, weak Mg\\,{\\sc ii} $\\lambda$4481 lines, and H lines with cores typical of early to late A-type stars, but often with shallow wings.  Relative to the Balmer line cores the K and metal line types are too early - the stars are metal weak (Gray 1988).  Quantitative measurements suggest that CNO abundances are approximately normal while Fe, Mg, Ca and other elements are underabundant by up to 2 dex (Venn \\& Lambert 1990; St\\\"{u}renburg 1993). ",
        "conclusions": ""
    },
    "9805/astro-ph9805011_arXiv.txt": {
        "abstract": "The long-term evolution of stellar orbits bound to a massive centre is studied in order to understand the cores of star clusters in central regions of galaxies. Stellar trajectories undergo tiny perturbation, the origin of which is twofold: (i)~gravitational field of a thin gaseous disc surrounding the galactic centre, and (ii)~cumulative drag due to successive interactions of the stars with material of the disc. Both effects are closely related because they depend on the total mass of the disc, assumed to be a small fraction of the central mass. It is shown that, in contrast to previous works, most of the retrograde (with respect to the disc) orbits are captured by the central object, presumably a massive black hole. Initially prograde orbits are also affected, so that statistical properties of the central star cluster in quasi-equilibrium may differ significantly from those deduced in previous analyses. ",
        "introduction": "This paper extends previous studies on interaction between stars and an accretion disc near a massive galactic nucleus. Relevant references are, in particular, Syer, Clarke \\& Rees (1991, these authors estimate time-scales for the evolution of stellar orbital parameters in the Newtonian regime), and Vokrouhlick\\'y \\& Karas (1993, relativistic generalization dealing with individual trajectories). Pineault \\& Landry (1994) and Rauch (1995) studied statistical properties of stellar orbits in a dense cluster near a galactic core with an accretion disc.  Observational evidence and theoretical considerations suggest that many galaxies harbour very massive compact cores ($M_\\BH\\approx10^6$--$10^9M_\\odot$), presumably black holes. In particular, high energy output, variability, spectral properties, and production of jets in active galactic nuclei (AGN) can be understood in terms of the model with a supermassive central object surrounded by an accretion disc (e.g., Courvoisier \\& Mayor 1990; Urry \\& Padovani 1995). However, linear resolution of present observational techniques corresponds at best to several hundreds of gravitational radii of the hypothetical black hole. The innermost regions of these galaxies cannot thus be resolved and conclusions about their structure must be inferred from integral characteristics (integrated over the angular and temporal resolution of the device used for observation). Distribution of stars and gaseous material close to the galactic centre is one of the important tools in this respect because velocity dispersion and the corresponding luminosity profile of the nucleus reflect the presence and properties of the central massive object and the disc (Perry \\& Williams 1993; cf.\\ Marconi et al.\\ 1997 for recent observational results).  We will study the situation in which the central object is surrounded by an accretion disc and a dense star cluster. It is the aim of the present contribution to examine the role of periodic interactions of the stars with the disc material, {\\em{}simultaneously\\/} considering the gravitation of the disc. Mutual gravitational interaction of stars forming a dense cluster has been studied since the early works of Ambartsumian (1938) and Spitzer (1940) while the importance of star-disc collisions for the structure of galactic nuclei has been recognized in the early 1980s (Goldreich \\& Tremaine 1980; Ostriker 1983; Hagio 1987). Huang \\& Carlberg (1997) studied a related problem in the dynamics of galaxies.  The gravitation of accretion discs was neglected in previous works because its mass, $M_\\di$, is presumed very small compared to the mass of the central object ($\\mu\\equiv{}M_\\di/M_\\BH\\ll1$; $\\mu$ is a free parameter in our study). We also assume $\\mu\\ll1$ so that the gravity of the disc acts as a perturbation on the stellar motion around the central mass. We will show, independent of the precise value of $\\mu$, that the effect of the disc gravity on the long-term evolution of stellar orbits must be taken into account together with star-disc interactions. In particular, we will show that circularization of many of the orbits, evolution of their inclination, and stellar capture rate are visibly affected by the disc gravity. We will also argue that the physical reason for this fact is the existence of three different time-scales involved in the problem: (i)~the orbital period of the star around the central mass (short time-scale), (ii)~the period of oscillations in eccentricity and inclination of the orbit (medium time-scale, these oscillations are due to the disc gravity), and (iii)~the time for grinding the orbital plane into the disc (long time-scale, due to successive interactions of the stars with the disc). Effects which can be ascribed to the medium time-scale present a new feature discussed in this paper within the context of galactic nuclei surrounded by an accretion disc, although analogous effects of oscillations or sudden changes in orbital parameters are well-known from other applications (cf.\\ recent discussion on dynamics of planetary motion by Holman, Touma \\& Tremaine 1997; Lin \\& Ida 1997).  Details of the model are described in the next section. Then, in Sec.~\\ref{evolution}, long-term evolution of stellar orbits is examined. Finally, conclusions of our present paper are summarized in Sec.~\\ref{conclusions}. ",
        "conclusions": "\\label{conclusions} It has been recognized by previous works that statistical properties of central galactic clusters are influenced by an accretion disc surrounding the nucleus because of twice-per-revolution interaction which affects stellar motion. The main results of this paper can be summarized as follows:  \\begin{description} \\item[(i)] we demonstrated that any consistent model of the star-disc interaction has to take the influence of the disc gravity into account, in addition to the effects of direct collisions with gaseous material; \\item[(ii)] as a result of the disc gravity, individual stellar orbits exhibit evolution which is  different if compared to the situation when collisions are considered but gravity neglected. Most importantly, we found that a significant fraction of initially retrograde (i.e.\\ counter-rotating with respect to the disc material) orbits are captured by the central object. This is due to large oscillations in eccentricity which affect polar orbits. \\end{description}  We wish to note that our two findings mentioned above are to some extent different in their nature. The former one --- (i) --- is essentially a statement of consistency claiming that any reasonable model which involves the effects of direct star-disc physical interaction has to take disc gravity also into account. We argued that this claim is valid for all astrophysically reasonable objects expected in central clusters of galaxies: neutron stars, white dwarfs and stripped stars. The logic behind this result is due to the fact that both effects are controlled by the total mass of the disc. The latter finding --- (ii) --- then states how the model supplemented by effects of the disc gravity differs from previous simpler models. Gravity of the disc induces dynamical structures, libration and circulation zones of the argument of pericentre. Adiabatic change of quasi-integral quantities and related transitions of trajectories between the two zones are the essence of our results.  It is worth recalling that the above-mentioned results did show sensitivity on a particular model of the disc, especially on the radial gradient of the surface density. Indeed, while Rauch (1995), considering only star-disc collisions, reported his results to be insensitive to a particular profile of the surface density or even to the model of the star-disc interaction, we observed that the fraction of retrograde orbits captured by the central mass in the course of their evolution depends on details of both star-disc collisions and effects of the disc gravity. On the other hand, our results show only a weak sensitivity on the total mass of the accretion disc. This feature can be easily understood by realizing that the disc mass parameter $\\mu$ factorizes out (in the first order of approximation) from the averaged potential ${\\bar V}_{\\rm d}$. As a consequence, the value of $\\mu=10^{-3}$ taken in our examples in Sec.~3 is not essential, and similar results hold also for less massive discs.  \\begin{figure*} \\epsfxsize=\\hsize \\centering \\mbox{\\epsfbox[017 380 593 611]{fig9.eps}} \\caption{Long-term evolution of the orbit from Figure~\\protect\\ref{f8} projected onto the $(k,h)$-plane. Evolution in the inner zone of circulation is shown in panel (a), until the trajectory escapes to the outer zone of circulation. Form of the 8-shaped separatrix is indicated at the moment of transition to high eccentricity. Subsequent evolution continues in panel (b) with a steady decrease of eccentricity. The trajectory eventually approaches the origin of the $(k,h)$ plane. \\label{f9}} \\end{figure*}  Note: we have prepared a Java animation which illustrates long-term evolution of stellar orbits in the two zones of the $(k,h)$-plane (libration and oscillation in eccentricity); cf.\\ ``\\hp\\lb{2}karas/\\lb{2}papers/\\lb{2}discapplet.\\lb{2}html''. \\medskip  We are grateful to Richard Stark and the unknown referee for very helpful suggestions concerning the presentation of our article. We acknowledge support from grants GA\\,CR\\,205/\\lb{2}97/\\lb{2}1165 and GA\\,CR\\,202/\\lb{2}96/\\lb{2}0206 in the Czech Republic. V.~K. is grateful for kind hospitality of the International Centre for Theoretical Physics and International School for Advanced Studies in Trieste where this work was completed."
    },
    "9805/gr-qc9805056_arXiv.txt": {
        "abstract": " ",
        "introduction": "The detection of gravitational waves (GW) is a field of active research from the point of view both of the development of suitable detectors and of the study of possible sources and signal analysis. The detectors now operating as GW observatories are of the resonant-mass type and have a sensitivity to typical millisecond  GW bursts of $h\\approx 6\\times 10^{-19}$ ($h$ is the wave amplitude) or, in spectral units, $10^{-21}$ (Hz)$^{-1/2}$ over a bandwidth of a few Hz around 1 kHz  \\cite{amaldi}. The first bound is appropriate for describing the sensitivity to gravitational collapses  while the square of the second bound represents the input GW spectrum that would produce a signal equal to the noise spectrum actually observed at the output of the detector. With this sensitivity it is possible to monitor the strongest potential sources of GW in our galaxy and in the local group (distances of $\\approx 1$ Mpc). In order to improve the sensitivity of these instruments, more advanced transducers and amplifiers are under development as well as new resonant-mass detectors of spherical shape. Furthermore a huge effort is under way to build large laser interferometers. It is widely believed that in the near future, sensitivities of the order of $10^{-23}$ (Hz)$^{-1/2}$ over a bandwidth of several hundred Hz will be attained allowing the observation of GW sources up to distances of the order of $100$ Mpc \\cite{amaldi}. It thus seems that the detection of GWs is highly probable at the beginning of the new millennium. In addition to information of astrophysical interest, the detection of GW gives an opportunity to test the content of the theory of gravity. In fact, it has been shown that a single spherical resonant-mass detector \\cite{bccff}, or an array of interferometers \\cite{mshi}, have the capability to probe the spin content of the incoming GW.  One of the most intensively studied GW source is the inspiralling compact binary system \\cite{taylor} made of neutron stars or black holes. In the Newtonian regime, the system has a clean analytic behaviour and emits a wave-form of increasing amplitude and frequency that can sweep up to the kHz range of frequencies. In this paper we study the radiation emitted by this source in the framework of the Jordan-Brans-Dicke (JBD) theory. We consider this theory to be of particular interest, since the coupling between the scalar field and the metric has the same form of that of string theory, which is widely believed to give a consistent quantum extension of classical gravity. Our main motivation then comes from the attemp to explore a possible experimental signature of string theory as already discussed in \\cite{bccff}. Furthermore the results obtained here generalize to any theory with a JBD type coupling between matter and gravitation.  There has been much work in this domain in the past years. Before going to the plan of the paper, we shortly review it. In \\cite{wz} binary systems were first proposed as possible sources from which extract more stringent bounds on $\\omega_{_{BD}}$ (see (2.1) for its definition) than those obtained from solar system data. An analysys of spherically symmetric collapse of inhomogeneous dust was carried on in \\cite{mshi} and later confirmed in \\cite{sst}. The case of homogeneous dust was treated in \\cite{hcnn}. In \\cite{ssm} a test particle around a Kerr black hole was studied and results very similar to those of our section 5.1 for interferometers were found. In \\cite{de1} it was pointed out that deviations from general relativity can be much different in strong and weak gravity. In \\cite{de2} these deviations were parametrized in a two dimensional space and exclusion plots were drawn out of the available data. Finally in \\cite{no} spherical collapses were studied in a formalism which kept in account strong gravity effects.  The plan of the paper is the following: in Section two we describe the scalar and tensor GW solutions of the JBD theory. In Section three we compute the power emitted in tensor and scalar GW by a binary system. In Section four we concentrate on the scalar waveform. In Section five, we study the interaction between the scalar waveform and two types of earth-based detectors: interferometers and spherical resonant-mass detectors, giving limits for the detectability of the signals coming from typical binary sources. Eventually, in Section six, we draw some conclusions. ",
        "conclusions": "In this paper we have studied the waveforms emitted by a system of binary stars in the framework of the JBD theory and computed the power emitted in GW's for the tensor and scalar components. Eventually we derived limits for the detectability of such signals by interferometers and resonant mass detectors. In the former case we left aside the question of the detectability of the scalar component of the GW \\cite{mshi} and we have concentrated on waves impinging from the most favourable direction. We would now like to comment The detectability ranges obtained in Sections 5.1,5.2 for the scalar component of the GW's emitted by a binary system, vary from few tens to few hundreds of kpc's for masses ranging from those of typical neutron stars ($1.4 M_\\odot$) to those of typical black holes ($10 M_\\odot$). We remind the reader that for the purely tensorial component (in this case the results obtained in the framework of the JBD theory are practically the same of those of GR) the detectability range (for $1.4 M_\\odot$) is $r\\simeq 120$ Mpc for spherical detectors \\cite{cf} and $r\\simeq 300$ Mpc for interferometers \\cite{cuoco}. The expected rate of coalescence events is of the order of 1 per year up to 100 Mpc \\cite{phin}. We can thus conclude that binary systems look less promising than gravitational collapses \\cite{bbcfl} as sources of detectable scalar GW from the next generation of earth-based detectors.  \\newpage  \\renewcommand{\\theequation}{\\thesection.\\arabic{equation}}  \\appendix"
    },
    "9805/astro-ph9805361_arXiv.txt": {
        "abstract": "We present F555W ($V$), F439W ($B$), and F336W ($U$) photometry of 9507~stars in the central $2'$ of the dense, post core collapse cluster M30 (NGC~7099) derived from {\\it Hubble Space Telescope\\/} Wide Field/Planetary Camera~2 images.  These data are used to study the mix of stellar populations in the central region of the cluster.  Forty eight blue straggler stars are identified; they are found to be strongly concentrated towards the cluster center.  The specific frequency of blue stragglers, $F_{\\rm BSS}\\equiv{N}({\\rm BSS})/N(V<V_{\\rm HB}+2)$, is~$0.25\\pm0.05$ in the inner region of M30 ($r<20''$), significantly higher than the frequency found in other clusters: $F_{\\rm BSS}=0.05\\>$--$\\>$0.15.  The shape of M30's blue straggler luminosity function resembles the prediction of the collisional formation model and is inconsistent with the binary merger model, of Bailyn \\& Pinsonneault (1995, ApJ, 439, 705).  An unusually blue star ($B=18.6$, $B-V=-0.97$), possibly a cataclysmic variable based on its color, is found about $1\\farcs2$ from the crowded cluster center.  Bright red giant stars ($B<16.6$) appear to be depleted by a factor of~2--3 in the inner $r<10''$ relative to fainter giants, subgiants, and main sequence turnoff stars (95\\% significance).  We confirm that there is a radial gradient in the color of the overall cluster light, going from $B-V\\sim0.82$ at $r\\sim1'$ to $B-V\\sim0.45$ in the central $10''$.  The central depletion of the bright red giants is responsible for about half of the observed color gradient; the rest of the gradient is caused by the relative underabundance of faint red main sequence stars near the cluster center (presumably a result of mass segregation).  The luminosity function of M30's evolved stars does not match the luminosity function shape derived from standard stellar evolutionary models: the ratio of the number of bright giants to the number of turnoff stars in the cluster is 30\\% higher than predicted by the model ($3.8\\sigma$ effect), roughly independent of red giant brightness over the range $M_V=-2$ to $+2$. ",
        "introduction": "The {\\it Hubble Space Telescope\\/} ({\\it HST\\/}) is ideally suited for the study of individual stars in the crowded central regions of dense Galactic globular clusters.  The projected density of evolved stars alone ranges from about 0.3~arcsec$^{-2}$ in the core of the well studied dense cluster M13 to $\\gtrsim30$~arcsec$^{-2}$ at the center of M15, the Galactic cluster with the highest known central surface density (\\cite{randi}, hereafter Paper~VI). Resolving individual stars at these densities is beyond the current angular resolution limits of ground-based telescopes, but such observations are crucial in order to address several important astrophysical issues.  Globular clusters are excellent testing grounds for models of stellar evolution because the cluster members are coeval and have similar chemical composition (cf.~\\cite{bergbvan}).  Deviations from the natural process of (isolated) stellar evolution caused by stellar interactions in these environments of extreme star density can be explored by examining the radial dependence of the mix of stellar populations (cf.~\\cite{piotto88}; \\cite{djor91}).  Additionally, the dense cores of globular clusters are unique laboratories for studying the effects of two-body relaxation, equipartition of energy, and binaries on the dynamical evolution of dense stellar systems (cf.~\\cite{hut92}; \\cite{meyl97}).  A visible product of stellar collisions and mergers of binaries, blue straggler stars (BSSs), are preferentially found in the central regions of most globular clusters (\\cite{ferrarom3}; \\cite{bsstheory}).  This is the seventh in a series of papers describing {\\it HST\\/} observations of the centers of the nearest Galactic globular clusters with $\\vert{b}\\vert>15^{\\circ}$.  The main scientific goals of this program are to measure the shape of the density profile in clusters and to understand the nature of evolved stellar populations in very dense regions by probing the variation in the mix of stellar types as a function of radius (and hence stellar density).  Complementary programs targeting main sequence (MS) stars in globular clusters are being conducted independently by other groups to explore cluster dynamics (cf.~\\cite{massseg}) and models of stellar evolution (cf.~\\cite{PCK}).  In this paper we examine the evolved stellar populations of M30 (NGC~7099) using a set of techniques, developed in earlier papers (\\cite{ppr1}; \\cite{ppr2}; \\cite{ppr5}, hereafter referred to as Papers~I, II, and V, respectively), to: (1)~Build empirical point spread function (PSF) models using isolated bright stars, allowing for PSF variability across the image and using faint stars to reconstruct the saturated cores of brighter stars; (2)~Iteratively fit the PSF template to the stars on the image; (3)~Perform aperture photometry on each star after subtracting its neighboring stars with the best-fit PSF template; and (4)~Carry out detailed and realistic image simulations to assess the effects of crowding on photometric accuracy and sample completeness.  M30 is a prototypical ``post-core-collapse'' globular cluster based on its rising surface brightness profile at radii~$<3\\arcsec$.  Its central surface brightness in the $V$ band is $\\mu_V(0)=15.20$~mag~arcsec$^{-2}$ (\\cite{metll}), corresponding to a projected density of $\\sim5$~arcsec$^{-2}$ in post MS stars alone.  The cluster has a relatively low metallicity, $\\rm[Fe/H]=-2.13$ (\\cite{djor93}; \\cite{zinn}), a low line-of-sight reddening, $E_{B-V}=0.05$ (\\cite{burs}), and is located at a distance of about 9.8~kpc (\\cite{reid}).  Based on this recent distance determination, the age of M30 is estimated to be about 10~Gyr (\\cite{Sandquist}).  It has long been noted that M30 displays a central bluing trend like other post-core-collapse clusters (\\cite{willbach}; \\cite{chunfree}; \\cite{cord}; \\cite{peterson86}; \\cite{piotto88}; \\cite{burgbuat}).  In this paper we use {\\it HST\\/} Wide Field/Planetary Camera~2~(WFPC2) data to investigate the origin of M30's nuclear color gradient and to study its abundant population of BSSs; Yanny et~al.\\ (1994b, hereafter \\cite{letter}) used the same data set to examine the inner 20\\arcsec\\ of M30.  This paper is organized as follows: Sec.~2 contains a description of the data and reduction techniques; Sec.~3 contains results from our study of M30's evolved stellar populations, with particular emphasis on BSSs, radial population gradients, and the stellar luminosity function (LF); Sec.~4 contains a summary of the main points of the paper. ",
        "conclusions": "\\begin{itemize}  \\item[{\\bf 1.}]{\\it Hubble Space Telescope\\/} Wide Field/Planetary Camera~2 images of the dense globular cluster M30 (NGC~7099) in the F555W, F439W, and F336W filters have been analyzed.  Accurate stellar positions and photometry in the (Johnson) $UBV$ bands are presented for 9507~stars (and $BV$ photometry for an additional 433~stars) within a projected distance of 130\\arcsec\\ from the cluster center.  Color-magnitude diagrams based on the $UBV$ bands are presented showing clearly distinguished sequences of stellar types: red giant branch, (blue) horizontal branch, subgiant branch, blue straggler sequence, main sequence turnoff, and main sequence.  Typical photometric errors for main sequence stars ($B=19\\>$--$\\>$20) range from $\\sim0.1$~mag for the central 10\\arcsec\\ of M30 to $\\sim0.05$~mag in the rest of the cluster.  Incompleteness sets in a little fainter than the main sequence turnoff ($V\\gtrsim19$) for $r<5\\arcsec$ but the sample is complete to progressively fainter magnitudes at larger radii ($V\\sim20.5$ for $r\\gtrsim20''$).  \\item[{\\bf 2.}]An unusual star is found $1\\farcs2$~SSW of the cluster center. Its very blue color ($B-V=-0.97$, $B=18.6$) indicates that it might possibly be a cataclysmic variable.  \\item[{\\bf 3.}]Forty-eight blue straggler candidates are identified in the cluster on the basis of a ($B-V$,~$B$) color-magnitude diagram; the BSS classification for some of these is tentative.  The specific frequency of BSS in M30, $F_{\\rm BSS}\\equiv N({\\rm BSS})/N(V<V_{\\rm HB}+2$, is~$0.39\\pm0.08$ in the inner 10\\arcsec\\ and $0.14\\pm0.02$ over the full WFPC2 image ($r<130\\arcsec$).  The central BSS specific frequency in M30 is the highest measured for any globular cluster to date.  The BSSs are strongly centrally concentrated relative to other evolved stars ($>99\\%$ significance).  The observed M30 BSS luminosity function is compared to theoretical predictions: although the number of BSS is too small to draw definite conclusions, the collisional formation model seems to be a better match to the data than the primordial binary merger model.  \\item[{\\bf 4.}]The abundance of bright red giant branch stars and asymptotic giant branch stars with $V\\lesssim16$ (relative to fainter RGB stars and subgiants) appear to be a factor of~2--3 lower in the central 15\\arcsec\\ of M30 compared to further out in the cluster (2--2.5$\\sigma$ effect). Horizontal branch stars, the evolutionary descendants of bright RGB stars, show no significant central depletion relative to the distribution of faint RGB stars and subgiants.  \\item[{\\bf 5.}]The $B-V$ color of the integrated cluster light in M30 varies from 0.82 around $r\\sim1'$ to 0.45 within $r<10''$, a radial color gradient of ${\\partial(B-V)}/{\\partial\\log(r)}=0.3$~mag/dex.  The central deficiency of bright red giant/asymptotic giant branch stars is responsible for about a third of the color gradient; all evolved stellar populations taken together (RGB, AGB, HB, BSS, subgiants) are responsible for about half of the overall color gradient.  Mass segregation of faint red main sequence stars results in a smooth $B-V$ color gradient of about 0.15~mag/dex, corresponding to half the observed gradient.  \\item[{\\bf 6.}]The $V$-band surface brightness profile of the integrated starlight in M30 has a power law slope of $\\alpha=-0.5$ in the inner 10\\arcsec, and the slope is $\\alpha=-1$ for the light of faint~RGB stars alone; the latter is identical to the number-weighted measure of the projected density profile slope found in \\cite{letter}.  The difference between the central slopes of the overall vs faint~RGB light is a result of the central depletion of bright~RGB stars.  The integrated surface brightness profile slope steepens to $\\alpha=-2$ for $r>40\\arcsec$.  Main sequence stars contribute nearly 40\\%, bright~RGB/AGB/HB stars with $V\\lesssim16$ contribute 30\\%, and faint~RGB stars contribute about 30\\% of the overall cluster brightness in the central $5''$ of M30; the corresponding fractions are 50\\%, 40\\%, and 10\\% at larger radii ($r\\sim1'$).  \\item[{\\bf 7.}]The stellar luminosity function in the inner $2'$ of M30 shows an anomalously high abundance of red giants (relative to subgiant, main sequence turnoff, and main sequence stars), as has been previously noted further out in the cluster.  Giants with $M_V<+2$ in M30 are overabundant by 30\\% (3.8$\\sigma$ significance) relative to the prediction of suitably normalized standard stellar evolution models.  \\end{itemize}   \\bigskip \\bigskip"
    },
    "9805/astro-ph9805157_arXiv.txt": {
        "abstract": "We report the cumulative results of an on-going near-infrared search for redshifted \\han~emission from normal galaxies at z$>$2.  An infrared search reduces the bias due to reddening.  Using narrow-band imaging with the Near Infrared Camera on the Keck I 10-m telescope, a survey area of almost 12 square arcminutes has been covered. Target regions were selected by matching the redshifts of QSO emission and metal-line absorptions to our available filters.  The survey depth is 7 $\\times 10^{-17}$ergs/cm$^{2}$/s (3$\\sigma$) in \\han~and \\kp $\\simeq$ 22.  Eleven \\han-emitters, plus two Seyfert I objects, have been discovered.  The high density of galaxy detections, corresponding to a co-moving volume density of 0.0135$+{0.0055}-{0.0035}$~Mpc$^{-3}$, makes it unlikely that all of the \\han~flux in these objects is the result of active nuclei. There is a strong suggestion of clustering in the environments of metal-line absorbers.  Each candidate galaxy lies typically within a projected distance of 250kpc of the QSO line of sight and is resolved but compact.  The average Star Formation Rate inferred for the galaxies from the \\han~flux is 50 \\Msun /yr, significantly higher than current day star-forming galaxies, but consistent with other estimates for galaxies at high redshift. ",
        "introduction": "All observations in this survey were taken at the 10-m Keck I and II telescopes.  We obtained images of 23 fields with the Near IR Camera (NIRC), which has a 256$\\times$256 pixel InSb array, (Matthews \\et, 1994) and a field of view of 38\\arcs $\\times$ 38\\arcs; 0.15\\arcs/pixel.  Although the field of view is small, this instrument has high throughput and excellent depth under good seeing conditions. Deep B,V and I imaging of the same fields was obtained with the Low Resolution Imaging Spectrometer (LRIS), which has a 2048$\\times$2048 pixel CCD detector, with a field of view of 5\\arcm~by 7\\arcm~(Oke \\et, 1995).  The Near IR Camera has five narrow-band filters in the 2--2.5\\mic ~range.  For the broad-band, we use the K$^{\\prime}$~filter (Wainscoat \\& Cowie 1992). Table 1 lists the properties of each narrow-band filter, including their transmission as measured from our extensive photometry of object fields and calibration stars. The molecular hydrogen transition filters have significantly lower transmission than the other three, so we soon decided not to use them.  The filter centered at 2.24\\mic, called the ``K-continuum\" filter, is wider than the others, with $\\Delta \\lambda / \\lambda = 2.2\\%$.  The wider band pass makes this filter less attractive, since it increases the sky background without any increase in emission-line flux. However, the tradeoff between contrast and search volume was favorable for a number of fields on our target list that included multiple absorption line systems within the redshift range probed by that filter (z=2.42--2.46), and in those cases we opted to use it.  All IR observations were taken using the ``dither pattern'' technique, in which many small displacements (\\eg 3--5\\arcs) of the telescope are made between successive exposures on the same field (see McLean \\& Teplitz, 1996, among many others), and twilight flat-fields were used to divide out variations in detector response.  Observations were reduced as follows.  Each frame was divided by the flat-field, and then an individual sky frame was created for each exposure from a running median of the other object frames taken nearest in time to each observation.  This specific sky frame was then subtracted from the object frame.  In cases where a flat field was unavailable, the sky frames were generated first, then normalized to their mode and used as a flat, while sky subtraction was handled in aperture photometry. Detailed comparisons showed little appreciable loss in the signal-to-noise ratio when using the latter method.  All data reduction was performed using IRAF routines.  A similar technique was used for I-band observations.  The B and V band CCD observations, however, were taken without dithering, and they were divided by twilight flats.  Objects were identified with the SExtractor program for galaxy identification and analysis (Bertin \\& Arnouts, 1996) and the object list was checked by eye for each frame.  Photometry was performed in circular apertures of 10 pixels (1.5\\arcs) diameter (approximately 2.5 times the seeing disk), which corresponds to 12 kpc at z=2.5. The size of the aperture was chosen primarily to obtain optimal signal to noise (see Thompson \\et, 1995, and Howell 1989).  The mode of the pixels in a ``sky annulus'' around the aperture was subtracted, to ensure proper sky subtraction in addition to that performed in the initial reduction.  The same aperture size was used for both broad-band and narrow-band frames.  As a check on these results, we also performed photometry in elliptical apertures determined by Kron's ``first moment'' (Kron 1980) routine in the SExtractor software.  In that case, the apertures were determined in the broad-band frame and then used with the same size and orientation in the narrow-band frame.  The flux ratios for each method agreed to better than 10\\%.  A third check was performed using gaussian fitting.  A two dimensional gaussian profile was fitted to each object in the broad band, to determine the shape and amplitude.  The same gaussian was fitted to the narrow-band image with only the amplitude allowed to vary.  Figure 1 shows a comparison of this method with the aperture photometry.  \\subsection*{Error Analysis}  Since we are looking for apparent excess in the broad-band minus narrow-band color, accurate measurement of the photometric errors is essential.  Errors in photometry were calculated from the $1 \\sigma$~limiting magnitudes.  First, photometry was performed on several hundred randomly selected locations within the frame, using the same standard aperture adopted for objects and the same ``sky subtraction'' annulus.  This set of sky measurements was analyzed to obtain the standard deviation of the flux in the apertures.  To demonstrate the validity of this method the following procedure was used.  For each frame, we inserted fake objects and measured their photometry in the standard aperture.  Several hundred such tests were performed across each image. The deviation in these measurements from the flux inserted agreed to better than 5\\% with the ``blank sky'' photometric method of determining 1$\\sigma$~errors, as expected; except for the random placement of the apertures, the methods should be identical.  The next step in error analysis is to consider the propagation of individual photometric errors into the derived color value.  At first we considered adding the individual fractional photometric errors in quadrature.  This basic approach is traditionally derived from the first order term in a Taylor expansion of the function into which errors are being propagated.  However, the first order terms are not sufficient in the case where the errors are relatively large compared to the measurements themselves.  In our case, the large error regime is important, particularly in cases where an object may be detected more strongly in the narrow band than in the broad band.  In order to account for higher order terms, we chose to model the error analysis. We assumed a gaussian distribution of photometric errors and then generated a data set of 100,000 points based on a known ratio of broad to narrow band flux.  We remove objects that are less than 1$\\sigma$~(after adding the gaussian error) in the broad-band from the simulated population, as they would be undetectable and would not populate our measured color-magnitude diagrams. Using this simulation, we are able to define confidence intervals by simply counting the simulated measurements (see Figure 2).  We find, as expected, that for the regime where errors are relatively small, the traditional approach is accurate.  In the case of objects with flux levels less than 5 times the photometric error, the first order error analysis grossly overestimates the error in the color, due to the vanishing denominator.  Even in cases of 5-10 $\\sigma$~objects, there can be substantial deviation.  We adopt the approach of using our empirically derived confidence intervals in evaluating color-magnitude plots.  The derived confidence intervals can be described to a reasonable approximation in terms of a single noise characteristic, $\\Sigma$. If the broadband minus narrowband magnitude difference for featureless continuum emitters is given by $R_{0}$, then we find that the shape of the upper confidence interval is well approximated by: \\begin{equation} R(f_{b}) = R_{0} + 2.5*log(10^{1+\\Sigma/f_{b}}) \\end{equation} where $f_{b}$~is the broad band flux.  This empirical formula thus gives a good estimate of how large a magnitude difference is required to be significant at the 99.5\\% confidence level. In our analysis, we use the slightly more precise upper boundary defined by the Monte Carlo simulations.   \\subsection*{Target Selection and Control Fields}  Targets were selected from a search of NASA/IPAC Extragalactic Database (NED).  We looked for fields containing QSOs, both radio loud and radio quiet, or absorption line systems, both Damped \\lya ~Absorption or metal line absorption.  Objects south of -30\\deg ~were not considered.  We then selected fields at redshifts within $\\pm 0.3 \\%$ of the central redshift of the filters. Table 2 describes the observations by date and filter.  In order to determine the effect of our target selection on the density of detections, we need to obtain control observations.  We plan to begin such an experiment by observing targets at $z>3$ which have no absorption systems at redshifts to place \\ha in our narrow-band filters.  One preliminary observation of this kind detects no emission-line galaxies in one NIRC field around the QSO 1542+4744 with an integration time of 6480 seconds in the narrow band, during which time we reached a 3$\\sigma$~limiting flux of 6$\\times 10^{-17}$ ergs/cm$^{2}$/s (as deep as most of the z=2.5 target fields).  A shallow observation of the field around the QSO 0234+013 also detects no emission-line galaxies, down to a limiting flux of 9$\\times 10^{-17}$ ergs/cm$^{2}$/s.  Two similar observations were carried out by Pahre \\& Djorgovski (1995) also with NIRC on Keck targeted to known objects at $z>3$, finding no detections in two NIRC fields, with flux limits of 3$\\times 10^{-17}$ ergs/cm$^{2}$/s. ",
        "conclusions": "As discussed in MTM96, there is the possibility that all these galaxies have an active nuclear component that contributes to their line emission.  We will address this suggestion before drawing conclusions from the \\han-emitting sample as a whole.  \\subsection*{Are we detecting AGN?}  The observed space density of high redshift quasars is so much smaller than our space density of detections that they are very unlikely to be the same population. Warren \\et~(1994) calculate the Luminosity Function (LF) for bright QSOs at $z>2$.  This LF can be used to estimate the density of active galaxies at the redshifts covered by our survey.  For bright quasars, with continuum absolute magnitudes on the AB system of $M_{c}<-23$, the space density is $\\le 7\\times 10^{-6}$ / Mpc$^{-3}$.  Using the evolving luminosity function suggested for the quasars, we can extend this LF down to the average magnitude for our detected objects, $M_{c}\\sim -19$.  This calculation predicts a density of $2\\times 10^{-4}$/Mpc$^{-3}$; or, put another way, the QSO density predicts that we should have to observe more than 100 fields to find a single AGN.  Our observed density of detections is higher than this by several orders of magnitude.  If all of our objects were to prove to have active nuclei, their space density is three orders of magnitude higher than expected, so we are not simply probing the usual population of quasars that are well studied by numerous surveys.  The predicted density of AGN is more consistent, however, with that detected in the Lyman Limit searches, which typically find 10\\% of their objects to be AGN (as estimated from counting the number of AGNs reported in, for example, Steidel \\et, 1996).  Our detections may be, on average, redder than the LLD galaxies, and they probably have stronger line emission (many LLD galaxies have \\lya ~in absorption, not emission).  If LLD searches were to probe the entire galaxy population at $z>2$, while our survey were to be unable to find anything but AGN, one would conclude that we should arrive at a space density ten times {\\it smaller} that the LLD searches.  That is not what observations show, however, as our inferred space density may be as much as 5 times {\\it larger} than estimates based on LLD searches.  On the other hand, our space density of objects is consistent with other searches which have been conducted for Emission Line Galaxies in the same redshift range.  They, too, find clusters of ELGs.  In particular, some of the rare successes of \\lya~searches have discovered groups of emission-line galaxies.  The highest density of objects observed in a single field is the Pascarelle \\et, 1996a, survey which found 18 \\lya-emitting objects within 1 Mpc$^{2}$ and 300 km/s of the radio galaxy 53W002.  These objects have been interpreted as ``subgalactic clumps\" that will collapse into a L* or slightly larger galaxy.  However, like one of the objects in the present survey, those ELGs show strong \\lya ~emission, as well as CIV and NV in some cases.  The same authors later detected similar objects in parallel HST observations (Pascarelle \\et, 1996b).  A high density of \\lya- and CIV-emitting galaxies, some up to L*, would explain most of our detections.  Francis \\et~(1996, 1997) have detected a supercluster of \\lya-emitters at z=2.38, all of which appear to have AGN characteristics in their spectra.  They speculate that they are detecting a dust-reddened population of active galaxies that are the radio-quiet counterparts of the radio-galaxy population.  They further suggest that they are seeing a population of galaxies undetectable in other surveys due to their red colors.  While that comparison was suggested to exclude our survey based on MTM96, their I-K galaxies are not uniformly redder in I-K than objects in the present survey.  We report several galaxies with I-K$\\simgt$4, while Francis reports I-K=3.4--5.2 for various objects.  This allows the possibility that both surveys could be finding a similar population.  \\subsection*{Does the space density imply clustering?}  We detect a higher density of objects than is observed either at the present day or in the LLD galaxies.  We find up to $\\sim 1.2$~ galaxies/sq. arcmin. in volumes that are 1 or 2\\% deep in redshift (though the galaxies may occupy a smaller $\\Delta$z if they are clustered).  LLD galaxies are found with a density of 0.4--0.75/sq. arcminutes, but over a much larger redshift range.  We can also consider the comoving volume density. As a typical example of an LLD search, Madau \\et~(1996) finds field galaxies with $<z>=2.75$ in the Hubble Deep Field with a comoving number density of $3.6\\times 10^{-3}$Mpc$^{-3}$ down to L*.  Calculating the comoving volume density of our detections yields 0.0135$^{+0.0055}_{-0.0035}$ ~Mpc$^{-3}$, a factor of 3-5 times more than the LLD galaxies.  The errors are based on Poisson statistics (Gehrels 1986).  The density could be higher if these galaxies are not uniformly distributed in the surveyed redshift windows; for example, if the \\han-emitters were as close the the absorber in the radial direction as they are in on the sky, the density could be 20 times higher.  It is difficult to say whether all of these emission line galaxies would be found by the UV technique without more direct comparisons (see for example Bechtold \\et~1997) for \\ha observations of a galaxy with the characteristic spectrum of a UV-selected galaxy).  Alternately, we consider the density of \\ha-emitters excluding the most likely cluster (0953+549) and excluding the Seyfert 1 (2149+02).  In that case, we find a comoving density of 0.008 Mpc$^{-3}$, still twice that of LLD galaxies, though Poisson statistics show the difference is only at the 1.5$\\sigma$~level.  We can also compare our observed density of objects to the current day volume density. For example, the comoving density of present-day galaxies brighter than L$_{*}$~is $\\sim 3.5 \\times 10^{-4}$ Mpc$^{-3}$~(Loveday \\et~1992).  This comparison may only be lower limit, as there may be effects of luminosity evolution to consider. For example, Cowie \\et~(1996) find the normalization of the luminosity function, $\\phi_*$, approximately doubles between z=0.2 and z=1.0. However, the evolution of the LF out to $z>2$~is highly uncertain; different analysis of the Hubble Deep Field data set produce very different LFs (see Bershady \\et~1997 and the references therein).  We adopt as a second standard of comparison, the luminosity function of Sawicki, Lin \\& Yee (1997; SLY) which has two magnitudes of evolution in $L_*$~at $2<z<3$ and a steeper faint end slop in that redshift range.  SLY predict a factor of $\\sim 3$~more galaxies down to today's $L_*$, or a comoving density of 1.2$\\times 10^{-3}$ Mpc$^{-3}$ at $<z>=2.5$.  If all our candidate galaxies are confirmed, it could imply that we are seeing clustering in the environments of metal absorption line systems.  Even if the ELGs are all AGN, it would be unusual to find so many.  If we consider clustering to be a possibility, we must examine two cases -- either the galaxies are in a cluster, or they are simply correlated on the large redshift scales at which we are observing.  In the first case, we assume that the galaxies are at nearly identical redshifts.  Specifically, we assume that they are as close in redshift as they are in projected distance, which increases our space density of detected objects by a factor of $\\sim 20$.  Under this assumption, we compare our comoving density to the current day comoving density and we find that on the scale of a cubic NIRC field the correlation function, $\\xi(0.3Mpc)$, is 5 times its present value (not accounting for possible evolution) or $\\sim 1.5$~times the value implied by SLY at the same redshift.  We note that this comparison may be an lower limit on how strongly clustered the galaxies are, if there is a gravitational effect that reduces how much the cluster expands relative to the Hubble expansion.  In the second case we assume that the galaxies we detect are uniformly distributed along redshift in the window sampled by the narrowband filter.  In this case, we consider a radius equal to half of the long side ($\\delta z$) of our highly rectangular search window.  Comparing our comoving density to the current day, the correlation function inside that comoving volume, $\\xi(3.25Mpc)$ is then 15 times the current value (where we assume $\\xi_{current} \\propto (r/5Mpc)^{-1.77}$, Peebles 1973) or 5 times the value inferred from SLY.  The inferred clustering of our galaxies is somewhat stronger than what is seen between the quasar metal absorption-line systems themselves. Absorption systems have been seen to be correlated on radial scale corresponding to our $\\Delta$z windows.  Sargent, Boksenberg, \\& Steidel (1988) see $\\xi =$5---10 for CIV absorbers at scales from 200 to 600 km/s.  Similarly, Steidel \\& Sargent (1992) see correlations of $\\xi (600\\le \\Delta v \\le 5000 km/s) = 2.6$, for MgII absorbers at redshifts of 0.6$<$z$<$2.2.  Our narrowband filters with $\\Delta z$/z=1\\%, probe velocity differences on the order of several times 10$^{3}$km/s, so we may see more clustering on this scale than is expected for absorbers.  On the other hand, the density we find seems to be consistent with surveys of QSO environments at high redshift.  Ellingson \\et~(1991) find that the richest environments of radio loud QSOs at $z\\sim 0.5$ are well fit by the Schechter parameters $\\Phi=6.5$/Mpc$^{2}$ ~and M$_{r}$*=-22.6, assuming $\\alpha=-1.0$.  This predicts that down to L* we should see 0.6 excess galaxies per comoving Mpc$^{2}$. In our case, we see $\\sim0.4$ ~galaxies per comoving Mpc$^{2}$.  Hall \\& Green (1998) find several radio-loud QSOs at z$\\sim$1.5 reside in apparent rich clusters, detected as an excess of red (in r-K) objects.  Their models suggest these clusters are consistent with z$_{form}>4$, similar to the conclusions in MTM96.  Hutchings, 1995, finds an excess of galaxies around QSOs at z=2.3.  These counts are attributed to compact groups of starburst galaxies.  The detected excess galaxies, down to R=24, is 30--84 per Mpc$^{2}$.  While Hutchings' counts are not spectroscopically confirmed, they do suggest the presence of clustering at high redshifts, at least in quasar environments.  \\subsection*{Reddening}  We note that reddening could also explain the potential difference in density between LLD searches and our candidate ELGs.  However, to reconcile the number counts with {\\it no} clustering, we would need $<E_{B-V}>$ ~to be sufficiently large to make galaxies with SFR$\\sim30$\\Msun/yr unobservable spectroscopically in the optical. Depending on the reddening law, this requires $<E(B-V)>\\sim 0.7-0.9$~mag.  This amount of extinction seems incompatible with the observed g-R colors of the LLD galaxies, for any reasonable extinction law (Fitzpatrick 1986, Calzetti \\et~1994, etc.)  Even if the observed densities of LLD galaxies and \\han-emitters are similar (if, for example, our marginal detections prove false) reddening is still important in understanding the nature of these high redshift galaxies.  In particular, consider the SFR determined from our best candidates.  Even assuming (as a worst case) that there is a small (30\\%) AGN contribution to the line flux, we infer an average SFR $\\simgt 35$\\Myr.  This SFR is large compared to the uncorrected average SFR of the LLD galaxies (Steidel \\et~1996, Lowenthal \\et~1997)), as derived from the UV continuum.  More recent estimates from the LLD galaxies (\\cf Pettini \\et 1997) give a factor of 3--5 redding correction, which brings the LLD average within a factor of two of ours.  We can also contrast our \\han~star formation rates with those inferred from \\lya~selected objects.  Cowie \\& Hu (1998) present a dozen \\lya-emitters, which would have a maximum SFR=10\\Myr, in the absence of extinction.  In summary, we have shown that with the Keck telescope the narrow-band near-IR search technique for emission-line galaxies at $z>2$ provides an effective means of detecting groups of star forming galaxies.  We detect galaxies with a high comoving volume density, suggesting that it is unlikely that the line and continuum emission from these objects is dominated by active nuclei.  The density further suggests that we observe clustering in the environment of metal-line absorbers.  The inferred star-formation rates of these galaxies agree well with de-reddened estimates for UV-selected galaxies at similar redshifts."
    },
    "9805/astro-ph9805143_arXiv.txt": {
        "abstract": "A careful investigation of a CCD spectrum of the SB1 system $\\kappa$ Cnc in the spectral region 3800~\\AA\\ -- 8000~\\AA\\ resulted in the discovery of the lines of the secondary star.  We then analyzed several short-wavelength range Reticon spectra obtained at different orbital phases to find additional radial velocities. The mass ratio is $m_{\\rm A}/m_{\\rm B}$ = 2.2 $\\pm$ 0.1 and, from binary spectrum-synthesis, the ratio of radii is $R_{\\rm A}/R_{\\rm B}\\geq$ 2. ",
        "introduction": "\\label{intr}  $\\kappa$ Cnc (= HR 3623 = HD 78316), one of the best known and studied HgMn stars, is an SB1 spectroscopic binary. Further, a note in the Bright Star Catalogue (BSC) indicates that $\\kappa$ Cnc is a triple occultation system with the primary having a rather bright companion ($\\Delta$m =0.2 mag). A third star of 7.8 mag in V is at a distance of 0.3 mas from the primary.  During a study of $\\kappa$ Cnc's Mn\\,{\\sc ii} lines we noticed an asymmetry in the red wings of all Balmer lines on an echelle spectrum obtained at the 1m telescope of the Special Astrophysical Observatory (SAO).  A careful reanalysis of all CCD and Reticon spectra available to us resulted in the discovery of the secondary's lines in them. ",
        "conclusions": ""
    },
    "9805/astro-ph9805005_arXiv.txt": {
        "abstract": "Archival IUE SWP and HST FOS spectra show the presence of a relatively strong, broad Lyman $\\alpha$ emission line superposed onto the UV spectral continuum of the blazar 3C~279. As opposed to a factor $\\sim$50 variation of the continuum flux during eight years, the emission line did not exhibit significant intensity changes. Simultaneous IUE SWP and LWP spectra of 3C~279 in low emission state are fitted by power-laws of index $\\alpha_\\nu \\sim 1$ ($f_\\nu \\propto \\nu^{-\\alpha_\\nu}$), significantly flatter than measured during higher states.  Our observations suggest that the Lyman $\\alpha$ line is not powered by the beamed, anisotropic synchrotron radiation which produces the observed continuum in 3C~279, but rather by an unbeamed component characterized by slower and lower amplitude variability. The latter may account for the UV continuum observed during the very low state of January 1993. \\vspace {5pt} \\\\    Key~words: ultraviolet spectra; blazar emission lines; blazar emission mechanisms; accretion disks. ",
        "introduction": "The blazar 3C~279 ($z = 0.54$) is well studied and shows frequent large continuum flares from radio to gamma-ray wavelengths. Inverse Compton scattering of relativistic electrons off synchrotron or ambient photons is likely responsible for the emission at hard X- and gamma-ray energies. Clarifying the exact nature of the seed photons for this mechanism would explain the origin of the huge amplitude variations exhibited by 3C~279 at the highest energies (Maraschi et al. 1994; Hartman et al. 1996; Wehrle et al. 1998). There have been several multi-wavelength observations of this blazar, and hence there are many UV data available in the archives. In particular, 3C~279 was monitored on a nearly daily basis with IUE and ROSAT for three weeks between December 1992 and January 1993, simultaneously with gamma-ray observations by EGRET, and with coordinated optical observations. We present here a study of the correlated variability of the UV continuum and of the broad Lyman $\\alpha$ emission line intensity over eight years, and compare our results with the findings of simultaneous optical, UV and X-ray monitoring in the period 2-5 January 1993. ",
        "conclusions": "In eight years, the UV continuum of 3C~279 has varied by a factor $\\sim$50, while the Lyman $\\alpha$ line flux has remained nearly constant.  This suggests that the observed highly variable continuum, most likely due to beamed  synchrotron radiation from a relativistic jet does not contribute significantly in powering  the emission line.  In fact the line equivalent width in 3C 279 is smaller than observed in 'normal' quasars, where the observed continuum, probably thermal radiation from an accretion disk, is also the source of ionizing radiation. If the line emitting gas in the broad line region of 3C 279 is also ionized by an inner disk, its radiation, usually swamped by the beamed blazar continuum, may become observable in very low states.  \\begin{figure*}[h] \\begin{center} \\leavevmode \\centerline{\\epsfig{file=optuvx.eps,width=10cm}} \\end{center} \\caption{\\em De-extincted simultaneous optical (from Grandi et al. 1996), UV (IUE) and soft X-ray (ROSAT) energy distributions in 2-5 January 1993.  The power-law fits to the ROSAT data and the black body fit to the IUE LWP and SWP data are shown as solid and dashed curves, respectively (1-$\\sigma$ ranges are also reported for the power-laws).}  \\end{figure*}   The spectral flattening observed in the 1200-2700 \\AA\\ range in correspondence with a very low UV continuum level is an uncommon feature in blazars, which generally exhibit spectral hardening during brighter states, and might represent the signature of the putative thermal, isotropic component underlying the highly variable, beamed continuum of 3C~279, and photoionizing the line emitting gas (Fig. 4).   The low state soft X-ray spectrum, which is well described ($\\chi^2 \\sim 1$) by a power-law of index $\\alpha_\\nu \\sim 0.7-0.8$ may also contain a Seyfert-like component. Assuming a simple accretion disk model described by a single black body, this would have a temperature of $\\sim$20000 K and a size of $\\sim$1 light day. The observation of a still weaker flux in January 1995 at the shorter UV wavelengths (Fig. 3) suggests some variability of this thermal component, albeit modest. More sensitive observations in the hard X-ray / gamma-ray band during a low state would be needed to constrain this hypothesis.  The presence of intense line emission and the suggestion of a thermal component in 3C~279 are consistent with the scenario in which the seed photons for the inverse Compton mechanism producing the gamma-rays are external to the relativistic jet and provided either by the broad line region or by the inner accretion disk. However, the observed  large amplitude variability in gamma-rays accompanied by lower amplitude variability at lower energies requires not only changes in the energetic electrons in the jet, {\\it but also variations in the soft photon field}, at least in a simple one zone emission model. A possible scenario is proposed  by Ghisellini and Madau (1996), whereby some line emission is induced by radiation from the jet. It is interesting to ask whether this mechanism would have some observable consequences on the observed line emission.  Although intriguing and promising, our results on the UV continuum and Lyman $\\alpha$ line characteristics in 3C~279 must be taken with caution: the available IUE and HST spectra are too few and too sparse in time to yield a definitive proof of lack of correlated variability between continuum and emission line at the shorter (one day or less) time scales.  Moreover, the low signal-to-noise ratio of the data prevent a very accurate measurement of the spectral index in low UV emission state.  Further data are necessary to confirm our findings. An intensive and regular monitoring of the UV spectrum of 3C~279 and other blazars, along with a detailed sampling during low emission states would clarify the existence and role of an isotropic emission component in this class of active galactic nuclei, and possibly lead to a link between blazar and normal quasar properties.  This task can be pursued more favourably in the UV spectral range, where the high ionization emission lines are located (at low or intermediate redshift) and which is less diluted by the stellar contribution of the host galaxy. This open problem represents an important heritage of IUE and its solution could be attempted only by a UV observing facility with its same easy and flexible scheduling."
    },
    "9805/astro-ph9805233_arXiv.txt": {
        "abstract": "In the ROSAT all-sky survey 11 HgMn stars were detected as soft X-ray emitters (Bergh\\\"ofer, Schmitt \\&\\ Cassinelli 1996). Prior to ROSAT, X-ray observations with the {\\em Einstein Observatory} had suggested that stars in the spectral range B5--A7 are devoid of X-ray emission. Since there is no X-ray emitting mechanism available for these stars (also not for HgMn stars), the usual argument in the case of an X-ray detected star of this spectral type is the existence of an unseen low-mass companion which is responsible for the X-ray emission. The purpose of the present work is to use all available data for our sample of X-ray detected HgMn stars and conclude on the nature of possible companions. ",
        "introduction": "\\label{intr} In the ROSAT all-sky survey X-ray emission was detected in 11 HgMn stars (5 spectroscopic binaries, 4 double-lined spectroscopic binaries, and 2 HgMn stars without available radial velocity data). For 2 of 3 spectroscopic binaries (SB) additional observations obtained with the ROSAT High-Resolution Imager (spatial resolution of $\\approx$ 5 arcsec) confirmed X-ray sources at the position of two systems (Bergh\\\"ofer \\&\\ Schmitt 1994). Known visual companions could be discarded as most likely X-ray emitters. Previous X-ray observations (by the {\\em Einstein Observatory}) had suggested that stars in the spectral range B5--A7 are devoid of X-ray emission. Since there is no X-ray emitting mechanism available for these stars, the usual argument in the case of an X-ray detection is the existence of an unseen low-mass companion which is responsible for the X-ray emission. However, this hypothesis is not easily testable.  For our sample of X-ray detected HgMn stars we used all available data to conclude on the nature of the companion. We emphasize that some of our sample stars consist of two nearly equal B stars. The observed X-ray emission in these systems is also inconsistent with the secondary star and, thus, a third component must exist to explain the X-ray emission by a low-mass companion. Some of our sample stars show X-ray luminosities that exceed the X-ray output of normal late-type stars and, therefore, an active pre-main sequence companion (PMS) is required. This hypothesis is supported by the fact that a significant fraction of the HgMn stars found in the ROSAT survey belong to rather young stellar groups like the Pleiades supercluster or the Sco-Cen association. Recently, it has been shown that there is a population of pre-main sequence stars in the Pleiades supercluster, and that both cluster and non-cluster members range in age from about $2.6 \\cdot 10^6$ to $10^8$ yrs (Eggen 1995; Oppenheimer et al. 1997). Many of these stars exhibit high levels of stellar activity and strong lithium lines.  If there is ongoing star formation in these regions, the phenomenon demands further study and the possibility of protostars in multiple star systems has wide-ranging implications. ",
        "conclusions": "Here we describe our method to conclude on the nature of possible low-mass companions. In a first step stellar masses and ages were derived for our sample of HgMn stars. For this we used the stellar model grids provided by Schaller et al. (1992); the stellar distances were taken from the recently released Hipparcos catalog and the effective temperatures were compiled from the literature. We then assumed that the absence of a secondary in the optical spectrum implies a mass ration of M$_1$/M$_2 \\geq 1.5$ for the two binary components and all systems are formed coeval. A further criterion was the saturation limit of $\\log ({\\rm L}_x/{\\rm L}_{Bol}) \\approx -3$ known for late-type star X-ray emission (cf. Schmitt 1997); for the observed X-ray luminosities this relation provides upper limits for the bolometric luminosities of the possible secondaries. Together with these limits for the companions masses, luminosities, and ages, we used the pre-main-sequence evolutionary tracks provided by D'Antona and Mazzitelli (1994) to limit the range of possible companions of the 11 HgMn stars.  For all of our sample HgMn stars detected in the ROSAT all-sky survey we find that a companion of lower mass can provide a natural explanation for the observed X-ray emission. In 7 cases (HD 32964, HD 33904, HD 35497, HD 75333, HD 110073, HD 141556, and HD 173524) the detected X-ray emission can be explained by a main-sequence late-type star, whereas for the stars HD 27295, HD 27376, HD 29589, and HD 221507 a PMS star is required. Further investigations by means of radial velocity studies and high-resolution imaging (e.g., in the near IR) are needed to detect the predicted companions. According to the lower mass limits derived for possible companions in our sample of HgMn stars, spectral types are in the range late K-M4.  It is remarkable that in many cases when a spectroscopic binary with a late-B primary has a third, distant companion, the SB primary is a HgMn star (e.g., Hubrig \\&\\ Mathys 1995)."
    },
    "9805/astro-ph9805249_arXiv.txt": {
        "abstract": "Bright EUV sunspot plumes have been observed in eight out of eleven different sunspot regions with the Coronal Diagnostic Spectrometer -- CDS on SOHO. >From wavelength shifts we derive the line-of-sight velocity, relative to the average velocity in the rastered area, 120$\\arcsec \\times$ 120$\\arcsec$. In sunspot plumes we find that the motion is directed away from the observer and increases with increasing line formation temperature, reaches a maximum between 15 and 41 km~s$^{-1}$ close to log T $\\approx$ 5.5, then decreases abruptly. The flow field in the corona is not well correlated with the flow in the transition region and we discuss briefly the implication of this finding. ",
        "introduction": "Foukal et al. (1974) introduced the notation ``sunspot plumes'' to describe areas above sunspot umbrae that are ``the brightest features in an active region by an order of magnitude''. This led to the idea that sunspot plumes are regions within large magnetic loops, extending to altitudes of several thousand kilometers above the photosphere, in which the temperature is one to two orders of magnitude lower than in the corona of the surrounding active region (Noyes et al. 1985). In contrast, based on numerous sunspot observations with the Ultraviolet Spectrometer and Polarimeter (UVSP) on the {\\it Solar Maximum Mission (SMM)} Gurman (1993) found that sunspot plumes were nearly nonexistent. Most recently Maltby et al. (1998) observed sunspot plumes in five out of nine sunspot regions with the Coronal Diagnostic Spectrometer (CDS; Harrison et al. 1995) on the {\\it Solar and Heliospheric Observatory (SOHO)} and discussed briefly previous conflicting results. The CDS observations showed that sunspot plumes exist in the upper part of the transition region, occur both in magnetic unipolar and bipolar regions, and may extend outside the umbra and into the penumbra.  >From the energy requirements in sunspot loops Foukal (1976) suggested that rapid downflows occur in the plumes. Strong downflows over sunspots were reported by Brueckner, Bartoe, \\& VanHoosier (1977) and studied by Nicolas et al. (1982), while Kjeldseth-Moe et al. (1988) found that both upflows and downflows occurred.  Other investigations have confirmed and extended these observations, for a review see Maltby (1997). None of the observations above referred specifically to plumes and to our knowledge the velocity in sunspot plumes is not known. An investigation by Brynildsen et al. (1998) on the connection between line emission and wavelength shift in sunspot regions may, however, hold some relevance to this. In this paper we extend the CDS material to twelve sunspot regions and present the first measurements of velocities in sunspot plumes. ",
        "conclusions": ""
    },
    "9805/astro-ph9805280_arXiv.txt": {
        "abstract": "We examine images of the field of X1832--330, the luminous (${\\rm L_X\\sim10^{36}\\ erg\\ s^{-1}}$) X-ray burst source near the center of the globular cluster NGC\\,6652, in order to identify the optical counterpart for further study.  U and B ground-based images allow us to set a limit $M_{B_0}\\squiggeqmm3.5$ for the counterpart at the time of those observations, provided that the color is $(U-B)_0\\sim-1$, similar to the sources known in other clusters.  Archival {\\it Hubble Space Telescope} observations survey most but not all of the $1\\sigma$ X-ray error circle, and allow us to set limits $M_{B_0}>5.9$ and $M_{B_0}>5.2$ in the WF/PC and WFPC2 regions, respectively.  In the WF/PC images we do weakly detect a faint object with UV-excess, but it is located \\decsec{11}{7} from the {\\it ROSAT} X-ray position. This considerable ($2.3\\sigma$) discrepancy in position suggests that this candidate be treated with caution, but it remains the only reasonable one advanced thus far.  We measure for this star $m_{439}=20.2\\pm0.2$, $(m_{336}-m_{439})=-0.5\\pm0.2$, and estimate $M_{B_0}=5.5$, $(U-B)_0=-0.9$, similar to other known optical counterparts.  If this candidate is not the identification, our limits imply that the true counterpart, not yet identified, is probably the optically-faintest cluster source yet known, or alternatively that it did not show significant UV excess at the time of these observations. Finally, we assess the outlook for the identification of the remaining luminous globular cluster X-ray sources. ",
        "introduction": "While the X-ray properties of the luminous globular cluster X-ray sources have been studied for over two decades, only in the last five years has there been significant progress in the study of their optical counterparts.  During this period the situation has gone from the existence of only one identification of an unusually optically-luminous system in M\\,15, to confirmed or likely optical counterparts in five clusters (Deutsch et al. 1998 and references therein).  {\\it Hubble Space Telescope (HST)} has largely been responsible for this advance, principally due to the extreme crowding in these fields, which limits the utility of ground-based programs.  Globular cluster X-ray sources are interesting targets for study for many reasons.  It has been known for over two decades that globular cluster X-ray sources are overabundant with respect to those in the field (Katz 1975; Clark 1975); while globular clusters contribute only a tiny fraction to the total number of stars in the Galaxy, 10\\% of the known low-mass X-ray binaries (LMXBs) are found in globular clusters (van Paradijs 1995).  It is still not clear whether these systems are somehow different as a group from those in the field, or instead the globular cluster environment merely enhances their formation probability.  Indeed, it has long been suspected that close binaries may dominate the binding energy in globular clusters, and these exotic binaries may hold important clues to binary formation and interaction in these clusters.  The number and properties of clusters containing LMXBs have been used to test stellar interaction hypotheses (e.g., Verbunt \\& Hut 1987, Predehl et al. 1991).  In addition, luminosities and intrinsic colors may be determined far more accurately than for field sources, as the distances and reddenings to the host clusters can be readily determined.  Identification and further study of the optical counterparts of these sources provide many more opportunities to determine system parameters and unravel the nature of these LMXBs than can X-ray observations alone.  Here we present results on a search for the optical counterpart, for which there is as yet no candidate, to X1832--330, the luminous (${\\rm L_X\\sim10^{36}\\ erg\\ s^{-1}}$) X-ray burst source near the center of the globular cluster NGC\\,6652.  This paper extends the preliminary work presented by Deutsch et al. (1997).  This X-ray source was probably first detected as H1825--331 in the HEAO-1 survey.  However, as the 2.7 deg$^2$ 90\\% confidence error box contained the cluster, but was very close to the Galactic Center, Hertz \\& Wood (1984) conservatively allowed that the association with NGC\\,6652 was premature.  With the significantly-better spatial resolution of the {\\it ROSAT} All-Sky Survey (RASS), Predehl et al. (1991) reported a bright X-ray source which was indeed coincident with NGC\\,6652 to 1$'$.  Using reprocessed RASS data, Verbunt et al. (1995) estimate that the flux was as much as $\\sim10\\times$ higher (depending on spectral assumptions) than during the HEAO-1 detection.  Pointed observations with the {\\it ROSAT} PSPC 1.5 yr after the RASS find the source somewhat brighter still, and show \\squig20\\% variations on a time scale of a few hours (Johnston et al. 1996).  X-ray luminosity estimates inferred from these various observations span the range ${\\rm L_X=10^{35} - 10^{36}\\ erg\\ s^{-1}}$, all normalized to the 0.5--2.5 keV band and distance adopted below.  This luminosity variation prompted Verbunt et al. (1995) to label this source as a transient, although the variability observed thus far seems orders of magnitude less than that of sources indisputably called transients.  Most luminous globular cluster X-ray sources are known to be bursters, indicating that the primaries are neutron stars.  Just recently X1832--330 joined the ranks of known bursters when two Type I bursts were reported by in 't Zand et al. (1998).  Ortolani et al. (1994) carried out the first color-magnitude study of NGC\\,6652.  They derive $({\\rm m-M})_0=14.85$ (${\\rm d=9.3}$ kpc), E($B-V)=0.10$, and estimate [Fe/H$]\\approx-0.9$.  These values are similar to previous determinations except for the distance, which is 30\\% closer.  From the compilation of Trager et al. (1993), we adopt a core radius $r_c=$\\ \\decsec{4}{3}.  X1832--330 should be a relatively easy target for a search for an optical counterpart as NGC\\,6652 is neither heavily reddened nor extremely dense, whereas most of the clusters which harbor luminous X-ray sources do fall into one or both of these two categories.  However, as the cluster is near the Galactic center ($l=\\decdegmm{1}{5}$, $b=\\decdegmm{-11}{4}$) the contamination of the field by bulge stars is of some concern.  A brief study of this issue is presented by Ortolani et al.  Finally we note that in their color-magnitude diagram, Ortolani et al. indicate a sizable population of blue stragglers. ",
        "conclusions": "We have presented a search of the optical counterpart for X1832--330, the luminous globular cluster X-ray source in NGC\\,6652.  Using the GSC reference frame, we determine the optical position of the {\\it ROSAT} X-ray coordinates on a CCD image.  U and B ground-based images from the AAT allow us to set a limit $M_{B_0}\\squiggeqmm3.5$ for the counterpart at the time of those observations, provided the color is $(U-B)_0=-1$. Archival {\\it HST} WFPC2 exposures which subtend most but not all of the error circle allow us to infer $M_{B_0}>5.2$ for the counterpart if in that region, again provided at the time of observation the source is UV-excess $(m_{218}-m_{439})_0=-1.4$, like the counterpart in NGC\\,6441.  Archival WF/PC observations allow a more sensitive search; within the \\squig90\\% of the \\decsec{5}{3} radius $1\\sigma$ error circle about the X-ray coordinates contained in the WF/PC images, we detect no objects at $m_{439}<20.6$ ($M_{B_0}<5.9$) with colors compatible with the other known optical counterparts in globular clusters.  The region outside radius $4''$ is not completely imaged by these data, and therefore a faint UV-excess counterpart could have been missed with these {\\it HST} observations.  We do weakly detect a faint UV-excess object \\decsec{11}{7} from the {\\it ROSAT} coordinates.  This is a $2.3\\sigma$ deviation from the X-ray coordinates, and thus this object certainly should not be completely ruled out based on its position.  If it is indeed the correct identification, this object provides another example of the extremely underluminous optical counterpart seen in NGC\\,1851.  We measure for Star 49 $m_{439}=20.2\\pm0.2$, $(m_{336}-m_{439})=-0.5\\pm0.2$, and estimate $B_0=20.4$, $(U-B)_0=-0.9$, and $M_{B_0}=5.5$.  Should the X-ray coordinates prove to be accurate to better than $4''$, the likely conclusion is that the true optical counterpart, not yet identified, is the intrinsically faintest cluster source yet known, at least at the time of these observations, and Star 49 may be yet another example of a faint UV-excess cluster object of unknown nature (see, e.g., Deutsch et al. 1996, 1998).  Another possibility is that the optical light from the system was not dominated by the hot accretion disk at the time of these observations, but rather by the secondary, thereby rendering the object's color unremarkable in our color-magnitude diagram; however, such behavior has not been observed for the other identified cluster sources.  Clearly, deep F336W, F439W WFPC2 observations placing the X-ray coordinates in the center of the PC chip are desirable to search this field more thoroughly, and a more accurate X-ray position would reduce the number of objects which must be considered as possible candidates.  Identification of optical counterparts in the remaining clusters with luminous X-ray sources will be difficult using current techniques, as is seen in Fig. 5.  Here we have plotted the $m_{336}$ apparent magnitudes, either observed or predicted, of the optical counterparts of luminous X-ray sources in the cores of globular clusters.  The objects are ordered by right ascension, so the units on the abscissa are of no significance.  Clusters with boxed names are those with optical counterparts already identified (or tentatively suggested in this work); all but one required {\\it HST} observations.  The dashed vertical lines denote the 5~mag range of the luminosity dispersion implied by the current complement of identifications, adjusted for the distance and reddening of each cluster (cluster parameters principally from the compilation of Djorgovski 1993).  The filled squares are the observed $m_{336}$ magnitudes derived from our {\\it HST} photometry (Deutsch et al. 1998); the positions of the squares within the dashed lines indicate the luminosities compared with the luminosity range of all the known sources (e.g., the object in NGC\\,1851 appears at the bottom of the dashed line as it is the least luminous one).  For the six sources with no current identification, the six open diamonds indicate where each of the known identifications would fall if relocated to the target cluster, again with appropriate distance modulus and reddening.  The horizontal line at $m_{336}=23$ denotes the approximate flux reached by a typical short WFPC2 program, i.e., 10\\% photometric precision in a two-orbit multicolor exposure series in a moderately crowded field.  Of the ``easy\" clusters, i.e. those for which the optical counterpart can be expected to be detected in one or two {\\it HST} orbits, NGC\\,6652 is the last for which a candidate has been put forth.  Due to the considerable foreground extinction, the remaining unidentified luminous cluster X-ray sources will be difficult to identify with current techniques at the UV and blue wavelengths where these sources have in the past been studied.  A possible exception is NGC~6440, which could be adequately studied in a few orbits.  For the remaining clusters, the amount of time required to detect a low-luminosity counterpart in an F336W frame with the WFPC2 becomes prohibitive, although detection of counterparts at the high end of the luminosity range is feasible.  Variability searches in the infrared may afford a way to identify and study the remaining, heavily-reddened sources. Future searches will also be facilitated when arcsecond-accuracy X-ray coordinates become available via AXAF observations, thereby drastically reducing the number of optical objects which must be considered. Nonetheless, there still remains much to be learned from the current crop of optical counterparts, just discovered in the last few years."
    },
    "9805/hep-ph9805211_arXiv.txt": {
        "abstract": "\\noindent We numerically investigate the decay, via parametric resonance, of the inflaton with an $m^2 \\phi^2$ potential into a scalar matter field with a symmetry breaking potential. We consider the case where symmetry breaking takes place during inflation. We show that when expansion is not taken into account symmetry restoration and non-thermal defect production during reheating is possible. However in an expanding universe the fields do not spend sufficient time in the instability bands; thus symmetry restoration and subsequent domain wall production do not occur. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805348_arXiv.txt": {
        "abstract": "We analyze ROSAT HRI observations obtained from 1992 to 1996 of the globular cluster \\tuc. Identifications of two X-ray sources with HD~2072 and with a galaxy, respectively, are used to obtain accurate ($<2''$) positions of the X-ray sources in the cluster. We find possible optical counterparts, including the  blue objects \\vv1 and \\vv2, for three X-ray sources in the core of \\tuc, but note that the probability of chance positional coincidence is significant.  One of the five sources previously reported by us to reside in the cluster core is found to be an artefact of misalignment between subsequent satellite pointings. ",
        "introduction": "The cores of globular clusters harbour many interesting objects detected at different wavelengths, such as X-ray sources, ultraviolet and visual variables and blue stragglers, and radio pulsars. The sheer number density of stars in the cluster cores makes identification of sources detected in one wavelength range with those found at other wavelengths a daunting task, especially for X-ray sources whose positions are uncertain by more than an arcsecond at best.  As an example, X-ray sources detected in the core of \\tuc\\ (Hasinger et al. 1994, henceforth called Paper~1)\\nocite{hjv94} have been tentatively identified with a cataclysmic variable (Paresce et al.\\ 1992), with blue stragglers (Meylan et al. 1996), and with a remarkable ultraviolet variable (Auri\\`ere et al.\\ 1989, Minniti et al.\\ 1997). \\nocite{pmf92}\\nocite{mmp+96}\\nocite{ako89}\\nocite{mmp+97} These various options are possible due to the uncertainty in the absolute positions of the X-ray sources of about 5$\\,''$.  In this paper we analyse three new ROSAT HRI observations of the globular cluster \\tuc, and re-analyse two. With use of the detailed astrometric study by Geffert et al.\\ (1997) we try to obtain an absolute accuracy of the X-ray positions at the arcsecond level. \\nocite{gak97} In Sect.\\ 2 we describe the observations and data analysis, and in Sect.\\ 3 the results. A discussion follows in Sect.\\ 4. ",
        "conclusions": "We have detected five X-ray sources in the core of \\tuc, and noted possible optical identifications for three of them.  Before we discuss the possible nature of these sources, we adress the question how confident we can be about the identifications. To do this, consider an area of $20''\\times20''$, centered on the cluster center according to Guhathakurta et al.\\ (1992). From Fig.~\\ref{fcore} we learn that this area contains three X-ray sources and 22 blue or variable stars. (Note that entry 8 of Table~3 in Geffert et al.\\ (1997) almost coincides with \\ako6.) If we suggest identification for each blue object lying in a $4''\\times4''$ box centered on an X-ray source, then the X-ray sources cover 12\\%\\ of the search area, and we have 22 trials for probability 0.12. The probabilities of finding 0, 1, 2 or 3 identifications are 6, 18, 26 and 23 \\%, respectively. We conclude that the probability that all suggested identifications are accidential is quite high.  It may be argued that the suggested identifications are special also optically. If we consider the three objects \\vv1, \\vv2 and \\vv3 only, we have 3 trials for probability 0.12, and the probability of finding 0, 1 or 2 identifications are 68, 28 and 4 \\%. Even for this limited set, the probability that both identifications of \\vv1 with \\x9 and \\vv2 with \\x19 are due to chance is non-negligible. For the moment we conclude that our suggested identifications are possible, but not secure.  If we assume that \\vv1 and \\vv2 may be identified with \\x9 and \\x19, respectively, we learn from Fig.~\\ref{fcvs} that their ratio of X-ray to optical flux is rather high if they are cataclysmic variables, but as expected for soft X-ray transients in quiescence. The X-ray countrates of the cataclysmic variables in Fig.~\\ref{fcvs} have not been corrected for interstellar absorption; the correction is expected to be small for most systems, but not necessarily for all. For typical X-ray spectra of cataclysmic variables, the visual flux is affected more strongly by interstellar absorption than the X-ray countrate, and thus it is not expected that correction for absorption will increase the ratio of X-ray to optical flux for cataclysmic variables. We conclude that Fig.~\\ref{fcvs} provides another illustration of the argument originally made by Verbunt et al. (1984) \\nocite{vpe84} that some of the dim X-ray sources in the cores of globular clusters are too bright to be cataclysmic variables.  The X-ray flux of \\x9 is variable; that of \\x19 may or may not be variable. The range of variability in \\x9 is not unprecedented in soft X-ray transients in quiescence: the variations in the flux of Cen X-4 in quiescence, reported by Campana et al.\\ (1997) and shown in Fig.~\\ref{fcvs}, is of a similar magnitude as that observed in \\x9. Such variations in a quiescent soft X-ray transient are not expected to be accompanied by detectable optical variations, and thus the absence of optical variation in \\vv1\\ need not be in conflict with the suggested identification. It may be noted that similar variations in the X-ray flux without accompanying variations in the optical are probably also possible in cataclysmic variables. For example, the dwarf nova VW~Hyi was brighter in quiescence when observed with ROSAT in Nov 1990 than when observed with EXOSAT several years earlier (Wheatley et al.\\ 1996, their Fig.~7).\\nocite{wvb+96}  \\vv2 has been detected at a level about 4 magnitudes above its quiescent level twice; this magnitude difference is more indicative of a dwarf nova than of a soft X-ray transient, as noted by Paresce \\&\\ De Marchi (1994) and by Shara et al.\\ (1996).  The two constant X-ray sources \\x5 and \\x7 in the core of \\tuc\\ have no suggested optical counterparts. The level and the constancy of their X-ray fluxes are compatible with them being radio pulsars. For example, PSR$\\,$B$\\,1821-24$ in globular cluster M$\\,$28 and PSR$\\,$J$\\,0218+4232$, at comparable distances as \\tuc\\ (5.5 and $>$5.7 kpc, respectively compared to 4.6 kpc for \\tuc), have ROSAT PSPC countrates of the same order of magnitude as \\x5 and \\x7. Whether \\x5 or \\x7, or any of the four X-ray sources just outside the core, can be identified with any of the 11 radio pulsars in \\tuc\\ awaits further study of the radio pulsars, in particular determination of their positions, and of their period derivatives (so that the X-ray data can be folded on a known period). More accurate pinpointing of the X-ray positions will be possible with AXAF. Considering the large numbers of potential optical counterparts, optical or ultraviolet monitoring of the inner region of \\tuc\\ simultaneous with the X-ray observations would be very useful, as detection of simultaneous X-ray and optical variability would strenghten any identification based on positional coincidence only.  To summarize, we find possible optical counterparts for three of the five X-ray sources in the core of \\tuc, but note that all could be chance positional coincidences. The X-ray luminosities of \\x5, \\x7 and \\x9 are rather high for these to be cataclysmic variables, but compatible with soft X-ray transients in quiescence. \\x9 is a variable X-ray source, and its X-ray to optical flux ratio suggests that it is a soft X-ray transient, hitherto always observed in quiescence.  The steadier sources \\x5 and \\x7 may be either soft X-ray transients or recycled radio pulsars.  The sources \\x19 in the core, and \\x4, \\x6, \\x11 and \\x13, outside but near the core, have X-ray luminosities $L_{\\rm X}<10^{32}$erg/s, compatible with them being soft X-ray transients, cataclysmic variables, or recycled radio pulsars. If \\vv2 is indeed a cataclysmic variable, it is probably the best candidate counterpart hitherto suggested for an X-ray source in \\tuc."
    },
    "9805/astro-ph9805038_arXiv.txt": {
        "abstract": "Observations of the diffuse emission in the 8--22 keV energy range, elongated parallel to the Galactic plane \\cite{smp93} and detection of the strong 6.4 keV fluorescent line with $\\sim$ 1 keV equivalent width from some giant molecular clouds (e.g. Sgr B2) in the Galactic Centre region \\cite{koy94} suggest that the neutral matter of these clouds is (or was) illuminated by powerful X-ray radiation, which gave rise to the reprocessed radiation. The source of this radiation remains unknown.  Transient source close to the Sgr B2 cloud or short outburst of the X-ray emission from supermassive black hole at the Galactic Centre are the two prime candidates under consideration. We argue that new generation of X-ray telescopes combining very high sensitivity  and excellent energy and angular resolutions would be able to discriminate between these two possibilities studying time dependent changes of the morphology of the surface brightness distribution, the equivalent width and the shape of the fluorescent line in the Sgr B2 and other molecular clouds in the region. We note also that detection of broad and complex structures near the 6.4 keV line in the spectra of distant AGNs, which are X--ray weak now, may prove the presence of violent activity of the central engines of these objects in the past. Accurate measurements of the line shape may provide an information on the time elapsed since the outburst. Proper motion (super or subluminal) of the fluorescent radiation wave front can give additional information on the location of the source. Observations of the described effects can provide unique information on the matter distribution inside Sgr B2 and other giant molecular clouds. ",
        "introduction": "Prediction \\cite{smp93} and discovery \\cite{koy94,koy96} of the bright iron fluorescent $K_\\alpha$ line in the direction of the molecular cloud Sgr B2 and Radio Arc in the Galactic Centre region should not remain unnoticed by the astrophysicists planning in the nearest future launch of the sensitive X--ray spectrometers  on board {\\it AXAF, XMM, ASTRO--E, ABRIXAS, Spectrum--X--Gamma}. These spectrometers will provide high angular resolution from seconds to minutes of arc and spectral resolution from 5 to 140 eV near X-Ray lines of iron in the 6--7 keV energy band. The missions of the next millennium, starting with {\\it Constellation} (White, Tananbaum \\& Kahn, 1997) and {\\it XEUS} \\cite{tur97}, are to achieve energy resolution of 2 eV and better. Particularly relevant problem, which deserves further consideration,  is the illumination of a molecular hydrogen cloud with column density $N_H\\sim 10^{23} - 10^{24}~~cm^{-2}$ (i.e. $\\tau_T\\sim 0.1-1$) by a variable X--ray emission from a bright transient source inside the cloud or outside the cloud (e.g. short episode of the effective accretion onto Sgr A* due to the tidal star disruption). The solution of this problem allows one to study the time evolution of the spectrum emerging from the cloud after fading of the primary X--ray source. The radius of the Sgr B2 cloud is of the order of 20 pc (e.g. Lis \\& Goldsmith 1989), although the size of the dense core(s) is significantly smaller ($\\sim 0.3$ pc, e.g. de Vicente et al. 1997). Depending on the mutual location of the cloud and a primary source of the continuum emission substantial evolution of the morphology, flux, equivalent width and shape of the iron fluorescent line might be noticed on the time scale as short as 0.1--10 years (the value which is not incomparable with the life time of the best modern space observatories). Three years already passed since the moment of the first firm detection of the line from Sgr B2 \\cite{koy94,koy96}. Detailed observations would allow one to reveal the geometry of the problem (i.e. mutual location of the primary source and the cloud), time elapsed since fading of the primary source flux. Observations might also shed additional light on the mass of the cloud, it's uniformity and, provided energy resolution better than 1 eV, on the matter and velocity distribution inside the cloud.  As the first approximation we are considering below qualitatively how the effects of time delay, large opacity and scattering by bound electrons affect the appearance of the 6.4 keV line from a molecular cloud illuminated by a continuum X--ray emission. Throughout the paper we adopted the approximation of Morrison and McCammon (1983) for photoelectric absorption ($\\sigma_{ph}(E)$) in the neutral gas, having a normal abundance of heavy elements, an abundance of iron of $\\delta_{Fe}=3.3\\times 10^{-5}$ with respect to hydrogen, a cross section of photoabsorption from iron K-shell as $\\sigma_{Fe}(E)=3.53\\times 10^{-20}\\times (E/7.1~keV)^{-3} cm^2/atom$ and a $K_\\alpha$ fluorescent yield $Y=0.3$ (e.g. Bambinek et al., 1972).  For simplicity we are considering (unless stated otherwise) the simplest case  of a point source of continuum X-ray emission in the centre of a spherically symmetric cloud of neutral gas.  The formation of the $K_\\alpha$ line from the neutral matter illuminated by a continuum radiation was considered in many publications (e.g. Basko, Sunyaev \\& Titarchuk 1974, Fabian 1977, Basko 1978, Vainshtein \\& Sunyaev 1980, Inoue 1985, George \\& Fabian 1991, Matt et al. 1991, Awaki et al. 1991, Nandra \\& George 1994, Ghisellini et al. 1994). In the discussion below we present the arguments which are particularly relevant to the observations of the Galactic Centre region in the neutral iron fluorescent line.  This paper does not pretend to explain the nature of an X--ray emission from the Galactic Centre region. Instead we consider a number of simple effects which might play an important role in the environment like the central region of our Galaxy. These effects could be used by the missions like {\\it Constellation} and {\\it XEUS} to verify the hypothesis that molecular clouds near the Galactic Centre were exposed to outburst of hard X--Ray radiation: \\begin{itemize} \\item Dependence of the morphology of the 6.4 keV surface brightness distribution on the mutual location of the source and the cloud. \\item Apparent motion (sub or superluminal) of the features associated with propagation of the source radiation through the clouds. \\item Evolution of equivalent width and shape of the 6.4 keV line as an indicator of multiple scatterings and time elapsed since outburst. \\end{itemize}  The structure of the paper is as follows: in sections 2 and 3 we discuss temporal behavior of scattered flux and morphology of the scattered radiation associated with first scattering, in the subsequent sections 4--7 we argue that equivalent width and the shape of the fluorescent line (due to multiple scatterings) may be used as another indicator of  cloud illumination with powerful flares in the past, section \\ref{ssum} summarizes the results. In appendix a simplified derivation of the evolution of the equivalent width and shape of the line after multiple scatterings is given. ",
        "conclusions": "\\label{ssum} The value of the equivalent width and the shape of the 6.4 keV iron $K_\\alpha$ line, emerging from the neutral matter, illuminated from inside or outside by the X--ray continuum spectrum, contains information on the time history of the illuminating continuum flux. Assuming normal abundance of iron in the scattering media the value of an equivalent width of about 1 keV indicate that we are dealing with a scattered component. Values in excess of 1 keV can be due to strong photoabsorption or multiple Thomson scattering. Spectroscopic analysis of the 6.4 keV line profile and the shape of the 7.1 keV absorption edge can help to distinguish between these possibilities. When combined with broad band spectroscopic measurements it can be used to determine the position and flux history of the primary continuum source. A new generation of X--ray instruments should be capable of achieving the required sensitivity and energy resolution to accurately measure the detailed structure of the 6.4 keV line in the Galactic Centre region. Comparing the flux and shape of the line from different molecular clouds in the region, one can reconstruct the date and duration of the flare, responsible for observed reprocessed emission. Note also that observations, over a period of 5--10 years, may show the variability of the line flux, shape and morphology of it's surface brightness distribution.   The assumption that Sgr A$^*$ emission has caused the 6.4 keV line (and continuum) emission from Sgr B2 \\cite{smp93,koy96} implies that its flux vary by a factor of at least $10^3$ (and perhaps much more) on a time scale of hundreds of years. If nuclei in other galaxies also have similar behavior there is a good chance of detecting `delayed' scattered components from some of them. With the high sensitivity of the {\\it Constellation} mission one can search for very weak nearby and distant AGNs. The detection of an iron line of large equivalent width and complex shape would prove the past violent activity of the central engines of these sources.  This work was supported in part by the grants RBRF 96-02-18588 and INTAS 93-3364-ext. We thank Marat Gilfanov for useful discussions and anonymous referee for helpful comments."
    },
    "9805/astro-ph9805274_arXiv.txt": {
        "abstract": "This paper presents a comparative study of emission line ratios of the Narrow Line Region (NLR) of Seyfert 1 and Seyfert 2 galaxies. It includes a literature compilation of the emission line fluxes [OII]$\\lambda$3727\\AA, [NeIII]$\\lambda$3869\\AA, [OIII]$\\lambda$5007\\AA\\ and [NeV]$\\lambda$3426\\AA, as well as 60$\\mu$m continuum flux, for a sample of 52 Seyfert 1's and 68 Seyfert 2's. The distribution of the emission line ratios [OII]/[NeIII] and [OII]/[NeV] shows that Seyfert 1's and Seyfert 2's are statistically different, in the sense that Seyfert 1's have values smaller than those of Seyfert 2's, indicating a higher excitation spectrum. These and other emission line ratios are compared with sequences of models which combine different proportions of matter and ionization bounded clouds and also sequences of models which vary only the ionization parameter. This comparison shows that the former models reproduce better the overall distribution of emission line ratios, indicating that Seyfert 1's have a smaller number of ionization bounded clouds than Seyfert 2's. This difference, together with other results available in the literature, are interpreted from the point of view of four different scenarios. The most likely scenario assumes that Seyfert 1's have NLR's smaller than those of Seyfert 2's, possibly due to a preferential alignment of the torus axis close to the host galaxy plane axis in Seyfert 1's. ",
        "introduction": "The observation of broad polarized lines in the spectrum of the Seyfert 2 galaxy NGC1068 (Antonucci \\& Miller 1985) showed that Seyfert 2's can be Seyfert 1's where the direct view of the central engine is blocked. This is the basis for the Unified Model of AGN's, which assumes that objects of different activity class, like Seyfert 1's and Seyfert 2's, are the same kind of object, surrounded by a dusty molecular torus. The orientation of this torus relative to the line of sight determines whether the AGN is observed as a broad lined object (Seyfert 1), where the nuclear engine is seen through the torus opening, or as a narrow lined object (Seyfert 2), where our view of the central engine and consequently the broad lines, is blocked by the torus.  Several pieces of observational evidence supporting this scenario have been obtained during the last decade, the strongest one being the observation of polarized broad emission lines in the spectrum of several Seyfert 2 galaxies (Antonucci \\& Miller 1985; Miller \\& Goodrich 1990; Kay 1994; Tran 1995). The observation of collimated radiation escaping the nuclear region of Seyfert 2's, seen as cone like emission line regions (Pogge 1988a,b, 1989; Schmitt, Storchi-Bergmann \\& Baldwin 1994, Schmitt \\& Storchi-Bergmann 1996, and references therein), or linear radio structures (Ulvestad \\& Wilson 1984a,b, 1989), also suggest that the direct view of the central engine is blocked in these objects. More direct evidence for the obscuration of the central engine in Seyfert 2's comes from the analysis of X-ray spectra, which show large absorbing column densities in these objects (Mulchaey, Mushotzky \\& Weaver 1992). Also, the observation of H$_2$O masers very close to the nucleus of some Seyfert 2's, like NGC1068 and NGC4258 (Miyoshi 1995, Gallimore 1996, Greenhill 1996), show the presence of large concentrations of molecular gas, hiding the central engine.  Recent papers, however, present some results suggesting that not only the orientation of the circumnuclear torus relative to the line of sight, but also its orientation relative to the host galaxy may be important in the AGN classification. It was known since Keel (1980), that there is a paucity of Seyfert 1's with edge-on host galaxies. This result was later confirmed by Maiolino \\& Rieke (1995) and Simcoe et al. (1997), who suggested that, in some cases, dust along a Seyfert 1 galaxy disk may be responsible for the obscuration of the broad lines (making it appear as a Seyfert 2). Moreover, Schmitt et al. (1997) presented a comparison between the linear radio structure of Seyfert galaxies, with their host galaxy major axis. They found that the radio structures are more likely to be aligned close to the host galaxy plane axis in Seyfert 1's, but can have any direction in Seyfert 2's, confirming the result by Maiolino \\& Rieke (1995). Another result that corroborates this scenario is the observation that the NLR of Seyfert 1's usually is much smaller than that of Seyfert 2's, when they are compared as if Seyfert 2's were observed pole-on, in the same way as Seyfert 1's (Schmitt \\& Kinney 1996). The smaller Seyfert 1 NLR's can be understood if these objects have their torus axis preferentialy aligned close to the host galaxy plane axis, where there is less gas to be ionized.  The above results show differences between the NLR of Seyfert 1 and Seyfert 2 galaxies and point towards older papers, where some other differences have also been detected. Heckman \\& Balick (1979) and Shuder \\& Osterbrock (1981) showed that the ratio [OIII]4363/5007 is larger in Seyfert 1's than in Seyfert 2's. This result indicates that the [OIII] zone of Seyfert 1's, when compared to Seyfert 2's, have larger temperatures and/or densities. Yee(1980) and Shuder (1981) showed that the emission lines, [OIII], [OII] and [OI], are more luminous in Seyfert 2's relative to Seyfert 1's of similar optical luminosity, consistent with the torus blocking part of the continuum light in Seyfert 2's. Shuder \\& Osterbrock (1981) and Cohen (1983) showed that the emission line ratios [FeVII]/H$\\beta$ and [FeX]/H$\\beta$ are larger in Seyfert 1's than in Seyfert 2's, indicating that Seyfert 1's have higher excitation. Yet another interesting result was obtained by De Robertis \\& Osterbrock (1986 and references therein), who showed that the FWHM of forbidden lines are well correlated with the ionization potential in Seyfert 1's, but not with the critical  density for de-excitation, while in Seyfert 2's the opposite happens. They have also showed that these lines have smaller FWHM in Seyfert 1's than in Seyfert 2's, and that the [OI] line profiles show evidence of two components in Seyfert 2's, probably formed in two different regions.  This paper presents a compilation of literature data of the emission line fluxes [OII]$\\lambda$3727\\AA, [NeIII]$\\lambda$3869\\AA, [NeV]$\\lambda$3426\\AA\\ and [OIII]$\\lambda$5007\\AA\\ ([OII], [NeIII], [NeV] and [OIII], hereafter), as well as 60$\\mu$m continuum fluxes for a sample of 52 Seyfert 1 and 68 Seyfert 2 galaxies. These lines are used to compare the excitation of the NLR gas in Seyfert 1's and Seyfert 2's, through the analysis of different emission line ratios.  A simple interpretation of the Unified Scheme would suggest that the spectrum of the NLR of Seyfert 1's and Seyfert 2's should have similar degrees of excitation. However, as shown by the above papers, this may not be true. Effects like the possible obscuration of parts of the NLR by the torus, or the smaller NLR size in Seyfert 1's, could influence the average NLR excitation in these two classes of objects.  The paper is organized in the following way, Section 2 presents the sample, the reasons for the choice of these emission lines and a discussion of the possible selection effects. Section 3 shows the results of the comparison between Seyfert 1's and Seyfert 2's. Section 4 shows the comparison between the data and photoionization models and discusses possible interpretations of the results, while Section 5 gives the summary. ",
        "conclusions": "\\subsection{Photoionization Models}  The results presented in the previous section show that the average excitation of the NLR of Seyfert 1's is larger than that of Seyfert 2's. This result is interpreted using diagnostic diagrams involving the emission line ratios studied in this paper. Figures 7a,b,c show the diagrams  Log~[OII]/[NeV]$\\times$Log~[OII]/[NeIII], Log~[NeIII]/[NeV]$\\times$Log~[OII]/[NeIII] and Log~[OII]/[OIII]$\\times$Log~[OII]/[NeIII], respectively. It can be seen that Seyfert 1's are more concentrated towards the lower left side in these diagrams, which correspond to a higher excitation, confirming the results obtained from Figures 3, 4, 5 and 6. These distributions of emission line ratios are compared with photoionization models, to analyze the possible origins of this difference in excitation.  These results can be interpreted from the point of view of models that combine different proportions of matter and ionization bounded clouds. In these models the matter bounded clouds produce most of the high excitation lines ([NeIII], [OIII] and [NeV]) and little of low excitation lines ([OII] and [NII]), while the ionization bounded clouds produce most of the low ionization lines and little of high excitation lines.  The use of such models was proposed by Viegas \\& Prieto (1992) to explain the emission line region of 3C227. Later Binette, Wilson \\& Storchi-Bergmann (1996) (BWSB96, hereafter) used models of this kind to study the extended NLR of Seyfert galaxies, showing their efficacy in the reproduction of  high excitation lines, like [NeV]$\\lambda$3426\\AA\\ and HeII$\\lambda$4686\\AA, as well as the [OIII] temperature, which has always been a problem for the traditional photoionization models which use sequences of ionization parameter.  Sequences of models adding different proportions of matter and ionization bounded clouds were calculated using the photoionization code MAPPINGS (Binette et al. 1993a,b), following the description given in BWSB96. The models were calculated using a power law ionizing spectrum of the form F$_{\\nu}\\propto\\nu^{\\alpha}$, and two different values of $\\alpha$ were tested, --1.3 and --1.5.  The matter bounded clouds are ionized by this spectrum and the calculation stops when 40\\% of the incident spectrum is absorbed. The output, reprocessed spectrum from the matter bounded clouds, is the one that ionizes the ionization bounded clouds. The models also assume that the ionization bounded clouds leak some of the input radiation, in order to avoid overproduction of low ionization lines, like [OII] and [NII]. In the case of $\\alpha=-1.3$, it is assumed that the ionization bounded clouds allow 3\\% of the ionizing radiation to escape, while for $\\alpha=-1.5$ this value is 10\\%.  The models were calculated considering an isobaric prescription, where the pressure is constant within any matter bounded or any ionization bounded cloud.  The ionization parameter adopted for the matter bounded spectrum was U=0.04. Nevertheless, for the ionization bounded clouds of the A$_{M/I}$ sequence (see below), instead of specifying the ionization parameter, their pressure was fixed at 20 times that of the matter bounded clouds, as done by BWSB96. The adopted density was n=50 cm$^{-3}$ and the gas metal abundance was solar (Z=1). It is also assumed that the gas is mixed with a small quantity of dust $\\mu=0.015$\\footnotemark and the abundance of metals in the grains is depleted from the gas. Notice that this is a very small amount of dust and, according to Binette et al. (1996), higher amounts of dust produce only minimal changes in the output spectrum of ionization bounded clouds but can have larger effects on the matter bounded clouds. However, the matter bounded clouds are not expected to have a large quantity of dust, because it can be easily destroyed by the radiation field. The only independent parameter in these models is the ratio between the solid angle subtended by matter bounded clouds, relative to the solid angle subtended by the ionization bounded clouds (A$_{M/I}$). Larger values correspond to a larger contribution from matter bounded clouds relative to ionization bounded clouds and vice-versa. This parameter was varied in the range 0.01$\\leq$A$_{M/I}\\leq634$, in steps of 0.2 dex.  These models are represented as a solid line in Figures 7a,b,c, with the value of $\\alpha$ indicated beside the line.  \\footnotetext{$\\mu$ is the dust to gas ratio of the clouds, in units of the solar neighborhood dust-to-gas ratio}  In order to test the effects of other physical and chemical conditions, two other sequences of A$_{M/I}$ models were calculated. In the first one $\\alpha=-1.5$, with the same parameters as above, but for gas with twice the solar metalicity (Z=2). In the second set of models $\\alpha=-1.5$ and Z=1, but the density is 500 cm$^{-3}$. These models have the same range of A$_{M/I}$ as above, are shown as a long dashed lines in Figure 7a,b,c and are identified as Z=2 and n=500 cm$^{-3}$, respectively.  Just for comparison with the above models, MAPPINGS was also used to calculate traditional sequences of models, varying only the ionization parameter.  The parameters of the models were, power law ionizing spectrum with $\\alpha=-1.3$, constant density n=50 cm$^{-3}$, metalicity Z=1 and dust content $\\mu=$0.015. As for the A$_{M/I}$ models, two other sequences, one with n=500 cm$^{-3}$ and Z=1, and the other with n=50 cm$^{-3}$ and Z=2, were also tried. It was assumed that 3\\% of the ionizing radiation escape from the clouds, in order to avoid the overproduction of low ionization lines. The ionization parameter was varied in the range $-4\\leq$Log~U$\\leq-0.8$, in steps of 0.2 dex. The three sequences of models are very similar, the only exception being the sequence of models with Z=2 in the diagram Log([OII]/[OIII])$\\times$Log([OII]/[NeIII]) (Figure 7c), which are very similar to the A$_{M/I}$ sequence with Z=2. Due to this fact, only the sequence with n=50 cm$^{-3}$ and Z=1 is presented as a dotted line in Figures 7a,b,c.  It can be seen in Figures 7a and b, that the A$_{M/I}$ sequences of models cover very well the observed distribution of values. In the case of Figure 7c, the diagram Log([OII]/[OIII])$\\times$Log([OII]/[NeIII]), these models have some problem to reproduce the observed distribution of values. It would be necessary to change other parameters, like the amount of dust in the models, the pressure jump between the matter and the ionization bounded clouds, or the amount of radiation which leaks from the ionization bounded clouds, in order to better reproduce the observed distribution of values.  The fact that Seyfert 1's have [OII]/[NeIII] and [OII]/[NeV] values smaller than Seyfert 2's, can be interpreted as due to a smaller contribution from ionization bounded clouds, relative to matter bounded clouds, to the spectra of those objects.  This comparison also shows that the A$_{M/I}$ models with $\\alpha=-1.3$ are not as good a representation for the observed values as the ones with $\\alpha=-1.5$, because they produce too much large fluxes of the higher excitation lines, like [NeV].   The comparison with the traditional U sequence of models, shows that they are a poor representation of the data points, even when varying parameters like the gas abundance or density.  Only in Figure 7c, where the A$_{M/I}$ sequence of models has some problems to represent the observed distribution of points, these models could be a better representation for the data. However, they require unconventionally large ionization parameters (U$>$0.01).  \\subsection{Possible interpretations}  Four possible interpretations for the above result are studied here.  {\\it 1-) Part of the matter bounded clouds (which produce most of [NeIII], [OIII] and [NeV]), is hidden by the circumnuclear torus in Seyfert 2's}. A similar problem was found by Jackson \\& Browne (1990) in the comparison of Quasars with Radio Galaxies. They show that the [OIII] emission of Quasars is much stronger than that of Radio Galaxies, proposing that part of the [OIII] emission is obscured by the torus in the latter objects. Hes, Barthel \\& Fosbury (1993) showed that, when comparing the [OII] emission of Quasars and Radio Galaxies, which comes from a less obscured, lower excitation region, both classes of objects have very similar distributions, corroborating the obscuration scenario.  While the obscuration scenario can be the solution for Radio Galaxies and Quasars, it may not be the general case for Seyfert 2 galaxies. Assuming that the [OII] emission in Seyfert 2's is similar to that of Seyfert 1's and not blocked by the torus, we can calculate, using the average values given in Table 3, that $\\approx$40\\% of the [NeIII] emission, $\\approx$55\\% of the [NeV] emission and $\\approx$25\\% of the [OIII] emission should be blocked by the torus in Seyfert 2's. This could happen for some of the Seyfert 2's in the sample, but notice that these are large values and go against the fact that Seyfert 2's have lower excitation lines (like [OII]) more luminous than Seyfert 1's of similar optical luminosity (Yee 1980; Shuder 1981). Also, Seyfert 2's usually have extended NLR's (Pogge 1989; Schmitt \\& Kinney 1996). Another fact that goes against the obscuration scenario being the general case is that, if part of the high excitation emission line region is hidden by the torus, we would expect to see considerable amounts of polarized [OIII] emission in Seyfert 2's. As shown by Goodrich (1992), with a small number of exceptions, Seyfert 2's do not have high degrees of polarized [OIII] emission.  {\\it 2-) We see a smaller number of ionization bounded clouds in Seyfert 1's, because they are seen from the back and are extincted.} Since the ionization bounded clouds are responsible for most of the [OII] emission and very little of the [NeIII], [NeV] and [OIII], this would imply a reduction of the ratios [OII]/[NeIII], [OII]/[NeV] and [OII]/[OIII] in Seyfert 1's, relative to Seyfert 2's.  >From the analysis of the X-ray spectra of Seyfert 1's (Reynolds 1996; Weaver, Arnaud \\& Mushotzky 1995), it is known that they usually have small column densities of absorbing material (N$_{HI}<10^{21}$ cm$^{-2}$). Assuming a standard dust-to-gas ratio (A$_V=5\\times10^{-22}$N$_{HI}$), it is possible to estimate a typical value of extinction from the above N$_{HI}$, which is A$_V<0.5$ (E(B-V)$\\approx$0.2). In the case of E(B-V)=0.1, the [OII] emission of the ionization bounded clouds would be reduced by $\\approx$35\\%, which could explain the difference between Seyfert 1's and Seyfert 2's. However, this scenario only works when the ionization bounded clouds do not block the direct view of the matter bounded clouds, otherwise the high excitation lines would also be obscured.   {\\it 3-) There is a smaller number of ionization bounded clouds in Seyfert 1's, possibly due to the orientation of the circumnuclear torus relative to the galaxy plane.} In thisscenario Seyfert 1's have their circumnuclear torus axis preferentially aligned closer to the host galaxy plane axis, while in Seyfert 2's the torus can have any orientation. In this way, Seyfert 1's would have smaller NLR's, because their ionizing radiation would shine out of the galaxy disk and find only a small number of clouds to be ionized, thus resulting in a smaller number of ionization bounded clouds in these objects. On the other hand, since the Seyfert 2's torus axis can have any orientation relative to the host galaxy disk, there is a larger chance for the ionizing radiation to cross the galaxy disk in this objects, which would result in a larger quantity of gas clouds to be ionized. The clouds closer to the nucleus filter the ionizing radiation and the more distant clouds are ionized only by this fainter and filtered continuum. Due to the larger number of clouds along the disk, the nuclear radiation ionizes a larger number of clouds, and this effect is similar to be seeing a larger number of ionization bounded clouds in Seyfert 2's.  Some of the results available in the literature, discussed in the introduction, corroborate this scenario. Seyfert 1's have higher [OIII]4363/5007 ratios than Seyfert 2's, which could be explained as higher [OIII] temperatures, or higher densities. If the higher [OIII]4363/5007 ratios of Seyfert 1's are in fact due to a higher [OIII] temperature, this is consistent with a smaller proportion of ionization bounded clouds in these objects, as shown by Binette et al. (1996) models. This interpretation can also explain the results obtained by Schmitt \\& Kinney (1996), that Seyfert 1's have much smaller NLR's than Seyfert 2's (when they are compared in a similar way, as if they were seen pole-on). Kraemer et al. (1998) confirmed this to the individual case of the Seyfert 1 galaxy NGC5548, showing that this galaxy have a compact NLR, with a size of the order of 70pc.  The above results imply that the NLR of Seyfert 1's have less gas than the NLR of Seyfert 2's, which can be explained if the Seyfert 1's torus axis is aligned closer to the host galaxy plane axis. This scenario is supported by the observation of a lack of Seyfert 1's in edge-on galaxies (Keel 1980; Maiolino \\& Rieke 1995; Simcoe et al. 1997) and by the relative orientation between linear radio  structures and the host galaxy major axis in Seyfert 1's (Schmitt et al. 1997).  {\\it 4-) Seyfert 2's are more associated with circumnuclear star formation (high metallicity HII regions) than Seyfert 1's.} Since high metallicity HII regions are strong emitters of [OII] and weak emitters of [NeIII], if the nuclear emission of Seyfert 2's is more likely to be mixed with HII regions than Seyfert 1's, this would explain the fact that their NLR's show less excited gas. Some evidence for the existence of circumnuclear regions in Seyfert 2's is given by Heckman et al. (1995), Heckman et al. (1997), Thuan (1984). However, this evidence is restricted to a small number of galaxies and it would be necessary to study the stellar population of a complete sample of Seyfert 1's and Seyfert 2's, in order to see if there is any difference between these two classes of objects and if Seyfert 2's in fact have more circumnuclear star formation. One such attempt was done by Schmitt, Storchi-Bergmann \\& Cid Fernandes (1998), who synthesized the nuclear stellar population of 20 Seyfert 2's, showing that young stars usually contribute with less than 5\\% (less than 1\\% in more than 50\\% of the sample) to the light of these galaxies at $\\lambda$5870\\AA."
    },
    "9805/astro-ph9805042_arXiv.txt": {
        "abstract": " ",
        "introduction": "Blue compact dwarfs galaxies (BCDG) are still experiencing a strong star formation event.  Their low metallicity suggests that these objects are unevolved. The nature and the age of the most underabundant ones are still controversial: are they ``young'' galaxies forming stars for the first time or older systems which have evolved very slowly ? Despite of extensive searches, no local galaxy with a metallicity lower than 1/50 $Z\\odot$ has been found, nor massive primordial HI clouds, without optical counterpart, at low redshift.  These facts could indicate that these objects are not ``young'' objects experiencing their first episode of star formation. ",
        "conclusions": ""
    },
    "9805/astro-ph9805332_arXiv.txt": {
        "abstract": "The Molonglo Observatory Synthesis Telescope (MOST) has been monitoring the candidate Galactic black hole binary system \\object{{\\rm GX}\\,339$-$4} at 843 MHz since 1994 April.  We present the results of this program up to 1997 February and show a possible correlation between radio and X-ray light curves. ",
        "introduction": "The low mass X-ray binary {\\gx} was discovered by the OSO-7 satellite in 1973 (\\cite{mar73}). It has been classified as a black hole candidate primarily because of the similarity of its X-ray emission to that of the canonical black hole system \\object{{\\rm Cyg}\\,X-1} (bimodal X-ray states: high/soft and low/hard) and because of rapid variability in its X-ray and optical emission (e.g., \\cite{mak86}; \\cite{miy92}; \\cite{now95}).  {\\gx} exhibits four distinct X-ray states, three of which were initially identified by Markert et al.\\ (1973): high, low and off. The high state is characterized by an extremely soft spectrum ($kT$ = 1--2 keV) accompanied by a hard power-law tail, the low state is described by a single power-law hard spectrum, and the off state is in fact a very weak hard state (\\cite{mot85}).  Recently an intermediate state between the low and the high states has been reported (\\cite{men97}).  The optical counterpart was identified as an 18 mag star (\\cite{dox79}; \\cite{cow91}) which was found to be highly variable, with $V$ ranging from $\\sim$15.4 to $>$20. Photometric data revealed a 14.8 hour modulation which has been attributed to the orbital period (\\cite{cal92}). Emission from the accretion disk has dominated the spectra making it difficult to obtain a definitive estimate for the mass of the compact object (\\cite{cow87}).  Distance estimates vary from 1.3 kpc (\\cite{pre91}) to $\\sim$4 kpc (\\cite{mak86}).  Simultaneous optical and X-ray observations have shown quasi-periodic oscillations (QPOs) at mean periods of $\\sim$10\\,s and $\\sim$20\\,s (\\cite{mot83}) in the X-ray low state (see \\cite{t&l95} and references therein).  The relationship between optical and X-ray fluxes is not well understood. For example, in 1981 Motch et al.\\ (1983) found anticorrelation between the 1--13 keV X-ray and optical fluxes, but correlation between the 13--20 keV X-ray and optical.  The discovery of a variable radio counterpart (\\cite{s&cw94}) initiated the monitoring program at 843 MHz undertaken with the Molonglo Observatory Synthesis Telescope (MOST) which is the subject of this paper.  Fender et al.\\ (1997) observed {\\gx} at high resolution in 1996 July with the Australia Telescope Compact Array (ATCA) at a wavelength of 3.5 cm and reported the detection of a jet-like extension to the west of the source. Subsequent observations in 1997 February failed to confirm this extension (\\cite{cor97}). Both Fender et al.\\ (1997) and Corbel et al.\\ (1997) report a flat or inverted radio spectrum.   \\begin{table} \\caption{Journal of MOST observations of GX\\,339$-$4 and final calibrated 843 MHz flux densities. \\label{tbl1}} { \\begin{tabular}{lccclr} \\hline \\\\[-3mm] Sequence &Date & N$_{\\rm samp}$ & TJD & \\multicolumn{2}{c}{S$_{843}$} \\\\ \\multicolumn{1}{l}{number} & \\multicolumn{1}{c}{ }& \\multicolumn{1}{c}{ } & \\multicolumn{1}{c}{(mid-obs.)}  & \\multicolumn{2}{c}{(mJy)} \\\\ \\hline \\\\[-3mm] 1659483\\rlap{$^*$}  & 94/04/25 & 1796 & 49498.123 & 9.81 & \\llap{$\\pm$}0.70 \\\\ 1659484      & 94/06/01 &  898 & 49504.976 & 6.04 & 0.99 \\\\ 1659481      & 94/06/02 &  690 & 49506.003 & 7.87 & 0.97 \\\\ 1659482      & 94/06/03 & 1795 & 49507.097 & 6.93 & 0.75 \\\\ 1659485      & 94/10/04 & 1019 & 49629.867 & 7.08 & 1.36 \\\\ 1659486      & 94/10/17 & 1101 & 49642.821 & 6.24 & 1.19 \\\\ 16594810     & 95/04/30 & 1101 & 49837.281 & \\llap{$<$}2.0$^{\\rm a}$ & ---\\\\ 16594811\\rlap{$^*$} & 95/06/02 & 1151 & 49871.099 & 5.26 & 0.66 \\\\ 16594815     & 95/08/25 & 1796 & 49954.983 & 4.22 & 0.68 \\\\ 16594813     & 95/10/14 &  981 & 50004.734 & 3.11 & 0.65 \\\\ 16594816     & 96/03/15 & 1796 & 50158.313 & 3.21 & 0.72 \\\\ 16594819\\rlap{$^*$} & 96/03/31 & 1796 & 50174.269 & 2.22 & 0.67 \\\\ 16594820\\rlap{$^*$} & 96/04/06 & 1796 & 50180.253 & 2.98 & 0.65 \\\\ 16594821     & 96/04/19 & 1796 & 50193.217 & 2.47 & 0.71 \\\\ 16594822     & 96/04/30 &  898 & 50204.187 & \\llap{$<$}2.0$^{\\rm a}$ & --- \\\\ 16594823     & 96/05/08 & 1796 & 50212.166 & 2.00 & 0.70 \\\\ 16594824     & 96/05/17 & 1796 & 50221.141 & 2.47 & 0.65 \\\\ 16594825\\rlap{$^*$} & 96/05/20 & 1795 & 50224.133 & 2.31 & 0.60 \\\\ 16594826     & 96/05/24 & 1796 & 50228.122 & 2.28 & 0.66 \\\\ 16594827     & 96/06/07 & 1617 & 50242.108 & \\llap{$<$}1.9$^{\\rm a}$ & ---  \\\\ 16594828     & 96/06/13 & 1796 & 50248.067 & \\llap{$<$}1.9$^{\\rm a}$ & --- \\\\ 16594829     & 96/06/23 & 1791 & 50258.040 & 1.76 & 0.59 \\\\ 16594830\\rlap{$^*$} & 96/07/13 & 1791 & 50277.985 & 6.47 & 0.58 \\\\ 16594831     & 96/07/28 & 1796 & 50292.944 & 6.70 & 0.68 \\\\ 16594832\\rlap{$^*$} & 96/08/06 & 1796 & 50301.920 & 6.02 & 0.56 \\\\ 16594833     & 96/08/24 & 1619 & 50319.896 & 3.46 & 0.66 \\\\ 16594834\\rlap{$^*$} & 96/09/15 & 1796 & 50341.810 & 3.47 & 0.65 \\\\ 16594836\\rlap{$^*$} & 97/02/04 & 1795 & 50483.423 & 7.04 & 0.66 \\\\ 16594837\\rlap{$^*$} & 97/02/11 & 1794 & 50490.404 & 6.25 & 0.65 \\\\ 16594838\\rlap{$^*$} & 97/02/18 & 1795 & 50497.385 & 6.14 & 0.71 \\\\ \\hline \\end{tabular} } \\begin{list}{}{} \\item[$^{\\mathrm{a}}$]3$\\sigma$ upper limit. \\item[$^*$]Co-added to form image in Fig.\\ \\ref{fig1}. \\end{list} \\end{table} ",
        "conclusions": ""
    },
    "9805/astro-ph9805104_arXiv.txt": {
        "abstract": "The spin-down of Be stars due to angular momentum transport from star to disc has been considered. This has been prompted by empirical studies of observed optical and IR line profile studies indicating that the disc is rotating in a Keplerian fashion. It is found that substantial spin-down may occur, especially for late B stars throughout their main-sequence lives for the ``strongest'' discs (most dense $\\sim 10^{-11}$g cm$^{-3}$ with high radial velocity $\\sim 1\\kms$ at their inner edge and with large opening angle $\\sim 15^\\circ$). This is in conflict with studies of rotational velocity distributions for different luminosity classes, which show no significant evolution. The implications of this for Be star discs are considered. ",
        "introduction": "Be stars are now widely accepted to have two distinct regions of circumstellar matter : a diffuse polar stellar wind and a dense equatorial ``disc'' (Dachs 1987, Slettebak 1988). One of the major objectives of Be star research is to develop a theory which describes both of these components. The fast diffuse polar wind is well described by standard line-driven wind theory (Castor, Abbott \\& Klein 1975, Friend \\& Abbott 1986, Kudritzki \\etal\\ 1989). However, it has been much more difficult to describe the equatorial disc. Empirical models of the disc structure have been presented by e.g. Marlborough (1969), Waters (1986) and Hanuschik (1996), whilst theoretically driven models have been developed by Poe \\& Friend (1986), Chen \\& Marlborough (1992), Bjorkman \\& Cassinelli (1993),  Willson (1986), Ando (1986) and Lee, Saio \\& Osaki (1991). These involve phenomena such as magnetic winds, latitudinal variation of driving lines, wind compression, stellar pulsation and viscous excretion. The most promising disc theory for several years, Bjorkman \\& Cassinelli's (1993) wind compressed disc model, has been shown to be incapable of reproducing observed discs by Owocki, Cranmer \\& Gayley (1996) and Porter (1997) via different routes.  One feature of empirical studies of line profiles in Be star discs is that they imply a rotationally supported disc, and that the rotation falls off in an Keplerian fashion (e.g. Dachs \\etal\\ 1986, Hanuschik 1989, 1996). The half-line width is typically larger than $v {\\rm sin}i$ (Hanuschik 1996).  Also, the current model for V/R variations (described in e.g. Dachs 1987) assumes that $m = 1$ perturbations arise in a Keplerian disc (Papaloizou \\etal\\ 1992). Excretion disc models proposed by Lee, Saio \\& Osaki (1991) provide naturally this sort of rotationally supported disc -- the disc material is rotating at its Keplerian speed, and drifts outward due to the effect of viscosity. These models require that angular momentum is supplied at the inner boundary of the disc. Given a prescription for the viscosity (e.g. '$\\alpha$' from Shakura \\& Sunyaev 1973), then the surface disc density, and disc scale height may be integrated from the equations conserving angular momentum and mass (e.g. see Pringle 1981). A similar model has been used by Pringle (1991) applied to the cessation of accretion (``decretion'') by a young stellar object.  In this investigation, assuming that the disc is indeed rotationally supported, the spin-down of the central star is calculated. In \\S2 the evolution of a star's rotational velocity is derived given that it is supplying angular momentum to the disc. Estimates of spin-down times are presented in \\S3 across the B star range of stellar parameters. This is discussed in \\S4 and conclusions are presented in \\S5. ",
        "conclusions": "Under the assumption that Be stars maintain discs throughout their main-sequence lives, spin-down is found to occur for certain parameters of the circumstellar disc if the disc rotates in a Keplerian fashion. This is in conflict with observational studies of the distribution of rotational velocity for Be stars of different luminosity classes. The simplest resolution to this paradox is that the majority of Be star discs must occupy parameter space in which $P~=~v_0\\theta\\rho_{-11} \\ltappeq 5\\times 10^{-3}$. For the range of opening angles from observational studies (e.g. Porter 1996), and densities (e.g. Waters \\etal\\ 1987) this implies very low radial velocities of $v_0~\\ltappeq~0.01$\\kms."
    },
    "9805/astro-ph9805110_arXiv.txt": {
        "abstract": "The High Energy Transmission Grating (HETG) Spectrometer on the Advanced X-ray Astrophysics Facility (AXAF) (scheduled for launch in August, 1998) will provide a new tool for the study of absorption in the X-ray spectra of high redshift quasars due to the material along the line of sight. In this paper we try to explore the possibility of using AXAF HETG to detect resonance absorption lines from the Damped Lyman-$\\alpha$ (DLA) clouds. ",
        "introduction": "AXAF HETG \\footnote{See web page http://asc.harvard.edu/} is designed to provide high resolution spectroscopy up to E/$\\Delta$E$\\sim1000$ (for point source) between 0.4 keV and 10 keV. If the DLA systems contain a sufficient column density of highly ionized metals, the high spectral resolution of the HETG will permit detection of resonance lines. For specificity, we consider Q2223-052, a high redshift quasar at z = 1.404 which has a DLA cloud at $z=0.484$. X-ray observations with $Einstein$ \\cite{wilkes} show its flux to be $9.5\\times10^{-12} ergs \\ s^{-1}cm^{-2}$ between 0.16-3.5 keV. This implies a count rate of 0.36 counts/sec with AXAF MEG in the first order. Given the observation time and the instrument resolving power we can calculate the minimum detectable equivalent width of an absorption feature and the required ion column density. In Table 1 $N_{1}$ is the required column density. Here $\\tau$ is the optical depth at the line center, and the velocity dispersion $b=200 km/sec$. All energies are in the observer frame. We assume the spectrum of the high-z quasar has the form of a power law + Galactic absorption + an assumed resonance line from the DLA system due to Si~{\\sc xiii} at $1.26 keV$ in the observer frame, then fit it with a model not containing the absorption line. In the $\\chi^{2}$ plot we can clearly find the line. \\\\  \\begin{figure} \\centerline{\\vbox{ \\psfig{figure=taotaoF1.ps,height=3.cm,angle=270} }} \\caption[]{Q2223-052, the residual $\\chi^{2}$ plot} \\end{figure}  \\begin{center}  \\begin{tabular}{|l|l|l|l|l|l|l|l|} \\multicolumn{6}{c}{{\\bf Table 1.} Two Resonance Lines in AXAF Band} \\\\ \\hline ion&E(keV)&EW(eV)&$\\tau$&$N_1(\\times 10^{16})$&$N_2(\\times 10^{16})$\\\\ \\hline Mg XI&0.91&1.4&3&6&9.8 \\\\ \\hline Si XIII&1.26&1.2&1.2&3.2&4.6 \\\\ \\hline \\end{tabular}  \\end{center} ",
        "conclusions": ""
    },
    "9805/astro-ph9805326_arXiv.txt": {
        "abstract": "I use the most recent cosmic microwave background (CMB) anisotropy measurements to constrain the leading cold dark matter models in the $\\om - \\ol$ plane. A narrow triangular region is preferred. This triangle is elongated in such a way that its intersection with even conservative versions of recent supernovae, cluster mass-to-light ratios and double radio source constraints, is small and provides the current best limits in the $\\om -\\ol$ plane: $\\ol = 0.62 \\pm 0.16$ and $\\om = 0.24 \\pm 0.10$. This complementarity between CMB and other observations rules out $\\ol = 0$ models at more than the $99\\%$ confidence level. ",
        "introduction": "\\label{sec:introduction}  The main goal of CMB measurements and the two new satellite missions MAP and Planck Surveyor is to determine a host of cosmological parameters at the unprecedented accuracy of a few percent (Jungmann \\etal 1996, Zaldarriaga \\etal 1997, Bond \\etal 1997). As part of this goal it is important to keep track of what can already be determined from the CMB without conditioning on certain families of models or on certain values of parameters within these families. In this {\\it Letter}, the analysis of CMB anisotropy measurements is expanded to include the most popular families of cold dark matter (CDM) models such as flat, flat-$\\Lambda$ and open models, as well as the less popular open$-\\Lambda$ models. Figure 1 presents an overview of this parameter space. Open-\\l models are considered here because they subsume all of the above models and thus  provide a parameter space in which the most popular models can be directly compared.  The popularity of non-zero $\\ol$ models has waxed and waned over the years; for excellent reviews see Felten \\& Isaacman (1986) and Carroll \\etal (1992). $\\ol$ was introduced by Einstein (1917) to solve the discrepancy between an apparently static universe and the dynamic cosmology of general relativity. Since this inauspicious beginning $\\ol$ has been invoked many times and seems to be a surprisingly multi-purpose cure-all for theory-observation mismatches. Several recent papers (Turner 1991, Ostriker \\& Steinhardt 1995, Roos \\& Harun-or-Rachid  1998), have pointed out the  effectiveness of $\\ol$ in resolving apparent conflicts between various observational constraints. Recently, $\\ol$ has been invoked to solve the discrepancy between globular cluster ages and the age of the Universe inferred from measurements of Hubble's constant.  Recent supernovae results in $\\ol = 0$ models yield $\\om$ values so low that they are unphysical:  $\\om = -0.4 \\pm 0.5$ (see column 2 of Table 1, error bars and limits in this {\\it Letter} are $68\\%$ confidence levels unless stated otherwise). Not only are they unphysically low but the highest $\\om$ values allowed by the error bars are in  strong disagreement with the high values of $\\om$ preferred by the CMB in these same models (Lineweaver 1998). In Lineweaver \\& Barbosa (1998b) we report a 99.9\\% confidence lower limit of $\\om > 0.31$ (see Figure 1). This supernovae/CMB inconsistency is strong motivation to use CMB data to explore a larger parameter space which includes $\\ol$. If the inconsistency is caused by the incorrect assumption that $\\ol = 0$, then such an analysis will show it. The result of the analyis presented here is that $\\ol > 0$ can resolve this inconsistency.  Testing large parameter spaces is important to minimize the model dependence of the results. For example, in Lineweaver \\& Barbosa (1998a) the CMB data favored $\\h= 0.30^{+0.18}_{-0.07}$ ({\\bf if} $\\om = 1$, {\\bf if} $\\ol = 0$ and {\\bf if} all the other assumptions made are valid). In Lineweaver \\& Barbosa (1998b), hereafter LB98b, we dropped the $\\om = 1$ assumption and still found low $\\h$ values ($\\h = 0.40$  but with large error bars: $0.26 < \\h < 0.97$) and $\\om > 0.57$. Thus the  CMB data prefer high $\\om$ values ({\\bf if} $\\ol=0$). These may be big {\\bf if} 's. The purpose of this paper is to make these {\\bf if} 's smaller by exploring a still larger region of parameter space.  Other workers have used CMB observations to constrain cosmological parameters in CDM models ( e.g. Bond \\& Jaffe 1997, deBernardis \\etal 1997, Ratra \\etal 1997, Hancock \\etal 1998, Lesgourgues \\etal 1998, Bartlett \\etal 1998, Webster \\etal 1998, White 1998). The previous work most similar to this Letter is White (1998). White (1998) combined supernovae results with the Hancock \\etal (1998) estimate of the position in $\\ell-$space of the peak in the CMB power spectrum. In Section 5, I compare my results to White (1998) and other work. ",
        "conclusions": "The results presented here are largely observational but are model dependent. In a series of papers (Lineweaver \\etal 1997, Lineweaver \\& Barbosa 1998a, 1998b), and now in this work, we have looked at increasingly larger regions of parameter space. Each time the $\\chi^{2}$ minimum has been found to lie within the new region. This might be taken as a sign of caution not to take the currently favored region too seriously. On the other hand, our choice of new parameter space to explore has been guided by independent observational results.  I have used the most recent CMB data to constrain the leading CDM models in the $\\om - \\ol$ plane. A narrow triangular region is preferred. This triangle is elongated in such a way that its intersection with even conservative versions of other constraints is small and provides the current best limits in the plane: $\\ol = 0.62 \\pm 0.16$ and $\\om = 0.24 \\pm 0.10$. This complementarity between CMB and other observations rules out $\\ol = 0$ models at more than the $99\\%$ confidence level.  Until recently observations could not discriminate between a zero and a non-zero cosmological constant. However a wide range of observations have indicated that $\\om < 1$ and the most recent observations appear to favour $\\ol > 0$. The addition of the CMB constraints presented here to these other cosmological observations strengthens this conclusion substantially.  I gratefully acknowledge discussions with Saul Perlmutter and Brian Schmidt about the supernovae data sets and I acknowledge Uros Seljak and Matias Zaldarriaga for providing the Boltzmann code. I am supported by a Vice-Chancellor's fellowship at the University of New South Wales, Sydney, Australia.  \\clearpage {\\normalsize    %"
    },
    "9805/astro-ph9805056_arXiv.txt": {
        "abstract": "We show how previous works (Fuhrmann et al., van 't Veer-Menneret \\& M\\'egessier) demonstrate the efficiency of the use of Balmer line profiles for effective temperature determination. In agreement with them, we insist on the physical interest of this method based on the behaviour of these lines with the variations of the parameters involved in the treatment of the convective transport. The comparison between Fuhrmann's results and ours, independently obtained, exhibits a quite good agreement. We show new results of effective temperature, gravity and metallicities for a few of our programme stars, ranging from solar to overabundant metallicities. ",
        "introduction": "\\label{intr} The first step, in a detailed abundance analysis, is to obtain an accurate effective temperature. The quality of the other fundamental parameters, gravity and metallicity, will follow. Reliable fundamental parameters are also required for internal structure models used for computing diffusion processes or for asteroseismology.  The wings of the first four members of the Balmer line series are very sensitive to effective temperature (hereafter $T_{\\rm eff}$), and almost insensitive to gravity changes for $T_{\\rm eff}$ less than 8500~K, with a limit around 5000~K, from where Stark broadening becomes inefficient. Recent works (Fuhrmann et al., 1993, 1994; van~'t Veer-Menneret and M\\'egessier, 1996, hereafter VM) have shown that Balmer Line Profiles (hereafter BLP) are also sensitive to the temperature structure of the models used to interpret them. This means that BLPs depend on the treatment of the convection transport and on the metallicity entering the Opacity Distribution Functions (ODF).  In section 2 we describe and comment on these previous works. In section 3 we present our results, and compare them to Fuhrmann's ones. ",
        "conclusions": "We conclude by stressing the remarkable internal consistency in the results of Fuhrmann and collaborators, likewise in our results, and the significant agreement between these two completely independant works. As a consequence, the value $\\alpha = 0.5$  for the mixing length parameter, can be considered as the recommended value to be introduced in the models used for $T_{\\rm eff}$ determination of all stars with $T_{\\rm eff}$ between 8500~K and 5000~K, when H$_{\\alpha}$ is not available."
    },
    "9805/astro-ph9805307_arXiv.txt": {
        "abstract": "All previous attempts to understand the microlensing results towards the Large Magellanic Cloud (LMC) have assumed homogeneous present day mass functions (PDMFs) for the lensing populations. Here, we present an investigation into the microlensing characteristics of haloes with spatially varying PDMFs and anisotropic velocity dispersion tensors. One attractive possibility -- suggested by baryonic dark cluster formation in pregalactic and protogalactic cooling flows -- is that the inner halo is dominated by stellar mass objects, whereas low mass brown dwarfs become more prevalent on moving outwards. The contribution to the microlensing rate must be dominated by dark remnants ($\\sim 0.5\\,\\msun$) to recover the observed timescales of the microlensing experiments. But, even though stellar remnants control the rate, they do not dominate the mass of the baryonic halo, and so the well-known enrichment and mass budget problems are much less severe. Using a simple ansatz for the spatial variation of the PDMF, models are constructed in which the contribution of brown dwarfs to the mass of the baryonic halo is $\\sim 55 \\%$ and to the total halo is $\\sim 30 \\%$.  An unusual property of the models is that they predict that the average timescale of events towards M31 is shorter than the average timescale towards the LMC. This is because the longer line of sight towards M31 probes more of the far halo where brown dwarfs are the most common constituent. ",
        "introduction": "Recent results from gravitational microlensing experiments indicate that at least part of the dynamically-dominant dark Galactic halo has now been detected~(e.g., Evans 1997). Analyses of the first few years of data from the MACHO and EROS experiments~(\\cite{aubourg93,alcock97}) seemingly show that a substantial fraction of the halo comprises compact objects which induce microlensing variations on timescales~\\footnote{In this Letter, timescale refers to the Einstein diameter crossing time} of between 30 and 130 days.  However, the nature of the lenses remains mysterious. Statistical analyses, which assume an isothermal halo distribution function populated by objects with a universal mass function, typically yield lens masses between $0.1~\\msun$ and $1~\\msun$ and halo fractions between $20\\%$ and $100\\%$, implying a large population of low-mass stars or stellar remnants~(\\cite{alcock97}).  Such inferences contrast sharply with other observational and theoretical evidence. In particular, star-count studies undertaken with the Hubble Space Telescope (HST) place stringent limits on the numbers of low-mass stars and on the age and spatial density of stellar remnants~(\\cite{bahcall,gf,sge}). Again, the abundance of metals in the interstellar medium limits the contribution of dark remnants to below that suggested by the microlensing analyses~(\\cite{gm,fms}).  However, this apparent inconsistency between microlensing on the one hand and deep star-count and metallicity data on the other may be an artifact of the underlying assumptions in the adopted models.  In this Letter, we introduce halo models in which key microlensing parameters -- namely the lens mass, Galactocentric distance and transverse velocity -- are correlated. How might such correlations arise? Some of the more promising ``cooling-flow'' theories for the formation of baryonic dark clusters~(e.g., \\cite{keith,bernard,depaolis}) predict a spatial gradient in the present day mass function (PDMF). Here, the inner halo comprises partly visible stars, which are associated with the spheroid globular cluster population, whilst the outer halo comprises mostly low-mass stars and brown dwarfs.  Let us remark that there is evidence for similar such spatial gradients in the Galactic disc~(\\cite{taylor}). Correlations between the lens velocity and Galactocentric distance are also possible if the velocity distribution is anisotropic (c.f., Markovic \\& Sommer-Larsen 1997).  There is strong theoretical motivation for halo brown dwarfs, both from formation arguments (e.g., Ashman 1990, Tegmark et al. 1997) and from constraints on other candidates (e.g., Carr 1994). But, powerful arguments using the virial theorem (Gyuk, Evans \\& Gates 1998) have shown that the timescales of the microlensing events preclude the lenses being brown dwarfs. If the PDMF is the same everywhere, then this also prevents them making a substantial contribution to the halo.  The aim of this Letter is to show that the microlensing data-set is consistent with a baryonic component dominated by brown dwarfs. The crucial point is that, if the PDMF varies with position, the microlensing rate may be dominated by one mass scale, but the mass density may be dominated by another one entirely. ",
        "conclusions": "The assumption of a uniform present day mass function (PDMF) throughout the Galactic halo has been made in all previous studies of the microlensing dataset. Of course, it is natural enough to make the simplest assumption, but there is no evidence to suggest that it is correct. The main thrust of this Letter is to point out that the impasse of the microlensing results can be overcome by discarding this unwarranted assumption.  Although our own PDMF is not likely to be correct in detail, its general form does receive physical support from baryonic dark cluster formation theories~(e.g., \\cite{keith,bernard,depaolis}). These suggest that there may be a gradation of masses of baryonic objects in the halo, with larger mass objects in the inner regions and abundant brown dwarfs in the far halo.  Such models can reproduce the microlensing observables within the uncertainties. Such models can be made consistent with Hubble Space Telescope limits on faints stars. The important point is that in models with spatially varying mass functions, {\\it brown dwarfs need not dominate the microlensing rate but they can still dominate the mass of the baryonic halo}. In such models, the lenses are predominantly dark stellar remnants, but these now comprise a much smaller fraction of the halo. So, this evades the embarrassing mass budget and the chemical enrichment problems that occur when the halo has a uniform PDMF~(\\cite{gm,fms})."
    },
    "9805/astro-ph9805288_arXiv.txt": {
        "abstract": "We present 1.3 and 3.3 mm polarization maps of Orion-KL obtained with the BIMA array at approximately $4\\arcsec$ resolution.  Thermal emission from magnetically aligned dust grains produces the polarization. Along the Orion ``ridge'' the polarization position angle varies smoothly from about $10\\arcdeg$ to $40\\arcdeg$, in agreement with previous lower resolution maps.  In a small region south of the Orion ``hot core,'' however, the position angle changes by $90\\arcdeg$.  This abrupt change in polarization direction is not necessarily the signpost of a twisted magnetic field.  Rather, in this localized region processes other than the usual Davis-Greenstein mechanism might align the dust grains with their long axes parallel with the field, orthogonal to their normal orientation. ",
        "introduction": "Magnetic fields play many roles in the star formation activity in molecular clouds (see review by \\cite{mck93}). Polarization of radiation is one of the important signatures of interstellar magnetic fields (\\cite{hil88}). Spinning dust grains in the interstellar medium become partially aligned with the magnetic field, generally with their long axes perpendicular to the field (\\cite{dav51}; hereafter DG). Thus, the thermal emission from grains at far infrared and millimeter wavelengths is partially linearly polarized, with polarization direction perpendicular to the magnetic field.  As the nearest region of OB star formation, the Orion molecular cloud has been studied intensively (see review by \\cite{gen89}).  In the vicinity of the Kleinmann-Low Nebula (KL) the cloud contains at least two massive stars, IRc2 and the Becklin-Neugebauer object (BN), embedded within a flattened ``ridge'' of molecular gas which extends along a position angle of 30$\\arcdeg$.  Polarization from aligned dust grains in Orion has been mapped at 100 $\\micron$ with 35--40$\\arcsec$ resolution (\\cite{nov89}; \\cite{gon90}; \\cite{sch98}), at 350 $\\micron$ and 450 $\\micron$ with 18$\\arcsec$ resolution (\\cite{sch97}; \\cite{sch98}), at 800 $\\micron$ with 14$\\arcsec$ resolution (\\cite{ait97}), and at 1.3 mm with 30$\\arcsec$ resolution (\\cite{lea91}).  In all these maps the polarization vectors are roughly parallel with the molecular ridge, indicating that the large scale magnetic field is perpendicular to the ridge.  The uniformity of the polarization direction across the region suggests that the field is quite strong, of order 1 mG (\\cite{gon90}; \\cite{lea91}).  The fractional polarization ranges from 4\\% to 8\\% except along the line of sight directly through KL, where it is significantly lower.  Higher angular resolution is needed to probe the polarization pattern near KL.  Heretofore, this has been possible only at near- and mid-IR wavelengths, where the competing effects of absorption, emission, and scattering all influence the polarization direction.  Some of the most secure results are obtained by mapping the polarization of the 2~$\\micron$ S(1) line of H$_2$ (\\cite{hou86}; \\cite{bur91}; \\cite{chr94}).  The line emission (assumed unpolarized)  originates from shock-excited H$_2$ in the bipolar outflow from IRc2.  Absorption by aligned grains in front of the outflow produces the polarization. Aitken et al.\\ (1997) also mapped the continuum polarization at 12.5 and 17 $\\micron$; spectropolarimetry suggests that it, too, is mostly attributable to absorption.  These studies all find that the polarization vectors are twisted near KL.  The authors argue that this is evidence for a toroidal magnetic field in a disk-like structure centered on IRc2.  In this paper we present the first interferometric polarization maps of Orion at millimeter wavelengths, where the polarization arises unambiguously from emitting grains.  These high resolution maps confirm the abrupt change in polarization direction near IRc2 first detected in the 2 $\\micron$ data.  We argue that in this small region the bipolar outflow from IRc2 might align the grains with their long axes parallel with the magnetic field, orthogonal to their usual orientation.  Thus, it is not certain that the change in the polarization direction arises from a twisted magnetic field. ",
        "conclusions": "We have made the first interferometric observations of the polarization from dust emission in the Orion BN/KL region, at wavelengths of both 1.3 mm and 3.3 mm.  We find that the decrease in fractional polarization toward KL previously seen with single dish observations is due to polarization structure which is averaged out by larger beams.  In particular, our maps show that the polarization direction changes abruptly by 90$\\arcdeg$ in a small region south of the Orion hot core. If the grains are aligned everywhere by the Davis-Greenstein mechanism, then the magnetic field in this anomalous region is almost orthogonal to the large scale field in the Orion ridge.  However, it is plausible that the grains in this region are aligned by a wind from IRc2, and that the magnetic field is relatively straight.  These results suggest that one should be cautious in using polarization data to infer the magnetic field structure around young stars, because of the possibility that the grains are mechanically aligned by outflows."
    },
    "9805/astro-ph9805241_arXiv.txt": {
        "abstract": "{ We compare the spectral properties of the millisecond and slow pulsars detected in the Parkes 70 cm survey. The mean spectral index for the  millisecond pulsars (MSPs) is --1.9$\\pm$0.1 whereas the mean spectral  index for the slow pulsars is a surprisingly steep --1.72$\\pm$0.04.  A Kolmogorov-Smirnov test indicates that there is only a 72\\% probability  that the two distributions differ. As a class, MSPs are therefore only  fractionally steeper-spectrum objects than slow pulsars, as recent literature would suggest. We then model the expected distribution of millisecond pulsars in the Galaxy and find that high-frequency surveys, with  sensitivities similar to the current Parkes multibeam survey, are likely to detect MSPs in large numbers. The observed distribution of MSPs will be  much less isotropic than that resulting from low-frequency surveys,  with 50\\% of detectable MSPs residing within 11\\arcdeg of the Galactic  plane in an all-sky survey.} ",
        "introduction": "In the early 1980s, attention was focussed on the enigmatic radio complex at the position of 4C21.53. In particular the object to the west of this position, named 4C21.53W, was resolved into two components,  an extended flat-spectrum source $\\sim$1 arcsec north of a compact steep-spectrum object. The steep spectrum and interplanetary scintillation of the compact source suggested that it was a radio pulsar. Several searches, sensitive only to  periods greater than a few milliseconds, failed to detect any pulsations,  but Backer \\etal (1982) \\nocite{bkh+82} announced that the source was the first millisecond pulsar (MSP), PSR B1937+21, with a rotation period of just 1.5 ms. Spurred on by this exciting discovery, Hamilton, Helfand \\& Becker (1985) \\nocite{hhb85} made a spectral survey in 12 nearby globular clusters for unresolved objects which might be MSPs. Their best candidate, in the core of M28, was shown to be a highly linearly polarized, steep-spectrum object (\\cite{embh87}). This steep-spectrum source was later found to be the fourth MSP (PSR B1821--24) and the first globular cluster pulsar discovered (\\cite{lbm+87}). A search aimed at a further 24 nearby globular clusters resulted in the discovery  of yet another steep-spectrum pulsar, PSR B1620--26, in M24 (\\cite{lbb+88}).  The first detailed spectral study of MSPs was made by Foster, Fairhead \\& Backer (1991; FFB)\\nocite{ffb91}.  Their study included the three pulsars mentioned above, as well as PSR B1855+09. The latter was discovered in a 430 MHz pulsar survey conducted at Arecibo (\\cite{srs+86}; \\cite{fru89}). Except for PSR B1855+09, all the MSPs had spectral indices steeper than --2.3. Although the sample of four pulsars was not very statistically significant, this paper helped reinforce the growing belief that MSPs had spectra that were significantly steeper than their slower counterparts.  Since then there have been a number of large-scale surveys for pulsars (e.g., \\cite{bl92}; \\cite{clj+92}; \\cite{fcwa95}; \\cite{cnt93}). As a result there are now some 60 known pulsars with periods less than 20 ms. The high-frequency surveys of Clifton \\& Lyne (1986) and Johnston \\etal (1992) \\nocite{cl86,jlm+92} were mildly sensitive to  MSPs but found none in their surveys of the Galactic plane.  This restricted the population of very luminous high-frequency MSPs,  but offered little insight into the number of low-luminosity MSPs  detectable at high frequencies. On the other hand, the Parkes 70 cm survey (\\cite{mld+96}; \\cite{lml+98}) detected 19 MSPs, 17 of which were new discoveries. The history of MSP searching with its focus on steep-spectrum objects, and the comparative success of low-frequency compared with high-frequency surveys  might lead one to conclude that there are good reasons to  conduct MSP searches exclusively at low frequencies.  The major works on pulsar spectra are those of Lorimer \\etal (1995) \\nocite{lylg95} and Kramer \\etal (1998) \\nocite{kxl+98}. Lorimer \\etal derived a mean spectral index for millisecond and slow pulsars of --2 and  --1.6 respectively. Their study included all of the pulsars regularly  observed from Jodrell Bank, and did not differentiate between  those  found in high- or low-frequency surveys. Kramer \\etal attempted to reduce observational biases by selected spectra only from pulsars out to a distance of 1.5 kpc. They found the average spectra of MSPs and slow pulsars to be essentially the same, with mean indices of --1.6 and --1.7 respectively. As noted by Kramer \\etal the  Parkes survey, with its large number of detections of both millisecond and slow pulsars,  provides an excellent sample from which to discuss the spectral properties of pulsars in a more unbiased way.  All of the 19 MSPs detected in the Parkes survey have been either timed regularly at Parkes or observed for polarization studies with the Caltech pulsar correlator (\\cite{nav94}). The correlator provides accurate fluxes suitable for a spectral study. The aim of this paper is to compare the spectra of the millisecond and slow pulsars detected in the Parkes survey. In \\S 2 of this paper we describe the observations and data-reduction techniques used in obtaining the spectra. \\S 3 presents the average flux density measurements and spectral indices for the MSPs and in \\S 4 we compare  the spectral index distribution of these MSPs and the slow pulsars  detected by the survey. This demonstrates that MSPs found in a large-scale, low-frequency survey have spectra which are only slightly steeper than their slower counterparts, and supports the case for high-frequency surveys for MSPs. In \\S 5 we present the results of a simulated high-frequency search for MSPs near the Galactic plane which suggests that MSPs will be found in significant numbers by the current Parkes multibeam survey. ",
        "conclusions": "We have obtained reliable multi-frequency flux density measurements for the 19 southern MSPs discovered in the Parkes 70 cm survey, enabling us to determine the spectral index distribution for this sample. We have compared this spectral-index distribution with that of slow pulsars detected in the same survey. The spectral-index of the millisecond pulsars was found to have a mean of $\\alpha \\sim -1.9$, only slightly steeper than the mean $\\alpha \\sim -1.7$ for the slow pulsars. A Kolmogorov-Smirnov test suggests only a 72\\% probability of the distributions differing.  This result adds to the growing evidence that MSPs have spectral properties similar to slow pulsars.  Our simulation of high-frequency surveys similar to the Parkes multibeam survey, using the data presented in this paper, suggests that they will discover a large number of MSPs. High-frequency surveys are therefore likely to have a significant effect on the known population of millisecond pulsars in the near future.  \\newpage \\begin{deluxetable}{cccccc} \\tablecolumns{6} \\tablewidth{0pc} \\tablecaption{Flux Densities and Spectral Indices for Southern MSPs} \\tablehead{ \\colhead{PSR J}& \\colhead{$S_{436}$}& \\colhead{$S_{660}$}& \\colhead{$S_{1400}$}& \\colhead{$S_{1660}$}& \\colhead{$\\alpha$}\\\\ \\colhead{}& \\colhead{(mJy)}& \\colhead{(mJy)}& \\colhead{(mJy)}& \\colhead{(mJy)}& \\colhead{}   \\\\ } \\startdata 0034--0534 & 17(5) & 5.5(5) & 0.61(9) & 0.56(9) & --2.6(10) \\nl 0437--4715 & 550(100) & 300(20) & 137(3) & 115(3) & --1.1(5) \\nl 0613--0200 & \\nodata & 7.3(4) & 2.2(1) & 2.0(1) & --1.5(5) \\nl 0711--6830 &\\nodata & 11(2) & 3.4(5) & 2.1(3) & --1.7(14) \\nl 1024--0719 &\\nodata & 4.2(6) & 0.9(1) & 0.88(7) & --1.7(12) \\nl 1045--4509 & 15(3) & 8.5(3) & 1.9(1) &\\nodata & --2.0(5) \\nl 1455--3330 & 9(1) & 8(1) & 1.2(1) &\\nodata & --1.8(11) \\nl 1603--7202 & 21(2) & 17(1) & 2.9(2) & 2.3(2) & --1.8(5) \\nl 1623--2631 &\\nodata & 7.2(2) & 2.0(3) &\\nodata & --1.5(2) \\nl 1643--1224 &\\nodata & 16.0(5) & 3.3(1) & 3.0(1) & --1.9(3) \\nl 1730--2304 &\\nodata & 14(1) & 3.0(4) & 2.5(3) & --1.9(10) \\nl 1744--1134 & 18(2) & 16(2) & 2.0(2) & 1.7(3) & --1.8(7) \\nl 1804--2718 & 15(4) &  8(1) &  0.7(1) &\\nodata &  --2.9(15) \\nl 1823--3021A & 16(5) &  6.8(7) &  0.72(2) &\\nodata & --2.7(9) \\nl 1911--1114 & \\nodata & 7.1(5) & 0.7(1) & 0.63(4) & --2.6(7) \\nl 2051--0827 &\\nodata & 5.5(5) & 1.4(1) & 1.5(1) & --1.6(9) \\nl 2124--3358 & 17(4) & 7.8(7) & 2.6(2) & 1.9(2) & --1.5(8) \\nl 2129--5718 & 14(2) & 8.4(7) & 1.2(1) & \\nodata & --2.2(8) \\nl 2145--0750 & 100(30) & 36(4) & 7.0(9) & 5.4(8) & --2.1(11) \\nl \\enddata \\end{deluxetable}   \\newpage"
    },
    "9805/astro-ph9805077_arXiv.txt": {
        "abstract": "We report for the first time the detection of long-term X-ray variability in the bright bulge source GX\\,354--0 (=4U\\,1728--34) observed with the  All Sky Monitor (ASM) on board the {\\it Rossi X-Ray Timing Explorer} ({\\it RXTE}).  The 2-year {\\it RXTE} ASM database reveals significant power at $\\sim$\\,72 days.  Similar behaviour was seen in the 6-year {\\it Ariel 5} ASM database, but at a period of $\\sim$\\,63 days.  The timescales and light curves resemble the $\\sim$\\,78 days modulation seen in Cyg X--2 and we therefore interpret this modulation in GX\\,354--0 as a super-orbital effect. ",
        "introduction": "The low-mass X-ray binary (LMXB) GX\\,354--0 (=4U\\,1728--34) is a well-known X-ray burster (e.g. Basinska et al. 1984) which has been classified as an atoll source based on {\\it EXOSAT} data (Hasinger \\& van der Klis 1989).  Despite its detection in both soft and hard X-ray bands, its orbital period is still unknown.  Based on {\\it Einstein} HRI and infrared observations, Grindlay \\& Hertz (1981) claimed the association of GX\\,354--0 with a heavily reddened globular cluster. However, the existence of this cluster was not confirmed by later infrared observations (van Paradijs \\& Isaacman 1989).  Apart from the bursting activities, GX\\,354--0 also exhibits complex behaviour on short time scales. Quasi-periodic oscillations (QPOs) at 363 Hz (2.75 ms) during the burst itself were recently discovered by the {\\it Rossi X-ray Timing Explorer} ({\\it RXTE}) (Strohmayer, Zhang \\& Swank 1996; Strohmayer et al. 1996a; 1996b) providing the first evidence for a millisecond spin period in LMXB. A comprehensive discussion of these temporal and spectral characteristics can be found in Strohmayer et al. (1997).  GX\\,354--0 is likely to be a typical LMXB and can be considered as a burster involving a rapidly rotating neutron star.  However, with no optical or infrared  counterpart having yet been determined, because of the high extinction in this direction as well as the presumably large distance, there are few constraints on the nature or evolutionary state of GX\\,354--0. We therefore decided to exploit the long-term monitoring capabilities of the ASM on both the {\\it Ariel 5} and {\\it RXTE} satellites in order to study the X-ray behaviour of this source on timescales of weeks to months.  In this way we could search for long-term (super-orbital) periods similar to those seen in other bright LMXBs, such as Cyg X--2 and X1820--30 (e.g. Smale \\& Lochner 1992), and which might be related to either the accretion geometry, disc or neutron star properties. ",
        "conclusions": "We have detected a 71.7 $\\pm$ 0.5\\,d periodicity in the {\\it RXTE} ASM light curve and a 63 $\\pm$ 0.1\\,d periodicity in the {\\it Ariel 5} ASM light curve of the LMXB GX\\,354--0.  Such a long period modulation has hitherto been rather uncommon in LMXB. From the long-term variability survey carried out by Smale \\& Lochner (1992) using {\\it Vela 5B}, only 3 out of 16 LMXBs (X1820--30, X1916--05, Cyg X--2) were found to have long-term periods.  In particular, X1820--30 and X1916--05 have periods of 176 days and 199 days, respectively, while recent observations by {\\it RXTE} ASM confirm the long-term period of $\\sim$\\,78 days in Cyg X--2 (Wijnands et al. 1996).  These long-term periods are designated {\\it superorbital}, as the orbital periods are known for all three objects (11.3\\,mins, 50\\,mins and 9.8\\,d, respectively).  In massive X-ray binaries, super-orbital periods are more common.  Cyg X--1, SS433, LMC\\,X--4, Her X--1 and several other sources all have super-orbital periods in the range of 30--300 days (e.g. Priedhorsky \\& Holt 1987).  The cause of these super-orbital periods is still a subject of debate, and possible explanations include the precession of a tilted accretion disk, neutron star precession, mass transfer feedback and triple systems (see Priedhorsky \\& Holt 1987 and Schwarzenberg-Czerny 1992 for more details).  In the case of LMXBs the {\\it super-orbital} period is much more rare and there is a large spread in values of the ratio of super-orbital to orbital periods.  For X1820--30, X1916--05 and Cyg X--2, the ratio is 22100, 5750 and 8 respectively, whereas in massive systems it is in the range of 10 to 100 (see e.g. Wijers \\& Pringle 1998).  The long-term variation in LMXBs may instead be due to radiation driven warped accretion discs (e.g. Wijers \\& Pringle 1998) or a disc instability in the system (Priedhorsky \\& Holt 1987; Dubus et al. 1998).  Many orbital periods of bright galactic sources (including GX\\,354--0) remain unknown mainly due to the heavy optical extinction and/or crowded regions.  While the known orbital periods of LMXBs range from 11\\,mins to $\\sim$\\,10\\,d, Cyg\\,X--2 and X0921--63 actually have long orbital periods ($\\sim$\\,10 days). We note that the {\\it RXTE} ASM light curves of Cyg X--2 (Fig. 3a) and GX\\,354--0 are rather similar.  This similarity is enhanced by our analysis of the much more extensive {\\it RXTE} ASM database of Cyg X--2 that is available now. The Lomb-Scargle periodogram shows a 69 $\\pm$ 0.4\\,d period (Fig. 3b) in addition to the already noted $\\sim$\\,78\\,d period (Wijnands et al. 1996). Fig. 3c shows the folded light curve of Cyg X--2 with a period of 69 days, where phase zero is defined by the first data point. A recent study of GX\\,1+4 by Chakrabarty \\& Roche (1997) has suggested that the orbital period of this source may exceed 100 days or even 260 days.  However, since any orbital modulation should be uniquely stable, we conclude that our periods of 63\\,d and 72\\,d found for GX\\,354--0  are ``super-orbital'' and {\\it not} orbital.  Further optical/infrared campaigns are needed to reveal its orbital period and to make further progress in this area."
    },
    "9805/astro-ph9805182_arXiv.txt": {
        "abstract": "We have developed a model of the high-energy accretion region for magnetic cataclysmic variables and applied it to {\\it Extreme Ultraviolet Explorer} observations of 10 AM Herculis type systems. The major features of the EUV light curves are well described by the model. The light curves exhibit a large variety of features such as eclipses of the accretion region by the secondary star and the accretion stream, and dips caused by material very close to the accretion region. While all the observed features of the light curves are highly dependent on viewing geometry, none of the light curves are consistent with a flat, circular accretion spot whose lightcurve would vary solely from projection effects. The accretion region immediately above the WD surface is a source of EUV radiation caused by either a vertical extent to the accretion spot, or Compton scattering off electrons in the accretion column, or, very likely, both. Our model yields spot sizes averaging 0.06 R$_{WD}$, or $f \\sim 1 \\times 10^{-3}$ the WD surface area, and average spot heights of 0.023 R$_{WD}$. Spectra extracted during broad dip phases are softer than spectra during the out--of--dip phases. This spectral ratio measurement leads to the conclusion that Compton scattering, some absorption by a warm absorber, geometric effects, an asymmetric temperature structure in the accretion region and an asymmetric density structure of the accretion column are all important components needed to fully explain the data. Spectra extracted at phases where the accretion spot is hidden behind the limb of the WD, but with the accretion column immediately above the spot still visible, show no evidence of emission features characteristic of a hot plasma. ",
        "introduction": "AM Her stars are a class of interacting binaries consisting of a highly magnetic (typically $10-60$ MG) white dwarf (WD) primary and a red main sequence secondary that fills its Roche Lobe. Stellar material flows through the inner Lagrangian point (L1) and falls toward the primary forming an accretion stream. In the absence of a magnetic field, an accretion disk normally forms. However, in AM Her systems, the strong magnetic field captures the ionized material and channels it directly toward one or both of the magnetic poles of the WD primary, forming a hot accretion region. The interaction of the accretion stream with the magnetic field circularizes the orbits of both stars and synchronizes the WD rotation period with the binary orbital period. When viewed from the rotating frame, the primary, secondary, and magnetic field all appear static. The only motion is that of the in-falling material, its free-fall time being about one fourth the orbital period.  Conversion of the kinetic energy of the stream manifests itself in many forms. Electrons in the stream near the accretion region spiral around the field lines and emit highly polarized cyclotron radiation at optical and near infrared wavelengths. X-rays arise in a shock region where the supersonic stream gives up energy, becomes heated and flows subsonically onto the WD surface. Extreme ultraviolet (EUV) radiation is emitted from, or very close to, the heated WD surface; the consequence of reprocessed X-rays and the direct mechanical heating of the WD photosphere by the impacting material. See Liebert \\& Stockman (1985), Cropper (1990), and Warner (1995) for reviews on AM Her systems.   The field strengths of the white dwarfs in magnetic CVs are strong enough to focus accreted material onto their surface in a relatively small area ($f \\sim 1 \\times 10^{-3}$ of the WD surface) called the accretion region. Observational evidence for this situation comes from many sources, all of which provide convincing proof for the white dwarf spin to be locked with the binary orbit and for the compactness of the accretion region. The accretion onto a magnetized white dwarf is nearly a radial in-fall situation, making the problem almost one-dimensional in nature. Radial accretion onto a WD surface was first considered to explain X-ray emission in Sco X-1 (Cameron \\& Mock, 1967).  The general idea of magnetically focused accretion is as follows. Ionized material leaving the L1 point of the secondary is in free fall until at some point, called the coupling region, the kinetic energy associated with the angular momentum is overcome by the magnetic energy of the primary field, and the material gets funneled along the field lines to the WD surface. The flow can initially be assumed to be uniform from the secondary all the way to the primary star. We will see that this is likely {\\it not} to be the case but will serve as a working model for now. A strong shock is encountered by the accreting stream just above the surface where it decelerates by a factor of 3-4, converting much of its infall energy to short wavelength radiation. Post--shock material moves at subsonic velocities and settles onto the WD surface. Temperatures in the post--shock region are near kT$_{TB}~\\sim$ 10-25 keV and kT$_{BB}~\\sim$20-40 eV, with emission peaking in the X-ray and EUV regions. This short wavelength radiation leaves the accretion column easily, with about half escaping into space and half going towards the WD photosphere around the accretion region. The accretion area near the WD surface becomes heated to a few times $10^5$~K, giving rise to EUV emission.  The accretion region was originally modeled as a circular region (often referred to as an accretion spot) lying flat on the WD surface (Lamb and Masters, 1979). These authors also calculated the overall emitted spectrum as a sum of cyclotron emission, bremsstrahlung, and reprocessed black-body emission from the heated surface. Observations of AM Hers at high energies (X-rays), revealed a fascinating array of complexities in the accretion regions. By the late 1980's, new ideas for the size and shape of the accretion regions were emerging. Foremost among these was the work of Wickramasinghe and Meggitt (1985), Wickramasinghe, Ferrario \\& Bailey (1989), and Ferrario, Wickramasinghe \\& Tuohy, (1989). Their view gave the accretion spot an arc-like shape of length $\\sim \\frac{1}{4}$ R$_{WD}$, instead of a circular profile, regions of high and low density mass flow, and the spot had a small height ($\\sim \\frac{1}{10}$ R$_{WD}$) above the surface. Surrounding the entire spot was an extended corona-like halo of hot plasma .  The early models of a homogeneous flow with cylindrical symmetry in the accretion stream made the prediction that the emitted luminosities at high energy will be such that the hard X-ray emission will be greater than about 2 times the soft X-ray (EUV) emission. Observations in the EUV spectral region, particularly in recent years by the R\\\"ontgen Satellite ({\\it ROSAT}) and the Extreme Ultraviolet Explorer ({\\it EUVE}), have again caused accretion region models to be re-evaluated (eg. Ramsay et. al. 1994, Beuermann \\& Schwope 1994). These authors and others, have shown that soft X-ray excesses exist in most of the AM Herculis systems, unexplainable in the older models.  Discussions of the ``soft X-ray problem'' are given in Cropper (1990), King (1995), and Beuermann \\& Burwitz (1995) and references cited therein. New ideas including inhomogeneous flows, blobby accretion, buried accretion regions, and cyclotron cooling of the post--shock region have been pursued in an attempt to understand the observations. Blobby accretion was first discussed in Kuijpers \\& Pringle (1982) and followed up on by others. Litchfield (1990), allowed the blobs to bury themselves deep into the WD atmosphere before giving up their store of energy, causing increased local heating and giving rise to an ``excess'' of EUV photons. Litchfield also calculated a shape and size for the accretion region, and the lightcurve expected for 100~\\AA\\ observations (see \\S 7.2).  King (1995) has suggested that one can match the observations via an accretion region consisting of various sites of bombardment within the total effective radiating area, the latter being the area in which the blobs usually land. Each of the sites may become a depressed accretion region, buried beneath the WD surface by several atmospheric scale heights.  All of the models discussed above generally agree that the accretion region is relatively small and its ``size'' is wavelength dependent. The hard X-rays come from the central concentration, presumably located near the stream shock. Surrounding this central concentration, there is an extended region of EUV emission ($\\sim 1000$ km in extent), heated from above by X-rays from the shock and from below by thermal energy released by the impacting blobs. (see Cropper;1990, Stockman et. al.; 1994, \\& Schwope; 1995).   Disentangling the various radiations from their points of origin is difficult, especially in the optical where the photospheres of both stars as well as the accretion stream all contribute flux. Observations in the EUV (65\\AA\\  to 180\\AA) are particularly well suited to probing the accretion region since the source of radiation is often confined to a very small region on, or near, the WD surface. Furthermore, many AM Her systems are intrinsically bright around 100~\\AA . The peak energy release of the extended spot, and the pre- and post-shock regions of the stream are all sources of EUV emission.  The {\\it EUVE} satellite has performed a number of pointed photometric and spectroscopic observations of AM Herculis stars (cf., Craig et. al., 1997), and in this paper we present results for many of these stars. Using the high quality, high time-resolution EUV data, we have developed a new model for the accretion geometries of AM Her systems. This model, discussed in \\S 4, provides quantitative information on the size, shape, and variability of the accretion region. In addition, our model provides the first direct evidence of extended EUV emission above the WD surface, originating in either an extended halo, or the lower (pre--shock) portion of the stream itself. ",
        "conclusions": "We have analyzed EUV light curves for ten AM Her systems.  All show strong modulations as a function of orbital phase. Projection effects account for the overall bright and faint phases, but cannot explain the sharp transitions and dips seen in the light curves. In the systems where the accretion spot rotates behind the limb of the WD for some portion of the phase (category 1), the rise and fall phases of the light curve are very steep, and are symmetric about the EUV mid-phase. The observations imply two things: first, the shape of the steep rise and fall phases is dominated by vertical extent of the accretion spot, and second, any longitudinal structure of the accretion region must be less than 0.2 of the R$_{WD}$, since a larger accretion region would produce an asymmetric lightcurve during the rise and fall phases. For the eclipsing systems UZ For and HU Aqr, the short ingress and egress times directly set upper limits to the spot size of $\\le 0.23$ R$_{WD}$.  All the light curves show one or more dips. The eclipsing systems constrain the model well and yield the best results for orbital inclination, spot latitude and longitude, spot height, and spot size. Once the viewing geometry is established, the phasing of the dips restricts the spatial location of where the accretion stream material must reside in order to self-eclipse the accretion region. The accretion stream, when it is far from the WD surface, quickly crosses the line of sight to the accretion spot and can completely occlude the spot. These narrow dips last typically $< 0.1$ of the orbital phase, often saturate to zero flux, and show no wavelength dependence for the absorption. The broad dips typically last $\\sim 0.25$ of the orbit, are asymmetric in profile, occur well before EUV mid phase, never saturate to zero, and show a {\\it softer} spectrum than the non--dip phases. If the accretion column strikes the accretion region normally, and is concentric with the spot, the expected light curve would be symmetric and show a broad dip at phase 0 as predicted by ID83. The large asymmetries present in all the observed AM Her light curves are {\\it direct} evidence for asymmetries within the accretion regions. An elongated spot, with the dense portion of the accretion column eccentrically contacting the spot in the direction of binary rotation can explain the phasing, duration in phase, and depth of the broad dips. Furthermore, if the accretion spot shows a temperature gradient, (hottest towards the center, coolest at the edges), and the column width is less than the EUV emitting area (or has a non-uniform density, or both), the broad dips will never saturate to zero and their spectra will be softer when the dense portion of the column hides the hottest part of the spot.  Using detailed EUV photometric and spectroscopic observations of AM Herculis stars, we have constructed an appropriate model of the accretion regions which include effects of both the far-- and near--field accretion stream. Many of our results below apply equally well to observations of AM Herculis stars in other high energy bandpasses. Summarizing our relevant findings we have:  \\begin{itemize} \\item [1.]  A flat, circular geometric accretion spot model only accounts for the gross features of the EUV light curves, namely the faint phase, the rise to a maximum, and then the return to the faint phase. This model does not match the observations during the steep rise and fall phases (which depart from cosine behavior), for which a vertically extended structure is needed. We have chosen to model this simply as a hemispherical structure, but more complex geometries are possible.  \\item [2.]  All ten stars show large modulations in the form of dips in their EUV light curves that are inconsistent with cosine projection effects. All stars where the accretion region is known to lie in the northern hemisphere (thus guaranteeing that the accretion stream crosses the line of sight to the accretion region) show a far--field stream eclipse of the accretion region. In addition, the two highly inclined systems VV Pup and UZ For, whose spots are in the southern hemisphere, also show evidence for far--field stream eclipses indicating a non--zero width of the coupling region. Eight stars show additional dip features other than what can be accounted for by the far--field stream. The fraction of flux ``missing'' amounts to as much as 50\\% of the flux predicted by the non-absorbed geometric model (see Table 2).  \\item [3.] All AM Herculis stars with multiple EUV observations (UZ For, RE1149, QS Tel, HU Aqr), show significant long term variations in the features of their lightcurves, (eg. the duration of the bright phase, and the depth and position in phase of both the far--field stream dips, and the near--field broad dips). All stars observed also show short term orbit to orbit variations.  \\item [4.]  The cause of a majority of the features of the EUV light curves is highly dependent on viewing geometry (i.e., viewing angle to the spot and through the accretion stream). The remaining variations are due to the physical structure and properties of the accretion stream and near--field column themselves. Attributing the causes of the features in the light curves simply to various astrophysical effects is incorrect until all the geometrical effects have been taken into account and seperated from the physical effects.  \\item [5.] The low inclination systems (category 2) show more complex EUV light curves and dip features, attributable again mainly to geometric aspects of both the accretion stream and the near--field column. The additional complexity observed in these stars is probably caused by the fact that the angle subtended between the spot normal and the viewing direction is small and the accretion region is viewed nearly face on, through a long length of the column, for a large fraction of the binary phase ($\\sim 25$\\%). The accretion region in AM Her is seen nearly face--on as it crosses the central meridian which (along with its rather weak magnetic field) may account for it possessing the most complex lightcurve of the category 1 systems.  \\item [6.] Our geometric model matches the EUV observations of the AM Hers very well and yields spot sizes averaging 0.06 R$_{WD}$, or $f \\sim 1 \\times 10^{-3}$ the WD surface area, and average accretion spot heights around 0.023 R$_{WD}$. When the accretion spot undergoes self eclipse (category 1 systems) the raised mound model is useful for determining the system inclination to within 5 degrees, and the spot colatitude to within 7--15\\arcdeg . The near perfect symmetry of the rise and fall phase of the six systems shown in Figure 5 preclude any large scale ($> 0.2$ R$_{WD}$) structure of the accretion region. The height of the accretion region above the WD surface seems to be correlated with overall EUV flux (accretion rate), but the spot longitude $\\psi$ does not (see Table 2).  \\item [7.] All category 1 light curves show a small amount of EUV flux from above the accretion region. This flux can be isolated and observed during the short phase intervals when the main accretion region itself is just hidden from view behind the WD limb. The EUV spectra during these phases show no emission lines characteristic of a high temperature, corona-like plasma, but signal-to-noise available during these short phases is insufficient to completely exclude the possibility of weak emission features. Preliminary analysis of high signal to noise EUV spectra of the highly magnetic system AR UMa (Howell \\& Sirk, 1998) observed during outburst show no evidence of emission lines at any phase. This result adds weight to the growing body of negative evidence against the existence of a high temperature, low density plasma (corona) surrounding the accretion regions of AM Her stars.  \\item [8.] The far--field accretion stream causes narrow dips in the light curves that saturate to zero in 5 systems: UZ For, AN UMa, EF Eri, V834 Cen, and HU Aqr, and reduce the flux by a factor of 2 for VV Pup. The spectra extracted for UZ For and VV Pup during these narrow dip phases show no wavelength dependent variations or absorption features (lines or edges) compared with spectra obtained during the out--of--dip phases. Thus, in at least these two systems, the far--field accretion stream behaves like a grey absorber.  \\item [9.] A broad dip occurs at binary rotation phases where the far--field accretion stream is not interfering with our view to the accretion region. In addition, the broad dips do {\\it not} occur at EUV phase zero as predicted by assuming a circular spot impinged normally by a cylindrical accretion column (ID83). The cause for these broad dips must be from material very near the WD surface and appears to be almost entirely caused by Compton scattering. The broad dips tell us that either the column is highly inclined with respect to the WD surface, non-uniform in nature, or the source of maximum EUV radiation is not concentric with the densest portion of accretion column. Since polarimetry data show that the magnetic field is only slightly tilted with respect to the spot normal in several AM Her stars (Meggitt \\& Wickramasinghe, 1989), we are forced to conclude that the broad dips are caused by small scale ($ < 0.2$ R$_{WD}$) asymmetric structure in both the accretion spot and the near--field column.  \\item [10.] The broad dip / bright phase spectral ratio shows a wavelength dependence in that the broad dip phase spectra appear softer than the bright phase spectra. Absorption by a cold absorber (neutral hydrogen) is not possible due to the high state of, or complete ionization of H and He and would give a very different spectral slope. It thus appears that the broad dips are caused by some (unequal) combination of a warm absorber, Compton scattering, and geometric effects caused by the near--field stream. It is apparent that near the WD surface, the accretion column and region have a highly asymmetric structure which can significantly change on short timescales, days (WSV) to months and longer (see Fig 10, and Sirk et al., 1998).  \\end{itemize}  We have detailed the geometric nature of AM Herculis accretion regions via the use of high-quality EUV photometric and spectroscopic observations. Our few--component model provides good fits to the data, but also indicates some areas where a more complex structure is present. The magnetic field strength of the WD and the mass accretion rate are likely to be the dominant mechanisms which cause both the general similarities observed in many of the AM Hers, as well as the detailed differences between systems. The next step in understanding the details presented here is to use the presented model, along with the appropriate physical conditions likely to be present within the stream and accretion region itself, to confirm our results. We need to understand the roles played by geometry, Compton scattering, photoelectric absorption, magnetic field strength, mass accretion rate, and other physical properties, in order to form a better picture of the accretion regions in AM Hers.  The authors wish to thank the staff of the Center for Extreme Ultraviolet Astrophysics for their help throughout the EUVE mission. SBH and MMS wish to acknowledge partial support of this research by NASA EUVE grants NAG 5-3523 \\& 5-4241 and NASA ADP grants 5-2989 \\& 5-3379. MMS extends thanks to Patrick Sirk for Figure 11. Adrienne Cool provided many useful comments on an early version of the manuscript. Mark Cropper and Axel Schwope have also contributed useful comments. This manuscript is certified Cruelty Free; no graduate students were abused in its preperation. Not dishwasher safe.    \\clearpage      \\makeatletter \\def\\jnl@aj{AJ} \\ifx\\revtex@jnl\\jnl@aj\\let\\tablebreak=\\nl\\fi \\makeatother   \\begin{deluxetable}{lccrrcc} \\tablewidth{0pt} \\tablecaption{Observation Log} \\tablehead{ \\colhead{System } & \\colhead{Instrument$^a$ } & \\colhead{Starting Time } & \\colhead{Duration} & \\colhead{Exposure} & \\colhead{N Orbits$^b$} & \\colhead{Mean Countrate$^c$}\\nl & & (GMT) & (hours) & (ks) & & (s$^{-1}$) \\nl } \\startdata UZ For&ScB        & 1993 Oct 16  04:58 & 73 & 98  & 13 & 0.37 \\nl UZ For&DS         & 1993 Nov 18  18:47 & 78 & 85  & 11 & 0.93 \\nl UZ For&DS         & 1995 Jan 15  20:34 & 90 & 82  & 11 & 0.82 \\nl VV Pup&DS         & 1993 Feb 07  21:25 & 34 & 37  & \\ 6& 1.53 \\nl AM Her&DS         & 1993 Sep 23  17:57 &114 & 123 & 11 & 2.97 \\nl RE1149+28&DS      & 1993 Feb 22  18:50 & 53 & 64  & 12 & 0.48 \\nl RE1149+28&ScA     & 1994 Mar 08  01:56 &117 & 147 & 27 & 0.26 \\nl RE1149+28&DS      & 1994 Dec 26  06:06 &198 & 145 & 27 & 0.15 \\nl HU Aqr&DS         & 1996 May 29  02:13 & 123& 121 & 16 & 0.044\\nl RE1844-74&DS      & 1994 Aug 17  13:53 & 154& 188 & 35 & 0.65 \\nl EF Eri&DS\t  & 1993 Sep 05  13:42 &  99& 107 & 22 & 0.69 \\nl AN UMa&DS         & 1993 Feb 27  22:14 &  27&  33 & \\ 5& 0.32 \\nl V834 Cen&DS       & 1993 May 28  03:07 &  35&  37 & \\ 6& 0.76 \\nl QS Tel&DS         & 1993 Oct 06  07:51 & 105& 113 & 13 & 1.55 \\nl  \\tablenotetext{a}{Deep Survey (DS), Scanner A (ScA), and Scanner B (ScB).} \\tablenotetext{b}{Number of full binary orbits sampled.} \\tablenotetext{c}{The Scanner countrates have been multiplied by a factor of 2.2 to account for their smaller effective area compared with that of the the Deep Survey in the Lexan/Boron passband (Sirk et. al., 1997).}  \\enddata \\end{deluxetable}  \\clearpage      \\makeatletter \\def\\jnl@aj{AJ} \\ifx\\revtex@jnl\\jnl@aj\\let\\tablebreak=\\nl\\fi \\makeatother   \\begin{deluxetable}{lcccccccccccc} \\tablewidth{0pt} \\tablecaption{System Parameter Fits\\break The system inclination is $\\iota$. The angle between the rotational pole and the EUV accretion spot is $\\beta$. The radius and the height of the accretion spot are $r$ and $h$, respectively, in units of R$_{WD}$. The maximum height of the accretion column above the WD surface that shows significant flux is $h_{\\rm col}$ in units of R$_{WD}$. The longitude of the accretion spot is $\\psi$, the field strength of the primary magnetic pole is $B$. The final columns list the ratio of the absorber model flux to the un--absorbed model, and the absorption coefficient $\\tau$ found from the absorber model. Italic entries for $\\iota$, $\\beta$, $r$, $h$, and $h_{\\rm col}$ are the parameter solutions derived from our geometric model fits (roman entries indicate fixed parameters). Italic entries for values of the period and $\\psi$ are determined from {\\it EUVE} data. Non-derived entries (roman) are from Cropper (1990), except where noted. } \\tablehead{ \\colhead{System } & \\colhead{Inst. } & \\colhead{Date } & \\colhead{Period } & \\colhead{$\\iota$ } & \\colhead{$\\beta$ } & \\colhead{$r$ } & \\colhead{$h$ } & \\colhead{$h_{col}$ } & \\colhead{$\\psi$ } & \\colhead{$B$} & \\colhead{Ratio}& \\colhead{$\\tau$} \\nl & & & min & ($\\arcdeg$) & ($\\arcdeg$) & (R$_{WD}$) & (R$_{WD}$) & (R$_{WD}$) & ($\\arcdeg$) & (MG)& & \\nl } \\startdata UZ For&ScB        & Oct 93 & 126.5 &    80.2 & \\it114.1 & \\it.073 & \\it.018 & \\it.14 & \\it55 & 56 & \\it 0.47 & \\it 22\\nl UZ For&DS         & Nov 93 & 126.5 & \\it80.2 & \\it136.6 & \\it.060 & \\it.031 & \\it.15 & \\it49 & 56 & \\it 0.61 & \\it 23\\nl UZ For&DS         & Jan 95 & 126.5 & \\it81.7 & \\it136.5 & \\it.045 & \\it.021 & \\it.15 & \\it49 & 56 & \\it 0.61 & \\it 23\\nl VV Pup&DS         & Feb 93 & 100.4 & \\it73.1 & \\it147   & \\it.022 & \\it.011 & \\it.12 &    49 & 32 & \\it 0.65 & \\it 50\\nl AM Her&DS         & Sep 93 & 185.6 & \\it37.1 & \\it68    & \\it.045 & \\it.013 & \\it.033&    31 & 12 & \\it 0.62 & \\it 35\\nl RE1149+28&DS      & Feb 93 & \\it90.17& \\it70   & \\it142   & \\it .10 & \\it .035& \\it.12 & ---   & $43^a$ & \\it 0.89 & \\it 11 \\nl RE1149+28&ScA     & Mar 94 & \\it90.17& \\it70  & \\it136   & \\it.054 & \\it .025& \\it.12 & ---   & $43^a$  & \\it 0.78 & \\it 29 \\nl RE1149+28&DS      & Dec 94 & \\it90.17& \\it70   & \\it143   & \\it .10 & \\it .030& \\it.12 & ---   & $43^a$ & \\it 0.86 & \\it 27 \\nl HU Aqr&DS         & May 96 & 125.0  & \\it 81  & \\it 40   & \\it.036 & \\it .021& \\it .10& \\it 49 & 37$^b$  & --- & --- \\nl RE1844--74&DS     & Aug 94 & 90.10  & \\it 73  & \\it 144  & \\it.090 & \\it .025& \\it .12& ---   & 10$^c$  & \\it 0.94 & \\it 8\\nl EF Eri & DS       & Sep 93 & 81.01  & \\it 52  & \\it 33   & \\it.059 &     .025& ---    & ---   & 11$^d$  & ---      & \\it 12\\nl \\tablenotetext{a}{Schwope, A. D.,1997, private communication} \\tablenotetext{b}{Schwope et. al., (1993)} \\tablenotetext{c}{Ramsay et. al., (1996)} \\tablenotetext{d}{Paerels, F., 1995, private communication} \\enddata \\end{deluxetable}  \\clearpage   \\makeatletter \\def\\jnl@aj{AJ} \\ifx\\revtex@jnl\\jnl@aj\\let\\tablebreak=\\nl\\fi \\makeatother   \\begin{deluxetable}{ccccccccc} \\tablewidth{0pt} \\tablecaption{System Geometry Comparison} \\tablehead{ \\colhead{System } & \\colhead{Present Work } & \\colhead{} & \\colhead{Polarimetry} & \\colhead{} & \\colhead{Ref.} & \\colhead{Spectrophotometry} & \\colhead{} & \\colhead{Ref.} \\nl & $\\iota$ & $\\beta$ & $\\iota$ & $\\beta$ & & $\\iota$ & $\\beta$ & \\nl } \\startdata UZ For & 81 (2) & 136 (5) & 81 (2) & 150 (12) & 1 & 85 (3) & 155 & 2,3 \\nl HU Aqr & 81 (5) & 40 (10) & 80 (5) & 40 (10)  & 4 & 83 (3) & --- & 5    \\nl VV Pup & 73 (3) & 147 (5) & 74     & 147      & 6 &        &    &         \\nl AM Her & 37 (4) & 68 (8)  & 52     & 49       & 7 &        &    &         \\nl RE1844--74& 73 (5)& 144 (7)& 60    & 167      & 8 &        &    &         \\nl EF Eri & 52 (5) & 33 (10) & 55     & 35       & 6 &        &    &         \\nl \\enddata \\tablenotetext{}{Numbers in parentheses are the 1 $\\sigma$ errors. 1)From optical eclipse analysis by Bailey \\& Cropper, (1991). 2)Beuermann, Thomas \\& Schwope, (1988). 3)Schwope, Beuermann \\& Thomas, (1990). 4)Glenn et. al., (1994). 5)Schwope, Mantel \\& Horne, (1997). 6)Meggitt \\& Wickramasinghe, (1989). 7)Wickramasinghe et al., (1991). 8)Bailey et. al., (1995).} \\end{deluxetable}  \\clearpage"
    },
    "9805/nucl-th9805007_arXiv.txt": {
        "abstract": "{\\normalsize We use the Quasiparticle Random Phase Approximation to include the effects of low-lying Gamow-Teller and first forbidden strength in neutrino capture by very neutron-rich nuclei with $N$ = 50, 82, or 126.  For electron neutrinos in what is currently considered the most likely $r$-process site the capture cross sections are two or more times previous estimates.  We briefly discuss the reliability of our calculations and their implications for nucleosynthesis.} ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805196_arXiv.txt": {
        "abstract": "s{ Observations show that Type Ia Supernovae (SNe Ia) form a homogeneous class of objects.  They share similar spectroscopic evolution, light-curve shapes, and peak absolute magnitudes. The slight departures from homogeneity that are observed can be used to produce a ``calibrated candle'' with corrected magnitudes with even smaller dispersion. The existence of this intrinsically bright distance indicator has inspired two coordinated high-redshift supernova searches: the Supernova Cosmology Project and the High-z Supernova Search Team. To date $\\sim 100$ SNe Ia have been discovered by the two groups. The preliminary analysis of the first of these objects demonstrate how well SNe Ia can be used to measure the mass density of the universe $\\Omega_M$ and the normalized cosmological constant $\\Omega_\\Lambda \\equiv \\Lambda/3H_0^2$. } ",
        "introduction": "For the past several years, two independent groups have been discovering and following high-redshift supernovae ($z>0.3$) using telescopes from all over the world and beyond.  The lofty and imposing goal of these searches? To determine the ultimate fate of the universe!  But before I tell you what the answer is (so far), I should explain what makes supernovae so special and give you an idea of what's been observed to date.  Then comes the answer, along with a discussion of some of the systematic errors involved and how we can address them.  I conclude by presenting the scientific course we plan to take in the near future. ",
        "conclusions": "We can confidently say that our data disfavors an $\\Omega=1$, $\\Lambda=0$ universe, considering the huge amount of systematic error required to make the two consistent. For the future, we need to pursue two opposite directions in order to get a more precise value of the cosmological parameters. By finding more supernovae at $z>0.8$, we will attempt to reduce the length of the contour in Figure~\\ref{conf} to give a better simultaneous measurement of $\\Omega_M$ and $\\Omega_\\Lambda$.  By finding supernovae at $z < 0.2$ we will learn more about their intrinsic properties and give a larger sample with which we can study and hopefully reduce systematic effects. SCP is now dedicating a large part of its efforts in nearby supernova searching, and in collaboration a group of French scientists is planning to use the CFHT specifically for $z>0.8$ searches. (HIZ's most distant candidate was in fact found using the CFHT). Members of the HIZ team are already heavily involved in nearby searches and as a whole are pursuing higher redshifts. With all this focused activity, reduction of the systematic and statistical errors should not be far off in the future."
    },
    "9805/astro-ph9805369_arXiv.txt": {
        "abstract": "We present $N$--body models for triaxial elliptical galaxies or halos of galaxies, which are fully self--gravitating, have near constant axis ratio as a function of radius and a $r^{-1}$ central density cusp. Preliminary investigation suggests the model are stable and orbit analysis shows no indication of chaotic orbits. The models provide a starting point for investigations into the evolution of triaxial figures of equilibrium, response of triaxial figures to central black holes, external perturbations and interactions. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805019_arXiv.txt": {
        "abstract": "The no-longer-eclipsing system SS~Lac in the young open cluster NGC 7209 has been recently announced to show a constant radial velocity.  Puzzled by this finding, we have monitored the system during 1997 obtaining 24 Echelle+CCD spectra over 8 orbital revolutions. Our spectra reveal a nice orbital motion with periodic splitting and merging of spectral lines from both components.  An accurate orbit has been derived, together with individual masses of the stars. SS~Lac presents a moderately eccentric orbit and a probably full synchronization between stellar rotation and orbital revolution.  \\keywords {Binaries: eclipsing -- Binaries: spectroscopic -- Stars: individual (SS~Lac) -- Open Clusters: individual (NGC 7209)} ",
        "introduction": "According to recent photometry (cf. Vansevicius et al. 1997) SS~Lac belongs to the young open cluster NGC 7209 . Hoffmeister (1921), Dugan \\& Wright (1935), Wachmann (1936), Nekrasova (1938) and Kordylewski et al. (1961) described SS~Lac as an eclipsing binary of 14.4 days period, with significant eccentricity because the equal depth secondary eclipse occurred at phase 0.57.  However, more recent observations by Zakirov \\& Azimov (1990), Lacy (1992), Mossakovskaya (1993) and Schiller et al. (1996) show that the eclipses no longer occur. The end of the eclipsing phase is set around 1940 by Mossakovskaya (1993) and around 1960 by Lehmann (1991). The latter ascribes the rotation of the plane of the orbit to the presence of an unseen third body in the system, orbiting at a great distance the closer central pair.  Quite puzzling has been the report by Schiller et al. (1991) and Schiller \\& Milone (1996) that 1983-84 spectra of SS~Lac did not reveal the expected double--lined nature of the spectra. Very little, if any, variation in radial velocity was observed. Using available predictions for the rate of change of the SS~Lac orbital inclination, Schiller \\& Milone (1996) expected a semi-amplitude of the radial velocity curve of 150 km sec$^{-1}$, far in excess of their instrumental resolution. They suggested that SS~Lac could be a triple system suddenly become chaotic or that a close encounter with another NGC 7209 member could have ionized the binary (cf. Schiller 1996).  Stimulated by the Schiller et al. (1991) report, Etzel et al. (1996), Stefanik et al. (1996) and Etzel \\& Volgenau (1996) announced in IAU Circulars that, according to their preliminary spectroscopic observations, SS~Lac is a double--lined system, with indications of variability in the radial velocities. So far no investigation of the orbital motion of SS~Lac has appeared in literature.  As an ex-eclipsing binary in a well populated young open cluster, SS~Lac clearly deserved further investigations to clarify the whole issue. In this note we report about our spectroscopic monitoring of SS~Lac performed to the aim of confirming the binary nature and to derive the spectroscopic orbit. ",
        "conclusions": "\\subsection{Eclipse timing}  The {\\sl MDJ}$_\\odot$=716.315 spectrum in Figure~1 presents a perfect radial velocity superposition of the two spectra, thus it corresponds to eclipse conditions. The eclipse ephemeris given by Mossakovskaya (1993) \\begin{displaymath} Min \\ I \\ = \\ 2415900.76 + 14.41619 (\\pm0.00013) \\times E \\end{displaymath} predicts a +0.456 day shift for the {\\sl MJD}$_\\odot$=716.315 spectrum, which is comparable with the propagation of the uncertainty on the period. Imposing phase coincidence between Mossakovskaya's principal minimum and {\\sl MJD}$_\\odot$=716.315 spectrum, this leads to the improved ephemeris \\begin{equation} Min \\ I \\ = \\ 2450716.32(\\pm0.15) + 14.41638 (\\pm0.00010) \\times E \\end{equation} The slightly longer orbital period nearly coincides with those given by Brancewicz \\& Dworak (1980, P=14.4163 days) and by Dugan \\& Wright (1935, P=14.41629 days).  \\begin{table} \\tabcolsep 0.08truecm \\caption{ Orbital elements for SS~Lac. Where appropriate, the second value correspond to the component $b$ represented by a dotted line and open circles in Figure~2. The quoted errors are the formal errors of the orbital solution. The entry ``{\\sl deviation}\" is the mean weighted deviation of the observed radial velocities from the computed orbital solution (weight = err$^{-2}$ in Table~1).  \\ {\\sl MJD}$_\\odot$ = JD$_\\odot$ -- 2450000. The error on the masses is $\\sim$0.1 M$_\\odot$.} \\begin{tabular}{lllll} \\hline &&&\\\\ && \\multicolumn{1}{c}{star $a$} &&\\multicolumn{1}{c}{star $b$}\\\\ \\cline{3-3} \\cline{5-5} &&&\\\\ period               & (days)          & 14.41638           &&                   \\\\ baricentric velocity & (km sec$^{-1}$) & --21.2$\\pm$0.3     &&                   \\\\ semi-amplitude       & (km sec$^{-1}$) & 74.7$\\pm$1.1       && 77.6$\\pm$1.4      \\\\ eccentricity         &                 & 0.122$\\pm$0.019    &&                   \\\\ $a_i$sin$i$          & (AU)            & 0.098$\\pm$0.001~~~ && 0.102$\\pm$0.002~~~\\\\ $T_\\circ$            & (MJD$_\\odot$)   & 741.4$\\pm$0.2      &&                   \\\\ $\\Omega$             & (deg)           & 332$\\pm$9          &&                   \\\\ deviation            & (km sec$^{-1}$) & 0.87               && 0.70              \\\\ &&&\\\\ inclination          & (deg)           & 78$\\pm$5           &&                   \\\\ mass                 & (M$_\\odot$)     & 2.80               && 2.69              \\\\ \\hline \\end{tabular} \\end{table}   \\subsection{Spectral classification}  The absence of He~I absorptions, the paucity of metallic lines and the very strong Stark broadening of the wings of the Balmer lines suggest a spectral classification around A2~V, virtually identical for the two members (cf. Figure~1). A more refined classification however needs devoted observations at the head of the Balmer series. Component $b$ appears slightly less luminous than component $a$, as suggested by the relative depths of the central Doppler cores and contribution to the overall H$\\alpha$ line wings in Figure~1. A lower brightness agrees with the slightly smaller mass of component $b$ (cf. the semi-amplitudes and semi-major axes in Table~2) and the just slightly reduced amplitude of $b$ eclipses compared to $a$ eclipses (a few hundredth of a magnitude, cf. Dugan \\& Wright 1935, Mossakovskaya 1993).  \\begin{figure} \\centerline{\\psfig{file=fig2.ps,height=8cm,width=8cm}} \\caption[]{Orbital solution for SS~Lac. The curves give the orbital solution of Table~2. Solid curve and filled circles refer to component $a$ in Tables~1 and 2. The crosses mark spectra for which the too small velocity separation caused merging of the two components into a single line. The dashed line is the baricentric velocity (--21.2 km sec$^{-1}$). Orbital phase according to Eq.(2).} \\end{figure}  \\subsection{Orbital inclination}  The spectral classification of the two components can be turned into a physical radius of R=2.25 R$_\\odot$ from interpolation of tabular values given by Allen (1973). The orbital separation from Table~2 is {\\sl a\\,sin\\,i} = 0.200$\\pm$0.002 AU = 42.5$\\pm$0.5 R$_\\odot$. Lehmann (1991) has stated that the system presented total eclipses (central eclipses given the nearly identical size of the components) in 1900-1915, and afterward the eclipses monotonically reduced in depth until they disappeared around 1960. From separation and size of the two components it may be computed that at the end of the eclipsing season the inclination was $i \\sim 83^\\circ$.  The corresponding linear change in orbital inclination is \\begin{equation} \\frac{d{\\rm i}}{dt} = 0.13 \\pm 0.01 \\ \\ \\ \\ \\ deg \\  yr^{-1} \\end{equation} which is significantly lower than the {\\sl di/dt} = 0.18 deg yr$^{-1}$ obtained by  Lehmann (1991) on the assumption of a B7~V classification for both components (clearly ruled out by our spectra that show no evidence for HeI absorption lines). From Eq.(3) and the occurrence of central eclipses at the beginning of this century, the current orbital inclination is \\begin{equation} i_{(1998)} \\sim 78^\\circ \\end{equation} The above results indicate that the non-eclipsing season of SS~Lac lasts for about 1275$\\pm$90 years.  \\subsection{Rotational velocities}  The rotational velocity of the components of SS~Lac has been derived by comparison with standard stars from Sletteback (1975), which were observed under the same instrumental conditions. The velocity turned out to be the same for both components: \\begin{equation} V_{rot} = 10 \\ \\pm \\ 5 \\ \\ \\ \\ \\ \\ km \\ sec^{-1} \\end{equation} The dependence on inclination (see Eq[4]) has been removed under the assumption that the rotational and orbital axes are parallel.  A $V_{rot} = 10\\pm5$ km sec$^{-1}$ means that rotational and orbital motion are synchronized or close to, the orbital velocity amounting to V$_{orb} = 8$ km sec$^{-1}$.  It is worth noting that SS~Lac shows evidence for synchronization when the orbital circularization is still far from being achieved. The two phenomena seems therefore to evolve over quite different time scales, in agreement with the theoretical expectations (cf. Zahn 1977).  The expected synchronization time for a binary with the SS~Lac parameters may be estimated from the Tassoul's (1987) formalism to be: \\begin{eqnarray} t_{syn}(yrs)& \\geq & 2500 \\times [P(days)]^{\\frac{11}{4}}  \\\\ & \\geq & 4 \\ million \\ years  \\nonumber \\end{eqnarray} which is relatively short compared to the NGC 7209 cluster age (300 million yrs according to Lynga 1985). The synchronization status cannot therefore be used to decide if SS~Lac is or not a primordial binary of NGC 7209.  \\subsection{Individual masses}  A straightforward application of the Kepler's III law to Table~2 data gives 2.69 and 2.80 M$_\\odot$ for the masses of the two SS~Lac components.  The tabular mass from Allen (1973) for the A2~V spectral type is 2.78 M$_\\odot$, in excellent agreement.  \\subsection{Mass-Luminosity relation}  Vansevicius et al. (1997) have estimated a 1.0 kpc distance and a $A_V$=0.54 mag extinction to NGC 7209. Using a bolometric correction {\\sl B.C.}= --0.29 mag for the spectral type A2~V (Allen 1973), the bolometric magnitude of the SS~Lac components is \\begin{equation} M_{bol} = +0.25 \\ \\ \\ mag \\end{equation} corresponding to a luminosity $L = 65 \\ L_\\odot$. The mean mass for the two components is 2.745 M$_\\odot$. At this mass the {\\sl Mass-Luminosity} relation has the numerical form (cf. Bowers \\& Deeming 1984): \\begin{equation} log \\frac{L}{L_\\odot} = 0.479 + 2.91 \\ log \\frac{M}{M_\\odot} \\end{equation} which predicts $L \\sim 57 \\ L_\\odot$ for the SS~Lac components. The agreement between the observed and predicted luminosity is satisfactory in view of the uncertainties in the distance, reddening and the approximate bolometric correction.   \\begin{table} \\tabcolsep 0.08truecm \\caption{Heliocentric radial velocity of the weak component of the Balmer lines described in Sect.4.7. The errors may be quantified as $\\pm$1 km sec$^{-1}$ when $N$=3, $\\pm$2 for $N$=2 and $\\pm$4 for $N$=1. {\\sl MJD}$_\\odot$ = JD$_\\odot$ -- 2450000. $N$ = number of Balmer lines used.} \\begin{tabular}{cclcclccl} \\hline &&&&&\\\\ MJD & RV$_\\odot$&N~~~~& MJD &RV$_\\odot$ & N~~~~ & MJD &RV$_\\odot$ & N\\\\ &(km/sec)&&&(km/sec)&&&(km/sec)&\\\\ &&&&&&&\\\\ 659.537 & --16 &  1  & 732.317 & --27 &  1 & 768.345 & --7  &  1 \\\\ 660.562 & --22 &  3  & 734.264 & --19 &  3 & 770.223 & --41 &  1 \\\\ 675.362 & --29 &  1  & 748.391 & --23 &  2 & 771.205 & --19 &  1 \\\\ 714.378 & --20 &  1  &         &     &    &         &     &    \\\\ &&&&&&&\\\\ \\hline \\end{tabular}  \\end{table}  \\subsection{The unseen third body}  Lehmann (1991) ascribes the rotation of the plane of the orbit to the presence of an unseen third body in the system, orbiting at a great distance the closer central pair. A discussion of this hypothesis is beyond the scopes of the present paper. We may however point out that in our spectra when the Balmer lines show a clear split and the S/N ratio is high, a weak absorption is visible. The H$\\alpha$ component does not correspond to known telluric absorptions in this region, and the coincidence between the various Balmer lines favour an interpretation as a {\\it real} stellar line. To the benefit of future Investigators of SS~Lac we report in Table~3 the heliocentric velocity of this weak component as measured on our spectra (same weighting as described in Sect.2), that could in some way be related to the unseen third body."
    },
    "9805/astro-ph9805063_arXiv.txt": {
        "abstract": "We present a short review of the abundances in the atmospheres of SB2 systems with Mercury-Manganese (HgMn) primaries. Up to now a careful study has been made for both components of 8 out of 17 known SB2 binaries with orbital periods shorter than 100 days and mass ratio ranging from 1.08 to 2.2. For all eight systems we observe a lower Mn abundance in the secondary's atmospheres than in the primary's. Significant difference in the abundances is also found for some peculiar elements such as Ga, Xe, Pt. All secondary stars with effective temperatures less than 10000 K show abundance characteristics typical of the metallic-line stars. ",
        "introduction": "The study of abundances in atmospheres of SB2 systems with peculiar components may provide constraints on the origin of the abundance anomalies. It is natural to believe that both components in a binary system with a period less than 100 days form from the same protostellar cloud, therefore their initial abundances have to be identical. Any observed difference and its dependence on the mass and/or atmospheric parameters of the star may show us the development of abundance anomalies during stellar evolution. Among peculiar stars, the highest frequency of binary systems is observed for the non-magnetic HgMn and metallic-line (Am) stars. We give here a short review of the atmospheric abundances in SB2 binary stars with HgMn primaries. ",
        "conclusions": "\\begin{itemize} \\item{Abundances in the atmospheres of SB2 systems may differ significantly even when both stars have practically equal masses} \\item{A comparison of the abundances in $\\alpha$ And, $\\kappa$ Cnc, and 112 Her shows that the observed peculiarities do not correlate with the orbital periods or eccentricities} \\item{In three hot primaries of practically equal temperatures, a violation of the odd-even effect in Mn-Fe is observed in non-synchronized systems} \\item{The gallium abundance drops by about 1.5 dex within a narrow temperature range 11300-11600 K. It follows from the paper by Smith (1996) and it is supported by the abundances in 46 Dra. The only exception is HR 7775, which has \\teff=10650 K and a high gallium abundance} \\item{There is no correlation between Hg abundance in the primaries and their atmospheric parameters, masses or orbital parameters, which supports the same conclusion made by Smith (1997)} \\item{The detailed study of the atmospheres of HR 4072 B and 112 Her B shows that their abundance pattern is similar to abundances in Am and roAp stars. The only feature which allows to classify both secondaries as Am stars is the Co/Ni ratio: [Co/Ni]$\\approx$-1.0 in HR 4072 B, 112 Her B and in Am stars, while [Co/Ni]$\\approx$0.5 in all roAp stars with detailed abundance analysis} \\item{In view of the similarity of the abundances of the main elements in the atmospheres of the secondary stars with \\teff$<$10000 K, it seems to be possible to classify them as Am stars} \\end{itemize}"
    },
    "9805/astro-ph9805255_arXiv.txt": {
        "abstract": "\\baselineskip 4.5 true mm % I review the current observational status of the faint end of the optical luminosity function of field galaxies at low redshift. There is growing evidence for an excess number of dwarf galaxies that is not well fit by a single Schechter function. These dwarf galaxies tend to be of late morphological and spectral type, blue in colour, of low surface brightness and currently undergoing significant star formation. ",
        "introduction": "The galaxy luminosity function (LF) characterizes the number density of galaxies as a function of luminosity, or absolute magnitude. The faint end of the LF tells us the abundance of dwarf galaxies, accurate knowledge of which is important for constraining models of both galaxy formation and evolution.  The optical galaxy luminosity function has traditionally been fit by a Schechter \\cite{s76} function: \\begin{equation} \\phi(L)dL = \\phi^* \\left(\\frac{L}{L^*}\\right)^\\alpha \\exp\\left(\\frac{-L}{L^*}\\right) d\\left(\\frac{L}{L^*}\\right). \\label{eqn:schec} \\end{equation} For galaxies more luminous than a characteristic luminosity $L^*$, $\\phi(L)$ drops exponentially with luminosity; for galaxies fainter than $L^*$ $\\phi(L)$ approaches a power-law with slope $\\alpha$. The quantity $\\phi^*$ represents an overall normalisation of the luminosity function.  Most local galaxy surveys (eg. \\cite{eep88,lkslots96,lpem92,mhg94,rspf98}) have found a value of $\\alpha$ that lies in the range $-1.2 \\lsim \\alpha \\lsim-0.7$, giving a ``flat'' faint-end slope when the LF is plotted as a function of absolute magnitude. However, there is recent evidence that the faint end of the LF is not well fit by a Schechter function: \\begin{enumerate} \\item An analysis of the CfA redshift survey by Marzke \\etal\\ \\cite{mhg94} finds that although the best-fit Schechter function has $\\alpha = -1.0 \\pm 0.2$, there is a significant excess of galaxies above the Schechter fit at magnitudes $M_Z \\gsim -16$\\footnote{Throughout, absolute magnitudes will be quoted assuming a Hubble constant $H_0$ of 100 km s$^{-1}$ Mpc$^{-1}$}. \\item For the Las Campanas Redshift Survey, Lin \\etal\\ \\cite{lkslots96} measure a faint-end slope of $\\alpha = -0.70 \\pm 0.05$ in the Gunn-$r$ band. However, all of their data points fainter than $M_r \\approx -17.5$ lie above this Schechter fit. \\item In their analysis of the ESO Slice Project (ESP) redshift survey, Zucca \\etal\\ \\cite{z97} find that the $b_J$ luminosity function is best fit by a flat faint-end Schechter function for $M_{b_J} < -17$ and a power-law of slope $-1.57$ at fainter magnitudes. \\end{enumerate}  Given the importance of the number-density of local dwarf galaxies for theories of galaxy formation and evolution, it is extremely desirable to better constrain the faint-end of the luminosity function. The problem is that most galaxies in flux-limited surveys have $L \\sim L^*$ and measuring galaxy redshifts to fainter magnitudes does not increase the number of dwarf galaxies {\\em relative} to the number of bright galaxies. There is no substitute for surveying large numbers of faint galaxies to constrain  the faint-end of the LF. Thus we require either to carry out large redshift surveys, such as 2dF~\\cite{c98} and SDSS~\\cite{l96}, or to estimate galaxy redshifts from their photometric colours or clustering properties.  In the following section I will describe how one can constrain the space density of dwarf galaxies from counts of faint galaxies around bright galaxies of known redshift. I will then review the properties of galaxies which dominate the faint-end of the LF and discuss future prospects for measuring the field galaxy LF. ",
        "conclusions": ""
    },
    "9805/astro-ph9805125_arXiv.txt": {
        "abstract": "The isothermal gravitational collapse and fragmentation of a region within a molecular cloud and the subsequent formation of a protostellar cluster is investigated numerically. The clump mass spectrum which forms during the fragmentation phase can be well approximated by a power law distribution $dN/dM \\propto M^{-1.5}$. In contrast, the mass spectrum of protostellar cores that form in the centers of Jeans-unstable clumps and evolve through accretion and $N$-body interactions is described by a log-normal distribution with a width that is in excellent agreement with observations of multiple stellar systems. ",
        "introduction": "\\label{sec:intro}  Understanding the processes leading to the formation of stars is one of the fundamental challenges in astronomy and astrophysics.  With the advent of new observational techniques and instruments, especially in the IR and radio wavebands, a vast amount of astronomical data about star forming regions has been accumulated. However, on the theoretical side not much progress has yet been made. Analytical models of the star formation process are restricted to describing the collapse of isolated, idealized objects (for an overview see Whitworth \\& Summers 1985). Much the same applies to numerical studies (e.g. Bonnell \\& Bastien 1993, Boss 1997, Burkert et al.~1996, 1997, Nakajima \\& Hanawa 1996). Star formation is a complex self-gravitating, magneto-hydrodynamical problem, which includes the effects of heating and cooling, and feedback pro\\-cesses from newly formed stars. Furthermore, it is influenced by the galactic environment. Taking into account all these processes with high spatial resolution exceeds by far present computational capabilities.   Previous numerical simulations of the collapse and fragmentation of molecular cloud regions have shown that a large number of condensed objects can indeed form on a dynamical timescale as a result of gravitational fragmentation (e.g.~Larson 1978, Monaghan \\& Lattanzio 1991, Keto et al.~1991). In these studies, the clouds were treated as isolated gaseous spheres which collapsed completely onto themselves. Instead, we study a small region embedded in a large, stable molecular cloud complex where only the overdense regions are able to contract due to self gravity.  We assume the molecular cloud is supported on large scales by turbulence and/or other processes.  Previous numerical models were also strongly constrained by numerical resolution.  Larson (1978), for example, used just 150 particles in an SPH-like simulation.  Whitworth et al. (1995) and Turner et al. (1995) were the first who addressed star formation on larger scales in detail using high-resolution numerical models. However, they studied a different problem: fragmentation and star formation in the shocked interface of colliding molecular clumps.  While clump-clump interactions are expected to be abundant in molecular clouds, the rapid formation of a whole star cluster requires gravitational collapse on a larger scale which contains many clumps and dense filaments.   In this letter, we extend previous studies of the collapse of isolated objects to the regime of the isothermal collapse and fragmentation of a gravitationally unstable {\\em region} embedded in the interior of a molecular cloud. We present a high-resolution numerical model of the dynamical evolution and follow the fragmentation into dense protostellar cores.  The temperature and the density is chosen such that the region is highly gravitationally unstable and forms a hierarchically-structured protostellar cluster.  The results of this study, i.e. the properties of the dense clumps and of the newly formed protostellar cores are compared with observations. ",
        "conclusions": "\\label{sec:summary}  Since collapse and fragmentation in molecular clouds is an extremely complex and dynamical process, many authors have sought to understand the stellar initial mass function as resulting from a sequence of statistical events which may naturally lead to a log-normal IMF (see e.g.~Zinnecker 1984, Adams \\& Fatuzzo 1996; also Price \\& Podsiadlowski 1995, Murray \\& Lin 1996, Elemegreen 1997).  However, using numerical simulations, it is possible to identify underlying processes which may contribute to the form of the stellar initial mass function.  In the calculations presented here, we find several trends.  The ``protostellar'' cores that form first are generally formed in the clumps with the highest initial density, and tend to have the highest final masses.  Cores that form later, form from gas that was initially in low-density clumps or distributed gas which converged to form a higher-density clump before quickly collapsing.  Overlaid on these general trends, dynamical interactions between individual cores can act to terminate accretion on to a core by ejecting it from a clump, thus setting its final mass.  The excellent agreement between the numerically-calculated mass function and the observed IMF for multiple stellar systems (Kroupa et al.~1990) strongly suggests that these gravitational fragmentation and accretion processes dominate the origin of stellar masses.  In a subsequent paper, the results from calculations spanning a larger range of the parameter space relevant for molecular clouds shall be discussed in detail."
    },
    "9805/astro-ph9805313_arXiv.txt": {
        "abstract": "s{ Inflation predicts the generation of cosmological perturbations. Usually, the power spectra for the scalar and tensor modes are calculated with help of the slow roll approximation. In the case of power law inflation an exact result is available. We compare the predictions for the cosmic microwave background anisotropies from the slow roll approximation with the exact results from power law inflation. We find that the so-called consistency check from the slow roll approximation, $C_2^{\\rm T}/C_2^{\\rm S} \\approx -6.93 n_{\\rm T}$, may differ considerably from the exact result.} ",
        "introduction": "The inflationary scenario allows to solve the horizon and flatness problems and predicts the generation of density (scalar) perturbations and of gravitational waves (tensor perturbations) \\cite{MFB}. Due to the Sachs-Wolfe effect \\cite{SW} those perturbations can be observed in the cosmic microwave background (CMB). The COBE satellite has measured these anisotropies for the first time \\cite{COBE}. Forthcoming high precision observations, especially the MAP and PLANCK satellites \\cite{MP}, will determine the temperature correlations with a precision of a few percent. Therefore predictions from inflationary models should be made on the few percent level as well.  Almost all (analytical) predictions for perturbation spectra from inflation rely on the slow roll approximation \\cite{S,L}. So far, no systematic, quantitative analysis on the error of the slow roll approximation has been performed, neither for the power spectra, nor for the temperature two-point correlations. We consider this work as a first step in such an error analysis. We compare the results from slow roll inflation, i.e.~$a(t) \\sim \\exp(H t), H \\sim $ const, with the exact solutions from power law inflation, i.e. $a(t) \\sim t^p, p=$const.  In a previous work \\cite{MS}, in response to contrary claims, we showed that the contribution of tensor perturbations with respect to scalar perturbations to the CMB anisotropies is small for the equation of state $\\rho \\approx - p$ during inflation. However, inflation occurs already if $\\rho + 3 p < 0$, which does not necessarily lead to slow roll inflation.  Power law inflation provides exact solutions for the time evolution of cosmological perturbations and inflation can occur although the slow roll conditions are violated. It is therefore interesting to investigate the difference of the exact predictions with the slow roll predictions.  Here, we concentrate on the so-called consistency check, which relates the scalar and tensor CMB quadrupole. ",
        "conclusions": "The exact result and the slow roll result differ by a factor $F(n_{\\rm T})/(1 - n_{\\rm T}/2)$. For inflation $p > 1$, which translates into $0 > n_{\\rm T} = - 2/(p-1) > - \\infty$ we find that the slow roll result might differ considerably, e.g.~for $p=2$ the error is a factor $F(-2)/2 \\approx 0.34$. Only for $p > 100$, i.e. for $1>n_{\\rm S}>0.99$, the error of the slow roll approximation is less than $1 \\%$.  For small numbers $l$ the cosmic variance introduces an uncertainty of $\\Delta C_l = \\sqrt{2/2l+1} C_l$, for the quadrupole $\\Delta C_2 = 0.63 C_2$. Thus, even at the largest scales the error from the slow roll approximation might be as big as the cosmic variance. For intermediate values $l < 30$ the error from the slow roll approximation is even more important.  Our main conclusion is that the consistency check, Eq.~(\\ref{TSsr}), is not valid for an arbitrary inflationary model. When the slow roll approximation does not apply, as for power law inflation, we expect significant modifications."
    },
    "9805/gr-qc9805072_arXiv.txt": {
        "abstract": "s{We study the influence of reheating on super-horizon density perturbations and gravitational waves. We correct wrong claims \\cite{G} about the joining of perturbations at cosmological transitions and about the quantization of cosmological perturbations.} ",
        "introduction": "The aim of these proceedings, based on Ref.~2, is to clear up the recent controversy on the relative contribution of density perturbations and gravitational waves to the CMBR today. \\par The background model is taken to be a spatially flat FLRW model. We restrict our considerations to density perturbations and gravitational waves. The most general line element reads: \\begin{equation} \\label{0} {\\rm d}s^2 = a^2(\\eta )\\{- (1+2\\phi){\\rm d}\\eta ^2 + 2B_{|i}{\\rm d}x^i {\\rm d}\\eta + [(1-2\\psi)\\gamma _{ij}+2E_{|i|j}+h_{ij}^{\\rm TT} {\\rm d}x^i{\\rm d}x^j\\} \\ . \\end{equation} In the synchronous gauge, without loss of generality, the same line element can be written as: \\begin{equation} \\label{1} {\\rm d}s^2=a^2(\\eta )\\{-{\\rm d}\\eta ^2+[(1+hQ)\\delta _{ij} +\\frac{h_l}{k^2}Q_{,i,j}+h_{\\rm gw}Q_{ij}]{\\rm d}x^i{\\rm d}x^j\\}. \\end{equation} The scalar function $Q$ satisfies the Helmholtz equation and $Q_{ij}$ is a symmetric, transverse and traceless spherical harmonic. $k$ is the comoving wave number. In the following we use this form of the line element to make contact with previous works \\cite{G}. It is convenient to define ${\\cal H}\\equiv a'/a$, which is related with the Hubble constant $H$ by the relation $H={\\cal H}/a$. A prime denotes a derivative with respect to conformal time. The quantity $\\gamma $ is defined by the expression: $\\gamma (\\eta )\\equiv 1-{\\cal H}'/{\\cal H}^2$. For the de Sitter space-time $\\gamma$ vanishes. \\par Let us now consider the equations of motion for the perturbed metric. For density perturbations (in the case of a vanishing anisotropic pressure), all relevant quantities can be expressed in terms of the variable $\\mu $ defined by: $\\mu \\equiv a(h'+{\\cal H}\\gamma h)/({\\cal H} \\sqrt{\\gamma })$, except in the de Sitter case, which must be treated separately. For gravitational waves, the relevant quantity is $\\mu _{\\rm gw}\\equiv ah_{\\rm gw}$. Using the perturbed Einstein equations, one can show that both types of perturbations obey the same class of equation, i.e. the equation of a parametric oscillator: \\begin{equation} \\label{mu} \\mu ''+[k^2-U(\\eta )]\\mu =0 \\ , \\end{equation} with $U_{\\rm dp}=(a\\sqrt{\\gamma })''/(a\\sqrt{\\gamma })$ and $U_{\\rm gw}=a''/a$. But there is a fundamental difference: the presence of the factor $\\sqrt{\\gamma }$ in the effective potential of density perturbations. \\par Equation (\\ref{mu}) is valid for any model. However, it is important to consider cases where exact analytical solutions can be found. This happens for power law inflation where the scale factor is given by $a(\\eta ) =l_0|\\eta |^{1+\\beta }, \\beta \\le -2$. We will consider a model with three epochs in succession: inflation, radiation-dominated era, and matter-dominated era. For this simple model, the function $\\gamma (\\eta )$ is a constant during each epoch and is given by: $\\gamma_{\\rm i}=(2+\\beta )/(1+\\beta ), \\gamma_{\\rm r}=2, \\gamma_{\\rm m}=3/2$. Therefore the constant factor $\\sqrt{\\gamma }$ drops out of $U_{\\rm dp}(\\eta )$, and the solutions for density perturbations and gravitational waves are given by the same Bessel functions: \\begin{equation} \\label{10} \\mu (\\eta )=(k\\eta )^{1/2}[A_1J_{\\beta +1/2}(k\\eta )+ A_2J_{-(\\beta +1/2)}(k\\eta )]. \\end{equation} The aim is now to compute the amplitude of both types of perturbations during the matter-dominated era. In order to perform this calculation, two questions must be addressed: \\par 1) The initial conditions must be fixed. This amounts to choose the coefficients $A_1$ and $A_2$. This will be done with the help of quantum mechanical considerations. \\par 2) The way the solutions are matched between different epochs must be specified. This is no problem for gravitational waves since the effective potential is well-defined (although discontinuous). This is more tricky for density perturbations. $\\gamma $ is a Heaviside function. This means that the effective potential of the density perturbations is not defined in the sense of distributions at the different transitions. Recently, there was a controversy on this point \\cite{G,DM}. One purpose of Ref.~2 and this paper is to clear up this question. This question arises because we consider simple models which allow analytical solutions. In reality, the transition is smooth and the effective potential is never ill-defined. ",
        "conclusions": ""
    },
    "9805/astro-ph9805035_arXiv.txt": {
        "abstract": "Three topics are discussed: 1) Photometric observations of the rapidly oscillating Ap stars have shown that the pulsation amplitude drops dramatically as a function of wavelength from the blue to the red. A theoretical derivation, plus modelling, indicate s that this is because the vertical wavelength of the pulsation mode is short compared to the scale height of the atmosphere; in fact, it indicates that we are seeing a pulsation node in the observable atmosphere. Radial velocity observations, and theoretical calculations now support this. The implication for other research on CP stars is that this can provide observational constraints on the atmospheric structure independent of traditional spectral analysis. 2) Luminosities of roAp stars can be determine d from asteroseismology. A recent comparison of such asteroseismic luminosities with HIPPARCOS luminosities is shown. This suggests that roAp stars have lower temperatures and/or smaller radii than previous models have used, or that the magnetic fields in these stars alter the frequency separations. 3) The latest results of our long-term monitoring of the pulsation frequencies in certain roAp stars are discussed. There is a clear cyclic variability to the pulsation cavity, hence the sound speed and/or sound travel time (radius) of these stars. This might be indicative of magnetic cycles at a level that magnetic measurements cannot currently detect, although there is no theoretical support for such an idea. ",
        "introduction": "High in the mountains of Angola arises the Kavango river. It flows down across the deep sands of the Kalahari desert in Botswana where it spreads out over 200 km to form the Okavango Swamps - one of the greatest wildlife refuges in Africa. In the Swamps the water meanders through papyrus-choked channels dotted with small desert islands. The sand filters the water to spectacular clarity and purity. Hippos wallow in the main channels; crocodiles are common. On the biggest island, Chief's Island, the ``real'' Africa of the imagination comes alive: There are elephants, lions, cape buffalo, impala, warthogs, and the ``Swamp Specials'', Tssessebe and Lechwe - buck specially adapted to swamp living.  Many years ago five friends and I from Cape Town flew deep into the Swamps where we joined three Batswana guides in dugout canoes (called Mekoros) for a 10-day camping trip through the Swamps. We had no one common language between our three guides and the six of us, but amongst us all we spoke enough of a mixture of Tsetswana, Zulu, Xhosa, Fanikalo, English and Afrikaans to communicate. Each night, around the campfire over a shared dinner, we told stories of Africa.  On our first night we six from Cape Town all crawled into warm, goose-down sleeping bags - the latest and best in Western camping gear - snug against the sub-freezing cold of a winter's cold snap. We lay on our groundsheets oriented radially away from the fire with our feet at a safe distance from the sparks which might damage the expensive sleeping bags, and our heads out in the darkness where we could see the spectacular African Sky through the foliage of the bushveld trees. Our three guides had only two blankets for the three of them. One they put down in the sand to sleep on; they then curled up together for warmth and put the other blanket over them. Their heads were almost in the fire, and a stack of wood for stoking the fire during the night was nearby.  Even with the cold we asked them, ``Why do you sleep with your heads so close to the fire? Aren't you afraid of being burned by the sparks?'' And they explained, ``We have lived here all our lives. When a hyena or lion comes out of the night and tries to bites us, we would rather it bit at our feet, than our heads. We will then sit up and hit it with this Panga [machete] to scare it off.'' ``Ha, ha, ha'', we laughed, ``Listen to our guides trying to scare us city-slickers; well, you can't scare us! We are experienced campers.''  As we were trying to go to sleep there came the nearby roar and clatter of a train approaching! How could this be? There are no trains in the Okavango Swamps. Our guides patiently explained that it was a herd of Lechwe fleeing at high speed through the shallows of the swamp, possibly being chased by lions. We fell asleep thinking ``there they go - trying to scare us again. Ha Ha.''  But then, in the night, the lion roars came. They sounded like they were just beyond the shadows of the firelight - very near to our exposed heads. When you hear lions roar nearby, there is a louder-than-possible, low-frequency, or even sub-sonic, rumble which shakes your insides and turns them to jelly. It says, ``Be Frightened!'' And you are frightened. When we awoke the next morning all six Capetonians were sleeping with feet away from the fire and heads nearly in it; and that is how we slept the rest of the trip.  One night around the campfire we were exchanging our multiglottal stories. Our guides told us of a giant snake which lives far up near the headwaters of the Kavango River in Angola, a snake so big that it can swallow a Mekoro and its three occupants whole in a single swallow! When I recovered from laughing uproariously at this ridiculous claim, I asked ``Have $you$ ever seen one of these snakes?'' ``Wellllll, actuallllllly, hmmmm, no we haven't! .... BUT! We have reliable friends who have seen them, and we $know$ they are there.''  I just poo-pooed the idea. I said it was absurd. There is no snake in the world anywhere near that big.  Then I decided to tell the guides about the Great White Sharks of False Bay near Cape Town. In the Okavango Swamps lives the terrifying and terrific Tigerfish, with its razor teeth and mouth big enough to take off the hand of an unwary or foolish fisherman. Our guides knew this fish well; they all knew about the huge sea, too, although they hadn't seen it. False Bay is famous for its 5-m Great Whites. So I told our guides that in the sea near Cape Town we have fish so big that it can swallow a person in only two bites!  ``Hahahahahahahahahahahahahahahahahahahahahahahahaha!!!!!!!!!!!!''  All three guides were rolling on the ground in the sand with tears streaming down their faces. When they finally recovered from laughing uproariously at this ridiculous claim, one looked at me, trying to control involuntary outbreaks of more laughing, and asked ``Have $you$ ever seen one of these fish?''  I was rather taken aback by the question. I thought about it, then said in all honesty, ``Wellllll, actuallllllly, hmmmm, no I haven't! .... BUT! I have reliable books! I have seen reliable films! I have reliable friends who have seen them, so I $know$ they are there.''  They just poo-pooed the idea. They said it was absurd; there is no fish in the world anywhere near that big.  Both ``the fish'' and ``the snake'' were undoubtedly-real and obviously-absurd, depending on our preconceptions and viewpoints. In science a major task is to discriminate between the fish and the snake. Either may seem plausible - indeed, may seem probable - and either may turn out to be absurd in the end. Great care and hard work are needed to learn which is true, and which absurd. ",
        "conclusions": ""
    },
    "9805/astro-ph9805345_arXiv.txt": {
        "abstract": "We present a model for the different X-ray spectral states displayed by Galactic Black Hole Candidates (GBHC). We discuss the physical and spectral implications for a magnetically structured corona in which magnetic flares result from reconnection of flux tubes rising from the accretion disk by the magnetic buoyancy (Parker) instability. Using the observations of one of the best studied examples, GX339-4, we identify the geometry and the physical conditions characterizing each of these states. We find that, in the Soft state, flaring occurs at small scale heights above the accretion disk. The soft thermal--like spectrum, characteristic of this state, is the result of heating and consequent re-radiation of the hard X-rays produced by such flares. The hard tail is produced by Comptonization of the soft field radiation. Conversely, the hard state is the result of a phase in which flares are triggered high above the underlying accretion disk and produce X-rays via Comptonization of either internal cyclo--synchrotron radiation or of soft disk photons.  The spectral characteristics of the different states are naturally accounted for by the choice of geometry: when flares are triggered high above the disk the system is  photon--starved, hence the hard Comptonized spectrum of the hard state.  Intense flaring close to the disk greatly enhanced the soft--photon field with the result that the spectrum softens.  We interpret the two states as being related to two different phases of magnetic energy dissipation. In the Soft state, Parker instability in the disk favours the emergence of large numbers of relatively low magnetic field flux tubes. In the hard state, only intense magnetic fields become buoyant and magnetic loops are able to rise and expand in the coronal atmosphere. The model can also qualitatively account for the observed short timescale variability and the characteristics of the X-ray reflected component of the hard state. ",
        "introduction": "Galactic X-ray sources are classified as black holes candidates (GBHC) if either the measured binary mass function indicates the presence of an object with $M \\approxgt 3\\Msun$ (for a review see Tanaka \\& Lewin 1995) or their high energy X-ray spectra and temporal variability are similar to other GBHC.  For many years GBHC have been known to radiate in five different spectral states, defined by the observed spectral components and flux level typically in the $1-10 \\keV$ band. Systems in the {\\it Hard/Low state} emit most of their energy in a hard tail which can be represented as a power law with a photon index $\\Gamma \\sim 1.3-1.7$ and an exponential cut--off at about 100 \\keV (Tanaka \\& Lewin 1995, Zdziarski et al. 1997). Most of these objects are also observed in a {\\it Soft/High State}, when most of the energy is emitted in a blackbody component with characteristic temperature in the range of $0.6- 1 \\keV$. In addition to this thermal component the spectrum comprises a power law, characterized by a slope $\\Gamma=2.0-2.5$, which dominates above $\\sim$ few keV.  It was originally thought that the GBHC systems were much more luminous in the soft than in the hard state, but recent observations of GX339-4 and Cyg X-1 have shown that despite their dramatic spectral changes the bolometric luminosity changes only slightly.  Although the hard and soft states are the most common ones, occasionally GBHC have been observed in three other states (these occur as transitional events and have been observed rather infrequently).  The {\\it Intermediate State} is seen during transitions between the low and high state. In the {\\it Very High State}, GBHC have the highest luminosity: the high-energy power law component has a flux comparable to the soft blackbody one and the high energy emission does not show any sign of a cut-off.  In the {\\it Off State}, spectra are completely dominated by a power-law ($\\Gamma=1.7$) component with a flux level lower than in the hard state by a few orders of magnitude.  Another important signature of the different source behaviour in the different spectral states is the variability pattern.  In the soft state the normalization of the power law has been observed to vary whereas the soft blackbody component is very stable.  The emission in the hard state instead shows extreme variability on time scales as short as $10^{-3}$ s. Short timescale variability is also observed in the power law of the very high state.  Although temporal and spectral behaviors of GBHC have been widely studied in the last few years, the nature of the different spectral states and in particular the mechanism driving the transition from one to the other are largely unknown. The thermal radiation, characteristic of the high, very high and intermediate states is generally modeled to be blackbody emission from a (standard) optically thick accretion disk.  The power law component is generally attributed to inverse Compton radiation from an optically thin corona sandwiching the accretion disk.  The presence of such a hot medium was initially hypothesized in the context of Seyfert galaxies (e.g. Haardt \\& Maraschi 1993) and because of the similarity between the spectra of Seyfert galaxies and those of GBHC in the hard state, the same model was naturally extended to the galactic objects too.  In the Seyferts scenario it has also become necessary, in order to account for the different ratios of UV vs X-ray luminosity and the extremely short variability time scales observed, to assume that such a corona actually consists of localized active regions (e.g. Haardt, Maraschi \\& Ghisellini 1994, Stern et al. 1995, Nayakshin \\& Melia 1997a,b). Such regions could be the end result of impulsive magnetic energy dissipation (i.e. reconnection) in flux tubes emerging from the disk by buoyant instability (as originally suggested by Galeev, Rosner \\& Vaiana 1979).  In GBHC different geometries for the spatial distributions of the hot and cold matter associated with the accretion flow have been proposed. Some models (Dove et al. 1997, Gierlinsky et al. 1997, Poutanen, Krolik \\& Ryde 1997) deduce the geometry of the emitting regions from the spectral analysis of the different states.  In particular, the authors of such models argue that because the reflection component in the hard state of GBHC is much less prominent then in Seyferts, the inner radius of the disk must be far away from the black hole, in order to subtend a small solid angle to the X-ray emitting region. Most of the energy would be dissipated in a thermally hot central cloud/corona--like structure which, by Comptonization, produces the observed spectrum.  In the soft state the optically thick cool disk is postulated to move inwards and the majority of the dissipated energy would emerge in the form of a blackbody-like spectrum.  In these scenarios, though, no physical mechanism is provided to explain the origin of such drastic changes in the geometry of the inner region between the two states.  A possible interpretation has been discussed by Esin, McClintock and Narayan (1997) in the context of advection-dominated accretion flow (ADAF) solutions. In these models the inner compton cloud in the hard state is identified with an advection dominated zone (with still an outer thin disk at larger radii). As the accretion rate increases an ADAF is no longer allowed and it shrinks in size. The outer thin disk moves inwards and the spectrum changes form hard to soft.  It should be noticed that although the geometry of the X-ray emitting region in both classes of models (the Compton cloud and the advection-dominated) predicts the observed lack of reflected emission, it can not explain the fast variability time scales observed in the hard states of GBHC.  In this paper we show that the different spectral components and the different spectral states of GBHC can be easily reproduced within the context of a magnetically structured corona by considering flares triggered at different scale heights above an accretion disk. We expect such a corona to form because the strong magnetic fields, continuously generated by the dynamo action in an accretion disk, are strongly buoyant and are forced to invade the region sandwiching the disk itself.  Once outside the disk the magnetic flux tubes can reconnect efficiently and dissipate part of the accretion energy in localized active flares.   Here, we shall, at first, constrain the geometry and the energy dissipation distribution of the emitting regions in the different states exclusively from the detailed broad band spectral information provided by the observations.  In Section 2 we will give an overview of our model, and specify the parameters which can be derived almost directly from observations and those related to the system geometry. In particular, we will consider the spectral implications of having localized active regions at different scale heights above the accretion disk.  In section 4 we will obtain the parameter spaces which characterize the different spectral states by applying our model to GX339-4 (we summarize the data collected from the literature in section 3).  In section 5 we will relate the properties of the active coronal regions derived from the physics of radiation, to those of a magnetically structured corona. We will show that our model is completely consistent with the idea that X-ray flares are the after effect of reconnection of magnetic flux tubes rising from the accretion disk due to magnetic buoyancy. The different spectral states can then be identified with different phases of flare activity. Such phases would, in turn, be regulated by the physical conditions in the accretion disk relevant for the onset of buoyancy instability. Finally, in Section 6 we briefly discuss our model predictions for the variability and the reflection properties in the different states. ",
        "conclusions": "We have discussed the implications on the spectral states observed in GBHC of a magnetically--structured corona model, in which magnetic flares may result from reconnection of flux tubes rising above the cold accretion disk due to magnetic buoyancy.  Using a simple geometrical representation which takes into account the scaling of the different quantities with height of the corona above the disk, we have shown that the hard X-ray state in GBHC can be produced by flares located high above the underlying disk via Comptonization of either the cyclo-synchrotron radiation or the photons produced in the disk. The soft state is instead due to flaring of active regions close to it. The soft blackbody component is in this case dominated by the re-radiated X-ray field and the hard tail is produced via Comptonization of this same soft radiation.  Depending on whether the flares are high above the disk or close to it the spectrum changes from being hard to soft respectively (and therefore gives rise to the hard state or the soft). When flares are triggered high above the disk the system is highly photon starved and therefore the Comtonized spectrum very hard. Conversely when a large numbers of flares are triggered very close to the disk the the soft photon field is highly enhanced; this naturally accounts for the a hotter blackbody component (due to reprocessing in the disk itself) and the very soft Comptonization spectrum typical of the soft state. By varying the typical scale height and scale size of the dissipative regions, the spectra for all of the five states observed in GX339-4 can be qualitatively reproduced.  Applying the model to GX339-4 has allowed us to determine constraints both on the geometrical and physical properties of such active regions.  We have qualitatively integrated the geometrical description of the different states in the physical context of magnetic buoyancy. We have shown that in the soft state, buoyancy instability in an accretion disk (which only dissipates a small fraction of the total energy) naturally favors the emergence of possibly a large number of relatively weak magnetic field flux tubes (maybe when the dynamo action is initiated in the disk). Because such tubes have a low internal pressure their expansion is basically halted as soon as they enter the coronal regions.  This is the reason why the flaring occurs really close to the accretion disk itself.  Conversely, the hard state could be due to a typically different phase of magnetic energy dissipation (maybe as the dynamo action builds higher magnetic fields).  This would take place in a small number of very intense magnetic field loops.  The reason why flares would reach higher scale heights, in this model, can then be ascribed to the intrinsically more intense magnetic field loops which would lead the loops to grow and expand before efficient reconnection takes place. Also, large loops would naturally form from small scale ones. Reconnection between small loops can drive a time--dependent inverse cascade process which leads to the formation of larger structures and cause spectral changes.   While the model described is certainly too schematic both in its geometrical/uniformity and physical assumptions it seems to account qualitatively for other observed properties of GBHC, namely the variability and characteristics of the reflected X--ray component.  In other models proposed to account for the spectral states of black hole candidates the amount of power dissipated in the accretion disc varies greatly from one state to the other, so that it is almost negligible in the hard state while dominating in the soft one. The model examined here on the contrary postulates that a constant fraction of the total power is radiatively dissipated in the disk. Most of the energy is instead released in magnetic coronal flares at different heights above the disk. The amount of heating and re--radiation from the accretion disk itself during the triggering of flares can naturally give rise to all of the spectral components observed in the different states of the source."
    },
    "9805/astro-ph9805109_arXiv.txt": {
        "abstract": "Several phenomena in high energy astrophysics have been recently related to clusters of galaxies and to cosmic ray interactions occurring inside these structures. In many of these phenomena the observable effects depend on the energy density of cosmic rays confined in the Intra Cluster (IC) medium, which is a poorly known quantity. We propose here that useful indications about this quantity can be obtained from present and future observations of galaxy clusters in the radio and hard X-ray frequency ranges. ",
        "introduction": "Clusters of galaxies are the largest gravitationally bound structures in the universe. They are exceptionally useful laboratories for {\\it cosmology} and {\\it high energy astrophysics}.  From the cosmological side these structures probe:  \\noindent {\\it i)}  the amplitude and shape of the primordial fluctuation spectrum, because the mass (or temperature) function of rich clusters is strongly sensitive to the value of the fluctuation power-spectrum, $\\sigma(M,z)$. Thus it is possible to measure $\\sigma(M,z)$ on the cluster scales fitting the mass or temperature function to the local data (see, e.g., \\cite{cmv97} \\cite{neta} and references therein);  \\noindent {\\it ii)}  the evolution of baryons, both condensed in the form of galaxies and diffuse in the form of IC gas, which is abundantly present within the cluster potential wells. In fact, galaxies are responsible for the chemical enrichment of the IC gas at a level $Fe/H \\simgt 0.3$ of the solar value, during their life-cycles. Thus,  high-quality X-ray spectral observations of galaxy clusters allow to study in details the physical state of the IC gas and the feedbacks from galaxy evolution \\cite{c97};  \\noindent {\\it iii)} the overall structure of the universe, using the possibility to measure directly $H_0$ and $\\Omega_0$ through X-ray and Sunyaev-Zel'dovich effect (in the radio and sub-mm bands) cluster observations \\cite{yr}.  Clusters of galaxies are also relevant for high energy phenomena in large scale structures because they can be regarded as the {\\it largest particle accelerators in the sky}. In the following, we will discuss some aspects in which galaxy clusters can yield important insights for high energy astrophysical phenomena. ",
        "conclusions": ""
    },
    "9805/astro-ph9805279_arXiv.txt": {
        "abstract": "The LINER galaxy NGC 2639 contains a water vapor megamaser, suggesting the presence of an edge-on nuclear accretion disk or torus. This galaxy is thus a good candidate for revealing absorption by the torus of any compact nuclear continuum emission. In this paper, we report VLBA radio maps at three frequencies and an ASCA X-ray spectrum obtained to search for free-free and photoelectric absorptions, respectively. The radio observations reveal a compact ($<$ 0.2 pc) nuclear source with a spectrum that turns over sharply near 5 GHz. This turnover may reflect either synchrotron self-absorption or free-free absorption. The galaxy is detected by ASCA with an observed luminosity of $1.4 \\times 10^{41}$ erg s$^{-1}$ in the 0.6 -- 10 keV band. The X-ray spectrum shows emission in excess of a power-law model at energies greater than 4 keV; we interpret this excess as compact, nuclear, hard X-ray emission with the lower energies photoelectrically absorbed by an equivalent hydrogen column of $\\simeq$ 5 $\\times$ 10$^{23}$ cm$^{-2}$.  If we assume that the turnover in the radio spectrum is caused by free-free absorption and that both the free-free and photoelectric absorptions are produced by the same gaseous component, the ratio $\\int n_{e}^{2} dl/\\int n_{H} dl$ may be determined. If the masing molecular gas is responsible for both absorptions, the required ionization fraction is $\\gtrsim 1.3 \\times 10^{-5}$, which is comparable to the theoretical {\\it upper} limit derived by Neufeld, Maloney \\& Conger (1994) for X-ray heated molecular gas. The two values may be reconciled if the molecular gas is very dense -- $n_{H_{2}} \\gtrsim 10^{9}$ cm$^{-3}$. The measured ionization fraction is also consistent with the idea that both absorptions occur in a hot ($\\sim$ 6,000K), weakly ionized (ionization fraction a few times 10$^{-2}$) atomic region that may co-exist with the warm molecular gas. If this is the case, the absorbing gas is $\\sim$ 1 pc from the nucleus. We rule out the possibility that both absorptions occur in a fully ionized gas near 10$^{4}$K. If our line of sight passes through more than one phase, the atomic gas probably dominates the free-free absorption while the molecular gas may dominate the photoelectric absorption. ",
        "introduction": "Over the last decade, it has become clear that many, perhaps all, active galactic nuclei, are surrounded by dusty accretion disks or tori on the pc or sub-pc scale (e.g. Antonucci 1993). When viewed equatorially, these tori hide emission from the central regions and they are believed to be responsible for the observational difference between broad-line objects (Seyfert 1s, broad-line radio galaxies and quasars), in which the torus is viewed near pole-on, and narrow-line objects (Seyfert 2s, narrow-line radio galaxies), in which the torus is viewed near edge-on. This model is supported by a wide range of observational results on narrow-line objects -- polarized broad lines (e.g. Tran 1995), reddened nuclei at optical wavelengths (Mulchaey, Wilson \\& Tsvetanov 1996), broad infrared recombination lines (e.g. Goodrich, Veilleux \\& Hill 1994), large gas columns to the nuclei as inferred from photoelectric absorption of soft X-rays (e.g. Turner et al. 1997) and bi-cones of ionized gas aligned with the radio ejecta (e.g. Wilson \\& Tsvetanov 1994).  The obscuring material, which appears to take the form of gas clouds in a geometrically thick torus or a warped, thin disk, is illuminated by the central UV and X-ray continuum source. Calculations of the physical and chemical properties of the gas have been made by Krolik \\& Lepp (1989) and Neufeld, Maloney \\& Conger (1994, hereafter NMC). The UV and soft X-rays are expected to be absorbed in thin layers at the surface of the clouds, but hard X-rays may penetrate and heat the interiors. This X-ray heated gas may possess a two phase structure, in which an atomic phase at $T \\approx$ 5,000 -- 8,000K coexists with a molecular phase at $T \\approx$ 600 -- 2,500 K (e.g. NMC). The ionization fraction is expected to be $\\lesssim$ 10$^{-5}$ in the molecular region and a few times 10$^{-2}$ in the atomic region, but the exact values are sensitive to the X-ray flux and gas density and to details of the chemical and physical processes in the gas. Clearly, an observational determination of the ionization fraction would be of value and is the goal of the present paper.  The ionization fraction may be estimated, in principle, by comparing the emission measure, $\\int n_{e}^{2} dl$ (from measurements of free-free absorption of the nuclear radio emission) with the total equivalent hydrogen column, $\\int n_{H} dl$ (from measurements of photoelectric absorption of nuclear X-rays), both through the disk or torus. To provide maximum path length for these absorptions, the disk needs to be oriented close to edge-on. Studies of H$_{2}$O megamaser emission from the nucleus of NGC 4258 reveal that the maser emission arises in a thin Keplerian accretion disk which is viewed very close to edge-on (Watson \\& Wallin 1994; Miyoshi et al. 1995). More recent VLBI mapping of other H$_{2}$O megamasers shows that, in almost all cases, the maser emission traces a line on the sky with kinematics consistent with an edge-on, rotating disk (Greenhill \\& Gwinn 1997; Greenhill, Moran \\& Herrnstein 1997; Trotter et al. 1998). Thus galaxies with detected H$_{2}$O megamaser emission provide the best opportunity for measuring the ionization fraction of the circumnuclear disk or torus.  NGC 2639 has a LINER-type nucleus (e.g. Ho, Filippenko \\& Sargent 1993) and H$_{2}$O megamaser emission (Braatz, Wilson \\& Henkel 1994). Although a VLBA map of the maser emission is not yet available, the systemic maser emission has been found to drift redward at a similar rate to that seen in NGC 4258 (Wilson, Braatz \\& Henkel 1995). In NGC 4258, this redward drift is known to result from the centripetal acceleration of clumps of masing gas on the near side of the edge-on accretion disk, as the gas passes in front of the nuclear radio source (Herrnstein et al. 1997). By analogy with NGC 4258, it may be argued that the redward drift in NGC 2639 arises in the same way and that it too contains an accretion disk viewed very close to edge-on. We therefore chose NGC 2639 for a search for absorption by the putative disk.  Throughout this paper, we adopt a velocity of NGC 2639 with respect to the microwave background radiation of V$_{3K}$ = 3,434 km s$^{-1}$ (de Vaucouleurs et al. 1991) and a Hubble constant of 75 km s$^{-1}$ Mpc$^{-1}$, giving a distance of 45.8 Mpc and a scale of 222 pc per arc second. ",
        "conclusions": "We have found that the spectrum of the nuclear radio source in NGC 2639 turns over below $\\sim$ 5 GHz and shown that this effect may result from either synchrotron self-absorption or free-free absorption. If the latter process is responsible, the implied value of $\\int n_{e}^{2} dl$ may be combined with the equivalent hydrogen column, $\\int n_{H} dl$ (derived from the X-ray spectrum), to provide the ionization fraction of the gas in terms of the electron temperature and hydrogen density, assuming the same gas is responsible for both kinds of absorption. The lower limit to the ionization fraction obtained by assuming the absorptions occur in the masing molecular gas is comparable to the upper limit to the ionization fraction derived theoretically by Neufeld, Maloney \\& Conger (1994) for warm molecular gas heated by a nuclear hard X-ray source, implying that the molecular gas must be dense ($\\gtrsim 10^{9}$ cm$^{-3}$) if it is to provide the free-free absorption. The required ionization fraction is consistent with that expected in the atomic phase at 5,000 -- 8,000 K which may co-exist with the molecular component. If our line of sight passes through {\\it both} phases, the hot atomic phase would likely dominate the free-free absorption while molecular gas could dominate the photoelectric absorption. We rule out the hypothesis that both types of absorption occur in a fully ionized gas at 10$^{4}$ K. Planned VLBA imaging of the H$_{2}$O megamaser will help distinguish between these various possibilities.  We thank J. H. Krolik and D. A. Neufeld for comments on an early draft of the manuscript and P. R. Maloney for advice on the expected ionization fractions. The National Radio Astronomy Observatory is a facility of the National Science Foundation, operated under cooperative agreement by Associated Universities, Inc. This research was supported by NASA through grants NAG 53393 and NAG 81027, by NSF through grant AST9527289 and by NATO through grant SA.5-2-05 (GRG. 960086) 318/96."
    },
    "9805/hep-ph9805272_arXiv.txt": {
        "abstract": "We present an exact analytic solution to the neutrino evolution equation in matter with periodic step-function density profile and discuss in detail the parametric resonance of neutrino oscillations that can occur in such a system. Solar and atmospheric neutrinos traversing the earth pass through layers of alternating density and can therefore experience parametric resonance of their oscillations. Atmospheric neutrinos can undergo parametrically enhanced oscillations in the earth when their trajectories deviate from the vertical by about $26^\\circ - 32^\\circ$. Solar neutrinos traversing the earth can experience a strong parametric resonance of their oscillations in a wide range of zenith angles. If the small mixing angle MSW effect is the solution of the solar neutrino problem, the oscillations of solar neutrinos crossing the core of the earth {\\em must} undergo strong parametric resonance; this phenomenon should facilitate significantly the observation of the day-night effect in oscillations of solar neutrinos. If observed, the enhanced day-night effect for core crossing neutrinos would therefore confirm both the MSW solution of the solar neutrino problem and the parametric resonance of neutrino oscillations. ",
        "introduction": "It is well known that the resonantly enhanced neutrino oscillations in matter, the Mikheyev-Smirnov-Wolfenstein (MSW) effect \\cite{MS,W}, have a simple analogue in classical mechanics: oscillations of two weakly coupled pendulums with slowly changing frequencies \\cite{MS2,Wein}. One then naturally wonders if there are any other resonance phenomena in mechanics which might have analogues in neutrino physics.  One such phenomenon is parametric resonance. The parametric resonance can occur in dynamical systems whose parameters vary periodically with time. A textbook example is a pendulum with vertically oscillating point of support \\cite{LL,Ar}. Under certain conditions topmost, normally unstable, equilibrium point becomes stable. The pendulum can oscillate around this point in the upside-down position. Another example, familiar to everybody, is a swing, which is just a pendulum with periodically changing effective length. It is the parametric resonance that makes it possible to swing on a swing.  What would be an analogue of the parametric resonance for neutrino systems? Since matter affects neutrino oscillations, periodically varying conditions can be achieved if a beam of oscillating neutrinos propagates through a medium with periodically modulated density. For certain relations between the period of density modulation and oscillation length, the parametric resonance occurs and the oscillations get enhanced. The probability of neutrino transition from one flavor state to another may become close to unity even for small mixing angle.  This phenomenon is very different from the MSW effect. Indeed, at the MSW resonance the neutrino mixing in matter becomes maximal ($\\theta_m=\\pi/4$) even if the vacuum mixing angle $\\theta_{vac}$ is small. This leads to large-amplitude neutrino oscillations in a matter of constant density equal (or almost equal) to the resonance density, or to a strong flavor conversion in the case of matter density slowly varying along the neutrino path and passing through the resonance value. The situation is quite different in the case of the parametric resonance. The mixing angle in matter (the oscillation amplitude) does not in general become large. What happens is an amplification of the transition probability because of specific phase relationships. Thus, in the case of the parametric resonance it is the {\\em phase} of oscillations (rather than their amplitude) that undergoes important modification.  The parametric resonance of neutrino oscillations has a very simple physical interpretation: the average value of the transition probability, around which the oscillations occur, drifts. This may lead to large probabilities of flavor transitions. We shall discuss this interpretation in the last section of this paper.  The possibility of the parametric resonance of neutrino oscillations was suggested independently in \\cite{ETC} and \\cite{Akh}. In ref. \\cite{ETC} an approximate solution for sinusoidal matter density profile was found. In \\cite{Akh} an exact analytic solution for the periodic step-function density profile was obtained. However, in ref. \\cite{Akh} the results were presented only for a simplified case of small matter effects on the oscillations length and mixing angle, $l_m\\approx l_{vac}$, $\\theta_m\\approx \\theta_{vac}$. Parametric effects in neutrino oscillations were further studied in \\cite{KS} where combined action of parametric and MSW resonances and possible consequences for solar and supernova neutrinos were discussed.  Recently, there has been an increasing interest in parametric resonance of neutrino oscillations. It was pointed out \\cite{LS} that atmospheric neutrinos traversing the earth travel through layers of alternating density and can therefore undergo parametrically enhanced oscillations. The same was also shown to be true for solar neutrinos passing through the earth \\cite{P}. Interestingly, this situation very closely corresponds to a periodic step-function density profile studied in \\cite{Akh}. We therefore believe that it would be useful to present the exact analytic solution for this case in full, without assuming smallness of matter effects. We also study the evolution of oscillating neutrinos in the earth and derive an exact analytic expression for the neutrino evolution matrix in the constant-density-layers model of the earth structure. Although in this case neutrinos travel only over ``one and a half'' periods of density modulation, parametric resonance effects are possible and can be very large in this case, too. We discuss these effects and their possible manifestations for solar and atmospheric neutrinos.  This paper is organized as follows. The evolution equation for a system of oscillating neutrinos in matter is reviewed and the solution for the periodic step-function density profile is derived in Sec. 2. Various limiting cases are considered and the parametric resonance of neutrino oscillations is studied in Sec. 3. Parametric enhancement of oscillations of solar and atmospheric neutrinos in the matter of the earth is considered in Sec. 4. The results are discussed and a simple physical interpretation of the parametric resonance in neutrino oscillations is given Sec. 5. ",
        "conclusions": "Neutrino oscillations can be parametrically enhanced in a medium with periodically varying density. The periodic step-function (``castle wall'') density profile is a very simple example; it allows for an exact solution of the neutrino evolution equation and therefore is very illuminating. In addition, this example is of practical importance since to a good approximation the density profile of the earth can be considered as a step-like function with nearly constant densities of the mantle and core.  The parametric resonance realizes very special conditions for neutrino oscillations, leading to a possibility of a striking increase of the transition probabilities. Typically, neutrinos have to travel over many periods of density modulation in order for the parametric enhancement to manifest itself. However, under certain conditions even for neutrino evolution during one period the parametric effects can be quite sizeable. Of course, in this case the neutrinos are not exposed to a periodic potential, and so one may wonder how the parametric enhancement of neutrino oscillations can occur. This can be explained as follows. The parametric resonance implies that the changes of the oscillation phase and matter density profile with the coordinate along the neutrino path are correlated in a very special way. This ``synchronization'' allows the transition probability to overbuild after every half-period of density modulation (see below); if the parameters $A_1$ and $A_2$ defined in eq. (\\ref{AB1}) are of opposite sign, and in addition the oscillation amplitudes at the densities $N_1$ and $N_2$ are not too small, a considerable increase of the transition probability is possible even for one period. The parametric enhancement for one and a half periods of density modulation is then even larger.  It is instructive to examine under what conditions a complete neutrino flavor conversion over one period of density modulation is possible. Though rather contrived, this case clarifies the essence of the parametric resonance of neutrino oscillations. {}From  eq. (\\ref{prob3}) we find that a complete conversion over period $T$ would require \\be 2(\\theta_1-\\theta_2)=\\pm \\pi/2\\, \\label{cond1} \\ee in addition to the conditions (\\ref{princ}). It is easy to see that, when eq. (\\ref{cond1}) is satisfied, the transition probability (\\ref{prob3}) is equal to the sum of the amplitudes of the oscillations in matter of constant densities $N_1$ and $N_2$, i.e. \\be \\sin^2 (2\\theta_1-2\\theta_2)=\\sin^2 2\\theta_1+\\sin^2 2\\theta_2\\,. \\label{cond2} \\ee Since at the resonance $\\phi_1=\\phi_2=\\pi/2$, we have the following pattern of neutrino oscillations in this case. During the first part of the period, $0\\le t\\le T_1$, usual oscillations in matter of constant density $N_1$ occur. At the time $t=T_1$ maximal possible in this case transition probability is achieved, which is $\\sin^2 2\\theta_1$. If the density remained constant, the transition probability $P$ would start decreasing after that and would return to zero at $t=2T_1$. However, the density changes to $N_2$ and, as a consequence of the parametric resonance, the transition probability continues growing instead of starting decreasing. From eqs. (\\ref{cond2}) and (\\ref{cond1}) it follows that in the time interval $(T_1, T_1+T_2)$ the transition probability $P$ undergoes one more half-period increase (with the amplitude $\\sin^2 2\\theta_2$), but starting from the initial value $\\sin^2 2\\theta_1$ and not from zero. Thus, in this idealized case the parametric resonance places one half-wave piece of the transition probability on the top of the other: the probability never decreases until it reaches the maximal value equal to unity. If the condition (\\ref{princ}) is only approximately satisfied, there is some decrease of the transition probability after the first ``half-period'', but $P$ does not reach zero, and the decrease is followed by another increase. What happens is essentially that the average value around which the transition probability oscillates starts drifting towards larger values. In this way $P$ can become quite large even if the amplitude of the oscillations around the average value is small. This is illustrated by Fig. 15, where the dependence of the transition probability on the coordinate along the neutrino trajectory is shown.  Neutrino oscillations in the earth can undergo a strong parametric enhancement. Neutrinos coming to the detector from the lower hemisphere from a source with the zenith angle in the interval $\\Theta=180^\\circ \\pm 33.17^\\circ$ traverse the earth's mantle, core and then again mantle and so pass ``one and a half'' periods of density modulation. The possibility of parametric enhancement of atmospheric neutrino oscillations in the earth was pointed out in \\cite{LS} where the case $\\delta=\\Delta m^2/4E\\ll V_1, V_2$ and the oscillations in the $\\nu_{\\mu}\\leftrightarrow \\nu_s$ channel were discussed. It was shown that the parametric resonance can modify the zenith angle distribution of the atmospheric neutrino events. As follows from our consideration, the parametric resonance can occur for oscillations of atmospheric neutrinos in both $\\nu_{\\mu}\\leftrightarrow \\nu_{e}$ and $\\nu_{\\mu}\\leftrightarrow \\nu_s$ channels for large enough nadir angles, $\\Theta_n\\simeq 26^\\circ - 32^\\circ$,  The energy width of the parametric resonance $\\Delta E/E_0$ is typically $\\sim 2 - 3$. Since the atmospheric neutrino anomaly has been observed in a rather wide range of neutrino energies (sub-GeV and multi-GeV), it seems unlikely that the parametric resonance can alter considerably the gross features of the atmospheric neutrino oscillations, such as the allowed values of neutrino parameters $\\Delta m^2$ and $\\sin^2 2\\theta_{vac}$. In any case,  numerical analyses of the atmospheric neutrino data should have automatically taken the parametric enhancement into account. Still, we believe that it may be worth re-analyzing the data paying more attention to the parametric resonance and looking for its possible manifestations. It should be also stressed that the parametric resonance can occur in oscillations of atmospheric neutrinos even in those channels that are not responsible for the atmospheric neutrino anomaly; such effects are potentially observable and of considerable interest.  The parametric resonance may be very important for oscillations of the solar neutrinos in the earth. For small $\\sin^2 2\\theta_{vac}$, in a wide range of zenith angles almost completely covering the earth's core, the $\\nu_e\\leftrightarrow \\nu_{\\mu,\\tau}$ oscillations exhibit a strong parametric resonance with the peak at $\\delta\\simeq (1.7 - 1.9)\\times 10^{-13}$ eV. For typical energy of $^{8}$B solar neutrinos $E\\simeq 8$ MeV these values of $\\delta$ give $\\Delta m^2\\simeq (5.5 - 6.1)\\times 10^{-6}$ eV$^2$. Amazingly, this almost exactly corresponds to the center of the allowed interval of $\\Delta m^2$ for the so called ``small mixing angle'' MSW solution of the solar neutrino problem (for recent discussions see, e.g., \\cite{LMP,BK,FoLiMo}). Strong parametric enhancement of the probability of the oscillations of solar neutrinos in the earth, by up to a factor of 7, occurs in the whole allowed region of $\\Delta m^2$ and $\\sin^2 2\\theta_{vac}$. The small mixing angle MSW solution gives the best fit of the available solar neutrino data and so is the most likely solution of the solar neutrino problem. The parametric resonance can also occur in the $\\nu_e\\leftrightarrow \\nu_s$ oscillations in the earth, but the effect is smaller in this case. The resonance values $\\delta\\simeq (6.8 - 10)\\times 10^{-14}$ eV correspond to $\\Delta m^2\\simeq (2.2 - 3.2) \\times 10^{-6}$ eV$^2$ (for $E=8$ MeV). These values are close to the lower bound of the allowed interval of $\\Delta m^2$ for the small mixing angle solution in this channel of oscillations. For large mixing angle the parametric resonance is also possible, but only in a rather limited interval of zenith angles of neutrino trajectories. In addition, the resonance values of $\\Delta m^2$ in this case are below the allowed interval.  In both cases of $\\nu_e\\leftrightarrow \\nu_{\\mu,\\tau}$ and $\\nu_e \\leftrightarrow \\nu_s$ oscillations the parametric effects are strongest at neutrino energies which are between the energies corresponding to the MSW resonances in the core and in the mantle of the earth. This is illustrated by Fig. 16; one can see that the parametric peak at $\\delta=1.5\\times 10^{-13}$ eV is between the maxima of $\\sin^2 2\\theta_1$ and $\\sin^2 2\\theta_2$ which are due to the MSW resonances in the mantle and in the core respectively. In other words, the MSW resonance density for neutrinos with $\\delta\\simeq 1.5\\times 10^{-13}$ eV is about $7.8~g/cm^3$; this density is between the core density and the mantle density. The value of the transition probability at the parametric peak exceeds the amplitudes of neutrino oscillations in the mantle $\\sin^2 2\\theta_1$ and in the core $\\sin^2 2\\theta_2$ by more than a factor of six.  The enhancement of the transition probability in matter in the case when the minimal and maximal densities satisfy $N_1<N_{MSW}<N_2$ would be of no surprise if the periodic density profile was, e.g., of the sinusoidal type. Then somewhere the matter density would be equal to the MSW resonance density, and one would expect an increase of the transition probability \\cite{KS}. However, in the case of the ``castle wall'' density distribution the MSW resonance condition is not satisfied anywhere because the density profile is discontinuous: the only allowed values of density are $N_1$ and $N_2$. It is quite remarkable that in this case, too, there is a sizeable enhancement of the transition probability compared to the case of matter of constant density. This effect cannot be explained by the usual MSW-type increase of the oscillation amplitude; it is related to a specific phase relationships which are due to the parametric resonance.  One of the most promising ways of testing the MSW solution of the solar neutrino problem is to look for the day-night effect due to regeneration of solar $\\nu_e$'s in the matter of the earth during the night \\cite{MS3,d-n,numer}. For the most probable small mixing angle solution, however, this effect is expected to be rather small -- at the level of $(1 - 3)\\%$. The parametric resonance must increase significantly the day-night effect for neutrinos crossing the earth's core. Unfortunately, due to their geographical  location, the existing solar neutrino detectors have a relatively low time during which solar neutrinos pass through the core of the earth to reach the detector every calendar year. The Super-Kamiokande detector has a largest fractional core coverage time equal to $7\\%$. In \\cite{GKR} it was suggested to build a new detector close to the equator in order to increase the sensitivity to the earth regeneration effect; this would also maximize the parametric resonance effects in oscillations of solar neutrinos in the earth.  The parametric enhancement of oscillations of solar neutrinos in the earth for neutrinos that travel significant distances in the earth's core was discovered numerically (but not recognized as the parametric resonance) in \\cite{numer,Nic,GKR,LMP}. The authors of \\cite{LMP} found that the day-night effect in the oscillations of solar neutrinos in the earth can be enhanced by up to a factor of six when solar neutrinos cross the earth's core. These results were quite surprising as the largest effect resulted for neutrinos of energies which correspond to the MSW resonance at the densities of about $(6 - 8)~g/cm^3$, the values lying between the core and mantle densities. For neutrinos of those energies the MSW resonance is inoperative~ \\footnote{Strictly speaking, the density profile of the earth is not discontinues; the densities $(6 - 8)~g/cm^3$ can be achieved on the border of the mantle and core where the density jumps from $5.5$ to $9.9~g/cm^3$. However, the jump occurs in a very thin layer and so the MSW resonance does not develop.}. The possibility of explaining these findings in terms of the parametric resonance was pointed out in \\cite{P}, where also a different interpretation was suggested which the author of \\cite{P} seems to prefer. Our results support the parametric--resonance interpretation.   It is remarkable that the parametric resonance conditions can be fulfilled for neutrino oscillations in the earth. An important observation here is \\cite{GKR,LS,LiLu} that the potentials $V_1$ and $V_2$ corresponding to the matter densities in the mantle and core, inverse radius of the earth $R^{-1}$, and typical values of $\\delta\\equiv \\Delta m^2/4E$ of interest for solar (and atmospheric) neutrinos, are all of the same order of magnitude -- ($3\\times 10^{-14}$ -- $3\\times 10^{-13}$) eV. It is this surprising coincidence that makes appreciable earth effects on oscillations of solar and atmospheric neutrinos possible. Even more remarkable fact is that the very specific conditions on the phases of neutrino oscillations in the mantle and core of the earth, $\\phi_1\\simeq \\phi_2\\simeq \\pi/2$, which lead to the parametric resonance of neutrino oscillations, are also satisfied in many cases of interest. Those include atmospheric neutrinos, and, most importantly, solar neutrinos.  For small vacuum mixing angle, the parametric resonance occurs for the values of $\\delta=\\Delta m^2/4E$ which exactly correspond to the allowed values for the small mixing angle MSW effect in the sun. If the small mixing angle MSW effect is the solution of the solar neutrino problem, the oscillations of solar neutrinos in the earth therefore {\\em must} undergo strong parametric resonance. The parametric resonance provides a natural interpretation for the increase of the night-time regeneration effect in the case of core-crossing neutrinos found numerically in \\cite{numer,Nic,GKR,LMP}. By selecting only those neutrinos, one can improve the prospects of observing the day-night effect (though at the expense of reduced statistics). If observed, the enhanced day-night effect for core crossing neutrinos would therefore confirm both the MSW solution of the solar neutrino problem and the parametric resonance of neutrino oscillations.  The author is grateful to Q.Y. Liu and A.Yu. Smirnov for useful discussions and to S.T. Petcov for informing him about the results of the paper \\cite{P} prior to its publication."
    },
    "9805/astro-ph9805092_arXiv.txt": {
        "abstract": "The origin of a regular, 600 pc-long arc of young stars and clusters in the Constellation III region of the Large Magellanic Cloud is considered. The circular form of this arc suggests that the prestellar gas was uniformly swept up by a central source of pressure. In the center of the arc are six $\\sim30$ My old A-type supergiant stars and a Cepheid variable of similar age, which may be related to the source of this pressure. We calculate the expansion of a bubble around a cluster of this age, and show that it could have triggered the formation of the arc at the right time and place. Surrounding the central old stars and extending well outside the young arc is the LMC4 superbubble and giant HI shell. We show how this superbubble and shell could have formed by the continued expansion of the 15 My old cavity, following star formation in the arc and the associated new pressures. The age sequence proposed here was not evident in the recent observations by Olsen et al. and Braun et al. because the first generation stars in the center of the LMC superbubble are relatively faint and scarce compared to the more substantial population of stars less than 15 My old that formed throughout the region in a second generation. These considerations lead to an examination of the origin of the LMC4/Constellation III region and other large rings in the LMC and other galaxies. Their size and circularity could be the result of low galactic shear and a thick disk, with several generations of star formation in their interiors now too faint to see. ",
        "introduction": "Two large arcs of young stars and clusters are prominently situated in the northeast corner of the Large Magellanic Cloud (LMC) (see Fig. 1). One is large and thick, being a quarter segment of a ring (designated the \"Quadrant\" here), and usually identified with Shapley's Constellation III. The other, slightly to the southwest, is smaller and contains brighter clusters shaped like one-sixth part of a ring (\"Sextant\"). The Quadrant consists of Lucke \\& Hodge (1970) associations LH 65, 77 and 84 and is often referred to as LH 77; the Sextant includes LH 51, 54, 60 and 63 and is not generally considered noteworthy, being often masked by bright HII regions (cf.  Table 1). The main peculiarity of both arcs is their very regular circular form.  The region surrounding these arcs is well studied as a possible example of propagating star formation. Westerlund \\& Mathewson (1966) noted on a UV plate \"the great arc of the bright blue stars,\" saying that \"Shapley called this arc Constellation III\"; they also suggested this arc is connected with the supergiant HI shell that surrounds the most prominent HI void in the LMC, noted by McGee \\& Milton (1966). The inner extent of this shell, as determined by Kim et al. (1997), is shown in figure 1 by a circle.  \\begin{figure*} \\vspace {7.5in} \\special{psfile=elmegreenfg1.ps hoffset=560 voffset=680 angle=180} \\caption{Photographic image of the Constellation III region of the LMC with a circle indicating the inner radius of the HI shell and other identifications discussed in the text. The diameter of the circle is 1.32 degrees. The original photograph is from Boyden Observatory, courtesy of Harlow Shapley, Harvard College Observatory, and kindly provided in tiff form by Knut Olsen, Paul Hodge, and Don Brownlee. } \\end{figure*}  Westerlund \\& Mathewson further suggested that \"the annular shell of neutral hydrogen and stars in the region of Constellation III\" is the remnant of a \"super-supernova\" outburst (following  Shklovsky 1960). They believed that gas was swept up by the expanding shell; the part of this shell that was moving to the center of the LMC would accumulate matter more quickly and be the first to become stationary.\tStars would condense out, forming the arc of stars of Constellation III.\tThey believed also that the southern part of the HI supershell is missing and that \"this gap is filled by the arc of bright blue stars forming Constellation III.\"  This latter suggestion is not confirmed by the HI data, which shows the southern part of the HI shell clearly (Domg\\\"orgen et al. 1995; Kim et al.  1997).  The enormous elongated ring of HII regions around the HI void, designated superbubble LMC 4, was found later by Meaburn (1980). The HI void, the surrounding HI shell, and the HII regions and clusters of the superbubble LMC 4, were all considered by Dopita et al. (1985) to be the result of supernovae and O-type stars near the center of LMC 4. They believed they saw an age gradient and determined the velocity of the star formation front, propagating outward from the center, to be 36 km/s, which was the same as the velocity of the HI shell that they measured. Subsequent investigations have not confirmed the age progression or HI velocity (Reid et al. 1987; Domg\\\"orgen et al. 1995; Olsen et al. 1997; Braun et al. 1997), though many still believe that LMC 4 and the ring of young clusters is connected with triggered star formation. Olsen et al. (1997) found that the age of LH72, which is near the center of Constellation III, is about the same as the ages of the young clusters along the LMC4 ring, although the age spread in LH72 is large, with stars ranging between $5\\times10^6$ and $1.5\\times 10^7$ years old. Olsen et al. concluded that their data are consistent with the alternate model by Reid et al. (1985), in which star formation proceeded locally for a long time after an initial trigger from the LMC4 superbubble shock. Braun et al. (1997) found no signs of an age gradient in the region, obtaining ages of field stars within a J-shaped strip inside LMC4, including Quadrant but not LH72.  Domg\\\"orgen et al.  (1995) studied the origin of the superbubble.  They considered stellar winds and supernovae in the central association, the collision of a high velocity cloud with the LMC disk, and the large-scale propagation of star formation, concluding that the latter process is more probable.  There is a third, larger, arc to the south of Quadrant, open to the northeast and touching Quadrant at the northwest. It is rather dispersed, and the age spread along it is large, but at the center of curvature of this arc there is a massive cluster, NGC 2041, that is older than most of the clusters within the arc. Members of this arc are listed in Table 1. It may be another example of triggered star formation, but it is less well defined than Quadrant and Sextant, so we do not discuss it further. All three arcs were noted and sketched by Hodge (1967), who also found similar features in NGC 6946.  Here we consider the possibility that the Quadrant and Sextant arcs formed by instabilities in a swept-up ring and shell, respectively. We also show how the giant HI ring surrounding Quadrant could have been rejuvenated by new pressures from these stars after the first generation pressures subsided. We then discuss why the LMC4 region appears so unusual compared to shells and rings in other galaxies. We do not comment on the origin of the first generation of stars (e.g., see de Boer et al. 1998), but only point out some likely members based on catalogs of supergiants and Cepheid variables. ",
        "conclusions": "Two arcs of star clusters inside and on the rim of the superbubble LMC 4, including what is commonly referred to as Constellation III, may have been formed by the self-gravitational collapse of gas in swept-up pieces of rings or shells. The dimensions and time sequences for these triggerings are reasonable, as is the energetics. This triggering model differs qualitatively from others in which pre-existing clouds are squeezed into star formation by passing shock fronts. The regular form of the Quadrant and Sextant arcs resembles more a piece of a ring or shell than a random arrangement of pre-existing clouds.  The conditions in the outer part of the LMC and other dwarf galaxies, and in the outer spiral arms of giant spiral galaxies, are favorable for the formation of giant gas shells and rings in which the first generation of stars is so old and dispersed that it is barely visible anymore. All that is required for this is a low rate of shear and a relatively large disk thickness.  A recent study suggests that Gamma Ray Bursts from old star-forming reigons might also play a role in enlarging supernova cavities beyond the size of the disk thickness (Efremov, Elmegreen \\& Hodge 1998). This would not affect the general triggering scenario discussed here, but it would affect the estimate for the number of stars that led to either the first or the second generation pressures.  We are grateful to the referee for valuable suggestions and to P.Bogdanovsky, whose computer program helped in searches for triggering stars. Dr. K. Olsen kindly provided a tiff image for the star field in figure 1."
    },
    "9805/astro-ph9805167_arXiv.txt": {
        "abstract": "I argue that the weight of the available evidence favours the conclusions that galaxies are unbiased tracers of mass, the mean mass density (excluding a cosmological constant or its equivalent) is less than the critical Einstein-de~Sitter value, and an isocurvature model for structure formation offers a viable and arguably attractive model for the early assembly of galaxies. If valid these conclusions complicate our work of adding structure formation to the standard model for cosmology, but it seems sensible to pay attention to evidence. ",
        "introduction": "In cosmology we attempt to draw large conclusions from limited and often ambiguous data. I am impressed at how well the enterprise is succeeding, to the point that we have an established standard model for the hot expanding universe (Peebles {\\it et al.} 1991). Which elements to include in the standard model is a matter for ongoing debate, of course. I am inclined to take a conservative line if only to avoid giving misleading impressions to our colleagues with deconstructionist tendencies. For example, the adiabatic cold dark matter model for structure formation has been more successful than I expected, and as a result is rightly the model most commonly used in studies of how structure might have formed. Simon White calls this model a paradigm, which I take to mean a pattern many find useful and convenient in their research. I like this use of the term, provided we agree to distinguish it from a well-established standard model.\\footnote{In Kuhn's (1970) picture ``normal science'' is done within the framework of a paradigm. I hope we can agree that those of us who study models for structure formation outside the set of ideas in the adiabatic cold dark matter paradigm are not necessarily doing abnormal science.} I think we cannot count the adiabatic cold mark matter paradigm as part of the standard model for cosmology because, as argued here, there is a viable and perhaps even more attractive alternative.  I organize this discussion around the issues of the weight of the evidence on whether galaxies are good tracers of mass, what we are learning from the cosmological tests, and the elements of the standard model for structure formation on the scale of galaxies and larger. I begin with another question, whether Einstein's introduction of the cosmological principle set a good example for research in our field. The hurried reader will find the main points summarized in \\S~6. ",
        "conclusions": "It is inevitable that the exciting rush of advances in this subject has left ideas unexplored. I have attempted to identify some roads not taken, less popular lines of thought that seem worth considering. The main points are summarized in the following questions.  \\subsubsection{Did Einstein set a good example?}  Einstein's brilliant success in establishing key elements of the standard cosmological model is an example of why we pay serious attention to elegant ideas even in the face of contrary empirical indications. But I think this is not an entirely edifying example: Einstein's intuition was not always so successful, and most of us are not Einsteins. In the present still crude state of cosmology it is better to be led by the phenomenology from astronomy and from particle physics (that may teach us the identity of the dark matter, for example).  Most of us agree that the Einstein-de~Sitter model is the elegant case, and it makes sense that the community has given it special attention despite the long-standing indication from galaxy peculiar velocities that the Einstein-de~Sitter density is too high. Now other lines of evidence are pointing in the same direction, as summarized in table~1, and I think there is general agreement in the community that we must give serious consideration to the possibility that Nature has other ideas about elegance. I count this as a cautionary example for the exploration of ideas on how the galaxies formed.  \\subsubsection{Why are galaxies thought to be biased tracers of mass?}  The galaxy two-point correlation function is quite close to a power law, $\\xi _{gg}(r)\\propto r^{-\\gamma}$, over three orders of magnitude of separation $r$ at low redshift, and the index $\\gamma$ is quite close to constant back to redshifts approaching unity. This is not true of the mass autocorrelation function $\\xi _{\\rho\\rho}(r)$ in the adiabatic cold dark matter (ACDM) model. Thus we have a measure of bias, $b(r,t)=\\left[ \\xi _{gg}(r,t)/\\xi _{\\rho\\rho}(r,t)\\right] ^{1/2}$ (eq.~[\\ref{eq:bias}]), that depends on position and time. Should we take this as evidence galaxies are biased mass tracers? Since the regularity is in the galaxies surely the first possibility to consider is that $\\xi _{gg}(r)$ is revealing a like regularity in the behaviour of the mass, that the bias is in the model. This reading is heavily influenced by a related issue: if much of the CDM is in the voids defined by normal galaxies where are the remnants of the void galaxies? Surely they are not entirely invisible?  I am impressed by the elegant simulations of the ACDM models Simon White presents in these Proceedings, and have to believe they reflect aspects of reality. But the curious issue of $b(r,t)$ leads me to suspect there is more to the story. Would the isocurvature variant do better? That awaits searching tests by numerical simulations of the kind that that have been applied to the adiabatic case.  \\subsubsection{What is the purpose of the cosmological tests?}  One often reads that it is to determine how the world ends. But should we trust an extrapolation into the indefinitely remote future of a theory that we know can only be a good approximation to reality? For a trivial example, suppose the universe has zero space curvature and the present value of the density parameter in matter capable of clustering is $\\Omega = 0.2$, with the rest of the contribution to $H_o{}^2$ in a term that acts like a cosmological ``constant'' $\\Lambda$ that is rolling toward zero (Peebles \\& Ratra 1988; Huey {\\it et al.} 1998). If the final value of $\\Lambda$ is identically zero then the world ends as Minkowksi spacetime (after all the black holes have evaporated). If $\\Lambda$ ends up at a permanent negative value, no matter how close to zero, the world ends in a Big Crunch. Should we care which it is? I would consider a bare answer an empty advance, because the excitement of physical science is in discovering the interconnections among phenomena. Perhaps the excitement of knowing how the world ends will be in what it teaches us about how the world began.  The classical cosmological tests, that probe spacetime geometry, have been greatly enriched by tests based on the condition that the cosmology admit a consistent and observationally acceptable theory for structure formation. The structure formation theory in turn tests ideas about what the universe was like before it was well described by the classical Friedmann-Lema\\^{\\i}tre model, and may eventually allow us to enlarge the standard model to include the story of how the world begins and ends.  \\subsubsection{What is the standard model for structure formation?}  Generally accepted elements are the gravitational growth of small primeval departures from homogeneity, that may be described as a stationary isotropic random process, in a universe with present mass that is dominated by CDM and maybe a term that acts like a cosmological constant.  The most striking piece of evidence for the gravitational instability picture is the agreement between the primeval density fluctuations needed to produce the CBR anisotropy and the present distribution and motion of the galaxies. Precision measurements in progress should allow us to fix many of the details of this gravitational instability picture, but within present constraints we cannot say that the primeval density fluctuations are Gaussian, or adiabatic, because we have a viable alternative, the non-Gaussian isocurvature model mentioned in \\S~4.  The main piece of evidence for the CDM is the mismatch between the baryon mass density in the standard model for the origin of the light elements and the mass density indicated by dynamical analyses of relative motions of the galaxies. Our reliance on hypothetical mass is embarrassing; a laboratory demonstration of its existence would be an exceedingly valuable advance.  \\subsubsection{Should we expect surprises from the next generation of surveys?}  It is a sign of the growing maturity of our field that we can pose questions that are motivated by specific theoretical issues and can be addressed by feasible observations. But I think our subject still is immature enough that we should be quite prepared for surprises. My favorite example is Shaver's (1991) demonstration that the radio galaxies within $50h^{-1}$~Mpc distance are close to the plane of the Local Supercluster, even though the plane is not apparent in the general distribution of galaxies at this depth. If the clusters and radio sources were produced by a pancake collapse why do we not see it in the general galaxy distribution? Maybe a better picture is that in the early universe a nearly straight cosmic string passed by, piling mass in its wake into a sheet that fragmented into the seeds of engines of active galaxies.  I think the most surprising outcome of the new surveys would be that there are no major corrections to what we think we know."
    },
    "9805/astro-ph9805217_arXiv.txt": {
        "abstract": "We present an {\\it ASCA} observation of the broad line radio galaxy 3C~111.  The X-ray spectrum is well described by a model consisting of a photoelectrically-absorbed power-law form.  The inferred absorbing column density is significantly greater than expected on the basis of 21-cm measurements of Galactic {\\sc H\\,i}.  Whilst this may be due intrinsic absorption from a circumnuclear torus or highly warped accretion disk, inhomogeneities and molecular gas within the foreground giant molecular cloud may also be responsible for some of this excess absorption.  We also claim a marginal detection of a broad iron K$\\alpha$ line which is well explained as being a fluorescent line originating from the central regions of a radiatively-efficient accretion disk.  This line appears weak in comparison to those found in (radio-quiet) Seyfert nuclei.  We briefly discuss the implications of this fact. ",
        "introduction": "Recent advances in X-ray spectroscopy, currently exemplified by the {\\it ASCA} satellite (Tanaka, Holt \\& Inoue 1994), have allowed new insights into the astrophysics of the central engines of active galactic nuclei (AGN).  Due to their proximity, and comparatively high brightness, the most complete studies have focussed upon Seyfert galaxies.  Whereas the underlying emission in Seyfert nuclei over the intrinsic {\\it ASCA} band (0.5--10\\,keV) is found to be well approximated by a power-law form, most Seyfert nuclei display additional spectral complexity.  Substantial photoelectric absorption by cold and/or ionized circumnuclear material is seen in half of all Seyfert 1 nuclei (Fabian et al. 1994; Reynolds 1997) and all Seyfert 2 nuclei.  This circumnuclear material may be related to the putative molecular torus, a warped accretion disk, and/or a disk-wind. In addition, almost all Seyfert 1 nuclei are seen to possess a broad, skewed iron K$\\alpha$ emission line (Mushotzky et al. 1995; Tanaka et al. 1995; Nandra et al. 1997; Reynolds 1997).  This line is thought to be due to fluorescence from the inner regions of the black hole accretion disk that occurs when the disk is strongly irradiated by the X-ray emitting plasma (Tanaka et al. 1995; Fabian et al. 1995).  These lines are typically seen to have a full width at half maximum (FWHM) of $100\\,000\\kmps$ and equivalent widths (EW) of $250\\eV$.  A Compton backscattered continuum (the so-called reflection continuum) accompanies these lines at approximately the strength expected on the basis of the X-ray reprocessing model (George \\& Fabian 1991; Matt, Perola \\& Piro 1991).  Detailed X-ray spectroscopic studies of radio-loud AGN are rather less complete than that of Seyfert galaxies.  This is primarily due to the comparative rarity of radio-loud objects (and, thus, the fact that we have to observe them at greater distance).  However, this is a subject of some importance: if the radio-loud/radio-quiet dichotomy reflects fundamental differences in central engine structure, we would expect those differences to be most directly revealed in the X-ray band.  The {\\it ASCA} spectrum of the powerful broad line radio galaxy (BLRG) 3C~109 reveals a fairly strong and broad iron line with FWHM$\\sim 120\\,000\\kmps$ and EW$\\sim 300\\eV$ (Allen et al. 1997).  This suggests a fundamental similarity between the central engine of this radio galaxy and typical Seyfert nuclei, i.e. that the central parts of the accretion flow are in the form of a geometrically-thin, radiatively-efficient accretion disk.  However, 3C~109 might be the exception rather than the rule.  Zdziarski et al. (1995) and Wo\\'zniak et al. (1997) have shown that BLRGs tend to have slightly weaker iron lines, and appreciably weaker reflection continua, as compared to Seyfert nuclei.  In this paper, we report the results of an {\\it ASCA} observation of 3C~111 ($z=0.048$).  This is an X-ray bright BLRG which is classified as an FR-II source with a double-lobe/single-jet morphology (Linfield \\& Perley 1984). The inner jet has been seen to display superluminal motion, with apparent outflow velocities of $3.4c$ (Vermeulen \\& Cohen 1994).  Assuming that the pattern speed is intrinsically sub-luminal, the standard theory of superluminal motion (e.g., see Blandford 1990) implies an upper limit on the angle between the line of sight and the jet axis of $\\theta<32\\degmark$.  Section 2 reviews the previous X-ray observations of this source.  In Section 3 we describe the basic {\\it ASCA} data analysis and present the results of our spectral fitting.  Section 4 discusses the astrophysical implications of our findings.  Section 5 draws together our conclusions. We assume that $H_0=50\\kmpspMpc$ and $q_0=0.5$ throughout this work. ",
        "conclusions": "Our {\\it ASCA} data for 3C~111 clearly reveal the presence of absorption in excess to that suggested by 21-cm measurements of the Galactic {\\sc H\\,i} column.  Whilst this may be evidence for absorption intrinsic to the 3C~111 system, we suggest that inhomogeneities and molecular material in the foreground (Galactic) giant molecular cloud may also be responsible for this excess absorption.  Molecular absorption line studies of this cloud lend strength to this possibility.  We also claim the marginal detection of a broad iron K$\\alpha$ line.  The detection of a broad line at all suggests the existence of a geometrically-thin, radiatively-efficient accretion disk within $r\\approxlt 20GM/c^2$ of the black hole.  However, the weakness of this line (and the associated backscattered continuum) as compared to typical Seyfert galaxies suggests that either X-ray beaming is important, or that the innermost regions of the disk are in a state incapable of producing the X-ray reflection signatures.  One possibility is that the disk undergoes a transition to an ADD state in its innermost regions."
    },
    "9805/gr-qc9805066_arXiv.txt": {
        "abstract": "s{This article gives an elementary review of gravitational collapse and the cosmic censorship hypothesis. Known models of collapse resulting in the formation of black holes and naked singularities are summarized. These models, when taken together, suggest that the censorship hypothesis may not hold in classical general relativity. The nature of the quantum processes that take place near a naked singularity, and their possible implication for observations, is briefly discussed.} ",
        "introduction": "After a star has exhausted its nuclear fuel, it can no longer remain in equilibrium and must ultimately undergo gravitational collapse. The star will end as a white dwarf if the mass of the collapsing core is less than the famous Chandrashekhar limit of 1.4 solar masses. It will end as a neutron star if the core has a mass greater than the Chandrashekhar limit and less than about 3-5 times the mass of the sun. It is often believed that a core heavier than about 5 solar masses will end, not as a white dwarf or as a neutron star, but as a black hole. However, this belief that a black hole will necessarily form is not based on any firm theoretical evidence. An alternate possibility allowed by the theory is that a naked singularity can form, and the purpose of the present article is to review our current understanding of gravitational collapse and the formation of black holes and naked singularities.  A black hole has been appropriately described by Chandrashekhar as the most beautiful macroscopic object known to man. Only a few parameters suffice to describe the most general black hole solution, and these objects have remarkable thermodynamic properties. Further, excellent observational evidence for their existence has developed over the years~\\cite{rees}. Thus, there can be no doubt about the reality of black holes, and the gravitational collapse of very many sufficiently massive stars must end in the formation of a black hole.  However, the following question is still very much open. If the collapsing core is heavy enough to not end as a neutron star, does this guarantee that a black hole will necessarily form? The answer to this question has to come from the general theory of relativity, and unfortunately this remains an unsolved problem.  What we do know from general relativity about gravitational collapse is broadly contained in the celebrated singularity theorems of Geroch, Hawking and Penrose. It has been shown that under fairly general conditions, a sufficiently massive collapsing object will undergo continual gravitational collapse, resulting in the formation of a gravitational singularity. The energy density of the collapsing matter, as well as the curvature of spacetime, are expected to diverge at this singularity.  Is such a singularity necessarily surrounded by an invisible region of spacetime, i.e. has a black hole formed? The singularity theorems do not imply so. The singularity may or may not be visible to a far away observer. If the singularity is invisible to a far away observer, we say the star has ended as a black hole. If it is visible, we say the star has ended as a naked singularity. We need to have a better understanding of general relativity in order to decide whether collapse always ends in a black hole or whether naked singularities can sometimes form.  Given this situation, Penrose was led to ask~\\cite{pe} whether there might exist a cosmic censor who forbids the existence of naked singularities, `clothing each one of them with a horizon'? Later, this led to the cosmic censorship hypothesis, which in broad physical terms states that the generic singularities arising in the gravitational collapse of physically reasonable matter are not naked. Till today, this hypothesis remains unproven in general relativity, neither is it clear that the hypothesis holds true in the theory. What is of course true is that the hypothesis forms the working basis for all of black hole physics and astrophysics. If cosmic censorship were to not hold, then some of the very massive stars will end as black holes, while others could end as naked singularities. As we will argue in Section 3, these two kinds of objects have very different observational properties.  There are various very important reasons for investigating whether or not cosmic censorship holds in classical general relativity. As we have mentioned above, the hypothesis is vital for black hole astrophysics. Unfortunately this fact is rarely appreciated by the astrophysics community. The hypothesis is also necessary for the proof of the black hole area theorem. It is not clear what the status of this theorem will be if the hypothesis were to not hold. If naked singularities do occur in classical relativity, they represent a breakdown of predictability, because one could not predict the evolution of spacetime beyond a naked singularity. Such singularities would then provide pointers towards a modification of classical general relativity, so that a suitable form of predictability is restored in the modified theory. Further, naked singularities might be observable in nature, if they are allowed by general relativity. Undoubtedly then, it is important to find out if the censorship hypothesis is valid.  We wish to make two further remarks. Firstly, while a great deal is known about the properties of stationary black holes, we know very little about the process of black hole {\\it formation}. In fact we know as little about the formation of black holes as we do about the formation of naked singularities. Secondly, it has sometimes been remarked that a theory of quantum gravity is likely to get rid of the singularities of classical general relativity, irrespective of whether these singularities are naked or covered. Why then does it eventually matter whether or not cosmic censorship holds? The answer to this legitimate objection is the following. A quantum gravity theory is expected to smear out a classical singularity and replace it by a region of very high, albeit finite, curvature. If the classical singularity is hidden behind a horizon (i.e. is a black hole), this quantum smeared region remains invisible to an external observer. However, if the classical singularity is naked, the smeared region of very high curvature will be visible to far away observers, and the physical processes taking place near this smeared region will be significantly different from those taking place outside the horizon of ordinary astrophysical black holes. Hence, from such an experimental standpoint, quantum gravity has little bearing on the question of cosmic censorship. To put it differently, quantum gravity is not expected to restore the event horizon, if the horizon is absent in the classical theory.  Since a theorem proving or disproving the hypothesis has not been found, attention has shifted to studying model examples of gravitational collapse, to find out whether the collapse ends in a black hole or a naked singularity. While specialised examples such as have been studied are nowhere near a general proof, they are really all that we have to go by, as of now. However, there does seem to be an underlying pattern in the results that have been found in these examples, which gives some indication of the general picture. It is interesting that all models studied to date admit both black hole and naked singularity solutions, depending on the choice of initial data. In the next section, we give a summary of what has been learnt from these examples and what they probably tell us about cosmic censorship. In the third section we will address the question of whether naked singularities might occur in nature, and if so, what they would look like to an observer.  The reader is also invited to study a few other excellent reviews~\\cite{revi} on the subject of cosmic censorship, which have appeared in recent years. Some of these reviews emphasize aspects other than those presented here, and some arrive at conclusions on cosmic censorship other than those given here. I would also like to draw attention to an earlier review of mine on this topic, which is somewhat more detailed, though a little dated~\\cite{tps}. Also, for a detailed bibliography the reader is requested is look up this earlier review; the references in the present article are not exhaustive, and largely confined to the more recent papers. ",
        "conclusions": ""
    },
    "9805/astro-ph9805021_arXiv.txt": {
        "abstract": "Recent observations of the spectral line profiles and intensity ratio of the O VI 1032 {\\AA} and 1037.6 {\\AA} doublet by the Ultraviolet Coronagraph Spectrometer (UVCS) on the Solar and Heliospheric Observatory (SOHO), made in coronal holes below 3.5 $R_s$, provide evidence for Doppler dimming of the O VI 1037.6 {\\AA} line and pumping by the chromospheric C II 1037.0182 {\\AA} line. Evidence for a significant kinetic temperature anisotropy of O$^{5+}$ ions was also derived from these observations. We show in this Letter how the component of the kinetic temperature in the direction perpendicular to the magnetic field, for both isotropic and anisotropic temperature distributions, affects both the amount of Doppler dimming and pumping. Taking this component into account, we further show that the observation that the O VI doublet intensity ratio is less than unity can be accounted for only if pumping by C II 1036.3367 {\\AA} in addition to C II 1037.0182 {\\AA} is in effect. The inclusion of the C II 1036.3367 {\\AA} pumping implies that the speed of the O$^{5+}$ ions can reach 400 km/s around 3 $R_s$ which is significantly higher than the reported UVCS values for atomic hydrogen in polar coronal holes. These results imply that oxygen ions flow much faster than protons at that heliocentric distance. ",
        "introduction": "The Ultraviolet Coronagraph Spectrometer (UVCS) on SOHO has proven to be a powerful tool for probing the physical conditions in the inner corona (Kohl et al. 1995, 1997, 1998; Habbal et al. 1997; Noci et al. 1997). These ultraviolet measurements extend to at least 3.5 $\\rm R_s$ in polar coronal holes, and to 10 $\\rm R_s$ in denser coronal plasmas such as streamers. As described by Kohl \\& Withbroe (1982), and Withbroe et al. (1982), determinations of proton and heavy ion flow speeds in the inner corona can be made using the Doppler dimming effect (see Hyder \\& Lites 1970). This method is applicable when ultraviolet coronal lines are formed primarily by the resonance scattering of ions by the chromospheric or transition region radiation in the same wavelength range, in addition to collisional excitation. As ions flow outwards in the corona, the fraction of the spectral line formed by resonance scattering becomes Doppler-shifted out of resonance with the disk emission. Subsequently, only the collisional component then remains. Of particular interest is the application of this technique to doublets such as the O VI 1032 and 1037.6 {\\AA} lines. In a static corona, the ratio of the line intensities 1032/1037.6 equals 4 when resonant scattering is dominant, and reduces to 2 when collisional excitation only is present.  Extending the concept of the Doppler dimming effect, Noci et al. (1987) showed that determinations of flow speeds in the range of 100 to 250 km/s could also be made. With the assumption of isotropic kinetic temperatures, that study was based on the possibility of scattering of the C II 1037.0 {\\AA} line by the coronal $\\rm O^{5+}$ ions when the ion flow speed is in the range of 100 to 250 km/s. When this occurs, the ratio 1032/1037.6 of the two line intensities is then expected to reach a minimum as the flow speed of the ions increases as a function of heliocentric distance.  Another important discovery from UVCS is the observation of broader than expected profiles of O VI 1032 and 1037.6 {\\AA} lines in polar coronal holes. Since these profiles are measured along the line of sight which is perpendicular to the radial direction at the point of closest approach to the Sun, their widths yield a measure of the $\\rm O^{5+}$ kinetic temperature in the direction approximately perpendicular to the magnetic field. UVCS observations also revealed that the O VI 1032/1037.6 intensity ratio reached a value of less than one at about 2.5 $R_s$ in coronal holes. If the kinetic temperatures parallel to the magnetic field were similar to the observed perpendicular temperatures, the O VI intensity ratio could not reach a value near unity. Hence, observations of broad line profiles and a decrease in the line ratios below a value of one led Kohl et al. (1997) to conclude that a significant kinetic temperature anisotropy exists for coronal $\\rm O^{5+}$ at a heliocentric distance of about 2.5 $R_s$.  Motivated by the UVCS observations, we illustrate in this Letter how the component of the kinetic temperature along the line of sight, for both isotropic and anisotropic temperature distributions, affect the derivations of outflow velocities from the O VI intensity ratio. However, by extending the effect of Doppler dimming and pumping of the O VI 1037.6 {\\AA} line to include another almost equally strong chromospheric C II 1036.3367 {\\AA} line in the vicinity of the C II 1037.0182 {\\AA} line (Warren et al. 1997), we show that the pumping by both C II lines can account for the O VI line intensity ratios recently observed by UVCS (Kohl et al. 1997). ",
        "conclusions": "This study demonstrates that the anisotropic kinetic temperatures of O$^{5+}$ implied by UVCS observations (Kohl et al. 1997) in the inner corona can significantly affect the Doppler dimming and pumping, and subsequently increase the observed O VI 1032/1037.6 line ratio expected for a given velocity. Another aspect of this study is that the pumping by both the C II 1037.018 {\\AA} and a second  C II line at 1036.3 {\\AA} is necessary to account for the observed intensity ratios. The inclusion of this second C II line, however, implies that the O$^{5+}$ ions can reach speeds of 400 km/s near 3 $\\rm R_s$. Consequently, the diagnostic techniques used in UVCS observations can be extended to much higher flow speeds than originally expected.  We also note that protons, which are strongly coupled to the neutral hydrogen, are expected to have the same flow speed as the neutrals (see Olsen et al. 1994, Allen et al. 1998). Hence, based on the recent UVCS observations, this study, together with the empirical model results by Kohl et al. (1998), shows that the O$^{+5}$ ions are flowing significantly faster than the protons around 3 $R_s$ in the fast solar wind. This is an indication that whatever mechanism is responsible for coronal heating and solar wind acceleration must preferentially accelerate minor ions such that they can flow faster than protons in the inner corona."
    },
    "9805/astro-ph9805370_arXiv.txt": {
        "abstract": "\\noindent Two dimensional calculations of the evolution of remnants generated by the strong mechanical energy deposited by stellar clusters in dwarf galaxies (M \\si $10^9 - 10^{10}$ \\msun) are presented. The evolution is followed for times longer than both the blowout time and the presumed span of energy injection generated by a coeval massive stellar cluster. The remnants are shown to end up wrapping around the central region of the host galaxy, while growing to kpc-scale dimensions. Properties of the remnants such as luminosity, size, swept up mass, and expansion speed are given as a function of time for all calculated cases.  The final fate of the swept-up galactic gas and of the matter processed by the central starburst is shown to be highly-dependent on the properties of the low density galactic halo. Superbubbles powered by star clusters, with properties similar to those inferred from the observations, slow down  in the presence of an extended halo to expansion speeds smaller than the host galaxy escape velocity. Values of the critical luminosity required for the superbubbles to reach the edge of the galaxies with a speed comparable to the escape speed are derived analytically and numerically. The critical luminosities are larger than those in the detected sources and thus, the superbubbles in amorphous dwarf galaxies must have already undergone blowout and are presently evolving into  an extended low density halo. This will inhibit the loss of the swept-up and processed matter from the galaxy. ",
        "introduction": "\\subsection{The ambient gas density distribution}  Massive galaxies ($M_{Gal}$ $\\geq$ 10$^{10}$ M\\sol), such as the Milky Way, are capable of retaining a hot ($T_{Gal}$ \\si a few 10$^6$K) and extended massive halo completely bound to their gravitational potential. This is frequently re-heated by a collection of  superbubbles, caused by aging OB associations, that blowout of the plane HI disk density stratification. Less massive galaxies, on the other hand, are unable to keep such massive and hot halos. Instead, the smaller the galaxy total mass, the cooler and less extended their possible resultant halos become. Thus these are usually idealized as a low density turbulent medium (or clouds) that extends up to a maximum radius R$_G$ at which its characteristic velocity dispersion exceeds the galaxy escape speed. As stated by Suchkov \\etal (1994, hereafter SBHL), \"in a real galaxy, much of the hydrostatic support of the interstellar matter is provided by the random motions of interstellar clouds\".  Following Li \\& Ikeuchi (1992), Tomisaka \\& Ikeuchi (1988, hereafter TI), Tomisaka \\& Bregman (1993, hereafter TB), and SBHL,  we model our dwarf galaxies with three different isothermal components. \\begin{equation} \\label{eq.1} \\rho_{g} = \\rho_{NM} +  \\rho_{IM} + <\\rho_{halo}> \\end{equation} where $\\rho_{NM}$, $\\rho_{IM}$ and $<\\rho_{halo}>$ are the gas densities of the neutral and ionized components and the mean density of the extended turbulent halo material, respectively. The model corresponds to the bound-cool type halo in the notation of Li \\& Ikeuchi (1992). The total gas pressure is then given by \\begin{equation} \\label{eq.2} P_{ext} = \\frac{1}{\\gamma} \\rho_{NM} C_{NM}^2 + \\frac{1}{\\gamma} \\rho_{IM} C_{IM}^2 + \\frac{1}{3} <\\rho_{halo}> C_{halo}^2, \\end{equation} where $C_{NM}$, $C_{IM}$ and $C_{halo}$ are the sound velocities of the neutral and ionized gas and the velocity dispersion of the extended halo component. $\\gamma$ is the ratio of specific heats.  Let $R_B$ be the maximum extent of the stars and dark matter, and $V_B$, the escape velocity at this surface. Following TI, and TB we assume that a fraction $(1-e^2)$ of the radial component of gravity is balanced by pressure gradients. Then the initial gas density distribution follows from the relation: \\begin{equation} \\label{eq.3} \\frac{\\rho_{g}}{\\rho_0} = \\alpha_{NM} \\exp{\\left[\\frac{\\gamma}{2}\\left(\\frac{V_B}{C_{NM}}\\right)^2 \\chi \\right]} + \\alpha_{IM} \\exp{\\left[ \\frac{\\gamma}{2}\\left(\\frac{V_B}{C_{IM}}\\right)^2 \\chi \\right]} + \\alpha_{halo} \\exp{\\left[\\frac{3}{2}\\left(\\frac{V_B}{C_{halo}}\\right)^2 \\chi \\right]} , \\end{equation} where $\\rho_0$ is the total gas density at the galactic center, $\\alpha_{NM}=\\rho_{NM,0}/\\rho_0$, $\\alpha_{IM}=\\rho_{IM,0}/\\rho_0$ and $\\alpha_{halo}=<\\rho_{halo,0}>/\\rho_0$ are the density ratios of the various components to the total density at the galactic center. The function $\\chi$ equals \\begin{equation} \\label{eq.4} \\chi = F(\\omega) - e^2 F(r) - (1-e^2) F(0), \\end{equation} where $r=\\sqrt{x^2 + y^2}$ and $\\omega = \\sqrt{x^2 + y^2 + z^2}$ are the cylindrical and spherical radii, and the function $F$ is defined by equation (\\ref{eq.9}). The rotation velocity of the ISM component comes from the relation: \\begin{equation} \\label{eq.4a} \\frac{V^2}{r} = e^2 (g_x^2 + g_y^2)^{1/2}, \\end{equation} where $g_x$ and $g_y$ are the x and y components of the gravity field.  We include a transition from a rotating disk component to a non-rotating spherical halo and radial gradients in the disk density distribution by assuming \\begin{equation} \\label{eq.5} e = e_0 / \\exp\\left[\\left(\\frac{z}{H_z}\\right)^2 + \\left(\\frac{r}{H_r}\\right)^2 \\right] , \\end{equation} where $H_r$ and $H_z$ are the characteristic scale heights along and perpendicular to the plane of the galaxy. Then (\\ref{eq.3}) is not self-consistent if a hydrostatic gaseous disk is assumed, but it is asymptotically  correct for a hydrostatic gaseous halo (TI, SBHL). Typical  dwarf galaxy  density distributions assuming a total mass M$_{B}$ = 10$^{10}$ M\\sol (models A and B) and 10$^9$ M\\sol (Models C) and a 10$\\%$ gaseous component are shown in the Figure 1 (see also Table 1). Note that the last contour drawed represents, in each case, the gaseous edge of the galaxy ($R_G$). \\ie material further out with a velocity dispersion $C_{halo}$ would exceed the escape velocity and thus it would not be bound to the system. Figure 2 gives a global impression of how the galactic ISM is distributed for different values of the assumed velocity dispersion of the extended component (also used to identify the models; see Table 1) and thus it shows how for larger values of $C_{halo}$, the smaller, or more compact that the bound gaseous component results. The value of P$_{ext}$($R_G$)/k varies in our models within a range 0.8 cm$^{-3}$ K (Model B60) - 5.4$\\times$ 10$^4$ cm$^{-3}$ K (Model A120). This pressure should balance the  extragalactic gas pressure and in fact covers the range often used in the  literature (eg. Babul \\& Rees 1992).  Several calculations were also made for different ratios of $M_{ISM}/M_{B}$, as well as for different dimensions  of the dark matter distribution (see section 5).  \\subsection{The gravitational field}  We have assumed that the main contribution to the gravitational field results from spherically symmetric stellar and dark matter components, disregarding the contribution of a stellar disk component, and the self-gravity of the gaseous matter. A King model was assumed for the background stars and a distribution inversely proportional to the square of the radius has been adopted for a dark matter halo: \\begin{equation} \\label{eq.6} \\rho_{tot}(\\omega) = \\frac{\\rho_1} {\\left[1 + \\left(\\frac{\\omega}{R_1} \\right)^2 \\right]^{3/2}} + \\frac{\\rho_2} {1 + \\left(\\frac{\\omega}{R_2} \\right)^2}, \\end{equation} where $R_1$ and $R_2$, $\\rho_1$ and $\\rho_2$ are characteristic scales and densities of the star and dark matter components, respectively.  The total mass of these components may be expressed as \\begin{eqnarray} \\label{eq.7} & & \\hspace{-0.5cm} M_B = 4 \\pi \\int_0^{R_B} \\rho_{tot}(\\omega)\\omega^2{\\rm d}\\omega = \\nonumber \\\\[0.2cm] & & \\hspace{-0.5cm} M_1 \\left[\\ln{(x_B + \\sqrt{1+x_B^2})} - \\frac{x_B}{\\sqrt{1+x_B^2}} \\right] + M_2 \\left[y_B - \\arctan{(y_B)}\\right], \\end{eqnarray} where $x_B = R_B/R_1$, $M_1 = 4 \\pi \\rho_1 R_1^3$, $y_B = R_B/R_2$, $M_2 = 4 \\pi \\rho_2 R_2^3$. We fix a constant at the gravity potential by assuming that it becomes Newtonian at the galactic boundary $R_B$. Then the gravity potential corresponding to the adopted total density distribution is expressed as follows: \\begin{equation} \\label{eq.8} \\Phi(r,z) = -\\frac{V_B^2}{2} F(\\omega), \\end{equation} where function $F(\\omega)$, is defined by the expression (see also SBHL) \\begin{eqnarray} \\label{eq.9} & & \\hspace{-0.5cm} F(\\omega) = 1 + \\frac{R_B}{R_2} \\frac{M_2}{M_{tot}}\\left[ \\frac{1}{2}\\left(\\ln{(1+y_B^2)} - \\ln{(1+y_{\\omega}^2)}\\right) + \\frac{\\arctan{(y_B)}}{y_B} - \\frac{\\arctan{(y_{\\omega})}}{y_{\\omega}}\\right] + \\nonumber \\\\[0.2cm] & & \\hspace{-0.5cm} \\frac{R_B}{R_1} \\frac{M_1}{M_{tot}}\\left[ \\frac{\\ln{(x_{\\omega}+\\sqrt{1+x_{\\omega}^2}})}{x_{\\omega}} - \\frac{\\ln{(x_B+\\sqrt{1+x_B^2}}}{x_B} \\right], \\quad \\omega \\le R_B, \\\\[0.2cm] & & \\hspace{-0.5cm} F(\\omega) = \\frac{R_B}{\\omega}, \\quad \\omega > R_B, \\nonumber \\end{eqnarray} where $x_{\\omega} = \\omega/R_1$, $y_{\\omega} = \\omega/R_2$. The value of $V_B$ is given by \\begin{equation} \\label{eq.10} V_B = \\sqrt{\\frac{2GM_B}{R_B}}, \\end{equation} whereas the escape velocity at the current position $(r,z)$ is \\begin{equation} \\label{eq.11} V_{esc} = V_B\\sqrt{F(\\omega)}. \\end{equation}  The $x,y,z$ components of gravity for $\\omega \\le R_B$ derived from (\\ref{eq.8}) and (\\ref{eq.9}) are \\begin{eqnarray} \\label{eq.12} & & \\hspace{-0.5cm} g_{x}=- \\frac{G M_1}{\\omega^2} \\frac{x}{\\omega} \\left[\\ln{(x_{\\omega}+\\sqrt{1+x_{\\omega}^2})} - \\frac{x_{\\omega}}{\\sqrt{1+x_{\\omega}^2}} + \\frac{M_2}{M_1}(y_{\\omega}-\\arctan{y_{\\omega}})\\right], \\\\[0.2cm] & & \\hspace{-0.5cm} g_{y}=- \\frac{G M_1}{\\omega^2} \\frac{y}{\\omega} \\left[\\ln{(x_{\\omega}+\\sqrt{1+x_{\\omega}^2})} - \\frac{x_{\\omega}}{\\sqrt{1+x_{\\omega}^2}} + \\frac{M_2}{M_1}(y_{\\omega}-\\arctan{y_{\\omega}})\\right], \\\\[0.2cm] & & \\hspace{-0.5cm} g_{z}=- \\frac{G M_1}{\\omega^2} \\frac{z}{\\omega} \\left[\\ln{(x_{\\omega}+\\sqrt{1+x_{\\omega}^2})} - \\frac{x_{\\omega}}{\\sqrt{1+x_{\\omega}^2}} + \\frac{M_2}{M_1}(y_{\\omega}-\\arctan{y_{\\omega}})\\right]. \\end{eqnarray} For larger distances $\\omega > R_B$ they are simply \\begin{eqnarray} \\label{eq.13} & & \\hspace{-0.5cm} g_{x}=- \\frac{GM_B}{\\omega^2} \\frac{x}{\\omega}, \\\\[0.2cm] & & \\hspace{-0.5cm} g_{y}=- \\frac{G M_B}{\\omega^2} \\frac{y}{\\omega}, \\\\[0.2cm] & & \\hspace{-0.5cm} g_{z}=- \\frac{G M_B}{\\omega^2} \\frac{z}{\\omega}. \\end{eqnarray} ",
        "conclusions": "Based on the scarce but definite observations of amorphous dwarf galaxies, which show that the HI distribution outgrows by large factors the optical size of these galaxies, we have constructed realistic models which account for an extended cool halo component bound to the system. Models of amorphous galaxies with an M$_{ISM} \\sim 10^9 \\, M_{\\odot}$ and an energy input up to $10^{41}$ erg  s$^{-1}$, as inferred from the observations, lead to blowout from the disk matter stratification, allowing the superbubbles to vent their hot matter into the galaxy halo. The outflowing gas immediately searches and follows the steepest density gradients affecting a large almost spherical volume. We have shown here that in the case of small systems, superbubbles wrap around and envelope the densest central region of the host galaxy. Such remnants are strongly decelerated as they evolve into the halo, sufficiently to remain bound to their host galaxy.   Our calculations confirm the results of De Young and Gallagher (1990), and De Young and Heckman (1994) during the early superbubble evolution. However, our ultimate conclusions are different as the final fate of the remnants  basically depends on the properties of the galactic halo and on the total energy of the starburst. The presence of the dark matter component also enhances the ultimate retention of the processed material. Under these circumstances total retention (i.e. expansion speeds $\\le$ escape velocity) of the remnants and thus of the matter processed by the exploding massive stars is unavoidable for galaxies with an ISM mass $\\ge$ than a few $10^9$ M$_{\\odot}$ and an energy input rate of up to 10$^{41}$ erg s$^{-1}$.  We have also derived analytical expressions for the critical luminosity required for a superbubble to be able to reach the outskirts of a galaxy and expulse the processed matter into IGM. The critical limit has been cross-checked with several numerical calculations for a large range of galaxy masses, dimensions and thus surface escape velocities.  >From the calculations, we then conclude that the thoroughly studied dwarf galaxies with a gaseous mass \\si 10$^9$ M\\sol and presenting giant remnants produced by the mechanical energy ($L \\leq 10^{41}$ erg s$^{-1}$) from massive stars (see Marlowe \\etal, 1995; Martin 1996; and Meurer 1994) are most likely to retain their ISM and their processed matter. The observed remnants present dimensions smaller than a few  kpc and expansion speeds  \\si 50 \\kms and thus based on our calculations, we predict that they have already undergone blowout from their disk matter distribution and that  presently are evolving into the extended low density halo of their parent galaxy. Our results also predict that these superbubbles will be ultimately dispersed by the turbulent motions in the extended galactic halo.  The consequences of these conclusions are important in a variety of fields and we shall analyse these in a forthcomming communication. Clearly, the kind of dwarf galaxies that we have analysed, are not responsible for the contamination of the IGM. Dwarf galaxies with an ISM mass of the order of 10$^9$ M\\sol will retain their metals unless they undergo an overwhelming burst of star formation much larger than those presently observed. We thus have to understand why do they present low metalicities. The secret must probably lies on the physics of mixing (see \\GTT 1996)."
    },
    "9805/astro-ph9805222_arXiv.txt": {
        "abstract": "We find the distribution of coalescence times, birthrates, spatial velocities, and subsequent radial offsets of coalescing neutron stars (NSs) in various galactic potentials accounting for large asymmetric kicks introduced during a supernovae.  The birthrates of bound NS--NS binaries are quite sensitive to the magnitude of the kick velocities but are, nevertheless, similar ($\\sim 10$ per Galaxy per Myr) to previous population synthesis studies.  The distribution of merger times since zero-age main sequence is, however, relatively insensitive to the choice of kick velocities.  With a median merger time of $\\sim 10^8$ yr, we find that compact binaries should closely trace the star formation rate in the Universe.  In a range of plausible galactic potentials (with $M_{\\rm galaxy} \\age 3\\times 10^{10} M_\\odot$) the median radial offset of a NS--NS mergers is less than 10 kpc. At a redshift of $z=1$ (with $H_0 = 65$ km s$^{-1}$ Mpc$^{-1}$ and $\\Omega = 0.2$), this means that half the coalescences should occur within $\\sim 1.3$ arcsec from the host galaxy.  In all but the most shallow potentials, ninety percent of NS--NS binaries merge within 30 kpc of the host.  We find that although the spatial distribution of coalescing neutron star binaries is consistent with the close spatial association of known optical afterglows of gamma-ray bursts (GRBs) with faint galaxies, a non-negligible fraction ($\\sim 15$ percent) of GRBs should occur well outside ($\\age 30$ kpc) dwarf galaxy hosts.  Extinction due to dust in the host, projection of offsets, and a range in interstellar medium densities confound the true distribution of NS--NS mergers around galaxies with an observable set of optical transients/galaxy offsets. ",
        "introduction": "The discovery of an X-ray afterglow by BeppoSAX (Costa et al.~1997; Piro, Scarsi, \\& Butler~1995) and subsequently an optical transient associated with gamma-ray burst (GRB) 970228 (van Paradijs et al.~1997) led to the confirmation of the cosmological nature of GRBs (Metzger et al.~1997).  The broadband optical afterglow has been modeled relatively successfully (M\\'esz\\'aros \\& Rees 1993; Wijers, Rees, \\& M\\'esz\\'aros~1997; Waxman 1997; Waxman, Kulkarni \\& Frail 1998) as consistent with an expanding relativistic fireball (Rees \\& M\\'esz\\'aros 1994; Paczy\\'nski \\& Rhoads 1993; Katz 1994; M\\'esz\\'aros \\& Rees 1997; Vietri 1997; Sari, Piran, \\& Narayan 1998; Rees \\& M\\'esz\\'aros 1998).  Still, very little is known about the nature of the progenitors of GRBs, and, for that matter, their hosts.  Broad-band fluence measures and the known redshifts of some bursts implies a minimum (isotropic) energy budget for GRBs of $\\sim 10^{52-53}$ ergs (Metzger et al.~1997; Kulkarni et al.~1998; Djgorovski et al.~1998).  The log $N$-log $P$ brightness distribution, the observed rate, $N$, of bursts above some flux, $P$, versus flux, indicates a paucity of dim events from that expected in a homogeneous, Euclidean space.  With assumptions of a cosmology, source evolution and degree of anisotropy of emission, the log $N$-log $P$ has been modeled to find a global bursting rate. Assuming the bursts are non-evolving standard candles Fenimore \\& Bloom 1995 found $\\sim 1$ bursts events per galaxy per Myr (GEM) to be consistent with the observed log $N$-log $P$.  More recently, Wijers et al.~1998 (see also, Totani et al.~1998; Lipunov, Postnov, \\& Prokhorov 1997) found the same data consistent with GRBs as standard candles assuming the bursting rate traces the star-formation rate (SFR) in the Universe; such a distribution implies a local burst rate of $\\sim 0.001$ GEM and a standard peak luminosity of $L_0 = 8.3 \\times 10^{51}$ erg s$^{-1}$ (Wijers et al.~1998).  Given the energetics, burst rate and implied fluences, the coalescence, or merger, of two bound neutron stars (NSs) is the leading mechanism whereby gamma-ray bursts are thought to arise (Paczy\\'nski 1986; Goodman~1986; Eichler et al.~1989; Narayan, Paczy\\'nski, \\& Piran 1992).  One quantifiable prediction of the NS--NS merger hypothesis is the spatial distribution of GRBs (and GRB afterglow) with respect to their host galaxies. Conventional wisdom, using the relatively long--lived Hulse-Taylor binary pulsar as a model, is that such mergers can occur quite far ($\\age 100$ kpc) outside of a host galaxy.  Observed pulsar (PSR) binaries with a NS companion provide the only direct constraints on such a populations, but the observations are biased both towards long lived systems, and systems that are close to the Galactic plane.  The merger rate of NS--NS binaries has been discussed both in the context of gravitational wave-detection and GRBs (eg.~Phinney 1991; Narayan et al.~1991; Tutukov \\& Yungelson 1994; Lipunov et al.~1995). Recently Fryer, Burrows, and Benz (1997), Lipunov, Postnov, \\& Prokhorov (1997), Portegies Zwart \\& Spreeuw (1996) studied the effect of asymmetric kicks on birthrates of NS--NS binaries, but did not quantify the spatial distribution of such binaries around their host galaxies.  Tutukov \\& Yungelson (1994) discussed the spatial distribution of NS--NS mergers but neglected asymmetric kicks and the effect of a galactic potential in their simulations.  Only Zwart \\& Yungelson (1998) have discussed the maximum travel distance of merging neutron stars including asymmetric supernovae kicks.  It is certainly of interest to find the rate of NS--NS coalescences {\\it ab initio} from population synthesis of a stellar population. This provides an estimate of beaming of GRBs, assuming they are due to NS--NS mergers, and hence an estimate of probable frequency of gravitational wave sources, providing a complementary rate estimate to those of Phinney 1991 and Narayan et al.~1991, which are based on long lived NS-PSR pairs only and are very conservative. It also provides an estimate of how the NS--NS mergers trace the cosmological star formation rate (SFR) of the Universe, if mean formation rates and binarity of high mass stars are independent of star formation environments such as metallicity.  Here we concentrate on estimating the spatial distribution of coalescing NS--NS binaries around galaxies.  To do so, both the system velocity and the interval between formation of the neutron star binary and the merger through gravitational radiation is found by simulation of binary systems in which two supernovae occur.  We explore the effects of different asymmetric kick amplitudes, and the resultant birthrates and spatial distribution of coalescing NS--NS binaries born in different galactic potentials.  In section 2 we briefly outline the prescription for our Monte Carlo code to simulate bound binary pairs from an initial population of binaries by including the effect of asymmetric supernovae kicks.  In section 3 we outline the integration method of NS--NS pairs in various galactic potentials.  Section 4 highlights the birthrates and spatial distributions inferred from the simulations.  Section 5 concludes by discussion the implications and predictions for gamma-ray burst studies. ",
        "conclusions": "A reconciliation with the expected distribution of presumed progenitors of GRBs and observed transient/host offsets is clearly required.  We find for all plausible galactic potentials that the median radial offset of a NS--NS merger is less than 10 kpc.  And in all but the most shallow potentials, ninety percent of NS--NS binaries merge within 30 kpc of the host. At a redshift of $z=1$ (with $H_0 = 65$ km s$^{-1}$ Mpc$^{-1}$ and $\\Omega = 0.2$), this means that ninety percent the coalescences should occur within $\\sim 4$ arcsec from the host galaxy.  Although the expected spatial distribution of coalescing neutron star binaries found herein is consistent with the close spatial association of known optical afterglows of gamma-ray bursts with faint galaxies, a non-negligible fraction ($\\sim 15$\\%) of GRBs {\\it should} occur well outside dwarf galaxy hosts if the NS--NS hypothesis is correct.  Otherwise, other models which keep progenitors closer to their host (eg.~BH--NS mergers, microquasars, or ``failed SN type Ib'') would be preferred.  As all the progenitor models mentioned are connected with high-mass stars, the true GRB afterglow rate as a function of redshift should trace the star-formation rate in the Universe.  However, environmental effects, such as dust obscuration, may severely bias the estimate of the true offset distribution.  Even in the NS--NS models where progenitors have a natural mechanism to achieve high spatial velocities, most will be closely connected spatially to their host. Redshifts derived from absorption in the afterglow spectra should be nearly always that of the nearest galaxy (Bloom et al.~1998b).  Rapid burst follow-up ($\\ale$ 1 hr), with spectra taken while the optical transients are bright should confirm some form of absorption from the host galaxy.  We have confirmed the strong dependence of birthrate of NS--NS binaries on kick velocity distribution and found the independence of the orbital parameters after the second supernova (and hence merger times and spatial velocity) on the choice of kicks.  The methodology herein can be extended to include formation scenarios of black holes. This could provide improved merger rate estimates for LIGO sources, and estimate the relative contribution of coalescences between neutron stars and low mass black holes to the event rate. Detailed modeling of the Milky Way potential would also allow predictions for the distribution of NS--PSR binaries observable in the Milky Way, which would provide an independent test of the assumptions made in these models.  \\vspace{-.5cm}"
    },
    "9805/astro-ph9805014_arXiv.txt": {
        "abstract": "The results of long-term surveys of radial velocities of cool Ap and Am stars are presented. There are two samples, one of about 100 Ap stars and the other of 86 Am stars. Both have been observed with the CORAVEL scanner from Observatoire de Haute-Provence (CNRS), France.  The conspicuous lack of short-period binaries among cool Ap stars seems confirmed, although this may be the result of an observational bias; one system has a period as short as 1.6 days. A dozen new orbits could be determined, including that of one SB2 system. Considering the mass functions of 68 binaries from the literature and from our work, we conclude that the distribution of the mass ratios is the same for the Bp-Ap stars than for normal G dwarfs.  Among the Am stars, we found 52 binaries, i.e. 60\\%; an orbit could be computed for 29 of them. Among these 29, there are 7 SB2 systems, one triple and one quadruple system. The 21 stars with an apparently constant radial velocity may show up later as long-period binaries with a high eccentricity. The mass functions of the SB1 systems are compatible with cool main-sequence companions, also suggested by ongoing spectral observations. ",
        "introduction": "The question here is that of the possible influence of a binary companion upon chemical peculiarity: does an Ap or Am star ``need'' a companion, or conversely, does it need to be single? If there is a companion, what is its nature? Finally, fundamental parameters like mass, radius and age are better constrained for the components of binary stars than for isolated ones.  The first systematic search for spectroscopic binaries among Bp-Ap stars was made by Abt \\& Snowden (1973) on a sample of 62 northern bright stars. They found a rate of binaries of only 20 percent among the magnetic Ap stars and of 43 percent among HgMn stars (the rate among normal stars is $47\\pm 5$ percent among normal B0 to M stars according to Jaschek \\& Gomez 1970). Aikman (1976) studied 80 HgMn stars (instead of 15 for Abt \\& Snowden) and confirmed a normal rate with 49 percent of binaries. The most recent and complete review is that of Gerbaldi et al. (1985), while North (1994) published some new results on cool Ap stars obtained with the CORAVEL spectrovelocimeter.  On the basis of a sample of 25 Am stars, Abt (1961) concluded that all are binaries, while Abt \\& Levy (1985) confirmed the high rate of binaries in a larger sample but also found some stars to be apparently single. For both Ap and Am stars, only bright objects had been examined, so it was judged worthwhile to extend the survey to fainter and more numerous ones.  This contribution is kind of a progress report on a survey of cool Ap and Am stars made with the CORAVEL scanner in the northern hemisphere. ",
        "conclusions": ""
    },
    "9805/astro-ph9805364_arXiv.txt": {
        "abstract": "We present optical identification and high-resolution spectroscopy of ROSAT sources in the field of the old open cluster M~67. For the first time it is possible to analyze coronal and chromospheric activity of active stars in a solar-age cluster, and to compare it  with  field stars.\tROSAT observed the high X-ray luminosity tail of the cluster sources. In agreement with what expected from studies of field stars, most of the detected X-ray sources are binaries, preferably with short periods and eccentric orbits. In addition, several of the M~67 ROSAT sources have peculiar locations in the cluster colour-magnitude diagram. This is most likely due to rather complex evolutionary histories, involving the presence of mass transfer\t or large mass losses.  The X-ray luminosity of the sources does not scale with the stellar parameters in an obvious way. In particular, no relationship is found between coronal emission and stellar magnitude or binary period. The Ca II K chromospheric flux from most of the counterparts is in excess to that of single stars in the cluster by one order of magnitude. The X-ray luminosity of the sources in the old M~67 is one order of magnitude lower than the most active active binaries in the field, but comparable to that of the much younger binaries in the Hyades.   \\keywords { Stars: binaries, Stars: coronae, Stars: chromospheres, open clusters: individual: M~67} ",
        "introduction": "The study of chromospheric and coronal activity has progressed impressively in the last two decades, thanks to the launch of X-ray satellites. Coronal sources have been detected in X rays all over the cool part of the H-R diagram (Vaiana et al. 1980) and active binaries of the RS CVn type  have soon been recognized as powerful X-ray emitters (Walter et al. 1980).  With the advent of the ROSAT satellite and its all-sky survey, a complete X-ray study of known active binaries  has been possible (Dempsey et al. 1993a,b). A large number of previously unknown coronal active stars is becoming available (see e.g. Metanomski et al. 1998), with interesting consequences on our understanding of the population of young stars in the solar neighborhood (Guillout et al. 1998).  While the general framework of dependence of X-ray emission on stellar rotational velocity (and then stellar age) has been established long time ago (Pallavicini et al. 1981), there still remain many uncertainties about which stellar parameters affect the coronal emission in late-type stars.  We know, for instance, that old stars may preserve a high level of X-ray emission if they are in binary systems, where a high rotational velocity can be maintained through tidal interaction, but it is not clear which role stellar mass, radius, orbital period and eccentricity play in determining the level of activity.  One of the most relevant limitations is that, since most studies are performed in field samples, it is difficult to determine precisely the characteristics of the studied objects. Stars with different (but poorly determined) masses, ages and possibly evolutionary histories are often compared.  The study of clusters, where ages, masses, and evolutionary status of the counterparts of X-ray sources can be well established, is a necessary step forward. In this framework, not only ROSAT observations of young clusters have provided new insight to the discussion on the age-activity relationship (see e.g. Randich and Schmitt 1995), but for the first time it has also been possible to detect coronal sources in old clusters, i.e. with ages comparable with the Sun.  With the ROSAT observations of M~67 and other intermediate age clusters, Belloni et al.(1993, 1996, 1997) allowed to extend the study of the evolution of coronal activity to systems with age up to $\\sim$ 6 Gyrs. By examining these coeval samples, these observations allow the study of active coronae from a different, unique perspective: it is possible to investigate which stars (or stellar systems) show the highest activity level in samples  so old that the emission from single, solar-type stars is expected to be very low.  The ROSAT observations require a follow up at different wavelengths, in order to :  \\begin{itemize}  \\item Confirm the identification of the optical counterparts. In these clusters the stellar density is rather high, and more than one source may be contained in the ROSAT error box. As an example of the relevance of this issue, the reader can compare the results of the present work (summarized in Table 1) with those of Belloni, Verbunt \\& Schmitt (1993), based only on X-ray and general optical information, but not on a detailed follow up.  \\item Study in more detail the characteristics of the sources, their binary nature, X-ray and chromospheric emission.  \\item Compare the observed characteristics with those of field stars.  \\end{itemize}  In this work, we present the identification and the study of the ROSAT counterparts of the old open cluster M~67 (Belloni, Verbunt \\& Schmitt 1993). A second, longer ROSAT pointing to M~67 has been performed (Belloni, Verbunt \\& Mathieu in preparation): where possible, the results from this observation have been included in the present paper.  M~67 is a very interesting target, not only because its age and metallicity are similar to those of the Sun, but also because the cluster has been subject to several photometric and spectroscopic studies which have led to detailed membership determinations (Sanders 1977, Girard et al. 1989), a large amount of data on binary stars (Latham et al. 1992), photometric variables, W UMa candidates (Gilliland et al. 1991), and blue stragglers (Mathys 1991).  This large body of  ancillary data makes M~67 one of the best studied clusters; one AM Her system and a hot white dwarf belonging to the cluster have already been confirmed within this identification project (Pasquini et al. 1994a). ",
        "conclusions": "The aim of this work is to discuss the nature of the coronal sources belonging to M~67 detected in the ROSAT observation of BVS. For this purpose we concentrate mostly on the firmly identified sources.  \\subsection{Which parameters determine X-ray emission in a 4Gyr Old Cluster?}  It is important to remember that the observations by BVS could only detect the high X-ray luminosity tail of the cluster. The longer, more recent ROSAT observations reveal more sources,  confirming that cluster X-ray emitters might exist at lower levels than the one analyzed here (Belloni, Verbunt \\& Mathieu, in preparation).  Of the 8 known members of Table 2, six are either known binaries or show clear signs of duplicity, like photometric modulation or displaced position in the colour-magnitude diagram of the cluster. Of the remaining two, S1077 is reported by BVS as a multiple system with short period, while S364 is not included in the current list of binaries. Jones and Smith (1984) found anomalous DDO colours for this star, but they concluded that this was probably due to due measurement uncertainties. Once again, we stress that this identification is only considered as `possible'.  {\\it The first conclusion is that in M~67 the strongest coronal sources are binaries}.\tThis was  expected from the results on field stars and from the age-coronal activity relationships: single, late type stars with  ages comparable to the Sun are not expected to emit X-ray in excess to $\\sim$10$^{29}$ erg/s in luminosity.  In Figure 3 the colour-magnitude diagram of the cluster is shown, with the X-ray identifications marked. It is striking to notice that a large variety of cases exists: one Blue Straggler (S1082), one peculiar object (S1063), evolved `normal' binaries (S1077, S999), main sequence binaries (S1019, S972), one Red Straggler (S1040), and possibly a red giant (S364). Among field stars such a comparison can hardly be done, due to the uncertainty in the fundamental stellar parameters.  \\begin{figure} \\psfig{figure=7307fig3.ps,width=6.0cm,silent=} \\caption[3]{ Colour-magnitude diagram of M~67. Filled squares show the positions of the firm X-ray counterparts. S759 (non member) and S364 (possible counterpart) are indicated respectively as an open square and a star. } \\end{figure}  {\\it The second conclusion is therefore that among the M~67 sources there exists a large variety in evolutionary status and composition of binary systems. In particular, several of the strong X-ray emitters are found among objects having peculiar location in the colour magnitude diagram.}  This variety makes it difficult to understand which parameters determine the X-ray emission in these binaries. From the study of field active binaries, it emerged that the main parameter determining X-ray luminosity is stellar radius. Therefore, the strongest X-ray emitters would be expected to be among the most luminous systems (Dempsey et al. 1993b). On the contrary, the most luminous X-ray sources in M~67 span over a range in visible luminosity of at least 4 magnitudes (cfr. Table 1). Although  S972, which has the lowest X-ray luminosity, is also the faintest established member of the cluster in the optical, other stars (like S1019 or S1063) show the highest X-ray emission, while being relatively faint in the optical.  {\\it The third conclusion is therefore that, in this coeval sample, the highest X-ray luminosity is not given by the most luminous stars.}  Another interesting point is the dependence of X-ray emission on the orbital (or rotational) period, since X-ray luminosity scales with stellar rotational velocity (Pallavicini et al. 1981), which in turn is related to stellar radius and orbital period. In most cases (and for main sequence stars), it is expected that short period binaries are synchronized, i.e. they have equal orbital and rotational periods. It is also expected that synchronization happens before orbit circularization (Zahn 1977). All 4 stars for which the rotational modulation is known, as inferred from the photometric variability, have rotational periods of less than 10 days. For S1063, only the orbital period is known, but this star may have a special history (see Section 5.3). Pending more data on the rotational period of the other objects, we can say that  {\\it Not only the X-ray sources in M~67 are binaries, but also most of them have short rotational periods.}  On the other hand, do all  M~67 short period systems show strong X-ray (or chromospheric) activity ? In M~67, 11 binaries have a measured orbital period shorter than 16 days (Latham et al. 1992), a canonical value for the definition of RS CVn systems, but most of them were not detected in the ROSAT\tobservations. Three of them (S1272, S1284, S1224) where out or at the very border of the ROSAT field; one (S999) was detected, but the remaining (S1045, S1234, S1024, S986, S1009, S1070 S1014, S810) were contained in the ROSAT field but not detected. These systems are likely active, but with X-ray luminosities L$_x \\le$ 30.3 (corresponding to the sensitivity of BSV),  as it is suggested by the likely detection of some of these stars in the deeper pointing (S1045, S1024, S1234 and S1070, Belloni, Verbunt \\& Mathieu, in preparation). Since some of the non-detected binaries, like S986, S1009 and S810, have periods of 10 days or shorter, the data collected up to now show that  {\\it Among the binaries of M~67, a short orbital period does not necessarily imply a high X-ray  luminosity}.   Since we are dealing mostly with evolved stars, for which short periods can coexist with eccentrical orbits, eccentricity could be a relevant parameter, because the non-synchronization would allow the possibility of having a rotational period shorter than the orbital period. Of the detected binaries, S1063 and S999 have highly eccentric orbits, but S1040 has an eccentricity comparable with 0. This is surprising, because if the photometric variability detected by Gilliland et al. (1991) really represents the rotational period, than this star would be circularized but not synchronized. On the other hand, S1040 had probably a very complex history, having had mass transfer in the past (Landsman  et al. 1997). Considering this star as a special case\twe, could argue that non-circular binaries are favored among the strongest X-ray emitters. However, in the list of Latham et al. (1992), there are five short-period binaries with eccentricities significantly different from 0 (S1284, S1272, S1234, S1224, S1014). Two of them (S1014 and S1234) were in the ROSAT field of view but were not detected. Therefore, it seems that also a high eccentricity does not represent {\\it per se} a condition for strong X-ray emission.  \\subsection{Chromospheric Activity}  The chromospheric flux of the Sun in the Ca II K line is of 3-5$\\times$10$^5$ erg cm$^{-2}$sec$^{-1}$ (Pasquini et al. 1988). For the main sequence G stars in M~67, being metallicity and age comparable with those of the Sun, values comparable with this are expected.  Recently, Dupree et al. (in preparation) studied the chromospheric emission in the Ca II lines for a sample of M~67 giants, using calibrations and spectra of similar quality to the ones presented here. In Figure 4 the Ca II chromospheric fluxes are plotted as a function of the (V-R) colour, both for the stars in Dupree et al. (open squares) and for the X-ray counterparts of Table 2 (filled squares). The ROSAT stars have a level of chromospheric activity one order of magnitude higher than the optically-selected giants in Dupree et al. The only exception is  S364, which fits extremely well within normal giants of similar spectral type. For this reason, despite the reasonable agreement with the X-ray position, this star is considered as a doubtful counterpart.  \\begin{figure} \\psfig{figure=7307fig4.ps,width=9.0cm,silent=} \\caption[4]{ Ca II chromospheric fluxes vs. (V-R) colour for M67 single stars (Dupree et al., in preparation: open symbols) and ROSAT stars (Table 2 of this work: filled symbols)} \\end{figure}   In order to perform an unbiased analysis of the relationship between chromospheric and coronal activity, we have to use indicators which express similar quantities, namely fluxes at the stellar surface. We computed the radii of the stars by using the same Barnes-Evans (1976) relationship used to calibrate the Ca II data: Log$\\phi$ = 0.4874 - 0.2V$_o$ + 0.858(V-R). Although we know that the resulting radii are probably incorrect (for instance, because the stars are implicitly assumed to be single), the fact that the same relationship is used for computing the Ca II and X-ray fluxes minimizes the presence of possible systematic effects in the comparison. We assumed an absorption A(V)=0.17  and a distance of 785 parsecs (Janes 1984). The resulting diameters and radii are given in Tables 2 and 3 (in units of milliarcseconds and 10$^{10}$ cm respectively). In Figure 5a, Log(F$_x$) is given as a function of Log F$_k$. The relationship is rather scattered and mostly the presence of the (doubtful) S364 hints to  the presence of a trend.  Since the sample contains systems whose evolutionary status is very different from each other, to further investigate this point we plot in Figure 5b the ratio between the X-ray and chromospheric fluxes versus apparent magnitude. Fainter (i.e. higher gravity) stars have much higher coronal to chromospheric flux ratios than more luminous (i.e. lower gravity) stars.  Although the number of objects is rather low, it appears that dwarfs and giants follow different trends, with dwarfs having higher F$_x$ for a given F'$_k$. This fact has two possible explanations:  \\begin{itemize}  \\item For comparable chromospheric fluxes, higher-gravity stars are more efficient in heating their coronae than lower-gravity stars. This could indicate the presence of different coronal structures between dwarfs and evolved stars.  \\item The assumption that the Ca II K fluxes are representative of the whole chromospheric losses may not apply  when comparing stars of different luminosity. Giants could for instance have a different balance in the different chromospheric lines than dwarfs.  \\end{itemize}  \\begin{figure} \\psfig{figure=7307fig5.ps,width=9.0cm,silent=} \\caption[5]{ 5a): Comparison between Ca II and X-ray fluxes for the X-ray counterparts. 5b) Coronal to Ca II fluxes as a function of stellar magnitude.} \\end{figure}   \\subsection{S1063 and S1113}  Two objects deserve a special attention, due to their peculiar position in the colour-magnitude diagram: S1063 and S1113.  S1113 was was outside the field of the observation by BSV. Because of the strong chromospheric activity observed, we would expect it to show rather strong X-ray emission, which is indeed detected in the new ROSAT pointing (cfr. Table 3).  These two stars are located below the giants branch: they are as red as single subgiants, but almost one magnitude fainter. S1063 is a known eccentric binary with a period of 18.3 days, while S1113 is a short period (2.82 d) circular binary (Latham et al. 1992, Mathieu et al. in preparation). The two stars are classified as  members in the proper motion studies of Sanders (1977) and Girard et al. (1989), with a probability higher than 90$\\%$. Their peculiar position in the colour-magnitude diagram and their high level of coronal and chromospheric activity make these two object very interesting. It is not possible to simply combine two M~67 stars and obtain the magnitudes and colours of S1063 and S1113. Some mass exchange, or large mass losses in the past history of the systems, possibly still going on, look unavoidable. The high-resolution Ca II spectrum of S1113 shows neither direct evidence of duplicity, nor strong asymmetries in the Ca II core typical of strong mass losses, but not much can be derived with only one optical spectrum for this star.  We stress that the conclusion that these two stars have suffered a special evolutionary history is made possible only by the fact that we they are members of a cluster (and therefore we can firmly position them in the colour magnitude diagram) and by the detailed optical follow-up. How many such systems exist among field binaries? Note that, in absence of detailed studies (e.g. accurate determinations of mass, metallicity and gravity), similar systems in the field cannot be distinguished by otherwise 'normal' RS CVn binaries. A more detailed study in other clusters and possibly the analysis of Hipparcos parallaxes of active binaries will help in understanding how common these systems are. Investigation of their binarity and orbital synchronization will also be crucial to model their possible evolution with time (see i.e. the discussion in Stepien 1995).   \\subsection{Comparison with ROSAT observations of field RS CVn}  One of the aims of this study is the possibility of comparing for the first time active stars in an old cluster with the field population of RS CVn. Dempsey et al. (1993a,b) studied the ROSAT detections of known RS CVn. Their sample, taken from the catalogue of Strassmeier et al. (1988), contains binaries of a large variety of ages, masses and  periods. We selected all the stars from Dempsey et al. (1993a) with known distance. We computed X-ray luminosities by converting the PSPC count rates (see also Dempsey et al. 1994) using the same conversion factor used for the M~67 stars. The distribution in luminosity is given in Figure 6 for M~67 and field RS CVn separately. Although the M~67 sources do overlap well with the main body of field RS CVn's, their emission is not as high as the most active RS CVn systems.  \\begin{figure} \\psfig{figure=7307fig6.ps,width=9.0cm,silent=} \\caption[6]{ ROSAT luminosity for the M~67 sources (broken line) and for RS CVn binaries with known distances (continuous line)} \\end{figure}  {\\it Figure 6 shows  that the high X-ray luminosity tail of binaries in the 4 Gyr old cluster M~67 is about 10 times lower than the high luminosity tail of field RS CVn's.}  The reasons for this difference are at the moment not very clear: it could be an effect due to the old age of the cluster and/or to the evolutionary status of the sources, or to a statistical effect caused by the fact that we can only sample a few hundred stars within the cluster. Identifications in clusters of different ages, as well as a detailed analysis of the strongest sources in the field, will help in understanding this open question. It is interesting to note that in the analysis of the much younger Hyades, Stern et al. (1995) found that binaries are the strongest X-ray emitters in this cluster. Similarities with M67 exist also in that the strongest X-ray source in the Hyades (V471 Tau) is a peculiar system. Finally, Hyades binaries have X-ray luminosities similar or lower than those observed among the  M67 sources.  This could indicate that as far as the high-luminosity tail of the X-ray luminosity function is concerned, age is not the primary parameter to determine the X-ray emission in binaries, as pointed out also by Ottmann et al. (1997) in their analysis of Pop II binaries.  We would therefore argue that the extremely active i.e. (L$_x$ $\\sim$ 10$^{32}$ erg cm$^{-2}$ sec $^{-1}$) RS CVn systems seem to be very rare and possibly limited to quite exceptional cases. Metanomski et al. (1998), in their study of hundred stars identified over a large area of the ROSAT All-Sky Survey, find results similar to ours: no such a high luminosity coronal source is indeed present in their sample."
    },
    "9805/astro-ph9805152_arXiv.txt": {
        "abstract": "The X--ray light curve and broad band spectral properties of the millisecond X--ray pulsar/burster SAX J1808.4-3658 are reported. In the course of RXTE observations in April--May 1998 the 3--150 keV luminosity of the source decreased  by a factor of $\\sim 100$ from the peak value of  $\\sim 9\\times 10^{36}$ erg/s (for a 4 kpc distance). However, the spectrum was remarkably stable and maintained a roughly power law shape with a photon index of $\\sim 2$ without a strong high energy cut-off below 100 keV, similar to that sometimes observed in the low spectral state of X--ray bursters. An approximation of the averaged spectrum with an exponentially cut-off power  law with a superimposed reflected component yields the 90\\% lower limit on the e-folding energy $E_{f}\\ga 270$ keV. We speculate that Comptonization on the bulk motion in the radiation dominated shock might be a possible mechanism of spectral formation. The decaying part of the X--ray light curve was cut off abruptly at  luminosity  $\\sim {\\rm few}\\times 10^{35}$ erg/s. Such behavior might be due to centrifugal inhibition of accretion (transition to the ``propeller'' regime) as well as to disk instability. In either case an upper limit on the neutron star magnetic field strength is $B\\la{\\rm few}\\times 10^7$ Gauss. ",
        "introduction": "The transient X-ray source SAX J1808.4--3658 was discovered in September 1996 using the Wide Field Camera (WFC) aboard BeppoSAX (\\cite{saxpaper}).  The source was not detected by WFC in August and October 1996 or by  MIR--KVANT/TTM (N.Alexandrovich \\& K.Borozdin, private communication)  7 years before in March and August--September 1989 with the upper limits  of $\\sim   1/50$ and $\\sim 1/10$ of the September 1996 peak flux.  Two Type I X--ray bursts were detected by WFC in September 1996. The analysis of the X--ray bursts gave 4 kpc source distance for which the 0.4--10 keV  persistent luminosity was $6\\cdot10^{36}$ erg/s. According to ASM/RXTE data the 1996 outburst lasted for $\\sim20$ days.   Recently the Rossi X--Ray Timing Explorer detected a new outburst from SAX J1808.4--3658 (Marshall et al. 1998). Coherent pulsations with a period of 2.49 msec (Wijnands \\& van der Klis 1998) and Doppler shift due to orbital motion with a period of $2^h$ (Chakrabarty \\& Morgan 1998) were discovered. The analysis of the RXTE/HEXTE data from  April 11 and 13  revealed a remarkable power law spectrum of the source in the 15--120 keV band with a photon index of $\\approx 2$ (\\cite{hexte}). The optical counterpart of SAX J1808.4--3658 brightened by $>3.4^m$ and reached $m_V=16.6$ on April 18 (\\cite{optical1}). \\cite{optical2} detected a roughly sinusoidal modulation of $0.12^m$  in the V--band with a $2^h$ binary system period.     \\begin{table*} \\small \\caption{SAX J1808.4--3658. Parameters of spectral approximation of PCA and HEXTE data by a power law model \\label{fit}} \\tabcolsep=0.19cm \\begin{tabular}{lcrrccccccc} \\hline Date& Time, UT&\\multicolumn{2}{c}{Expos.$^1$,sec}&\\multicolumn{2}{c}{PCA}&\\multicolumn{2}{c}{HEXTE}&\\multicolumn{3}{c}{PCA+HEXTE}\\\\ 1998& &\\multicolumn{3}{l}{~PCA~HEXTE$^2$~~~~~~~$\\alpha$}&$F_{3-25keV}^3$&$\\alpha$&$F_{20-150 keV}^{3,4}$&$\\alpha$&$F_{3-150keV}^{3,5}$&$\\chi^2_r$\\\\ \\hline 11/04 & 19:50--21:44 &  2974& 2041 & $2.00\\pm0.01$ & $245.4\\pm0.8$ &$2.13\\pm0.09$ &$138.8\\pm9.6$ &$2.01\\pm0.01$ &$450.0\\pm3.1$&1.50\\\\ 13/04 & 01:52--02:23 &  1375&  949 & $1.97\\pm0.01$ & $224.2\\pm0.8$ &$2.13\\pm0.12$ &$153.9\\pm4.2$ &$1.97\\pm0.01$ &$423.4\\pm3.3$&1.36\\\\ 16/04 & 17:19--23:01 &  9266& 8620 & $1.95\\pm0.01$ & $145.3\\pm0.5$ &$2.03\\pm0.06$ &$ 99.9\\pm4.8$ &$1.96\\pm0.01$ &$279.3\\pm1.8$&1.43\\\\ 17/04 & 01:42--04:39 &  5675& 4018 & $1.95\\pm0.01$ & $140.8\\pm0.5$ &$2.10\\pm0.08$ &$ 91.5\\pm5.7$ &$1.96\\pm0.01$ &$270.4\\pm1.9$&1.41\\\\ 18/04 & 03:09--09:19 & 13980& 9179 & $1.96\\pm0.01$ & $123.6\\pm0.4$ &$2.06\\pm0.05$ &$ 82.9\\pm3.7$ &$1.96\\pm0.01$ &$235.9\\pm1.6$&1.48\\\\ 18/04 & 12:34--01:03 & 25285&17056 & $1.96\\pm0.01$ & $118.5\\pm0.4$ &$2.08\\pm0.05$ &$ 77.5\\pm3.1$ &$1.97\\pm0.01$ &$225.5\\pm1.5$&1.54\\\\ 20/04 & 21:03--23:09 &  5120& 3557 & $1.98\\pm0.01$ & $ 98.1\\pm0.3$ &$2.11\\pm0.13$ &$ 59.5\\pm6.4$ &$1.99\\pm0.01$ &$182.7\\pm1.5$&1.49\\\\ 23/04 & 15:53--23:21 & 16493&11135 & $2.01\\pm0.01$ & $ 81.0\\pm0.3$ &$2.03\\pm0.09$ &$ 55.7\\pm4.0$ &$2.01\\pm0.01$ &$147.8\\pm1.1$&1.59\\\\ 24/04 & 16:08--23:22 &  8547&10426 & $2.03\\pm0.01$ & $ 77.6\\pm0.3$ &$2.12\\pm0.10$ &$ 47.1\\pm3.8$ &$2.04\\pm0.01$ &$138.5\\pm1.1$&1.49\\\\ 25/04 & 14:23--21:41 & 15045&10120 & $2.04\\pm0.01$ & $ 73.1\\pm0.3$ &$2.08\\pm0.11$ &$ 45.7\\pm4.0$ &$2.05\\pm0.01$ &$129.5\\pm1.0$&1.60\\\\ 26/04 & 16:01--23:28 & 15819&10748 & $2.05\\pm0.01$ & $ 48.5\\pm0.2$ &$2.11\\pm0.16$ &$ 27.6\\pm3.6$ &$2.05\\pm0.01$ &$ 85.6\\pm0.8$&1.36\\\\ 27/04 & 14:30--19:31 &  8765& 6006 & $2.11\\pm0.02$ & $ 24.8\\pm0.2$ &$2.15\\pm0.44$ &$ 15.8\\pm5.1$ &$2.11\\pm0.01$ &$ 41.8\\pm0.6$&1.22\\\\ 29/04 & 14:24--18:58 &  8536& 5735 & $2.10\\pm0.02$ & $ 19.9\\pm0.1$ &$2.39\\pm0.61$ &$ 12.7\\pm5.2$ &$2.10\\pm0.02$ &$ 33.8\\pm0.6$&1.04\\\\ 02/05$^6$ & 01:26--02:59 &3259&1988 & $1.86\\pm0.13$ & $  1.6\\pm0.1$ &$ 2.0^7  $ &$ \\la4.7^8$ &$1.86\\pm0.13$ &$  3.4\\pm0.6$&0.70\\\\ 03/05$^6$ & 19:10--21:01 &4072&2790 & $2.29\\pm0.08$ & $  3.3\\pm0.1$ &$ 2.0^7       $ &$ \\la4.6^8$ &$2.29\\pm0.07$ &$  4.9\\pm0.4$&0.70\\\\ \\hline \\end{tabular} The errors are 90\\% confidence; $^1$ -- dead time corrected; $^2$ -- on source, sum of two clusters of detectors; $^3$-- the energy flux in $10^{-11}$ erg/sec/cm$^2$; $^4$-- original HEXTE normalization $^5$-- the HEXTE normalization was adjusted to match the PCA data; $^6$-- low source count rate ($\\sim 5-10\\%$ of the PCA background, see text); $^7$-- fixed; $^8$-- 90\\% upper limit; \\end{table*} ",
        "conclusions": "Theories of the spectral formation in accreting pulsars with a strong magnetic field predict a change of the physical conditions near the surface of the neutron star at $L_X\\sim{\\rm few}\\times 10^{36}$ erg/sec. In the case of X--ray bursters strong spectral changes are associated with the transition from the soft to the hard spectral state at $L_X\\sim 10^{36}-10^{37}$ erg/sec (Fig.\\ref{spezoo}). We therefore expected to observe spectral evolution as luminosity decreased by a factor of  $\\sim 100$. Remarkably no significant spectral changes were detected (Table \\ref{fit}, Fig.\\ref{spezoo}). This, together with the detection of X--ray pulsations of roughly the same relative amplitude, implies that essentially the same mechanism is responsible for the spectral formation in a broad luminosity range.    \\begin{figure*} \\hbox{ \\epsfxsize=8.7cm \\epsffile{ae201_f2left.ps} \\hspace{2mm} \\epsfxsize=8.7cm \\epsffile{ae201_f2right.ps} } \\caption{{\\em Left:} The  averaged (April 11--25) spectrum of SAX J1808.4--3658 and the spectra of several X--ray bursters and black hole Cyg X--1 in the low spectral state. {\\em Right:} The spectra of SAX J1808.4--3658 obtained on April 11--13, 26--29 and May 2-3 in comparison with the spectra of other X--ray bursters at different luminosity: 4U1608--522 (Spring 1996 outburst) and Aquila X--1 (Spring 1997). All spectra were obtained using RXTE public data. \\label{spezoo}} \\end{figure*}    \\subsection{A rotation powered pulsar ?}  Stability and shape of the spectrum of SAX J1808.4--3658  suggests considering a possibility, that the 1998 outburst was powered by the standard pulsar emission mechanism. For  magnetic dipole emission the field strength required to generate X--ray flux $F_X$ [erg/s/cm$^2$] is $ B\\sim 2.2\\cdot 10^9 D_{\\rm kpc} R_6^{-3} P_{2.5}^4 \\left(\\frac{\\beta}{0.1}\\right)^{-1/2} \\left(\\frac{F_X}{4\\cdot 10^{-9}}\\right)^{1/2} $ Gauss, where $D_{\\rm kpc}$ is the distance in kpc, $R_6$ -- the neutron star radius in $10^6$ cm, $P_{2.5}$ -- rotation period divided by 2.5 msec, $\\beta$ -- the efficiency of the production of X--ray photons. On the other hand  the 1996 outburst was accretion powered as the Type I X--ray bursts were detected. For the accretion to proceed uninhibited by the centrifugal force due to the rotating magnetosphere of the neutron star the  accretion rate would be $ \\dot{M}\\ga 5\\cdot 10^{17} B^2_9 R^{16/3}_6 M_{1.4}^{-5/3} P_{2.5}^{-7/3}$ g/s i.e. $\\dot{M}\\ga 2.5\\cdot 10^{18} D_{\\rm kpc}^2$ for $B\\sim 2.2\\cdot 10^9~D_{\\rm kpc}$ Gauss. Therefore, independent of the source distance, the 2--10 keV flux observed in 1996 ($2\\cdot 10^{-9}$ erg/s/cm$^2$) would correspond to  $\\sim 10^{-4}$ of $\\dot{M}c^2$. Such a low efficiency of production of X--ray photons is unlikely for an accreting neutron star. Moreover, such a high $\\dot{M}$  would lead to a steady nuclear burning of the accreting matter, which contradicts the detection of the Type I X--ray bursts. Thus,  the standard pulsar emission mechanism can be ruled out.      \\subsection{The X--ray light curve.}  The X--ray light curve is  similar to the outburst profiles of two very different types of accreting sources -- the neutron star soft X--ray transients (e.g. \\cite{aql,aqlsax}) and Type B outbursts in dwarf Novae (\\cite{cvrev}). An abrupt cut-off of the light curve is also common for the black hole X--ray transients (\\cite{bhrev}). Two different interpretations of the abrupt decline of luminosity at the end of the outburst are possible in the case of SAX J1808.4--3658: \\begin{enumerate} \\item Closure of the centrifugal barrier at low accretion rate when the magnetosphere reaches the corotation radius $R_{co}\\approx 31 M_{1.4}^{1/3} P_{2.5}^{2/3}$ km and the source enters the ``propeller'' regime (\\cite{propeller}).  The cut-off luminosity provides an estimate of the magnetic field as: $B\\sim 3\\cdot 10^7 M_{1.4}^{1/3} R_{6}^{-8/3} P_{2.5}^{7/6} \\left(\\frac{L_X}{10^{35}}\\right)^{1/2} $ Gauss. \\item Disk instability. The decay time scale is defined by the propagation of the cooling wave. A rough estimate for the SAX J1808.4--3658 binary system parameters is $\\tau_d\\sim 1.3\\left(\\frac{0.1}{\\alpha}\\right)$ days, $\\alpha$ -- viscosity parameter (\\cite{cvrev}) which approximately accounts for the observed decay time scale (Fig.\\ref{lc}). The above expression for B is an upper limit in this case.  \\end{enumerate}   \\subsection{Spectral formation.}   A possible mechanism resulting in the formation of a power law spectrum with a slope $\\sim 2$ might be Comptonization on the bulk motion in a radiation dominated shock (Blandford \\& Payne 1981, \\cite{rdshock}). If, as indicated by the X--ray pulsation, there is non negligible disk--magnetosphere interaction, the accreting matter is funneled onto the polar caps with $\\sim$free fall velocity. If the radiation energy flux $q$ exceeds the critical value $q_c\\sim$local Eddington flux, the radiation dominated shock is formed near the surface of the neutron star at the polar regions (\\cite{rdshock1,bs}), where $\\ga 2/3$ of the accretion energy is released. A naive estimate -- assuming that the magnetosphere is close to corotation, the magnetic field inside the magnetosphere is not distorted and the in-falling matter fills the magnetic funnel -- gives the value of the critical luminosity $L_c\\sim {\\rm few}\\times 10^{36} - 10^{37}$ erg/sec, which exceeds by more than an order of magnitude the minimum luminosity observed for  SAX J1808.4--3658. However, an accurate calculation of $q/q_c$ requires detailed knowledge of the disk--magnetosphere interaction and the accretion flow pattern inside the magnetosphere. It is important to note that in order for this mechanism to be responsible for the spectral formation in SAX J1808.4--3658, self adjustment of the accretion funnel to maintain $q/q_c\\sim 1$ in a broad range of the accretion rates is required.   Comptonization on the bulk motion in the radiation dominated shock near the stellar surface in the equatorial area might also occur in the case of the ``classical'' X--ray bursters if the neutron star is smaller than the last marginally stable orbit (\\cite{smallns1,smallns}). The power law spectra with a slope $\\sim 2$ up to $\\sim 100$ keV are indeed observed for some of the ``classical'' X--ray bursters at low luminosity, $L_X\\la 10^{36}$ erg/sec (Fig.\\ref{spezoo})."
    },
    "9805/astro-ph9805258_arXiv.txt": {
        "abstract": "High resolution, deep imaging surveys are instrumental in setting constraints on semi--analytical structure formation models in Cold Dark Matter (CDM) cosmologies.  We show here that the lack of unresolved B--band ``dropouts'' with $V>25$ mag in the Hubble Deep Field (HDF) appears to be inconsistent with the expected number of quasars if massive black holes form with a constant universal efficiency in all CDM halos.  To reconcile the models with the data, a mechanism is needed that suppresses the formation of quasars in halos with circular velocities $v_{\\rm circ} \\lsim 50-75~{\\rm km~s^{-1}}$. This feedback naturally arises due to the photoionization heating of the gas by the UV background.  We consider several alternative effects that would help reduce the quasar number counts, and find that these can not alone account for the observed lack of detections.  If reddening by dust can be neglected at early epochs, consistency with the optical data also requires that the luminous extent of dwarf galaxies at high redshifts be larger than a few percent of the virial radii of their dark matter halos, in order not to overpredict the number of point--like B--band dropouts. Future deep observations in the $J$ and $H$ bands with NICMOS might reveal several $z\\gsim 5$ objects per field or provide even stronger constraints on the models than existing $B, V$, and $I$ data. ",
        "introduction": "One of the primary goals of models for structure formation in the Universe is to explore the formation process of luminous objects such as quasars or galaxies, and to explain the shape and evolution of their observed luminosity functions (LFs).  Recent theoretical work has focused on the formation of galaxies within the framework of hierarchical structure formation in CDM cosmologies, employing either three--dimensional numerical simulations (Gnedin \\& Ostriker 1997; Zheng~et~al.~1997), or semi--analytic modeling (Kauffman, Guiderdoni \\& White 1994; Cole~et~al.~1994; Baugh et al. 1997; Somerville \\& Primack 1998; Haiman \\& Loeb 1997, hereafter HL97).  These calculations have been quite successful in reproducing the observed Tully--Fisher relation in the local Universe, the faint galaxy LFs out to moderately high redshifts $z\\lsim3$, and the recently discovered clustering of Lyman--break galaxies at $z\\sim3$ (Baugh et al. 1997; Governato et al. 1998; Steidel~et~al.~1998).  Hierarchical models were also applied to the more complex problem of quasar formation (Efstathiou \\& Rees 1988; Haehnelt \\& Rees 1993; Small \\& Blandford 1992), and were shown to reproduce the observed evolution in the quasar LF at redshifts $z\\ga 2$ under a minimal set of simplifying assumptions (Haiman \\& Loeb 1998b, hereafter HL98; see also Haehnelt, Natarajan \\& Rees 1998).  The above models were not tested in detail at redshifts much higher than $z\\sim3$ because of the lack of related observational data.  Such tests are desirable from the theoretical point of view, since the model predictions are expected to be more robust at higher redshifts when the stars and quasars first condensed out of the well-defined set of initial conditions in the early Universe. The exponentially low abundance of collapsed objects at very early times reduces the probability for complicated merger events.  Moreover, the chemical and thermal conditions inside objects which assembled out of the pristine intergalactic gas are better defined before metal enrichment and thermal feedback from multiple generations of stars started to affect the appearance of quasars and galaxies.  To date, the deepest optical image of the Universe is the Hubble Deep Field (HDF, Williams~et~al.~1996), which surveyed a $2.3^\\prime\\times2.3^\\prime$ field in four bandpasses down to $\\sim 29$ magnitude.  Although this image has revealed a large number of galaxies of different morphological types, a surprisingly large fraction of the sky was found to remain dark at the HDF detection limit (e.g., Vogeley 1997). The Next Generation Space Telescope (NGST), scheduled for launch in 2007, is planned to have a sensitivity of 1 nJy in the 1--3.5$\\mu$m range, and should be able to detect globular cluster scale objects out to $z\\sim10$ (see Smith \\& Koratkar 1998).  The properties of faint extended sources in the HDF (Madau et al. 1996) have been well described by detailed semi--analytic galaxy formation models (Baugh et al. 1997).  On the other hand, the HDF has revealed only a handful of faint unresolved sources (M\\'endez et al. 1996), and none with the colors expected for high redshift objects.  The purpose of the present paper is to show that the lack of point--like sources in the HDF can be used to constrain models for the formation of cosmic structure at redshifts higher than previously possible ($z\\gsim4$).  The observational data and color--selection criteria are summarized in \\S~2 and \\S~3, and the simplest versions of the semi-analytic CDM models are outlined in \\S~4. In \\S~5, we compute the optical and infrared quasar number counts in these models, and show that the predictions from standard scenarios without a UV feedback appear to be inconsistent with the observed lack of high--$z$ sources.  We also show that models with a UV feedback are marginally consistent with the data.  In \\S~6, we describe possible modifications of these models that could reduce the predicted counts.  In \\S~7, we discuss the expected galaxy number counts. In \\S~8, we present our conclusions and argue that future infrared observations in the $J$ and $H$ bands with NICMOS may in principle be used to place even stronger constraints on this type of scenarios. ",
        "conclusions": "The observed lack of faint, unresolved, and red sources in the HDF down to $V\\approx 29$ requires models that are consistent with the data at the $2\\sigma$ level to predict no more than $2$ point--like high-$z$ objects to this limiting magnitude. Given that the observations are $\\sim50$\\% complete, this appears to be marginally consistent with the predicted number ($\\sim4$) of high redshift quasars in the simplest hierarchical models, which suppress the formation of black holes in CDM halos with circular velocities $v_{\\rm circ} \\leq 50~{\\rm km~s^{-1}}$.  Without a photoionization feedback, these models predict $\\sim 8$ quasars from $z>3.5$. In models without a UV feedback, the predicted quasar counts might be reduced by early structure formation (so that $M_* \\gsim 10^{11}{\\rm M_\\odot}$ by $z=3$), but only if one postulates that new black holes do not form in halos below the nonlinear mass--scale, $M_*(z)$. An example for such a model is the $\\Omega_0=0.35$ open CDM model, with a power spectrum slope $n=1.15$ and normalization $\\sigma_{8h^{-1}}=0.87$.  The $z>3.5$ number counts can be further reduced by introducing a black hole formation efficiency $M_{\\rm bh}/M_{\\rm halo}$ that depends on the halo mass as $\\propto M_{\\rm halo}^{0.5}$. Similarly, a shift of the big blue bump component to shorter wavelengths in the spectra of faint quasars could also help reduce the predicted counts.  Combined with a minimum circular velocity of $v_{\\rm circ}= 50~{\\rm km~s^{-1}}$, either of these latter two effects could reduce the number counts to less than $\\sim 2$ objects.  On the other hand, we find that without a photoionization feedback, neither of these effects would be sufficient by itself to explain the lack of detections.  The models predict $\\sim$ 5--35 times more $z>3.5$ galaxies than quasars at $V=29$. If the star formation rates in these galaxies are adjusted to produce an average Ly$\\alpha$ forest metallicity of $10^{-2}{\\rm Z_\\odot}$ by $z=3$, then more than $10\\%$ of all galaxies could appear unresolved at the 0.1\\H{} pixel size of the WFC, possibly leading to another discrepancy with the observed lack of detections. The situation changes if the star formation rate is lowered by an order of magnitude (normalized to produce a $10^{-3}{\\rm Z_\\odot}$ average IGM metallicity at $z\\sim 3$), since then almost all of the predicted $z>3.5$ galaxies are expected to be extended rather than point sources.  Future observations at longer infrared wavelengths such as the $J$ and $H$ infrared bands, would be better suited to the study of the $z\\gsim4.5$ universe as they allow bigger volumes to be probed. In Figure~\\ref{fig:qvijh}, we show the expected number counts of quasars in the $V$, $I$, $J$, and $H$ bands for the model which is consistent with the optical HDF data. An increasingly larger number of objects are predicted to be detected in the $I$, $J$, and $H$ bands as these are affected by intervening absorption only at extreme redshifts. These number counts could be somewhat reduced since faint, compact WFC sources lose signal--to--noise when observed by NICMOS, due to the poorer angular resolution of the infrared images. Figure~\\ref{fig:svijh} shows the expected number of galaxies in the infrared bands.  At 30 mag, the gain in observing galaxies redward of the $I$-band is not as significant as for quasars, because the intrinsic flux of unreddened galaxies, unlike those of quasars, decreases as a function of increasing wavelength at $z\\sim4$.  The redshifts of either type of sources can be obtained using a straightforward analogy to the $B$ dropout technique, applied to the $I-J$ vs.  $J-H$ color--color diagram shown in Figure \\ref{fig:ijjh}. Future deep observations with NICMOS in these bands of a field comparable in size to the HDF would be able to either reveal tens of ultra--high redshift quasars, or else place tighter constraints on quasar models than currently possible using $V$ and $I$ data."
    },
    "9805/astro-ph9805291_arXiv.txt": {
        "abstract": "We present the results of a study that uses numerical simulations to interpret observations of tidally disturbed satellites around the Milky Way.  When analysing the simulations from the viewpoint of an observer, we find a break in the slope of the star count and velocity dispersion profiles in our models at the location where unbound stars dominate.  We conclude that `extra-tidal' stars and enhanced velocity dispersions observed in the outskirts of Galactic satellites are due to contamination by stellar debris from the tidal interaction with the Milky Way.  However, a significant bound population can exist beyond the break radius and we argue that it should not be identified with the tidal radius of the satellite.  We also develop and test a method for determining the mass loss rate from a Galactic satellite using its extra-tidal population.  We apply this method to observations of globular clusters and dwarf spheroidal satellites of the Milky Way, and conclude that a significant fraction of both satellite systems are likely be destroyed within the next Hubble time.  Finally, we demonstrate that this mass loss estimate allows us to place some limits on the initial mass function (IMF) of stars in a cluster from the radial dependence of its present day mass function (PDMF). ",
        "introduction": "Accurate estimates of mass loss rates from the satellite galaxies and globular clusters in orbit around the Milky Way offer the possibility of better understanding the dynamical history of the Galaxy.  When integrated over the entire population, such measurements would tell us about the current accretion rate onto the Galaxy.  This in turn would indicate whether the Galaxy has grown significantly through the gradual process of tidal stripping and disruption of its satellites. The direct inference of mass loss rates from individual objects would place new constraints on detailed analytic or numerical dynamical models of these systems.  In addition, the measurements could be used to identify satellites which are likely to have well-populated streams of tidal debris associated with them.  In particular, the formation, evolution and ultimate fate of the Galactic globular clusters is a long standing puzzle in astrophysics. Dynamically, globular clusters are clean and relatively simple systems, yet on close inspection they reveal a wealth of complex behaviour, including relaxation and evaporation from internal dynamical interactions and mass loss in response to the tidal field of the Milky Way \\cite{me97,el87}.  Ideally we would like to be able to assess the importance of each of these processes for the evolution of individual globular clusters from observations.  Of particular interest for mass-segregated, relaxed clusters is the differential mass loss of stars due to the different relative densities of low and high mass stars at various radii, and the consequence of this loss for the evolution of the present day mass function (PDMF) away from the initial mass function (IMF). Understanding the form of the IMF of the current globulars is relevant for theories of star formation and globular cluster formation \\cite{fr77}, and the related issue of how the Galaxy was assembled. The mass function of a stellar system $dN/dM$ is defined as the number of stars $dN$ in the mass interval $(M,M+dM)$.  A common functional form chosen to approximate this distribution is \\begin{equation} {dN \\over dM} \\propto M^{-(x+1)} \\label{mf} \\end{equation} where $x$ is called the {\\it mass function index}.  The PDMF $x$ of a cluster has been found to be related to the Galactocentric radius $R_{\\rm GC}$ and height above the Galactic plane $Z$ of the cluster (Capaccioli, Ortolani \\& Piotto 1991). Using simple semi-analytic models Capaccioli, Piotto \\& Stiavelli (1993) showed that these correlations could be reproduced by evolving a population of globular clusters which all have the same initial IMF in the tidal field of the Milky Way.  Similarly, in a comparison of several globular clusters Piotto, Cool \\& King (1997) found the luminosity function of NGC 6397 to be much flatter than M15, M30 and M92, and concluded that this could be due to the extreme dynamical evolution implied by its orbit calculated from its proper motion \\cite{d96}.  In addition McClure et al. (1986) reported a dependence of the PDMF on metallicity, and Djorgovski, Piotto \\& Capaccioli (1993) successfully incorporated this together with the trends in $R_{\\rm GC}$ and $Z$ into a trivariate analysis.  In contrast, in a recent study of Pal 1, Rosenberg et al. (1997) found its mass function to be inconsistent with the trends in $R_{\\rm GC}$, $Z$, and metallicity and concluded that this could either be due to evaporation of low mass stars, or an intrinsically different IMF.  One step towards solving the puzzle of just how far the PDMF of a cluster differs from its IMF is determining how to interpret the signatures of tidal effects in observations of globular clusters and other satellites.  In the past, the limiting radii of Galactic satellites (which are assumed to correspond to the tidal radii $r_{\\rm tide}$ imposed by the Milky Way) have been interpreted using King's (1962) tidal radius formula \\begin{equation} r_{\\rm tide}=R_{\\rm GC} \\left({m_{\\rm sat} \\over 3 M_{\\rm GC}}\\right)^{1\\over 3}, \\label{rtide} \\end{equation} where $R_{\\rm GC}$ is the distance of the satellite from the Galactic centre, $m_{\\rm sat}$ is the mass of the satellite, and $M_{\\rm GC}$ is the mass of the Galaxy enclosed within $R_{\\rm GC}$.  For the observed $r_{\\rm tide}$ this formula has been used to: (i) find $m_{\\rm sat}$ from the current $R_{\\rm GC}$ \\cite{fl83}; (ii) estimate the pericentre of the satellite's orbit given the value for $m_{\\rm sat}$ implied by its internal dynamics (Oh, Lin \\& Aarseth 1995; Irwin \\& Hatzidimitriou 1995 -- hereafter IH85); and (iii) compare the Galaxy's tidal field with the satellite's internal field \\cite{ih95}. Such analyses have been complicated by the detection of `extra-tidal' stars around both dwarf spheroidal galaxies (IH95; Kuhn, Smith \\& Hawley 1996) and globular clusters (Grillmair et al. 1995 -- hereafter G95; for a summary of current observations of both Galactic and extra-galactic globular clusters, see Grillmair 1998), and these morphological features have been shown to be consistent with those seen in simulations of tidally disrupting systems \\cite{ola95,g95}.  Hut \\& Djorgovski (1992) also estimated destruction rates of globular clusters by finding the extent to which their distribution of half-mass relaxation times deviates from a power law and interpreting this difference in terms of evolution of the system.  An alternative to these very direct interpretations of observations is to build models that include some representation of all the expected dynamical effects on globular clusters and integrate the cluster's evolution over the lifetime of the Galaxy \\cite{aho88,ch96,go97,mw97,v97}. The success of these semi-analytic methods (as with those in the previous paragraphs) rests upon the representation of the physics involved, and there is some disagreement about the level of sophistication required for accurate predictions  -- one example is the theory of tidal shocking which has received renewed attention \\cite{w94,ko95,jhw98,gho98}. However, all these studies predict that the Milky Way's globular cluster system is currently undergoing significant evolution.  In this paper, we return to the earlier philosophy of trying to understand how much we can learn directly from observations of tidal signatures.  We approach the problem by using numerical simulations to assess how successfully such features can be interpreted.  Our study differs from earlier ones in several ways: (i) we include a mass spectrum, and begin with mass segregated models; (ii) there is a one-one representation of stars in e.g.  globular clusters; and (iii) the tidal field of a full Milky Way model is included rather than idealised as giving rise to a spherical tidal boundary or being represented by a one-component potential.  In our analysis, we explicitly distinguish between the bound and unbound stars in a satellite, which allows us to identify the characteristics of the extra-tidal population.  We use this analysis to develop and test methods for quantifying the mass loss rate and evolution of the mass function of clusters and satellites directly from current observations.  We present the simulation and analysis methods used throughout the paper in \\S 2.  The simulations are used to provide an overview of the characteristics of evolution in a tidal field in \\S 3.  We analyse our models from the viewpoint of an observer \\S 4.  In \\S 5, the interpretations developed in \\S 4 are applied to observations and used to measure the rate of destruction of the Milky Way's satellite system.  We summarise our results in \\S 6. ",
        "conclusions": "We summarise our main conclusions as follows: \\begin{enumerate} \\item Observations of extra-tidal stars and enhanced velocity dispersions in the outskirts of clusters can be attributed to tidally stripped material.  The star counts become dominated by unbound stars at the point where the slope of the surface density profile changes or the dispersion reaches a minimum.  This should not be identified with the tidal radius of the cluster since the edge of the bound population can still lie significantly beyond this radius. \\item The mass loss rate from a Galactic satellite can be directly estimated from the population of extra-tidal stars within a few of its tidal radii, the orbital time period and our line of sight. \\item Using current observations we calculated the mass-loss rate from dwarf spheroidal satellites and globular clusters and found that both systems will undergo significant evolution in the next Hubble time.  Future observations should be able to place much stronger limits on these destruction rates. \\item Mass-loss estimate together with measurements of the local and global PDMFs of a cluster can place limits on its IMF.  Our review of the literature suggests that it is currently observationally feasible to carry out this program. \\end{enumerate}"
    },
    "9805/astro-ph9805184_arXiv.txt": {
        "abstract": "s{A project to construct a 17 m diameter imaging air \\v{C}erenkov telescope, called the MAGIC Telescope, is described. The aim of the project is to close the observation gap in the $\\gamma$-ray sky extending from 10 GeV as the \\textit{highest} energy measurable by space-borne experiments to 300 GeV, the \\textit{lowest} energy measurable by the current generation of ground-based \\v{C}erenkov telescopes. The MAGIC Telescope will incorporate several new features in order to reach the very low energy threshold. At the same time the new technology will yield an improvement in sensitivity in the energy region where current \\v{C}erenkov telescopes are measuring by about an order of magnitude.} ",
        "introduction": "\\label{sec:intro} Currently the observations of electromagnetic radiation from astrophysical sources and high energy phenomena in the Universe are restricted to energies outside an observation gap extending from $\\approx $ 10 GeV to $\\approx $ 300 GeV. From the continuation of current measurements into this energy region we expect to find hints or answers to important physics questions in astrophysics, cosmology, and particle physics.  The reason for the observation gap lies in the flux limitation of space-borne instruments for $\\gamma$-astronomy due to very small collection areas ($\\mathcal{O}$(0.1) m$^2$) which limits the measurements to energies \\textit{below} 10 GeV. At the same time the not yet optimized technology of ground-based instruments for $\\gamma$-astronomy, i.e., the existing Imaging Air \\v{C}erenkov Telescopes (IACTs), limits the detection threshold of $% \\gamma$-showers to energies \\textit{above} 300 GeV.  In order to close this gap by the comparitively cheap ground-based technique, the 17 m $\\oslash$ MAGIC Telescope has been designed and important components have been developed during the last 2 years \\cite{magic}. ",
        "conclusions": "The 17 m diameter MAGIC Telescope has been designed to measure $\\gamma$-rays with energies above 10 GeV. Most of the new technology for this telescope has been developed during the last 2 years \\cite{magic}. Using innovative elements it will be possible to close the existing observation gap in the electromagnetic spectrum for about 1\\% (!) of the cost of a satellite experiment, which until now was believed to be necessary in order to do measurements in this energy domain. At the same time the sensitivity in the energy region of current \\Cerenkov telescopes will be improved by up to an order of magnitude. The innovative elements of the MAGIC Telescope technology will very likely be the basis for all IACTs of the next generation. We estimate the hardware-price of the telescope to be around 3.5 M\\$. The construction time will be 2.5 - 3.5 years."
    },
    "9805/nucl-ex9805008_arXiv.txt": {
        "abstract": "The neutron capture cross section of $^{26}$Mg was measured relative to the known gold cross section at thermonuclear energies using the fast cyclic activation technique. The experiment was performed at the 3.75 MV Van-de-Graaff accelerator, Forschungszentrum Karlsruhe. The experimental capture cross section is the sum of resonant and direct contributions. For the resonance at $E_{\\rm{n,lab}} = 220~{\\rm{keV}}$ our new results are in disagreement with the data from Weigmann {\\it et al}. An improved Maxwellian averaged capture cross section is derived from the new experimental data taking into account $s$- and $p$-wave capture and resonant contributions. The properties of so-called potential resonances which influence the $p$-wave neutron capture of $^{26}$Mg are discussed in detail. ",
        "introduction": "\\label{sec:intro} Neutron capture processes of neutron--rich light isotopes play an important role in astrophysical scenarios ranging from the so--called inhomogeneous Big Bang models to nucleosynthesis in stellar helium and carbon burning stages. In Inhomogeneous Big Bang Models the high neutron flux induced primordial nucleosynthesis bridges the mass 5 and mass 8 gap~\\cite{app88}. Subsequent neutron capture processes of neutron--rich isotopes may even lead to a primordial r--process~\\cite{cow91,rau94}. The efficiency for the production of heavy elements in such a scenario depends sensitively on the respective neutron capture rates for these light isotopes. Therefore the neutron capture cross sections have to be determined over a wide energy range up to about 1 MeV.  In massive Red Giant stars magnesium isotopes are mainly the products of hydrostatic carbon and neon burning. $^{25}$Mg and $^{26}$Mg have also appreciable abundances in the outer carbon layer as a result of the reactions $^{22}$Ne($\\alpha$,$\\gamma$)$^{26}$Mg, $^{22}$Ne($\\alpha$,n)$^{25}$Mg, and $^{25}$Mg(n,$\\gamma$)$^{26}$Mg~\\cite{woo95}. He--shell burning in low mass asymptotic giant branch (AGB) stars has been proposed as the site for the main component of the s--process \\cite{str95,gal97}. In AGB stars $^{26}$Mg is likely to be made by $^{22}$Ne($\\alpha$,$\\gamma$)$^{26}$Mg, as well as by $^{25}$Mg(n,$\\gamma$)$^{26}$Mg, by the decay of $^{26}$Al, and by $^{26}$Al(n,p)$^{26}$Mg with radioactive $^{26}$Al being synthesized through the previous H--burning shell. For the above discussed stellar scenarios neutron capture rates need to be known in the energy range between about 5 and 200 keV.  The reaction rate for $^{26}$Mg(n,$\\gamma$)$^{27}$Mg that is being investigated in this work is low through its small cross section. The reaction $^{26}$Mg(n,$\\gamma$)$^{27}$Mg is of interest in stellar nucleosynthesis, because (i) $^{26}$Mg is one of the most abundant isotopes in the cosmos, (ii) all Mg--isotopes show a strong deviation from the normal 1/v--behavior of their cross sections, and (iii) of the interplay between neutron production and captures for Mg--isotopes, and with the concurrential radioactive $^{26}$Al.  The cross section for neutron capture processes is dominated by the nonresonant direct capture process (DC) and by contributions from single resonances which correspond to neutron unbound states in the compound nucleus (CN). For calculating the different reaction contributions we used a simple hybrid model: the nonresonant contributions were determined by using a direct capture model, the resonant contributions were based on determining the resonant Breit Wigner cross section. In the case of broad resonances additional interference terms have to be taken into account which were neglected in the present work because the relevant resonances are relatively narrow.  In this work we also discuss the effects of potential resonances in direct capture. Potential resonances are well known in nuclear processes (see Ref.~\\cite{ohu91} and references therein), but their influence in numerical direct--capture calculations has to our knowledge not been investigated before in great detail. The reaction $^{26}$Mg(n,$\\gamma$)$^{27}$Mg provides an excellent example for discussing the artifacts such resonances can produce in direct--capture calculations.  Up to now the neutron capture cross section of $^{26}$Mg was measured at thermal energies in several experiments (see Refs.~\\cite{wal92,sel70,ryv70}, and references therein). However, at thermonuclear energies only the resonance properties of four resonances in $^{26}$Mg(n,$\\gamma$)$^{27}$Mg were determined by Weigmann {\\it et al.} \\cite{weig76}, and in that work the Maxwellian averaged capture cross section $< \\sigma > \\, = \\, < \\sigma \\cdot v > \\, / \\, v_{\\rm T}$, with $v_{\\rm T} = \\sqrt{2 k_B T / \\mu}$, was calculated by the sum of the $s$-wave contribution (extrapolated from the thermal capture cross section assuming the usual $1/v$ law) and of the contributions from the four measured resonances. The $p$-wave contribution of the DC was neglected in that work.  In the following we will present our experimental procedure and results in Sect.~\\ref{sec:exp}. In Sect.~\\ref{sec:theo} we analyze our new experimental data using the DC model and we discuss the properties of so-called potential resonances, in Sect.~\\ref{sec:astro} we show the improved Maxwellian averaged capture cross section $< \\sigma >$, and the results are discussed in Sect.~\\ref{sec:summ}. ",
        "conclusions": "\\label{sec:summ} There are three main results of this work: ($i$) the DC cross section of the reaction $^{26}$Mg(n,$\\gamma$)$^{27}$Mg was measured for the first time between the known resonances at $E_{\\rm{n,lab}} = 68.7~{\\rm{keV}}$ and $E_{\\rm{n,lab}} = 220~{\\rm{keV}}$ \\cite{weig76}, ($ii$) the properties of potential resonances are discussed in detail for the $p$-wave capture of $^{26}$Mg(n,$\\gamma$)$^{27}$Mg, and ($iii$) the resonance strength of the $E_{\\rm{n,lab}} = 220~{\\rm{keV}}$ resonance was determined to be $\\omega \\gamma = 1.34 \\pm 0.24~{\\rm{eV}}$ by a careful analysis of the neutron energy spectrum $\\Phi(E_{\\rm{n,lab}})$. This last result is in clear contradiction to the results of Ref.~\\cite{weig76}. Unfortunately, in Ref.~\\cite{weig76} the experimental information on the capture measurements is very limited, but it is stated in that work that the enriched $^{26}$Mg- and $^{25}$Mg-samples were analyzed in the same way. This fact makes the experimental results of Ref.~\\cite{weig76} for the capture reaction $^{25}$Mg(n,$\\gamma$)$^{26}$Mg at least questionable. A strongly reduced $^{25}$Mg(n,$\\gamma$)$^{26}$Mg reaction rate might have influence on the $s$-process nucleosynthesis because of the role of $^{25}$Mg as neutron poison. In this paper we do not want to speculate about astrophysical consequences of such a reduced reaction rate; instead, we think that primarily a new measurement of the resonance properties of $^{25}$Mg is necessary to obtain reliable reaction rates. Unfortunately, the reaction $^{25}$Mg(n,$\\gamma$)$^{26}$Mg cannot be analyzed using our activation technique because the residual nucleus $^{26}$Mg is stable."
    },
    "9805/astro-ph9805247_arXiv.txt": {
        "abstract": "Using the Hipparcos data, absolute magnitudes and thus the evolutionary status for the group of \\LB stars were derived. The origin of this small group of non-magnetic, chemically peculiar stars, still remains a matter of debate. Using new evolutionary tracks, an age determination could be provided to distinguish between the two competing theories - the diffusion/mass-loss and the accretion.  The results establish the members of this group as objects which are {\\it very close to the Main Sequence}. This is supported by Pre-Main Sequence evolutionary tracks as well as by observational results. This contradicts prior conclusions that most of these stars are in the middle of their Main Sequence lifetime. The new results strongly support the predictions of the accretion theory. ",
        "introduction": "Using the Hipparcos parallaxes, absolute magnitudes for members of the \\LB group were determined. The origin of these non-magnetic, A to F-type metal-deficient dwarfs still remains controversial. The two main competing theories involve diffusion, either in combination with mass-loss (Michaud \\& Charland 1986), or accretion of interstellar matter as in post-AGB stars (Turcotte \\& Charbonneau 1993). The latter model requires that \\LB stars are very close to the Zero-Age Main Sequence. The recent discovery of \\LB stars in the young Orion OB1 association and in NGC~2264 (Paunzen \\& Gray 1997) seems to support the predictions of the accretion theory. But an age determination of galactic field stars by Iliev \\& Barzova (1995, hereafter IB95), on the other hand, resulted in evolved members. These stars were recalibrated using the accurate new absolute magnitudes as well as new stellar evolutionary tracks (CESAM; Morel 1997). Furthermore, Pre-Main Sequence models (Palla \\& Stahler 1993) were used to confirm the results. ",
        "conclusions": "With the Hipparcos data, absolute magnitudes and evolutionary status (mass and age) for the group of \\LB stars were estimated. No systematic influence of the apparent distance, effective temperature, metallicity and rotational velocity was found on the difference between the photometrically calibrated and the ``new'' absolute magnitudes, thus proving the validity of the ``standard'' photometric calibration (e.g. in the Str\\\"omgren system) for these (chemically peculiar) stars.  It turned out that six stars (e.g. \\LB itself) are definitely {\\it very close to the Main Sequence}, and this is also true, with a high probability, of three additional programme stars. These results contradict the conclusions of IB95 and support the accretion theory. The small number of \\LB stars (statistical effect due to the short ``lifetime'' on the Pre-Main Sequence) and the lack of them in open clusters older than 10$^{7}$ years further strengthen the accretion hypothesis."
    },
    "9805/gr-qc9805036_arXiv.txt": {
        "abstract": "A relativistic model of a heat conducting collapsing star, which includes thermal pre-relaxation processes, is presented. Particular attention is paid to the influence of a given parameter defined in terms of thermodynamic variables, on the outcome of evolution. Evaluation of the system when passing through a critical value of the aforesaid parameter, does not yield evidence of anomalous behaviour. ",
        "introduction": "The behaviour of dissipative systems at the very moment when they depart from hydrostatic equilibrium has been recently studied \\cite {Herrera97,HeMa97a,visco}.  In \\cite{Herrera97} it appears that a parameter $\\alpha $ formed by a specific combination of thermal conductivity coefficient $\\kappa$, relaxation time $\\tau$, temperature $T$, proper energy density $\\rho$ and pressure $p$, \\[ \\alpha =\\frac{\\kappa T}{\\tau \\left( \\rho +p\\right) }, \\] may critically affect the evolution of the object. Specifically, it was shown that in the equation of motion of any fluid element, the inertial mass density term vanishes for $\\alpha =1$ (critical point) and is negative beyond that value.  In some cases (pure shear or bulk viscosity) \\cite{visco}, the critical point is well beyond the point where the causality requirements are violated and therefore forbidden.  In others (pure thermal conduction) \\cite{Herrera97,HeMa97a}, later requirements are violated slightly below the critical point.  In the general case ( heat conduction plus viscosity) it appears that causality may break down beyond the critical point \\cite{visco}.  However, it should be stressed that the critical point (as well as causality conditions) is obtained in the context of a linear perturbative scheme, where the system is evaluated immediatly after leaving the equilibrium and therefore time derivative of radial velocity as well as dissipative variables are considered small quantities, such that only linear expressions of them are kept. On the other hand the vanishing, at the critical point, of the inertial mass density term, leading to accelerated fluid elements in the absence of total radial forces, might suggest that linear approximation is not longer reliable there \\cite{HeMa97a,visco}.  The question arises then, whether a physical system may reach the critical point without exhibiting a clear unphysical behavior or if, on the contrary, for any physical system reaching that point the march of physical variables becomes physically unacceptable as suggested by the linear approximation. In other words we want to see how the physical meaning of the above mentioned parameter $\\alpha $, as implied by the linear approximation, carries over to nonlinear regimes.  In order to elucidate this question we shall consider here an exact numerical model of an evolving dissipative star (without viscosity). The system is forced to evolve through the critical point, and fundamental variables are monitored to detect any anomalous behaviour. Modeling is performed by means of the HJR method \\cite{HeJiRu80}, a brief resume of which is given in next section .  The heat conduction equation is given in section 3, and the model is described in section 4. Finally, a brief analysis of our results is given in the last section. ",
        "conclusions": "As mentioned before, our purpose here is to clarify the physical significance (if any) of the critical point in the evolution of a dissipative self-gravitating system.  In particular we want to find out if at the critical point the system, described by the full theory (without any approximation), exhibits an anomalous behaviour, as might suggest a linear approximation approach.  The three profiles of $\\alpha$ considered are displayed in figure 1. In one case the system never reaches the critical point, in other case it is at the critical point at some moment of its evolution, and in the third case the system goes beyond the critical point before returning to equilibrium.  Figures (3)-(6) show the evolution of temperature, temperature gradient, radius and surface velocity for the three different profiles of $\\alpha$.  It is apparent from these figures that even though increasing values of $% \\alpha$ are associated with more unstable configurations (in the sense of faster collapse), nothing strange seems to happens at or beyond the critical point. Figure (7), showing the evolution of the ratio of the neutrino mean free path to the radius of the star, indicates that the diffusion approximation is valid during most part of the emission process.  To reinforce this conclusion we have performed another numerical simulation with the profiles of $\\alpha$ given in figure (2). The evolution of relevant physical variables displayed in figures (8)-(11), confirms the conclusion emerging from figures (3)-(6). The validity of diffusion approximation is corroborated in fig.(12).  To conclude, we may say that the increasing of instability (in the sense mentioned above) with higher values of $\\alpha$, seems to confirm the decreasing of the inertial mass density factor, predicted in the linear approximation.  However, nothing dramatic appears at the critical point in the exact modeling, indicating that the later approximation is not reliable at (or close to) the critical point, as suggested before \\cite{HeMa97a,visco}."
    },
    "9805/astro-ph9805071_arXiv.txt": {
        "abstract": "Slowly rotating collapsing spherical shells have flat spaces inside and the inertial frames there rotate at $\\omega_s(t)$ relative to infinity.  As first shown by Lindblom \\& Brill the inertial axes within the shell {\\it rotate rigidly without time delays} from one point to another.  Although the rotation rate of the inertial axes is changing the axes are inertial so, relative to them, there is neither an ${\\dot{\\mbox {\\boldmath $\\omega$}}}_s \\times {\\bf r}$ (Euler) fictitious force nor any other.  However, Euler and other fictitious forces arise in the frame which is at rest with respect to infinity. An observer at the center who looks in one direction $(\\phi = 0, \\; \\theta = \\frac{\\pi}{2}$ say) fixed to infinity will see that the sky appears to rotate and can compare its apparent rotation with those of the local inertial frame and of the shell itself.  By contrast in the electromagnetic analogue there is a time delay in the propagation of the magnetic field inside a rotating collapsing charged shell in flat space.  We demonstrate this time delay by devising a null experiment in which the Larmor precession of a charged oscillator would be exactly cancelled by the rotation of the inertial frame but for the delay.  In the \\lq\\lq combined\" problem of a collapsing charged shell we show that, due to the coupling of electromagnetic and gravitational perturbations, the instantaneous rotation of inertial frames inside the shell can be caused by pure electric currents in a non-rotating shell. ",
        "introduction": "Gravity is a tensor field and electromagnetism a vector field but there are striking similarities between them.  At the classical level there is the clear analogy between their inverse square force laws. At the post-Newtonian level moving charges generate magnetic fields while moving mass currents generate gravomagnetic\\footnote{We have adopted this more sonorous shorter form of Ciufolini \\& Wheeler's \\cite{1} ``gravitomagnetic''.  For definitions of ${\\bf B}_g$ in strongly relativistic stationary space-times, see Lynden-Bell \\& Nouri-Zonoz \\cite{15}.} fields.  These fields $B_g$ are the spatial components of the curl of ${\\mbox {\\boldmath $\\xi$}} / |{\\mbox {\\boldmath$\\xi$}} |^2$, ${\\mbox {\\boldmath $\\xi$}}$ being the time-like Killing vector.  When the space-time is stationary, they give in different language a description of the effects that are often called \\lq\\lq the dragging of inertial frames\" by relativists, but the advantages of the gravomagnetic field concept have recently been stressed by Rindler \\cite{2}.  The main purpose of this paper is to demonstrate an interesting difference in behavior between the gravomagnetic field which gives the rotation of the inertial frame within a slowly rotating massive collapsing spherical shell and the magnetic field within a rotating collapsing shell of charge.  Whereas the gravomagnetic field has {\\it no time delay} and is uniform within the shell in the gravity problem, nevertheless the magnetic field travels inwards from the shell, {\\it has a time delay} and is therefore non-uniform in the electromagnetic problem.  The origin of this difference can be traced to the fact that gravity as a tensor field does not carry any dipolar waves, so a collapsing slowly rotating shell generates none of them.  Thus the dipole contributions in gravity are governed by an \\lq instantaneous' equation not a wave equation.  The uniform ${\\bf B}$ inside a rotating shell is dipolar, and thus reacts instantaneously in a frame in which its center is at rest.  In electricity the monopole contributions are governed by the instantaneous Gauss's theorem and remarkably that instantaneous relationship between $\\int{\\bf E}\\cdot d {\\bf S}$ and the included charge holds in all Lorentz frames!  However in electromagnetism there are dipolar waves and a collapsing rotating spherical shell generates them.  The increase in ${\\bf B}$ as the shell collapses and rotates faster travels inwards from the shell. This difference in the time delay behavior has therefore a good physical origin, nevertheless, all physicists shy away from instantaneous effects as acausal.  How can it be that the internal inertial frames rotate at once without waiting for a causal signal to travel inwards from the shell?  Whereas local inertial frames can be determined locally their rotations and accelerations relative to inertial frames at infinity depend on a convention of how such frames are extended inwards.  There is nothing wrong with conventions that travel with the \\lq speed of thought'.  One may consider the Andromeda Nebula at one moment and Jerusalem a second later, so thoughts can travel far faster than light.  Thus, when we find that the inertial frame within a collapsing rotating spherical shell rotates rigidly with no time delay between the center and the shell's inner edge, we should not claim some spooky faster than light influence, rather we should look at the way we define the rotation of a reference frame.  It is we who decide on the \\lq gauge' in which the rotation is measured - it is our thoughts that travel faster than light!  Our metric is stationary outside the shell with coordinates $t, \\phi$ such that observers at rest at infinity have $t$ as their proper time and remain with $\\phi$ constant.  We extend this $\\phi$ continuously into the shell's flat interior and with the appropriate time the inertial frame there rotates rigidly through the surfaces of constant $\\phi$.  The angular velocity is directly related to the observable effects analysed by Lindblom \\& Brill \\cite{3}.  In the time-slicing we use, the rotation of inertial frames is determined by the constraint equations and it is thus related to instantaneous values of the matter variables. ",
        "conclusions": "Our principal results are stated mathematically in Section 2.  Here we describe them qualitatively.  There are strong analogies between the gravitational effects of rotating and collapsing massive spherical shells and the electromagnetic effects of rotating and collapsing charged shells.  The precession of the inertial frame in the former is the analogue of the Larmor precession of a charged particle due to the magnetic field generated by the latter.  Thus the gravomagnetic field gives a good intuitive concept of the effects more normally ascribed to the ``dragging of inertial frames''.  However there are fundamental differences between the gravitational and electromagnetic problems because gravity carries no dipolar waves, so, in the natural gauge dipolar effects are instantaneous.  By contrast the electromagnetic field carries dipolar waves and when these are generated the fields suffer retardation.  Had we taken shells with quadrupolar distortions both the quadrupolar gravitational effects and the electromagnetic ones would have suffered retardation.  The formulae of Section 2 are expressed in Schwarzschild coordinates which make them simpler than those originally found in isotropic coordinates by Lindblom \\& Brill \\cite{3}.  A central observer at fixed orientation with respect to infinity may feel dizzy due to Coriolis force and will see the distant quasars rotating forwards with the whole sky, as the massive rotating sphere surrounding him falls inward.  (If it rotates without falling he merely sees the quasars in the `wrong' directions.)  Objects moving near him will be subject to both Coriolis and Euler forces and, as his orientation is not inertial, he needs a torque to keep him oriented.  By contrast a lazy inertial observer near the centre will rotate at a time dependent rate with respect to infinity, will suffer no Coriolis giddiness but will see the sky rotating backwards while the massive shell around him rotates forwards.  We distinguish between the fixed direction to a distant quasar in the world map at one time and its apparent direction from which the light is seen.  The latter is deviated by the gravomagnetic field (and in a time dependent way if the rotating sphere also falls)."
    },
    "9805/astro-ph9805301_arXiv.txt": {
        "abstract": "A rich galaxy cluster showing strong resemblance with the observed ones is simulated. Cold dark matter spectrum, Gaussian statistics, flat universe, and two components -- baryonic gas plus dark matter particles -- are considered. We have calculated the gravitational-wave output during the epoch of the fully nonlinear and nonsymmetric cluster evolution. The amplitudes and frequencies of the resulting gravitational waves are estimated. Since frequencies are very small --of the order of $10^{-17} \\ Hz$ -- a complete pulse cannot be observed during an admissible integration time; nevertheless, it is proved that these waves can produce  an interesting secular effect which appears to be observable with current technology. ",
        "introduction": "Any dynamical and nonspherical self-gravitating astrophysical -- or cosmological -- system of mass M and size R is a potential source of gravitational waves. The most powerful ones are those combining a very asymmetric shape together with a violent nonstationary evolution.   In present paper we are interested in the gravitational waves radiated by cosmological objects. Flattened structures as superclusters and walls show great departures from spherical symmetry, but they are evolving in the mildly nonlinear regime and, consequently, their evolution is slow. Gravitational waves from these structures will be considered elsewhere. Galaxy clusters evolve in the strongly nonlinear regime and their dynamics is expected to be faster than in the case of larger flattened structures; however, departures from spherical symmetry seem to be small (even if their corresponding initial conditions were fully asymmetric). In this case, high density contrasts -- of the order of $10^{3}$ -- are reached and, consequently, fully nonlinear numerical simulations are necessary. This paper concerns with the characteristics of the gravitational radiation released during galaxy cluster evolution.  As it is well known, clusters contain dark matter ($\\sim 90 \\% $) and a subdominant baryonic component ($\\sim 10 \\% $). Since the dynamics and spatial distribution of these components are different, the ratio between the gravitational luminosities coming from these two components is not expected to be constant. Numerical simulations including a baryonic component -- gravitationally coupled to the dark one -- are necessary to describe the evolution of that ratio.  Taking into account the above discussion, a two components cluster, mimicking the observed ones, has been simulated numerically, paying particular attention to estimate its gravitational luminosity and the features of the radiated gravitational waves. To this aim we have used the approach of considering that the sources satisfy the constraints of the so-called nearly Newtonian slow motion. As far as we know, similar calculations have not been performed up to now, probably due to some pessimistic point of view justified by the combination of great numerical difficulties (3D calculations) together with very disappointing small values expected for the gravitational-wave luminosity, the amplitude, et cetera. However, our results have been more encouraging than we could foresee. A new observable effect has been found.   Hereafter, $t$ stands for the cosmological time, $t_0$ is the age of the Universe. $\\dot{X}$ stands for the derivative of the function X with respect to the cosmological time. The present value of the Hubble constant is  assumed to be $50 \\ Km \\ s^{-1} \\ Mpc^{-1}$. Quantities $c$ and $G$  are the speed of light and the gravitational constant, respectively. Greek (Latin) indices run from 1 to 4 (1 to 3). ",
        "conclusions": "We have proved that gravitational radiation from clusters does not contribute significantly to the density parameter.  A new effect produced by the gravitational waves emitted by a cluster has been described. A cluster with the features of the observed ones has been simulated in the standard CDM model. In other cosmogonies, the effect produced by clusters having these features is expected to be comparable. This is because the clusters producing the secular effect are located near the observer and, consequently, their evolution --well inside the horizon-- is dominated by internal interactions. Simulations support this idea, see, e.g., Huss, Jain \\& Steinmetz (1998), who claim that the differences between individual cluster realizations of a given  model are more important than the differences within differing scenarios for a given realization. However, the properties of the spatial distribution of clusters depend on the chosen scenario strongly. A few words about detection and future work are worthwhile.  We could take advantage of the fact that the secular effect can be observed for many years. In order to do that, this effect should be measured during periods greater than the characteristic time of any time varying gravitational field acting on the detector. Thus, the pulses produced by these fields could be seen as local perturbations of the secular effect.   If the detector is pointing towards a certain direction, it is not receiving gravitational waves from an unique cluster; hence, the calculation of the total effect produced by a set of clusters is important. For the sake of briefness, we present a rough estimate based on idealized cluster distributions within a sphere of radius $R_{s}$. All the clusters are identical to that of this paper. The distance between any pair of clusters is constrained to be larger than a fixed distance $L$. No cluster correlations are considered. Given an observation direction, the contribution of each cluster to the relative variation of the distance between particles $A$ and $B$ (detector) is assumed to be $\\frac {1}{2} \\Psi \\Delta t$ with $\\Psi = 1.26 \\times 10^{-20}\\, (\\frac {100}{D})\\, \\kappa \\, cos \\alpha \\,\\, yr^{-1}$, where $\\alpha$ is the angle formed by the line of sight of the cluster and the observation direction. Since clusters do not radiate in a coherent way, their contributions to $\\Psi$ are not correlated. This fact is simulated by using the random number $\\kappa$ uniformly distributed in the interval (-1,1). The factor (100/D) has been introduced because the effect of a given cluster decreases with distance as $D^{-1}$ (see Eq. (\\ref {hij})). After superimposing the contribution of all the clusters, the mean $\\mathcal{M}$ and the standard deviation $\\sigma$ have been calculated from the $\\Psi$ values corresponding to many observation directions. For each pair ($R_{s}$, $L$), various simulations have been done. The mean $\\mathcal{M}$ changes from one to another, but $\\sigma$ is a very stable quantity. For the pairs ($R_{s}=600 \\ Mpc$, $L=50 \\ Mpc$), ($R_{s}=600 \\ Mpc$, $L=100 \\ Mpc$), ($R_{s}=500 \\ Mpc$, $L=50 \\ Mpc$), we have found $\\sigma=7.6 \\times 10^{-20}$, $\\sigma=2.7 \\times 10^{-20}$ and, $\\sigma=7.0 \\times 10^{-20}$, respectively. From these data one concludes that, for $R_{s}\\ge500$, quantity $\\sigma$ depends on $R_{s}$ weakly. This fact suggests that the use of radius $R_{s}>600$ is not necessary because the main part of the effect is produced by nearby clusters. Furthermore, the resulting $\\sigma$ values are greater than the value $1.6 \\times 10^{-20}$ --maximum value assumed for a single cluster and D=100-- even when large separations ($L=100 \\ Mpc$) are assumed. The large $\\sigma$ values given by simulations strongly  suggests the feasibility of anisotropy measurements. Improved simulations including cluster correlations should increase the anisotropy. The effect of realistic spatial distributions including correlations and clusters with different masses, sizes, et cetera will be studied elsewhere. Various parameters --as the proportions between baryonic and dark matter, the density parameter, et cetera-- would be involved in these simulations. The resulting anisotropy would depend on both cosmological and large scale structure parameters. Comparisons between measurements and predictions of the secular effect and its anisotropies could play a crucial role in Modern Cosmology."
    },
    "9805/astro-ph9805137_arXiv.txt": {
        "abstract": "A primary production mechanism for blue stragglers in globular clusters is thought to be collisionally-induced mergers, perhaps mediated by dynamical encounters involving binary stars.  We model the formation and evolution of such blue stragglers, and produce theoretical distributions of them in the colour-magnitude diagram.  We use a crude representation of cluster dynamics and detailed binary-single star encounter simulations to produce cross sections and rates for a variety of collisions. The results of the collisions are determined based on SPH simulations of realistic star models.  The evolution of the collision products are then followed in detail using the Yale stellar evolution code.  We present our results in the form of distributions in the observed colour-magnitude diagram.  We use our results to explore the effects of a variety of input assumptions on the number and kind of blue stragglers created by collisions.  In particular, we describe the changes in the blue straggler distribution that result from using realistic collision products rather than the ``fully-mixed'' assumption, and from changes in assumptions about the number and orbital period distribution of the primordial binary population.  We then apply our models to existing data from the core of M3, where the large blue straggler population is thought to be dominated by collision products.  We find that we have difficulty successfully modeling the observed blue stragglers under a single consistent set of assumptions.  However, if 3 particularly bright blue stragglers are considered to be part of a different observed population (as has been previously suggested for other clusters) and left out of the population, the remainder can be successfully modeled using realistic encounter products and assuming a 20\\% binary fraction with plausible period distribution.  Finally, we suggest a variety of routes towards a more comprehensive understanding of the blue straggler phenomenon. ",
        "introduction": "Globular clusters are important astrophysical laboratories for two different processes: stellar evolution and stellar dynamics. A population of stars with the same age, composition and distance has been an ideal place in which to test theories of stellar evolution, beginning with the basic evolutionary tracks and now including such details of stellar structure and evolution as diffusion of heavy elements or convective overshoot (e.g. Chaboyer \\etal\\ 1992, Brocato \\etal\\ 1997). A globular cluster can also be considered to be a dense system of $\\sim 10^5$ point masses which interact only under the influence of their mutual gravity. These simple systems are obvious choices in which to test N-body calculations and other dynamical simulations (Dull \\etal\\ 1997, Elson \\etal\\ 1987, Meylan \\& Heggie 1997 and references therein).  It is becoming apparent, however, that we can no longer study these two astrophysical processes independently if we are to completely understand globular clusters.  Stellar evolution tells us that contrary to the simple dynamic assumptions, stars are not point masses with infinite lifetimes, but instead have finite radii and have lifetimes which are shorter for more massive stars.  These properties can have a significant impact on the dynamical processes. For example, a result of equipartition of energy in the cluster is mass segregation, which causes the more massive stars to sink toward the centre of the cluster (Lightman \\& Shapiro 1978). These massive stars have shorter lifetimes than their less massive counterparts, and will evolve into stellar remnants more quickly. These remnants tend to have a smaller mass than their parent star, and so the distribution of masses in the core of the cluster will change as the stars evolve. This change in the mass function will affect the rate of core collapse and other dynamical processes (Chernoff \\& Weinberg 1990).  The presence of stars with finite radii allows for the possibility of physical stellar collisions or tidal interactions. These collisions will change the energy budget in the cluster, which can influence the timescales on which core collapse, mass segregation and other dynamical processes occur (Spitzer 1987). The collision products may be objects which cannot be explained by standard stellar evolution theory, such as blue stragglers, low mass X-ray binaries and millisecond pulsars (see Bailyn 1995 for a review). Since these stars have very different masses and lifetimes than normal globular cluster members, their presence can affect the mass function of a region of the cluster, which can then modify the timescales for dynamical processes.  Therefore, the study of stellar collision products can lead to an understanding of the interactions between stellar evolution and stellar dynamics in globular clusters.  One possible class of collision product to study is blue straggler stars.  Blue stragglers are stars which are bluer and brighter than the main sequence turnoff. They have the same surface gravity values as main sequence stars (Rodgers \\& Roberts 1995, Shara \\etal\\ 1997), and therefore must have been born more recently that the majority of stars in the cluster.  Blue stragglers are thought to be created by the merger of two normal globular cluster stars, either through the merger of the two components of a binary star system, or through direct stellar collisions (Stryker 1993).  Both creation mechanisms are expected to occur, and which mechanism is dominant depends on the binary population and the density of the environment.  Recent studies of the creation and subsequent evolution of collisional blue stragglers (Bailyn \\& Pinsonneault 1995, Sills \\etal\\ 1995, Ouellette \\& Pritchet 1996, 1998, Sandquist \\etal\\ 1997) used the results of SPH simulations of stellar collisions to predict the initial structure and composition of blue stragglers. A number of different cases were then evolved using a stellar evolution code.  The results were presented as a series of evolutionary tracks, or simple assumptions were made about the collision rate and luminosity functions were calculated (Bailyn \\& Pinsonneault 1995). In this paper, we extend these approaches, and combine the evolutionary tracks with detailed binary-single star encounter cross sections and crude cluster dynamics to model the distribution of blue stragglers in the colour-magnitude diagram of globular clusters.  In section 2, we present a method for calculating the total number and distribution of collisional blue stragglers in the colour-magnitude diagram. In section 3, we discuss our input assumptions, and in section 4 we produce theoretical colour-magnitude diagrams of blue stragglers for a variety of these assumptions. In section 5, we apply this method to the globular cluster M3 as a specific example. ",
        "conclusions": "\\subsection{Validity of Assumptions}  In order to study one component of the interaction of stellar dynamics with stellar evolution, we have made a number of simplifying assumptions about the cluster dynamics and populations. We attempted to make these assumptions as reasonable as possible while still retaining a tractable problem. In this section we attempt to characterize the effects of the assumptions on our conclusions.  The first of the simplifying assumptions we have made is that the density in the region of interest is constant. This is clearly not the case in real globular clusters. The density decreases by a factor of $\\sim 2$ between the centre and the edge of the core in a King model. By assuming that the core has a density equal to the central density, we have overpredicted the number of blue stragglers.  We have also assumed that all populations have the same density at the same position. Because of mass segregation, massive objects will be more centrally concentrated, so the relative number density of objects will vary radially. We expect that collisions involving binary stars will happen more frequently closer to the centre of the cluster, and that collisions involving more massive stars will happen closer to the centre. Therefore, the shapes of the predicted distributions will vary with radius in the cluster.  We have also assumed that the primary and secondary of each binary system was drawn at random from an initial mass function. Therefore, the mass function of the secondaries is independent of the primary mass function. Observational (Abt 1990) and theoretical (Kroupa 1995) evidence suggests that stars which form in aggregates of binary systems with flat period distributions and component masses paired at random will evolve dynamically to the field population seen today. Therefore, our assumption of randomly paired masses is appropriate for a young cluster, but we have neglected dynamical modification of the binary mass ratios. During encounters involving binary stars, higher mass secondaries are preferentially swapped into the binary systems (Heggie 1975, Hut \\& Bahcall 1983, Sigurdsson \\& Phinney 1995). Therefore, as the cluster evolves, the binary population will evolve towards higher mass ratios. The collisions involving these new binary stars will be more likely to produce more massive collision products, and therefore the distribution of the blue stragglers in the colour magnitude diagram will include more bright blue stragglers.  Since this paper concentrates on a method rather than detailed comparison with observations, we chose cluster parameters which are typical for globular clusters rather than trying to model M3 in particular. For example, we have used a velocity dispersion of 10 km/s in the core of the cluster. Observations (Gunn \\& Griffin 1979) suggest that a value of 6 km/s may be more realistic as a global velocity dispersion in M3.  The velocity dispersion of the cluster influences the distributions in two ways: it helps determine the flux of stars, and it affects the cross sections for collisions involving binary stars. The flux of stars is essentially a density multiplied by the velocity dispersion. Since we have assumed a uniform velocity dispersion for the entire region of interest, the velocity dispersion is simply a constant factor which determines the total number of blue stragglers.  As the velocity increases, more stars pass through a given point each second, and more collisions occur. The dependence of the cross sections on the velocity dispersion is rather more complicated. In general, if the velocity of the stars decreases, the cross section for each collision goes up since the stars are more likely to interact gravitationally if they do not fly past each other quickly. The nature of the collision can also change as the velocity of the stars changes, since the orbits of the stars around each other are determined by the stars' initial velocities.  Neglecting the effect of the velocity dispersion on the cross sections, a value for $<v>$ of 6 km/s instead of 10 km/s will reduce our total number of blue stragglers, bringing our predictions closer to the observed number of bright blue stragglers in the central 100'' of M3.  Since the velocity dispersion in the core of M3 is about 6 km/s, binaries with longer periods can exist in M3 than in the theoretical cluster with a larger velocity dispersion. The cutoff between hard and soft binaries occurs at longer periods as the velocity dispersion decreases. Soft binaries are disrupted early in the life of the cluster. Therefore, the theoretical distributions in this paper have included too little contribution from long period binaries than is applicable for M3. As is demonstrated in figure 6, if we add long period binaries to the distribution, the total number of stars is increased, but the overall shape of the distribution is not greatly affected.  We have also assumed that a single number, the velocity dispersion, is sufficient to model the velocity distribution of stars in the cluster. However, all stars are not moving with the same velocity, but will have a nearly Maxwellian velocity distribution which is characterized by the velocity dispersion. Therefore, we expect that each quantity which is affected by the stellar velocity will have a range of values. The flux of stars will follow a Maxwellian, and the cross section for each collision will be an appropriately weighted average of a range of cross sections. Both these effects should result in smoothing of the theoretical blue straggler distributions. This smoothing effect should be small for most globular clusters since the velocity dispersion is fairly small.  The metallicity of M3 is [Fe/H]=-1.57 (Harris \\& Racine 1979). The evolutionary tracks used in this work were more metal-rich, with a metallicity of [Fe/H]=-1.27 (Z=0.001). Therefore, all the evolutionary tracks are too red and a little too faint for M3. However, since the data are not calibrated to a standard system, we simply shifted the evolutionary tracks by an amount which made the main sequence turnoff match the turnoff for an M=0.8 M$_{\\odot}$ star. Since any differential shifts due to metallicity will be small, this change in metallicity will simply shift the distributions without changing their shape or the total number of predicted blue stragglers significantly.  \\subsection{Two Populations of Blue Stragglers?}  Currently it is thought that two main mechanisms are responsible for creating blue stragglers: direct stellar collisions, and the merger of the two components of a binary system (Stryker 1993). In this work, we have concentrated on the direct stellar collisions, but it is expected that both mechanisms occur in different environments, perhaps even in the same cluster. One piece of evidence for this has been the bimodal radial distribution of blue stragglers in M3, and the different luminosity functions of the inner and the outer populations (Bailyn \\& Pinsonneault 1995, but see also Sigurdsson \\etal\\ 1994). The results of this paper suggest that two different mechanisms may be at work in the same region of the cluster. We have shown that our theoretical distributions have difficulty in simultaneously explain both the brightest blue stragglers, and the width of the total distribution in the colour-magnitude diagram. It seems likely that the brightest three blue stragglers in the centre of M3 belong to a different population than the rest of the blue stragglers. The globular cluster NGC 6397 shows a similar population of bright, central blue stragglers which may be different from the rest of the blue stragglers in that cluster (Sills \\etal\\ 1995). In both clusters, our theoretical distributions cannot explain both the bright and the faint populations of blue stragglers using a consistent set of assumptions.  The possible second population in M3 is brighter than the other blue stragglers, and separated from the rest by a gap of about 0.5 magnitudes, and the stars in NGC 6397 have similar properties. It is also separated from the horizontal branch by about the same amount. It is unlikely that these stars are horizontal branch stars since the horizontal branch is well defined. Therefore, we assume that these stars are blue stragglers. Blue stragglers can be made bright in two ways: they can have a large mass, or they can have a high helium content. These blue stragglers could have a large mass because they are the result of a collision between three stars instead of between two. Triple collisions are expected to occur between three independent stars (a very low probability occurrence), or to occur during an encounter of a binary system with a single star, or during an encounter between two binary systems.  We included the triple collisions which occurred during encounters between binary and single stars our theoretical distributions.  The structure of the collision products was determined using the same algorithm as for the other `SPH' tracks. The evolutionary tracks of some of these products do indeed pass through the region in the colour-magnitude diagram in which these three stars exist. However, the triple collisions included in these calculations do not contribute significantly to the total number of blue stragglers (about 1\\% for case N). There are two reasons for this. First, the lifetimes of stars with masses of about $2 M_{\\odot}$ are much shorter than stars of $1.5 M_{\\odot}$ by a factor of about 3. Second, the cross sections for triple collisions are in general smaller than the cross sections for collisions between two stars. This cross section is small because triple collisions are difficult to cause when only three stars are involved since there is no other star available to carry away angular momentum from the system. The cross sections are also small because triple collisions occur preferentially with small binary systems, and the cross section is proportional to the binary semi-major axis.  In order to sustain the hypothesis that triple collisions are responsible for the bright population, we need a larger contribution of triple collisions to the blue straggler distributions.  It seems likely that we can increase the cross sections for triple collisions by considering encounters between two binary systems. We have not included binary-binary interactions in this work since the cross sections for the different kinds of products are not yet easily calculated for a global population of binary stars, in the same way that binary-single collisional cross sections can be calculated, using STARLAB. Binary-binary interactions should produce many more collision products for the same binary fraction, and the distribution of collision products could be quite different. It is not yet clear that the addition of binary-binary collisions can explain the bright population of blue stragglers while maintaining the correct total number of blue stragglers observed in clusters. Analysis of the cross sections for binary-single and binary-binary encounters suggest that unless the binary fraction is quite low, too many collisions take place (Bacon \\etal\\ 1996).  The second method for creating bright blue stragglers is to increase their helium content.  Blue stragglers can have a higher than expected helium content if the star is made of two (or three) stars which evolved significantly before the collision, and their central helium is mixed to the surface of the star. The most extreme amount of mixing allowed is represented by the fully mixed distributions presented in this work. We have shown that stars which are fully mixed do indeed become as bright and as blue as the three blue stragglers in question. We can even produce these stars with the collisions between two stars only, without having to invoke a third star, as long as the product is fully mixed. Since blue stragglers are not fully mixed during the collision or during the subsequent evolution, another process must be at work if these stars are indeed fully or significantly mixed.  One possible process is rotation of the stars. We have neglected the intrinsic rotation of the stellar collision products in this work because we have been dealing with head-on collisions only. However, when two stars collide in an off-axis collision, the resulting product can be rotating as fast as 70\\% of its break-up speed. Stars which rotate quickly are subject to a number of instabilities which can cause mixing of material, including meridional circulation (Tassoul 1978). Detailed modeling of the evolution of these rotating collision products will be necessary to determine if these processes produce a significant amount of mixing, enough to affect the position of the stars in the colour-magnitude diagram (although see Ouellette \\& Pritchet 1998). Any investigation of the evolution of rotating blue stragglers must also explain two seemingly incongruous facts: we expect that stars of all masses may undergo rotational mixing, suggesting a continuum of stellar properties, but yet the bright blue stragglers in both M3 and NGC 6397 are separated by a gap from the faint blue stragglers.  It is also possible that the merger of the two components of a binary system can result in a star which is significantly mixed. This star should be rotating rapidly also, so rotational mixing could play a role. Also, if the merger involves any sort of common envelope evolution, it is unclear whether the two cores of the stars remain intact during the spiral-in phase or whether some significant portion of the helium of these cores is mixed into the outer layer of the star. If some mixing does occur, this may be enough to push some stars into the region of the colour-magnitude diagram in question.  More detailed models of the process of binary mergers is necessary before this question can be answered.  \\subsection{Future Work}  The above discussion of the two possible populations of blue stragglers brings up a number of areas which still need to be explored: encounters between two binary systems, rotation of the non-head-on collision products, and binary mergers. We made a number of other assumptions about the cluster which should be relaxed in order to further understanding of the blue straggler phenomenon.  First of all, we have assumed that the inner region of this cluster is a uniform box with the same density and velocity dispersion everywhere. This is obviously not the case in real clusters. We should use either a static (King) model or an evolving dynamical model of the cluster to predict the density and velocity dispersion as a function of radius, and apply those values to our calculations in such a way as to take into account the stratification of the cluster.  We have neglected the evolution of the globular cluster in a number of respects. We have assumed that the parents of these stellar collision products are drawn from the current population of the cluster. We are choosing stars which have evolved for 15 Gyr. However, collisions have happened throughout the lifetime of the cluster, and so some of the parents will have evolved for less time than we have assumed. Collisions between stars with higher masses than the current turnoff mass will also have occurred. We have also neglected to include the effects of the collisions on the parent populations. For example, blue stragglers can themselves be involved in stellar collisions sometime after they were created. Also, interactions involving binary stars can change the characteristics of the binary population. In order to model the populations of collision products in a globular cluster properly, we need to incorporate the detailed evolutionary tracks used in this work into an N-body dynamical code, and follow the stellar collisions in some Monte-Carlo fashion.  Investigations of globular cluster dynamics with stellar collisions has been begun by Portegies Zwart \\etal\\ (1997a,b). They have included a simple prescription for normal stellar evolution in a detailed N-body dynamics code, and have attempted to model the evolution of a cluster in which stellar collisions occur. Their treatment of the stellar evolution is relatively crude, consisting of a table of masses, radii and luminosities. We have shown that the mass of the stellar collision product alone is not enough to determine its evolutionary state, and hence its effect on the cluster's evolution. However, Portegies Zwart's treatment of the dynamics of the cluster is far more detailed than anything we have attempted. We are approaching the same problem from opposite directions, and our two approaches must eventually be combined.  We should also expand this work to other globular clusters. This will allow us to explore the effects of cluster metallicity and age on the blue straggler population. We will also be able to make stronger statements about the effects of cluster densities and central concentrations on the populations in the cluster, and possibly comment on the primordial binary population of the cluster. In order to draw valid conclusions from theoretical predictions of a sample of clusters, we should have high quality data to which we can compare our models. These data should be photometrically complete down to the turnoff of the cluster (or at least within a magnitude). They should also be complete spatially, reaching into the core of the cluster since most of the collisions will occur in the core. For many clusters, particularly the dense (and hence interesting) ones, this will require the Hubble Space Telescope since ground-based observations of the cores of globular clusters are limited by crowding effects. To be able to compare our theory with the data in a statistically significant way, we also need on the order of 100 or more blue stragglers in each cluster. As clusters are surveyed more carefully and as the surveys reach into the cores, it is becoming clear that M3 is not alone in its large population of blue stragglers (e.g. 47 Tuc -- Edmonds 1998, M55 -- Mandushev \\etal\\ 1997). We feel confident that in a few years, a sufficient data set will exist to provide detailed comparison for our models."
    },
    "9805/astro-ph9805123_arXiv.txt": {
        "abstract": "This work presents a detailed analysis of Cosmic Microwave Background (CMB) radiation intensity observations. The CMB is a relic of the Big Bang and its study greatly enhances our knowledge of cosmology. This work has led to new values for the  best fit temperature $2.7356 \\pm 0.0038$~K (95\\%\\ CL) and the speed of our solar system relative to the CMB, and limits on the spectral distortion parameters: $\\mu$, $y$, and $Y_{\\rm ff}$. These in turn lead to tighter constraints on the thermal history and allowed energy release in the early universe and possible cosmological processes. These limits are approaching the order of magnitude required by known processes. ",
        "introduction": "\\subsection{The Cosmic Microwave Background Radiation}  Early nucleosynthesis calculations by Gamow, Alpher and Herman [\\cite{alpher88}] showed that, from the theory of a hot big bang, one can infer that a cosmic microwave background radiation (CMB) will be present today as a result of the primordial fireball. The spectrum was predicted to be of black body form with a temperature of a few Kelvin. In 1964 \\cite{penzias65} discovered the predicted relic radiation by measuring a universal ``background noise'' of about 3K. This discovery of the cosmic microwave background (CMB) led to the ascendancy of the hot big bang model of cosmology.  Thermal equilibrium was easily reached in the early universe as a result of the high rate of interactions between matter and radiation due to the high density and temperature. The Planckian distribution is invariant under the universal expansion when expressed in dimensionless frequency $x = h \\nu / kT = hc / \\lambda kT$. (After decoupling, the photon number is conserved, so the occupation number, $n=1/\\left(\\exp{\\frac{h\\nu}{kT}}-1\\right)$, and therefore $x = \\frac{h\\nu}{kT}$ is a constant for each quantum state.) Even when the matter and radiation are no longer in good thermal contact the Planckian shape is preserved. Therefore we expect the CMB spectrum to be Planckian today and, indeed, to a large extent it is.  The CMB spectrum would have a blackbody form if the simple, hot, Big Bang model is a correct description of the early universe, but will be distorted from that form by energy release for a redshift $z \\lsim3\\times10^6$ [\\cite{sunyaev80}]. % Such releases might arise from decay of unstable particles, dissipation of cosmic turbulence and gravitational waves, breakdown of cosmic strings and other more exotic transformations. The CMB was the dominant energy field after the annihilation of positrons and the decoupling of neutrinos until $ z \\gsim 3\\times 10^4$. % For energy release into the electron-proton plasma, between redshift of a few times $10^4$ and $10^6$ [\\cite{smoot88}] the number of Compton scatterings is sufficient to bring the photons into thermal equilibrium with the primordial plasma. During this time, bremsstrahlung and other radiative processes do not have enough time to add a sufficient amount of photons to create a Planckian distribution. The resulting distribution is a Bose-Einstein spectrum with a chemical potential, $\\mu$, that is exponentially attenuated at low frequencies, $\\mu = \\mu _0 e^{-2\\nu_b /\\nu}$ (where $\\nu_b$ is the frequency at which Compton scattering of photons to higher frequencies is balanced by bremsstrahlung creation at lower frequencies [\\cite{danese82}]). At redshifts smaller than $\\sim10^5$, the Compton scattering rate is no longer high enough to produce a Bose-Einstein spectrum. The resulting spectrum has an increased brightness temperature in the far Rayleigh-Jeans region due to bremsstrahlung emission by relatively hot electrons, a reduced temperature in the middle Rayleigh-Jeans region where the photons are depleted by Compton scattering, and a high temperature in the Wien region where the Compton-scattered photons have accumulated. ",
        "conclusions": "None of the distortions, $y, \\yff$ or $\\mu$, are larger than the current errors in measurement for the monopole and dipole CMB spectra. The monopole CMB spectrum is very close to a blackbody spectrum. This is strong evidence of the validity of the hot Big Bang model [\\cite{smoot96}] because there are roughly $10^9$ photons to each baryon in the universe, and it would be extremely difficult to produce the CMB in astrophysical processes such as the absorption and re-emission of starlight by cold dust, or the absorption or emission by plasmas, and still produce such a precise black body spectrum.  The dipole data did not significantly improve on the errors, but were consistent with and an independent check of the monopole data. Future measurements of the dipole could potentially contribute to tightening the limits or detecting distortions. The Planck Surveyor measurements of the dipole could, in principle, determine the distortions to approximately the same precision as the monopole measurements do now. Very substantial progress that would surely detect the predicted distortions and provide us with new information on the thermal history of the Universe will require a significant dedicated mission."
    },
    "9805/astro-ph9805315_arXiv.txt": {
        "abstract": "Massive  ($\\gsim 8\\Msun$) stars may end their lives in the molecular clouds in which they were born. O-type stars probably have sufficient photoionizing radiation and wind power to clear a region $>15$ pc in radius of molecular material. Early B stars (B1--B3 on the main sequence, or $8-12\\Msun$ stars) are not capable of this and may interact directly with molecular gas. Molecular clouds are known to be clumpy, with dense molecular clumps occupying only a few percent of the volume. A supernova remnant then evolves primarily in the interclump medium, which has a density $n_H=5-25$ H atoms cm$^{-3}$. The remnant becomes radiative at a radius of $\\sim 6$ pc, forming a shell that is magnetically supported. The structure of the shell can be described by a self-similar solution. When this shell interacts with the dense clumps, the molecular shock fronts are driven by a considerable overpressure compared to the pressure in the rest of the remnant. The expected range of clump sizes leads to a complex velocity distribution, with the possibility of molecular gas accelerated to a high velocity. Observations of the remnants W44 and IC 443 can be understood in this model. W44 has a shell expanding at $\\sim 150\\kms$ expanding into a medium with density $4-5$ cm$^{-3}$. The shock emission expected in such a model is consistent with the observed H$\\alpha$ surface brightness and the [OI] 63$\\mu$m line luminosity. The clump interaction is seen in OH maser emission, which shows a magnetic field strength that is consistent with that expected in the model. IC 443 appears to be expanding at a lower velocity, $100\\kms$, into an interclump medium with a higher density, $\\sim 15$ cm$^{-3}$. The interaction of the radiative shell with molecular clumps can produce the  molecular emission that is observed from IC 443. Both remnants are shell sources of radio synchrotron emission, which can be attributed to relativistic electrons in the cool radiative shell. If ambient cosmic ray electrons are further accelerated by the shock front and by the postshock compression, the radio fluxes and the flat spectral indices of W44 and IC 443 can be explained. The energetic electrons are in a high density shell and their bresstrahlung emission can approximately produce the $\\gamma$-ray fluxes observed by {\\it EGRET}. Molecular clouds have a significant uniform magnetic field component so that heat conduction is likely to be important in the hot interior and can explain the isothermal X-ray emission observed from the remnants. ",
        "introduction": "Twenty years ago DeNoyer (1979a,b) found clear evidence for the interaction of the supernova remnant IC 443 with molecular gas, based on observations of disturbed CO emission. IC 443 has become an especially rich target for molecular line observations, including H$_2$ (Burton et al. 1988; Richter et al. 1995b); CO and HCO$^+$ (Dickman et al. 1992; Tauber et al. 1994); HCN, CN, SiO, CS, SO, H$_2$CO, and C$_2$H (Turner et al. 1992; van Dishoeck et al. 1993). Weaker evidence for interaction was found in CO observations of W44 and W28 (Wootten 1977, 1981), which were earlier found to be sources of OH 1720 MHz line emission (Goss \\& Robinson 1968). OH 1720 MHz maser line emission has emerged as a signpost of molecular interaction with the detection of compact OH emission from 8 supernova remnants (Frail et al. 1996), including W44, IC 443, and 3C391. These remnants typically have a radius $\\sim 10$ pc.  The interaction of supernova remnants with molecular clouds can plausibly occur because massive stars are born in such clouds and have a relatively short lifetime. Stars with initial mass $\\gsim 8\\Msun$ are expected to end their lives as core collapse supernovae. Previous theoretical treatments of the interaction problem have concentrated on the interaction with very dense cloud material. Shull (1980) considered an ambient medium with density $n_o=10^5$ cm$^{-3}$ and predicted strong  IR (infrared) emission that decayed after 20 years. Wheeler, Mazurek, \\& Sivaramakrishnan (1980) found that evolution in a medium with $n_o=10^4$ cm$^{-3}$ leads to a high IR luminosity for $\\sim 100$ years, followed by a rapid drop. The remnant radius is $\\sim 1$ pc during its bright phase. Events of this kind have not been identified, but it is not clear what limits can be set from existing IR observations. The existing observations of remnant/cloud interactions require a different model.  The model I explore here is the interaction of a remnant with a clumpy molecular cloud. Observations of molecular clouds have shown them to be clumpy, with a filling factor for the dense molecular clumps of $\\sim 0.02-0.08$ (Blitz 1993). The interclump gas has a density of $5-25$ cm$^{-3}$. A supernova is more likely to explode in the interclump gas than in a dense clump, especially when the effects of presupernova winds and photoionization and stellar motion are taken into account. In \\S~2, I consider the likely environment for a supernova in a molecular cloud, including the effects of presupernova winds and photoionization as well as the structure of the molecular cloud. The remnant expansion in a molecular cloud is treated in \\S~3, on the assumption that the evolution in the interclump region can be separated from the interaction with clumps. In \\S~4, I compare this model to observations of three well-observed remnants: W44, IC 443, and 3C391. The model does appear to be compatible with the basic observations. Nonthermal particles and their emission are treated in \\S~5. This topic is of special interest in view of the probable detection of supernova remnants in molecular clouds with the {\\it EGRET} experiment on the {\\it CGRO} ({\\it Compton Gamma-Ray Observatory}) (Esposito et al. 1996). Possible future work is briefly discussed in \\S~6. ",
        "conclusions": "The aim of this paper has been to show that observations of a number of luminous supernova remnants can be interpreted in terms of interaction with a clumpy molecular cloud. The remnants become radiative in the interclump medium and the radiative shells then interact with molecular clumps. The magnetically supported radiative shells can be observed as cool HI shells and by the radiative shock emission. Because the extinction is frequently high to these remnants in dense regions, the study of infrared line emission can be especially fruitful. The [OI] 63$\\mu$m line has already been found to be a powerful probe of these remnants. A prediction of this model is that the high-energy $\\gamma$-ray emission detected from some of these remnants is from the radiative shell. This prediction can eventually be checked with experiments that have a higher spatial resolution than is currently possible, such as {\\it GLAST} ({\\it Gamma-ray Large Area Space Telescope}). A more detailed treatment of the expected emission spectrum for acceleration of ambient cosmic ray should also be carried out.  The interaction of the radiative shell with molecular clumps can give a number of features observed from these remnants, such as higher pressures than those in the radiative shell and the acceleration of molecules to relatively high velocities. Observations of molecular clouds have shown a great deal of small scale structure in the molecular material and this can be seen directly in H$_2$ observations of the remnant IC 443 (Richter et al. 1995b). There may also be instabilities associated with the interaction. Numerical computations of the interaction of a radiative shell with a dense clump should be carried out to elucidate the hydrodynamic features of the interaction. These calculations will be helpful in the interpretation of detailed molecular shock wave observations.  The hot interiors of the remnants are a source of X-ray emission. The observations show that the emission is close to isothermal, which suggests the action of heat conduction. However the evaporation of clumps does not appear to be a dominant process because of the small X-ray emitting mass and the fact that the radiative shells appear to be at the outer edges of the remnants. Evolutionary calculations that include heat conduction are needed for a detailed comparison with these remnants. After this paper was submitted, the paper by Shelton et al. (1998) on a detailed model for W44 became available. Shelton et al. (1998) present a model that is similar to the one presented here in that W44 is found to be a radiative remnant in a medium with density $\\sim 6$ cm$^{-3}$, the synchrotron emission is from compressed ambient cosmic ray electrons (although they do not include shock acceleration of these electrons), and the [OI] and H$\\alpha$ emission are from the radiative shock front. They go beyond the present work in showing how expansion into a medium with a density gradient can provide a detailed fit to the HI observations (Koo \\& Heiles 1995) and in calculating the X-ray emission for a hydrodynamic model with heat conduction. The model is significantly more successful in fitting the observed X-ray emission than the non-conduction model of Harrus et al. (1997), although the X-ray emission is still not as centrally peaked as is observed. A difference with the present model is that the OH maser emission is attributed to emission from the radiative shock front and Shelton et al. (1998) find no compelling evidence for interaction with a molecular cloud. The point of view taken here that the OH masers are associated with shock fronts in molecular clumps is supported by the recent observations of Frail \\& Mitchell (1998) who find that there are CO emitting clumps that are associated with the OH maser emission in W44.  Massive stars are expected to frequently have sufficient photoionizing radiation and winds to sweep out a region $\\gsim 15$ pc in size around the stars. It is only the massive stars at the low mass end ($\\lsim 12\\Msun$) that may interact directly with a molecular cloud. Although these are a minority of the remnants, they are bright because of the interaction with dense gas and so stand out in the observational sample. The further investigation of these objects can be expected to shed light on a number of physical processes, from molecular shock emission to the high-energy $\\gamma$-ray emission from cosmic rays."
    },
    "9805/astro-ph9805065_arXiv.txt": {
        "abstract": "Doppler imaging, a technique which inverts spectral line profile variations of an Ap star into a two-dimensional abundance maps, provides new observational constraints on diffusion mechanism in the presence of a global magnetic field. A programme is presented here with the aim to obtain abundance distributions of at least five elements on each star, in order to study how different diffusion processes act under influence of a stellar magnetic field. The importance of this multi-element approach is demonstrated, by presenting the abundance maps of helium, magnesium, silicon, chromium and iron for the magnetic B9pSi star CU Vriginis. ",
        "introduction": "The scientific goal of applying the Doppler imaging technique to Ap stars, is to derive observational constraints on the diffusion mechanism in the presence of a stellar magnetic field. Theoretical studies (Michaud et al, 1981; M\\'egessier, 1984) which predict that certain elements raise, sink or move horizontally in magnetic stellar atmospheres need to be verified. This can only be achieved if an unambiguous correlation between the elemental abundance structures and the magnetic field geometry is obtained. However, two major problems are responsible for difficulties in reaching that goal.  Firstly, Doppler imaging can (presently) only be applied to stars which have a fairly weak magnetic field and hence only the variation of the integrated longitudinal field strength (effective field, $B_{\\rm eff}$) is usually determined. This sets only weak limits to the `real' magnetic geometry. For objects with stronger fields the spectral line profiles are severely influenced by Zeeman broadening and are not mainly caused by abundance variations. This limitation can be avoided by obtaining all four Stokes parameters as a function of the stellar rotational phase. Then Zeeman Doppler imaging codes (Donati, 1995) can be applied to calculate the magnetic field strength and surface orientation in combination with the abundance distribution. But these measurements of the Stokes parameters are not yet available, though strong efforts are made to get them from polarimetry.  Secondly, in the past, only a few elements were mapped, mostly silicon, rarely iron and chromium, not even all of them for each object. Furthermore, these maps were based on inversions from single `unblended' spectral lines, which are difficult to find, especially in cooler Ap stars. This introduced uncertainties leading to a questionable reliability of the derived maps and weakening their relevance to theoretical modelling.  Considering this situation, we focused our programme on the production of abundance maps of up to five and more elements on the surface of each of our target stars. A brief report on the stars and the progress of our project is given hereafter, followed by the presentation of results obtained for the magnetic Ap star CU Virginis. ",
        "conclusions": ""
    },
    "9805/astro-ph9805253_arXiv.txt": {
        "abstract": "We present a detailed discussion of the redshift errors associated to the ESO Slice Project measurements. For a subsample of 742 galaxies with redshifts determined both from the absorption lines ($V_{abs}$) and from the emission lines ($V_{emi}$), we find an average difference $<V_{abs} - V_{emi}> \\simeq +100$ km/s. We find that a similar effect is present in another, deeper redshift survey, the Durham/Anglo--Australian Telescope faint galaxy redshift survey (Broadhurst et al. 1988), while is absent in surveys at brighter magnitude limits. We have investigated in detail many possible sources of such a discrepancy, and we can exclude possible zero--point shifts or calibration problems. We have detected and measured systematic velocity differences produced by the different templates used in the cross--correlation. We conclude that such differences can in principle explain the effect, but in this case the non--trivial implication would be that the best--fitting template does not necessarily give the best velocity estimate. As we do not have any {\\em a priori} reason to select a template different from the best--fitting one, we did not apply any correction to the ESO Slice Project velocities. However, as for a small number of galaxies the effect is so large that it is likely to have a physical explanation, we have also taken into account the possibility that the discrepancy can be partly real: in this case, it might help to understand the role of gas outflows in the process of galaxy evolution. In view of the future large spectroscopic surveys, we stress the importance of using different templates and making them publicly available, in order to assess the amplitude of systematic effects, and to allow a direct comparison of different catalogues. ",
        "introduction": "The understanding of the formation, evolution and present properties of the large--scale structure of the Universe is a key problem in modern cosmology (see Peebles 1980, 1993). One of the most important results of the first redshift surveys was the previously unexpected existence of coherent structures and voids at very large scales. Explaining these structures was a challenge for popular models of galaxy formation, but at the same time represented a problem for the interpretation of results obtained on small volumes which could not be representative of the Universe. Therefore, the need of a ``fair sample\" of the Universe, in order to understand the process of galaxy formation and evolution, led to an increasing number of deeper redshift surveys. Redshift surveys are now an ``industry\" with its own standards. Reduction of an ever growing number of data is based on software packages specially developed to this aim. The redshift $z= \\Delta \\lambda / \\lambda$, or, less rigorously, the ``recession velocity\" $V = cz$, is commonly determined using the wavelength shift of either absorption or emission lines appearing in the optical spectrum of a galaxy. Following the paper by Tonry \\& Davis (1979), most redshifts based on absorption lines are now obtained by cross--correlating galaxy spectra with one or more (or an average of) ``template'' spectra, while redshifts based on emission lines are measured by fitting the individual emission lines. Moreover, emission and absorption lines are produced in different environments. In normal galaxies, the former (such as the [OII]$\\lambda 3727$ line) are generated in HII regions associated with recent star--formation, while the latter (such as the calcium Ca II $K$ and $H$) are produced in stellar atmospheres and are related to the bulk of the star population. As a consequence, emission and absorption redshifts are not required to be exactly the same.  Despite the growing number of galaxy redshifts in the literature, most catalogues quote only the ``best'' estimate of the velocity of a galaxy, and take for granted the implicit and widespread assumption that, while for a given galaxy the absorption velocity $V_{abs}$ and the emission velocity $V_{emi}$ may differ, the average difference should be consistent with zero.  In the analysis of the ESO Slice Project (ESP; Vettolani et al. 1997; Zucca et al. 1997; Vettolani et al. 1998), we have devoted a particular effort to check the quality of our data, and in particular the precision of our absorption and emission redshift measurements which, as we have soon realized, present a puzzling discrepancy. Looking at the past and recent literature, we have also realized that this problem was not new, but was never discussed in a satisfactory way. We have therefore decided to study the effect in more detail, and we describe in this paper the results of our analysis and the possible explanations.  In section 2 we discuss the evidence of discrepancies in \\dv~ found in the past and in other surveys, and in section 3 we present the discrepancy detected in the ESP data. In section 4  we describe the tests we have performed on the ESP data, exploring instrumental and other effects which could in principle affect our results; in section 5 we analyse in detail the biases on velocity measurements due to the choice of the template spectra; in section 6 we discuss if such a discrepancy can be partly due to a real, physical effect; our conclusions are in section 7. ",
        "conclusions": "We have discussed in detail the errors associated to redshift determinations in the ESO Slice Project (Vettolani et al. 1997). We have found a systematic difference between absorption and emission line velocities of $\\sim +100$ km/s and we have shown that the same effect is present in the Durham/Anglo--Australian Telescope faint galaxy redshift survey (Broadhurst et al.1988). In the case of the ESP, we have excluded problems of zero--point error or calibration. Such a discrepancy has not been detected in large shallower surveys.  Shectman et al. (1996) briefly discuss a similar effect for the Las Campanas Redshift Survey, which they have corrected by using a different template for emission line galaxies.  We have generalized the suggestion by Shectman et al. (1996), who identify as the main cause of the discrepancy the systematic difference between the absorption line spectra of the standard templates and the typical emission line galaxy, particularly the blend between the Ca II $H$ and $H_\\epsilon$ lines, an effect already discussed by Sandage (1978). We find in fact systematic effects even from template to template, apparently due to the way each template fits the galaxy spectra; this implies that the choice of the template significantly affects redshift measurements.  For the ESP data, we decided to use the best--fit template (i.e. the one giving the smallest error), as using a different template might introduce unknown biases in the redshift measurement.  In the lack of a definitive explanation, the common assumption that the true galaxy redshift is given -- on average -- by the emission lines, is plausible, but not proven. It should be verified that the bias cannot be due, at least partially, to emission lines, and that the sample is not biased, for some reason, towards galaxies with outflows. One can for example speculate that other factors may contribute to the effect, such as sampling of different parts of the galaxies, the different mix of morphological types, and evolution with look--back time. A collection of high resolution data of a sample of galaxies and of different templates will be necessary to give a definitive solution.  We feel it is important to stress the existence of such an ``anomaly\": in view of  future, large surveys, the templates used should be carefully checked and made publicly available\\footnote{ESP templates are available at the following WWW address: http:/boas5.bo.astro.it/$^\\sim$cappi/esokp.html}, as already done by the LCRS group, to discover and quantify any systematic difference. Even if the amplitude of the effect is not large, it is quantitatively more important that typical zero--point shifts, and it is significant enough to affect for example measures of velocity dispersions and galaxy peculiar velocities, or the interpretation of results for very distant galaxies, as those which are reported by Cowie et al. (1995) and Steidel et al. (1997)."
    },
    "9805/astro-ph9805229_arXiv.txt": {
        "abstract": "The surface structure of the magnetic field in stars is one part of the information about the surface inhomogeneities like brightness, movement or chemical composition, which is contained in the integral radiation flux but distorted beyond recognition by losses of information on the topographical arrangement over the surface because of convolution processes and partial invisibility.  Therefore, the complicated processes, beginning with the representation of the map of the surface field distribution and ending with the resulting phase-curves of the integral magnetic field, are simulated by a computer program. We use this program as a tool for the analysis of the magnetic field structure out of the really observed data in phase diagrams by variation of the parameters and by fitting the calculated curves to the observed ones. ",
        "introduction": "{\\bf Mapping} of individual features on the surface of a star out of the integral radiation can be carried out only in the case where the linear convolution integral transform leading to the integral flux is complete and can be inverted. However, we have to cope with the so-called ill-posed inverse problem, which cannot be solved correctly. The method proposed here relates only to a limited number of parameters for the straightforward calculation, thus offering a way to overcome this problem.  The map of the {\\bf topographical arrangement} of the surface magnetic field with its vectorial character is therefore constructed by matrices. The matrix elements are defined by the spherical coordinates of the longitude and the latitude.  The {\\bf magnetic field vector} consists of three components with the unity vectors in direction of the radius of the star (normal vector), in direction of the longitude ($\\phi$-vector), and in direction of the latitude ($\\delta$-vector). A fourth component is added for a scalar magnitude, which can be used for different purposes (brightness, transparency, factor).  The calculation of the magnetic field components makes use of the fact that the linear aggregates of the potentials of {\\bf point-like field sources} are superposed linearly. Thus, the potential of a single source will be calculated by the transform of rectangular to spherical coordinates. Then the field vector is easily derived by the spherical gradient of the potential.  The advantage of the {\\bf linear superposition of potentials and vector fields} is obvious: the calculation is not limited to special source configurations. The individual treatment of {\\it monopoles} allows an arbitrary composition of configurations up to higher multipoles. In principle, any field you like can be represented by a row of ``monopole'' fields, the sum of charges being zero.  The {\\bf arrangement of the sources} may be anywhere in the interior of the stellar body. However, there could be magnetic sources placed even at the surface, representing stellar spots like sun spots. Sources outside the star could be positioned in companions, which influence the surface of the main star with their fields. ",
        "conclusions": ""
    },
    "9805/astro-ph9805190_arXiv.txt": {
        "abstract": "We present recent results from  our radial velocity (RV) studies of rapidly oscillating Ap (roAp) stars. Our measurements reveal that the pulsational amplitude of these stars depends on the spectral region that is examined. For one star, $\\gamma$ Equ, the pulsational RV amplitude depends on both line strength and atomic species. The elemental difference is most likely related to the abundance patches on the surface of these stars. The line strength difference is interpreted as arising form a height effect in the atmosphere. Ultimately, these measurements may provide valuable diagnostics of the atmospheres of these stars. ",
        "introduction": "The rapidly oscillating Ap stars (roAp)  represent a class of stars pulsating in nonradial p-modes with periods of 4--15 min. We have begun a programme of using precise stellar radial velocity (RV) measurements to study these pulsations. Data were acquired using the F3 focus of the 2-d coud{\\'e} spectrograph of McDonald Observatory's 2.7-m telescope ($R$ $\\approx$ 60,000). RV measurements were made using an iodine gas absorption cell placed in the optical path of the spectrograph (Cochran \\& Hatzes 1994). Exposures of 45-60 secs duration were made in rapid succession over  $\\sim$ 4 hrs for each of the stars presented here. ",
        "conclusions": "Our RV measurements show that the pulsational amplitude for roAp stars depends 1) on atomic species (e.g. in $\\gamma$ Equ, Cr has a higher amplitude than Fe), and 2) line strengths (weaker lines have a higher RV amplitude than stronger lines). The former effect is interpreted as arising from the inhomgenous distribution of elements on the stellar surface (if Cr is concentrated at the magnetic poles then its vertical motion is unhindered by the field lines). The line strength effect most likely arises from vertical structure of the pulsations since weaker lines are formed, on average, deeper in the stellar atmosphere."
    },
    "9805/astro-ph9805159_arXiv.txt": {
        "abstract": "s{ Several arguments suggest that quasars at $z > 4$ may be in the cores of future giant ellipticals, and forming at the very highest peaks of the primordial density field.  A strong bias-driven primordial clustering is then expected in these fields, which are naturally interpreted as the cores of future rich clusters.  High-redshift quasars can thus be used as markers of some of the earliest galaxy formation sites.  Recent discoveries of (proto)galaxy companions of $z > 4$ quasars, and hints of strong clustering among high-$z$ quasars give some support to this idea.  The frequency of companions is much higher than is expected for a general field, or the upper limits for comparable redshifts in the HDF.  Clustering of quasars themselves at $z > 3$ or 4 may reflect the primordial large-scale structure. } ",
        "introduction": "Primeval galaxies (PGs) have now been found, in a variety of forms and environments, using a variety of techniques: narrow-band Ly$\\alpha$ imaging~\\cite{Cow97}, Lyman-break method~\\cite{Ste96}, as DLA absorbers~\\cite{Djo96}, as quasar companions~\\cite{Djo85,Hu96}, as gravitationally lensed objects~\\cite{Fra97}, etc.  Statistical studies of the HDF~\\cite{Madau} have outlined a global history of galaxy formation in the general field. Moreover, we are starting to see evidence ~\\cite{Ste98} for large-scale structure formation at high redshifts, i.e., $z > 3$.  However, any single discovery method may bias the sample of objects found, and formative histories of galaxies in different environments may vary substantially.  For example, galaxies in rich clusters are likely to start forming earlier than in the general field, and studies of galaxy formation in the field may have missed possible rare active spots associated with rich protoclusters.  Diverse and complementary paths towards discoveries of first galaxies are still very much worth pursuing.  Our group is pursuing the use of quasars at $z > 4$ as markers of the early galaxy formation sites, by looking for PGs associated and/or clustered with quasar host galaxies themselves ($z_{gal} \\approx z_{QSO}$).  It is now believed that galaxy formation is an extended series of processes, which may have a broad peak at $z \\sim 2 - 3$.  However, it is of a considerable importance to catch the $first$ galaxy-size structures forming at high redshifts, i.e., the very onset of massive galaxy formation.  Any galaxies at $z > 4$ must be still very young, simply on account of the timing: they can be at most a few percent of the present galaxian age, or $\\sim 0.5 - 1$ Gyr, for any reasonable cosmology.  The evolutionary status of galaxies at $z \\sim 2$ or 3 is more ambiguous, with possible physical ages up to $\\sim 2 - 3$ Gyr, or $\\sim 10 - 20$\\% of their present age.  Going to $z > 4$ removes this age ambiguity.  We are conducting an observational program to discover and study such protogalaxies in the earliest stages of formation, clustered with the known quasars at $z > 4$.  This pushes beyond the territory now explored in the HDF and elsewhere, at the epochs when the universe was only a few percent of its present age. ",
        "conclusions": ""
    },
    "9805/astro-ph9805334_arXiv.txt": {
        "abstract": "Resonant relaxation is a novel form of two-body relaxation that arises in nearly Keplerian disks such as protoplanetary disks. Resonant relaxation does not affect the semimajor axes of the particles, but enhances relaxation of particle eccentricities and inclinations. The equilibrium state after resonant relaxation is a Rayleigh distribution, with the mean-square eccentricity and inclination inversely proportional to mass. The rate of resonant relaxation depends strongly on the precession rate of the disk. If the precession due to the disk's self-gravity is small compared to the total precession, then the relaxation is concentrated near the secular resonance between each pair of interacting bodies; on the other hand if the precession rate is dominated by the disk's self-gravity then relaxation occurs through coupling to the large-scale low-frequency $m=1$ normal modes of the disk. Depending on the disk properties, resonant relaxation may be either stronger or weaker than the usual non-resonant relaxation. ",
        "introduction": "\\label{sec:intro}  \\noindent The formation of planets from a disk of planetesimals is largely determined by two closely related processes: physical collisions and gravitational relaxation (e.g. \\cite{SW88}, \\cite{LS93}). Collisions result from close two-body encounters and drive the evolution of the mass spectrum of the planetesimals, while relaxation arises mainly from distant two-body encounters and drives the evolution of their phase-space distribution. The relative rates of the two processes are determined by the Safronov number, $\\theta\\sim (v_e/v_r)^2$, where $v_e$ is the surface escape speed and $v_r$ is the rms random velocity of the planetesimals; relaxation dominates when $\\theta\\gg 1$. Gravitational relaxation has two main effects on the random velocity of a particle in a disk: scattering drives a stochastic random walk towards larger values of $v_r$, while dynamical friction damps $v_r$. In general, scattering dominates for small particles while dynamical friction dominates for massive ones, as expected by equipartition.  The usual derivation of the rate of gravitational relaxation treats the evolution of the particle orbit as a sequence of uncorrelated two-body encounters with other particles. Such derivations neglect the effect of the gravitational field of the central star, but should be accurate in the early stages of planetesimal accumulation, so long as $v_r\\gg nr_H$, where $n$ is the mean motion and $r_H=r(m/\\mstar)^{1/3}$ is the Hill radius of the particle; here $m$ and $\\mstar$ are the masses of the particle and the central star, and $r$ is the orbital radius (\\cite{I90}).  This papers investigates a qualitatively different type of gravitational relaxation, resonant relaxation, which we have already investigated in the context of spherical stellar systems (\\cite{RT96}).  Resonant relaxation arises in addition to the usual (``non-resonant'') relaxation discussed above, when the potential in which the particles orbit is nearly Keplerian. To introduce the concept, let us imagine a time-exposure of a particulate Keplerian disk, in which the exposure time is much longer than the orbital period $t_{\\rm orb}$ but shorter than the characteristic precession time $t_{\\rm prec}$. Each particle appears in the photograph as an eccentric ellipse or wire. These wires exert torques on one another which remain roughly constant for $\\sim t_{\\rm prec}$; however, after several times $t_{\\rm prec}$ the configuration of the wires and the resulting torques will be quite different. These torques cause the angular momentum vectors of the particles---and thus their eccentricities and inclinations---to random walk, with the duration of one ``step'' in the walk being roughly $t_{\\rm prec}$. However, the energies or semimajor axes are unaffected because the potential from the wires is nearly stationary.  A closely related process is resonant friction, which damps the eccentricities and inclinations of massive particles without affecting their semimajor axes. The combined effects of resonant relaxation and resonant friction drive the phase-space distribution of particles towards the maximum-entropy state consistent with fixed particle energies and total angular momentum. The main goal of the paper is to estimate the resonant rates of excitation and damping of protoplanet eccentricities, to compare these briefly with the non-resonant rates, and to demonstrate that in some cases resonant relaxation dominates the eccentricity evolution.  Resonant relaxation is a close cousin of secular perturbation theory in celestial mechanics, which likewise averages the Hamiltonian over times much longer than $t_{\\rm orb}$. Thus an alternative name for the process we are examining would be the oxymoron ``secular relaxation''.  Nearly Keplerian disks can be parametrized by two dimensionless numbers \\be S(r)\\equiv -{g(r)\\over n(r)}{\\mstar\\over\\pi r^2\\Sigma(r)}=-{g(r)n(r)r\\over\\pi G\\Sigma(r)},  \\qquad Q(r)\\equiv 1.07{\\sigma_r \\over n(r)r}{\\mstar\\over \\pi r^2\\Sigma(r)}; \\label{eq:sdef} \\ee here $g=\\dot\\varpi$ is the apsidal precession rate, $\\Sigma$ is the surface density of the disk, and $\\sigma_r$ is the rms radial velocity, related to the mean-square eccentricity by $\\sigma_r^2=\\case{1}{2}r^2n^2\\langle e^2\\rangle$. The second number is Toomre's $Q$-parameter (\\cite{BT87}) and $Q>1$ is required for axisymmetric stability. If the precession rate is determined by the self-gravity of the disk, then generally $S\\sim 1$. For example, a disk with a power-law surface density $\\Sigma(r)\\propto r^{-\\beta}$ has \\be S=2{\\Gamma(2-\\case{1}{2}\\beta)\\Gamma(\\case{1}{2}+\\case{1}{2}\\beta)\\over \\Gamma(\\frac{3}{2}-\\case{1}{2}\\beta)\\Gamma(\\case{1}{2}\\beta)},\\quad 0<\\beta<3; \\label{eq:spow} \\ee in this case $S$ is independent of radius and varies only between 1 and 1.0942 in the range $1\\le\\beta\\le 2$.  More generally, if there are other sources of precession then $|S|\\gg 1$. In the rings of Saturn and Uranus, where the precession is dominated by the planetary quadrupole moment, $S\\sim -10^5$. In protoplanetary disks containing a thin disk of planetesimals embedded in a thicker disk of gas, $S$ will be roughly the ratio of gas mass to planetesimal mass, $S\\sim 10^2$. We shall find that resonant relaxation is quite different in disks with $S\\sim 1$ and $|S|\\gg 1$.  To simplify the discussion, we shall focus on the effects of resonant relaxation on planets, that is, on bodies much larger than the planetesimals with which they are interacting. Also, we shall concentrate on resonant relaxation of the eccentricities, and generally ignore inclination relaxation except for a brief discussion in \\S \\ref{sec:disc}.  The equilibrium distribution of the particle orbital elements in a resonantly relaxed disk is determined in \\S \\ref{sec:thermo}. Estimating the rate of resonant relaxation is a more challenging task.  A variety of tools can be used for this purpose, depending on the properties of the disk.  We shall employ secular perturbation theory in \\S \\ref{sec:lag} to provide analytic estimates of the rate of resonant relaxation and resonant friction in disks. We carry out similar calculations using WKB density-wave theory in \\S \\ref{sec:wkb}, and numerical normal-mode calculations in \\S \\ref{sec:num}. The results are summarized and discussed in \\S \\ref{sec:disc}. ",
        "conclusions": "\\label{sec:disc}  \\noindent We have investigated resonant gravitational relaxation, which arises in particle disks orbiting in a nearly Keplerian potential. Resonant relaxation does not affect semimajor axes, and tends to produce a Rayleigh distribution in eccentricity and inclination, with mean-square eccentricity or inclination inversely proportional to mass (eq. \\ref{eq:ray}).  The nature of resonant relaxation depends strongly on whether apsidal precession is dominated by the self-gravity of the disk, a condition which is parametrized by $S(r)$ (eq. \\ref{eq:sdef}). If other mechanisms dominate the precession ($|S|\\gg 1$) then resonant relaxation requires that a secular resonance is present in the disk, which in turn requires $m_P/m_D\\lesssim S$, where $m_P$ and $m_D$ are the planet and disk masses (eq. \\ref{eq:range}). If the resonance is present, the rate of damping of the planet's eccentricity from resonant friction is roughly (eq. \\ref{eq:sbig}) \\be {1\\over e_P}{de_P\\over dt}\\sim -n\\left(m_D\\over \\mstar\\right)^{4/3}S^{1/3}\\left(m_P\\over\\mstar\\right)^{-1/3}, \\ee and the rate of stochastic excitation of the eccentricity is \\be {1\\over\\langle e_P^2\\rangle} {d\\langle e_P^2\\rangle\\over dt}= {\\langle m^2\\rangle\\over\\mstar\\langle m\\rangle}\\left(\\mstar\\over m_P\\right)\\left|{1\\over e_P}{de_P\\over dt}\\right|. \\ee These formulae also require that (i) the planet mass is large enough that the secular resonance is separated from the planet by more than the characteristic disk thickness or epicycle size (eq. \\ref{eq:range}); (ii) the separation exceeds the characteristic width of the resonance caused by the disk self-gravity, $m_P/m_D \\gtrsim S^{-1/2}$ (eq. \\ref{eq:jjjj}).  If on the other hand the apsidal precession is determined by the disk mass ($S\\sim 1$), the planet interacts with the large-scale low-frequency normal modes of the disk rather than with particles near the secular resonance. Calculating the rate of resonant relaxation in this case is harder (cf. \\S \\ref{sec:num}), especially if the planet is embedded in the disk.  For heuristic purposes the rate of eccentricity decay from resonant friction may be written as (eq. \\ref{eq:relaxe}) \\be {1\\over e_P^2}{de_P^2\\over dt} \\sim -n{m_P\\over \\mstar}. \\label{eq:kkkk} \\ee A key point is that the damping rate is independent of the disk mass $m_D$ so long as the disk mass determines the precession rate.  A major uncertainty in equation (\\ref{eq:kkkk}) is whether the planetary precession rate is in resonance with one of the disk eigenfrequencies; this usually requires either a continuous spectrum of eigenfrequencies or that dissipative or evolutionary processes broaden the discrete disk eigenfrequencies so that they overlap. Because the characteristic frequency of the disk modes is of order $n(m_D/\\mstar)\\ll n$, even slow evolutionary or damping processes cause considerable broadening. An important next step will be to understand the low-frequency, large-scale $m=1$ eigenmodes of nearly Keplerian disks.  For comparison, the non-resonant eccentricity damping rate for a planet embedded in a disk may be written as (\\cite{SW88}, \\cite{LS93}) \\be {1\\over e_P^2}{de_P^2\\over dt} \\sim -n{m_P\\over \\mstar}{m_D\\over \\mstar}\\left(v_c\\over v_r\\right)^4\\ln\\Lambda, \\ee where $v_r$ is the rms random velocity of the planetesimals, $v_c$ is the circular speed, and $\\ln\\Lambda$ is the Coulomb logarithm. As described in \\S \\ref{sec:intro}, this result is valid so long as $v_r\\gg nr_H$ where $r_H$ is the planet's Hill radius (\\cite{I90}).  If a planet orbits at a distance $\\Delta a$ from a narrow ring, the non-resonant damping rate becomes (\\cite{GT80}, eq.  31) \\be {1\\over e_P^2}{de_P^2\\over dt} \\sim -n{m_P\\over \\mstar}{m_D\\over \\mstar}\\left(a_p\\over |\\Delta a|\\right)^5. \\ee  These crude formulae suggest that either resonant or non-resonant eccentricity damping can dominate in a given disk, depending on the disk mass, velocity dispersion, and other parameters.  In typical protoplanetary disks, non-resonant damping probably dominates if the planet is embedded in the disk. However if there is a gap $\\Delta a/a_P\\gtrsim (m_D/\\mstar)^{1/5}$ then resonant damping may be stronger.  Resonant inclination relaxation is similar to resonant eccentricity relaxation. The principal difference is that the precession due to the self-gravity of the disk is $g \\sim n m_D/\\mstar (v_c/v_r)$, stronger by a factor of order $v_c/v_r$.  Thus the minimum value of the parameter $|S|\\sim v_c/v_r\\gg 1$.  Resonant relaxation is a complex process whose efficiency depends sensitively on properties such as the precession rate, the presence of gaps around planets, etc. (although it is much less sensitive to the rms random velocity in the disk than non-resonant relaxation). Moreover many of the tools that are usually employed to investigate disk relaxation (two-body scattering, Fokker-Planck approximation, Hill's problem, etc.) are too crude to describe resonant relaxation. Nevertheless this process is likely to play an important role in the evolution of a variety of nearly Keplerian disks, including both protoplanetary disks and other objects such as stellar disks surrounding black holes at the centers of galaxies."
    },
    "9805/astro-ph9805102_arXiv.txt": {
        "abstract": "The axion is the most attractive candidate to solve the strong CP problem in QCD. If it exists, the inflationary universe produces axion fluctuations which are mixtures of isocurvature and adiabatic fluctuations in general. We investigate how large isocurvature fluctuations are allowed or favored in order to explain observations of the large scale structure of the present universe.  Generic flat universe models with mixed (isocurvature$+$adiabatic) density fluctuations are studied.  It is found that the observations are consistent with the mixed fluctuation model if the ratio $\\alpha$ of the power spectrum of isocurvature fluctuations to that of adiabatic fluctuations is less than $\\sim 0.1$. In particular, the mixed fluctuation model with $\\alpha \\sim 0.05$, total matter density $\\Omega_{0} =0.4$, and Hubble parameter $H_0=70$km/s/Mpc gives a very good fit to the observational data.  Since the height of the acoustic peak in the angular power spectrum of the cosmic microwave background (CMB) radiation decreases significantly when the isocurvature fluctuations are present, the mixed fluctuation model can be tested in future satellite experiments. Ratios of the amplitude at the peak location to that at the COBE normalization scale for various models are given.  Furthermore, we also obtain the amplitude of isocurvature fluctuations as a function of axion parameters and the Hubble parameter during the inflation. We discuss the axion fluctuations in some realistic inflation models and find that a significant amount of the isocurvature fluctuations are naturally produced. ",
        "introduction": "The axion is a Nambu-Goldstone boson associated with spontaneous breaking of the Peccei-Quinn $U(1)$ symmetry which is introduced in order to solve the strong CP problem in quantum chromodynamics (QCD). The breaking scale $F_{a}$ of the Peccei-Quinn symmetry is stringently constrained by accelerator experiments, astrophysics and cosmology, and it should lie in the rage $10^{10} - 10^{12}$GeV in standard cosmology. In particular, the upper limit is set by requiring that the cosmic density of the axion does not exceed the critical density of the universe. In other words, the axion should represent the dark matter of the universe if $F_{a}$ is $\\sim 10^{12}$GeV, which makes the axion very attractive.  In the inflationary universe, quantum fluctuations of the inflaton field result in adiabatic density fluctuations with a scale-invariant power spectrum which would account for the large scale structure of the present universe.  This natural generation of density fluctuations is one of successes of the inflationary universe.  However, if the axion exists, another kind of fluctuations is produced in the inflationary universe.  During inflation the axion has quantum fluctuations whose root mean square amplitude $\\delta a$ is given by $H/(2\\pi)$. ($H$ is the Hubble parameter during inflation.)  The quantum fluctuations of the axion are stretched by the inflation and become classical.  When the axion acquires a mass $m_{a}$ at the QCD scale, these axion fluctuations result in density fluctuations of the axion.  However, since the axion is massless during inflation, axion fluctuations with wavelength larger than the horizon size do not contribute to the total density fluctuations. For this reason, such density fluctuations are called ``isocurvature''.  Therefore, if indeed the dark matter consists of axions, it has both adiabatic and isocurvature fluctuations and may play an important role in the structure formation of the universe.  In a previous work~\\cite{KSY} the large scale structure formation due to the mixed (adiabatic $+$ isocurvature) fluctuations of the axion was studied with matter density parameter $\\Omega_{0} =1$.  It was found that the introduction of isocurvature fluctuations significantly reduces the amplitude of the power spectrum $P(k)$ of the density fluctuations, keeping $P(k)$ normalized by the COBE data.  Since it is known that the cold dark matter (CDM) model with pure adiabatic density fluctuations and $\\Omega_{0}=1$ predicts an amplitude of $P(k)$ that is too large on galaxy and cluster of galaxies scales, the mixed fluctuation model gives a better fit to the observations, although the shape $P(k)$ on larger scales does not quite fit the observations.  Furthermore, it was also pointed out that the height of the acoustic peak in the angular power spectrum of the cosmic microwave background (CMB) radiation decreases when the isocurvature fluctuations are added, which can be tested in the future satellite experiments. Similar studies were also done in Refs. 2) and 3).\\footnote{A similar result was obtained in the context of the hot dark matter model.~\\cite{Scherrer}}   However, previous works deal with only restricted cosmological models (i.e. $\\Omega_{0} =1$).  It is well known that the shape of the power spectrum inferred from the galaxy survey~\\cite{Peacock} favors a low matter density universe.  Therefore, in this paper we consider the structure formation with the mixed fluctuations in a generic flat universe (i.e.  $\\Omega_{0} + \\lambda_{0} = 1$, where $\\lambda_{0}$ is the density parameter for the cosmological constant) and investigate how large isocurvature fluctuations are allowed or favored in order to explain the observations.  The constraint on the amplitude of the isocurvature fluctuations is reinterpreted as a constraint on the Hubble parameter $H$ and the Peccei-Quinn scale $F_{a}$.  We also study the cosmological evolution of the axion fluctuations in some inflation models. Among many inflation models, we adopt a hybrid inflation model and a new inflation model. In the hybrid inflation model, which is most natural in supergravity, we find that a significant amount of isocurvature fluctuations of the axion are naturally produced if $F_a \\simeq 10^{15-16}$~GeV. We note that late-time entropy production may easily raise the $F_a$ up to $\\sim 10^{16}$GeV. In the new inflation model, we also find that isocurvature fluctuations are produced with a canonically favored value of $F_a \\sim 10^{12}$GeV without late-time entropy production.  In \\S\\ref{sec:obs} we investigate the structure formation with mixed fluctuations and compare the theoretical predictions with the present observations. The generation of isocurvature and adiabatic fluctuations of the axion is described in \\S\\ref{sec:axion}. We discuss some inflation models and isocurvature fluctuations of the axion in \\S\\ref{sec:inflation-models}.  \\S\\ref{sec:conclusion} is devoted to conclusions and discussion.  Throughout this paper, we set the gravitational scale $\\sim 2.4\\times 10^{18}$GeV equal to unity. ",
        "conclusions": "\\label{sec:conclusion}  In this paper we have studied density fluctuations which have both adiabatic and isocurvature modes and have discussed their effects on the large scale structure of the universe and CMB anisotropies. By comparing the observations of the large scale structure, we have found that the mixed fluctuation model is consistent with observations if the ratio $\\alpha$ of the power spectrum of isocurvature fluctuations to that of adiabatic fluctuations is less than $\\sim 0.1$. In particular, the mixed fluctuation model with $\\alpha \\sim 0.05$, total matter density $\\Omega_{0} =0.4$, vacuum energy density $\\lambda_{0} = 0.6$, and Hubble parameter $H_0=70$km/s/Mpc gives a very good fit to both the observations of the abundance of clusters and the shape parameter. (In recent observations of high-redshift supernovae, a similarly large value for $\\lambda_{0}$ has been obtained.~\\cite{Riess}) Therefore, the mixture model of isocurvature and adiabatic fluctuations is astrophysically interesting.  The CMB anisotropies induced by the isocurvature fluctuations can be distinguished from those produced by pure adiabatic fluctuations because the shapes of the angular power spectrum of CMB anisotropies are quite different from each other on small angular scales. The most significant effect of the mixture of isocurvature fluctuations is that the acoustic peak in the angular power spectrum decreases.  From observations of the CMB anisotropies in future satellite experiments, we may know to what degree the axionic isocurvature fluctuations are present.  The requirement $\\alpha \\lesssim 0.1$ leads to constraints on the Peccei-Quinn scale $F_a$ and the Hubble parameter during the inflation. Isocurvature fluctuations with $\\alpha \\simeq 0.05$ are produced naturally in the hybrid inflation model if we take the Peccei-Quinn scale $F_a \\simeq 10^{15-16}$~GeV. This $F_a$ exceeds the usual upper bound ($F_a \\lesssim 10^{12}$~GeV). However, this bound can be relaxed if there exists late-time entropy production, and in this case $F_a \\sim 10^{15-16}$~GeV is allowed. Furthermore, $F_a \\simeq 10^{16}$~GeV is predicted in the M-theory.~\\cite{Banks} Thus the existence of large isocurvature fluctuations may provide crucial support for the M-theory axion hypothesis.~\\cite{Kawasaki-Yanagida,KKSY}  In the new inflation model, we also have sufficiently large isocurvature fluctuations to be observed. In this case, we do not have to invoke late-time entropy production to dilute the axion density. In a simple chaotic inflation model, we have a larger value for $H$ during inflation than the above two cases. In this case, $F_a$ exceeds the upper limit, even considering the late-time entropy production. However, the Peccei-Quinn scalar field may have an extremely flat potential, and $F_a$ may be as large as the gravitational scale. In this case, we have an appropriate amount of isocurvature fluctuations again.  In all of these three realistic classes of inflation models, we have astrophysically interesting amounts of isocurvature fluctuations."
    },
    "9805/astro-ph9805272_arXiv.txt": {
        "abstract": "It has been recently shown (Tsiklauri \\& Viollier, 1998a) that the matter concentration inferred from observed stellar motion at the galactic center (Eckart \\& Genzel, 1997, MNRAS, 284, 576 and Genzel et al., 1996, ApJ, 472, 153) is consistent with a supermassive object of $2.5 \\times 10^6$  solar masses, composed of self-gravitating, degenerate heavy neutrinos. It has been furthermore suggested (Tsiklauri \\& Viollier, 1998a) that the neutrino ball scenario may have an advantage that it could possibly explain the so-called \"blackness problem\" of the galactic center. Here, we present a quantitative investigation of this statement, by calculating  the emitted spectrum of Sgr A$^*$ in the framework of standard accretion disk theory. ",
        "introduction": "The enigmatic radio source Sgr A$^*$ at the galactic center has been a longstanding puzzle. Observations of stellar motions at the galactic center (Eckart \\& Genzel, 1997; Genzel et al., 1996) and low proper motion ($\\leq$ 20 km sec$^{-1}$; Backer, 1996) of  Sgr A$^*$ indicate that, on the one hand, it is a massive $(2.5 \\pm 0.4) \\times 10^6 M_\\odot$ object dominating the gravitational potential in the inner $\\leq 0.5$ pc region of the galaxy. On the other hand, observations of stellar winds and other gas flows in the vicinity of Sgr A$^*$ suggest that the mass accretion rate $\\dot M$ is about $6 \\times 10^{-6}M_\\odot$yr$^{-1}$ (Genzel et al., 1994). This implies that the luminosity of the central object should be more than $10^{40}$ erg sec$^{-1}$, provided the radiative efficiency is the customary 10\\%. However, observations indicate that the bolometric luminosity is actually less than $10^{37}$ erg sec$^{-1}$. This discrepancy has been a source of exhaustive debate in the recent past. The broad-band emission spectrum of  Sgr A$^*$ can be reproduced either in the quasi-spherical accretion model (Melia, 1992, 1994) with $\\dot M \\simeq 2 \\times 10^{-4}M_\\odot$ yr$^{-1}$ or by a combination of disk plus radio-jet model (Falcke et al., 1993a, 1993b). As pointed out by Falcke and Melia (1997), quasi-spherical accretion seems  unavoidable at large radii, but the low actual luminosity of Sgr A$^*$ points toward a much lower accretion rate in a starving disk. Therefore, Sgr A$^*$ can be described by a model of a fossil disk fed by quasi-spherical accretion. Another successful model which is consistent with the observed emission spectrum of Sgr A$^*$ has been developed by Narayan et al., 1995, 1998 and independently by Manmoto et al., 1997. This model is based on the concept of advection dominated accretion flow, in which most of the energy released by viscosity in the disk is carried along with the gas and lost into the black hole, while only a small fraction is actually radiated off.  Recently, Tsiklauri \\& Viollier  (1998a) have proposed an alternative model for the mass distribution at the galactic center in which the customary supermassive black hole is replaced by a ball composed of self-gravitating, degenerate neutrinos. It was shown that a neutrino ball with a mass $2.5 \\times 10^6 M_\\odot$, composed of neutrinos and antineutrinos with  masses $m_\\nu \\geq 12.0$ keV$/c^2$ for $g=2$ or $m_\\nu \\geq 14.3$ keV$/c^2$ for $g=1$, where $g$ is the spin degeneracy factor, is  consistent with the current observational data.  The purpose of this paper is to present calculations of the spectrum emitted by Sgr A$^*$ in the framework of standard accretion disk theory, assuming that Sgr A$^*$ is a neutrino ball with the abovementioned physical properties, and to show that this could resolve the \"blackness problem\".  In the recent past, Viollier et al. have proposed that massive, self-gravitating, degenerate neutrinos arranged in balls, where the degeneracy pressure balances self-gravity, can form long-lived configurations that could mimic the properties of dark matter at the centers of galaxies (Viollier, 1994; Viollier et al., 1993; Viollier et al., 1992). Tsiklauri \\& Viollier (1996)  have shown that a neutrino ball could play a similar role as a stellar cluster in the 3C 273 quasar, revealing its presence through the infrared bump in the emitted spectrum. Tsiklauri \\& Viollier (1998b) further investigated the formation  and time evolution of neutrino balls via two competing processes: annihilation of the particle-antiparticle pairs via weak interaction and spherical (Bondi) accretion of these particles. Bili\\'c \\& Viollier (1997) showed how the neutrino balls could form via a first-order phase transition of a system of self-gravitating neutrinos in the presence of a large radiation density background, based on the Thomas-Fermi model at finite temperature. They find that, by cooling a non-degenerate gas of massive neutrinos below a certain critical temperature, a condensed phase emerges, consisting of quasi-degenerate supermassive neutrino balls. General relativistic effects in the study of the gravitational phase transition in the framework of the Thomas-Fermi model at finite temperature were taken into account in Bili\\'c \\& Viollier (1998a). A theorem was proven by Bili\\'c \\& Viollier (1998b) which in brief states that the extremization of the free energy functional of a system of self-gravitating fermions, described by the general relativistic Thomas-Fermi model, is equivalent to solving Einstein's field equations. ",
        "conclusions": "Results of our numerical calculations are presented in  Fig.1, where we plot the quantity $\\nu L_\\nu$, calculated using Eq.(7). Data points are taken from Table 1 in Narayan et al., 1998. The thick solid line corresponds to the case of a neutrino ball with $\\dot m=4 \\times 10^{-3}$, whereas the thin solid line corresponds to $\\dot m= 10^{-4}$. The short-dashed line represents  the calculation with a $2.5 \\times 10^6 M_\\odot$  black hole with $\\dot m= 10^{-4}$ and an accretion disk extending from from 3 to $10^5$ Schwarzschild radii. The long-dashed line corresponds to the case when $\\dot m$ is artificially brought down to $10^{-9}$. As we see from  Fig.1, and as also was pointed out by Narayan et al., 1998, the latter two curves provide a  poor fit to the observational data.  Actually, this is the major reason why the standard  accretion disk theory was abandoned as a possible candidate for the description of the emitted spectrum from Sgr A$^*$. However, as originally was pointed out in Tsiklauri \\& Viollier (1998a) in the neutrino ball scenario, the accreting matter  experiences a much shallower gravitational potential than in the case of the black hole with the same mass, and therefore less viscous torque will be exerted. The radius of a neutrino ball of total mass $2.5 \\times 10^6 M_\\odot$, which is composed of self-gravitating, degenerate neutrinos and antineutrinos of mass  $m_\\nu = 12.0$ keV$/c^2$ for $g=2$ or $m_\\nu = 14.3$ keV$/c^2$ for $g=1$, is $1.06 \\times 10^5$ larger than the Schwarzschild radius of a black hole of the same mass. In this context it is important to note that the accretion radius $R_{\\rm A}=2GM/v^2_{\\rm w}$ for the neutrino ball, where $v_{\\rm w}\\simeq 700$ km/sec is the velocity of the wind from the IRS 16 stars, is approximately 0.02pc (Coker \\& Melia, 1997), which is slightly less than the radius of the neutrino ball, i.e. 0.02545 pc (for $m_\\nu = 12.0$ keV$/c^2$ for $g=2$ or $m_\\nu = 14.3$ keV$/c^2$ for $g=1$). Therefore, in the neutrino ball scenario, the captured accreting matter will always experience a gravitational pull from a mass less than the total mass of the ball. One can see from  Fig. 1, that for this very reason the theoretical spectrum in the case of the neutrino ball with $\\dot m=4 \\times 10^{-3}$ gives a much better fit than in the case of a black hole for any (even unrealistically lowered) values of $\\dot m$. Discrepancies between the theoretical and observed spectra appear in the case of the neutrino ball  for frequencies $<40$ GHz and $\\geq 10^{14}$ Hz.  At the higher end ($\\geq 10^{14}$ Hz) of the spectrum, the discrepancy is due to the fact that  our model does not incorporate effects of Compton-scattered synchrotron radiation (which  causes the second peak on the left in  Fig.1 of Narayan et al., 1998).  Our model is based on the simple-minded assumption of a steady, geometrically thin accretion disk which radiates off the gravitational binding energy locally, according to the black-body radiation law. However, even in this simplified framework, our model gives a reasonable fit in the radio to near infrared part of the spectrum. Besides, it is important to note that, as it has been shown by Falcke \\& Melia (1997), the evolution of an accretion disk can be considerably influenced by the deposition of mass and angular momentum by an infalling Bondi-Hoyle wind. The major result of their paper is that the modification of the standard accretion disk model, by taking into account the contribution from the Bondi-Hoyle wind and considering the physical picture of accretion process in dynamics, yields significant changes in the emitted spectrum. In fact, it produces an infrared bump, in addition to the Big Blue Bump, due to the deposition of energy in the outer part of the fossil accretion disk. Our paper is based on the standard accretion disk model i.e. without modifications arising from taking into account effects from the wind. In our case the gravitational potential is shallower than in the case of a supermassive black hole with the same mass. Therefore, taking into account effects from the Bondi-Hoyle wind and considering the non-steady problem (as in the case of Falcke \\& Melia's paper), both bumps will be shifted into the lower frequency domain. Thus the incorporation of Falcke \\& Melia's model of the accreting flow into our scenario of the dark matter distribution at the galactic center would presumably produce a better fit in the $\\leq 40$ GHz part of the spectrum.  It is important to address the issue of consistency of our model with intrinsic source size versus frequency data. For the test we take the  data of emission wavelength $\\lambda =$ 7 mm (Bower \\& Backer, 1998) and 3.5 mm (Rogers et al., 1994; Krichbaum et al., 1994). The upper limits on the intrinsic source size are $< 4.1$ AU (Bower \\& Backer, 1998) for 7 mm and $< 1.1$ AU (Rogers et al., 1994) and $2.8 \\pm 1.2$ AU (Krichbaum et al., 1994) for 3.5 mm assuming a distance to the galactic center of 8.5 kpc. Now, we have to estimate the radial location of the circles of the accretion disk in our model, emitting at these two wavelengths. For this purpose we assume that the corresponding temperature of a circle can be determined by the Wien displacement law $\\nu_m \\approx 3 k_{\\rm b} T /h \\approx 6 \\times 10^{10} T$ Hz (Lang, 1974), i.e.  we assume that the maximal frequency (wavelength) in the brightness distribution given the black body law determines the temperature of the emitting region. This assumption seems reasonable recalling the sharpness of the maxima in the brightness distribution (see Fig. 1 in Lang, 1974). Therefore we obtain $T_{\\rm 7mm}=0.71 K$ and $T_{\\rm 3.5mm}=1.43 K$. To find out to which values of the radial distance in the accretion disk these two values correspond, we have to use Eq.(6), graphically depicted in the Fig.2. The values are $\\xi_{\\rm 7mm}=8.71 \\times 10^{-4}$ and $\\xi_{\\rm 3.5mm}=1.145 \\times 10^{-3}$. The final predictions of our model would be twice these values ({\\it diameter} of the emitting circle) which in dimensional units are 2.50 AU and 3.29 AU for 7 mm and 3.5 mm, respectively. Thus we conclude that for 7mm  our model is consistent with the observations by Bower \\& Backer (1998); the same applies to the  data by Krichbaum et al. (1994) for 3.5 mm. However, currently some of the VLBI observations at millimeter wavelength stand in conflict with each other (Bower \\& Backer, 1998). Therefore, the discrepancy of our estimates with the Rogers et al., (1994) data is a matter of debate. Another important requirement which our model does satisfy is the lower limit on the size derived from the scintillation experiments (Gwinn et al., 1991). These experiments imply that the source diameter should be $> 0.1$ AU for 0.8 mm wavelength. Our estimates  show that at this wavelength the source diameter is 34.11 AU which corroborates the validity of our model.  As we can conclude from the latter paragraph, our model satisfies the source versus frequency constraints. However, it seems unlikely that such low temperatures as required by our model, especially close to the center of the neutrino ball, are actually realized, as the Galactic center is immersed in a hot radiation field. So far, the lowest termperature of one of the (black body radiation) emitting components introduced by Zylka, Mezger \\& Lesh 1992, was estimated to be 30 K, as a possible alternative to the self-absorbed synchrotron emission. Thus, we do not claim that in the mm wave-length band (which corresponds to disk temperatures of order few Kelvin) our model is capable of explaning the emitted spectrum reliably. The fit of the spectrum at these frequences ($<40$ GHz), based on standard accretion disk theory for a disk immersed in the field of the neutrino ball, is anyway not good, and it seems that some other, presumably non-thermal mechanism is responsible for the radio  emission of Sgr A$^*$. Moreover, as  has been shown by Reynolds \\& McKee 1980, the compact radio source  at the Galactic center could be a pulsar, with total emitting power comparable to that of the Crab pulsar. These authors showed that Sgr A$^*$ can be understood in the framework of a number of dynamically self-consistent models of incoherent synchrotron sources which are energetically comparable with to the energy output of a few $\\times 10^{38}$ erg sec$^{-1}$ like the Crab pulsar. The radio pulsar, together with  a Shakura-Sunyaev disk embedded in the shallow gravitational potential of the neutrino ball, may be the key to the understanding of the emission spectrum of Sgr A$^*$. In this context, it is perhaps important to point out that the low proper motion of the central radio source ($\\leq$ 20 km sec$^{-1}$; Backer, 1996) could be explained by a slowly moving pulsar near the minimum of the gravitational potential of the neutrino ball, a possibility which of course would be excluded in a supermassive black hole scenario.  Apart from emission of high-energy radiation by the pulsar, and X-ray emission of the neutrino ball through radiative decay of the constituent neutrinos into light neutrinos, our model would be incapable of describing of X-rays and gamma-rays. First, even standard accretion disk theory around a central black hole cannot account for the emission of radiation above ultra-violet. The simplest model which might produce X-rays is the two-component plasma model by Shapiro, Lightman \\& Eardley 1976. However, in the case of a neutrino ball with the physical parameters mentioned above, it is impossible to get X-rays and gamma-rays from the accretion disk by {\\it definition}. In order to get an appreciable fraction of the rest mass of the electron converted into X-rays, the particles need to reach a sizable fraction of the velocity of light, which would be impossible in our scenario as  the escape velocity from the center of the neutrino ball is about 1400 km s$^{-1}$. This is, in fact, the reason why our model can explain the blackness problem of the Galactic center.  Let us now take an unprejudiced view on the high-energy data: In the 0.8-2.5 keV band the data  avaliable from ROSAT (Predehl \\& Tr\\\"umper 1994) have a resolution of $\\sim 20^{''}$; in the 2-10 keV band data available from  ASCA (Koyama et al. 1996) the resolution is $\\sim 1'$. The 35-150 keV data from  SIGMA (Goldwrum et al. 1994) have a resolution of $\\sim 15'$, while the EGRET data (Merck et al. 1996) from 30 MeV to 10 GeV have a resolution of $\\sim 1^\\circ$. The intrinsic size of the accretion disk (which is about the size of the neutrino ball) is about $0.65''$ which is much smaller than the resolution of current X-ray or gamma-ray detectors. Even the measurement by Predehl \\& Tr\\\"umper 1994, who established that the X-ray source is within a $10''$ (0.5 pc) distance from the Sgr A$^*$ is not conclusive. As discussed in detail in Tsiklauri \\& Viollier 1998a, a neutrino ball would also produce  X-rays via the radiative decay of the heavy into light neutrinos. Tsiklauri \\& Viollier 1998a have estimated that the luminosity of this emission line at energy $\\sim m_\\nu c^2/2$ (which would not be a sharp line due to scattering by the existing matter at the Galactic center) should be $L_\\gamma \\leq 1.45 \\times 10^{34}$ erg/sec. In fact, this luminosity is consistent with the observations by Predehl \\& Tr\\\"umper 1994.   Finally, we would like to emphasize that the idea that Sgr A$^*$ may be an extended object rather than a supermassive black hole is not new (see e.g. Haller et al., 1996; Sanders, 1992). To our knowledge all previous such  models  assume that the extended object is of a baryonic nature, e.g.  a very compact stellar cluster. However, it is commonly accepted that these models face problems with stability, and it has been questioned whether such clusters are long-lived enough, based on evaporation and collision time-scales stability criteria (for a different point of view see Moffat, 1997). Our model of Sgr A$^*$ is surprisingly simple while it satisfies all current observational constraints: First, a neutrino ball is a stable object quite alike an ordinary baryonic star, though much more massive, with the difference that its self-gravity is compensated by the degeneracy pressure of the neutrinos rather than thermal pressure as in the case of a baryonic star. Second, a neutrino ball with the abovementioned physical parameters is compact enough as to be virtually indistinguishable from a $2.5 \\times 10^6 M_\\odot$ black hole with current observational resolution ($\\approx 10^5$ Schwarzschild radii) of the observations of proper stellar motions (Eckart \\& Genzel, 1997; Genzel et al., 1996). Third, a neutrino ball of this mass  can  explain its low proper motion ($\\leq$ 20 km sec$^{-1}$; Backer, 1996). Fourth, as a bonus of our model, the neutrino ball is extended enough to provide a much  shallower gravitational potential than a $2.5 \\times 10^6 M_\\odot$ black hole for the accreting matter, thus producing a reasonable emission flux."
    },
    "9805/astro-ph9805208_arXiv.txt": {
        "abstract": "\\noindent This paper reviews the measurements of galaxy correlations at high redshifts, and discusses how these may be understood in models of hierarchical gravitational collapse. The clustering of galaxies at redshift one is much weaker than at present, and this is consistent with the rate of growth of structure expected in an open universe. If $\\Omega=1$, this observation would imply that bias increases at high redshift, in conflict with observed $M/L$ values for known high-$z$ clusters. At redshift 3, the population of Lyman-limit galaxies displays clustering which is of similar amplitude to that seen today. This is most naturally understood if the Lyman-limit population is a set of rare recently-formed objects. Knowing both the clustering and the abundance of these objects, it is possible to deduce empirically the fluctuation spectrum required on scales which cannot be measured today owing to gravitational nonlinearities. Of existing physical models for the fluctuation spectrum, the results are most closely matched by a low-density spatially flat universe. This conclusion is reinforced by an empirical analysis of CMB anisotropies, in which the present-day fluctuation spectrum is forced to have the observed form. Open models are strongly disfavoured, leaving $\\Lambda$CDM as the most successful simple model for structure formation. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805178_arXiv.txt": {
        "abstract": "A new x-ray transient, designated \\xj , was discovered with the Proportional Counter Array (PCA) on the \\it Rossi X-ray Timing Explorer (RXTE) \\rm in data from 12 August 1997.  Although it was the brightest source in the Galactic Center region while active (about $3.0 \\times 10^{-9}$ergs cm$^{-2}$s$^{-1}$ from 2 to 25 keV), it was only observed on that one day; it was not detectable nine days earlier or two days later.  There is no known counterpart at other wavelengths, and its proximity to the Galactic Center will make such an identification difficult due to source confusion and extinction. The x-ray spectrum and intensity suggest a giant outburst of a Be/neutron star binary, although no pulsations were observed and the outburst was shorter than is usual from these systems. ",
        "introduction": "\\begin{figure}[t!] \\centerline{\\epsfysize=2.5in \\epsfbox{figure1.eps}} \\caption{ Count rates from weekly PCA pointings toward \\onee .  Both instrumental and Galactic diffuse background have been subtracted.  The high point is the 12 August observation.  Statistical errors are smaller than the symbols used for the data points. } \\end{figure}  Most bright x-ray transients lasting days or weeks come from two kinds of systems: black hole binaries with a low-mass companion (x-ray novae) and neutron star binaries with a high-mass, Be-type companion (Be/NS).  In the latter systems, the neutron star is in a relatively wide, often eccentric orbit, and the outburst occurs at the orbital phase when the neutron star passes through the dense equatorial wind of the Be star, accreting wind material.  These systems can be distinguished from each other and from rarer kinds of transients by neutron star pulsations (if present), by optical observations of the companion (if available), by regularity of recurrence (which indicates a Be/NS binary), by a weak secondary outburst peak (which is a signature of an x-ray nova), or by their x-ray spectra (see sections 5 and 6).  We report a new x-ray transient in the Galactic Center region.  This source, \\xj , was discovered in an observation of the black hole candidate \\onee\\ with the Proportional Counter Array (PCA) on the \\it Rossi X-ray Timing Explorer (RXTE)\\rm .  The transient was initially reported in an IAU Circular (\\markcite{Sm97a}Smith et al. 1997a), with a corrected position reported later (\\markcite{Sm97}Smith 1997), unfortunately after radio observations of the original field (\\markcite{Hj97}Hjellming 1997).  \\begin{figure}[t!] \\centerline{\\epsfysize=3.0in \\epsfbox{figure2.eps}} \\caption{ Map of the Galactic center region showing well-known x-ray sources, the two scans from 12 Aug 1998, and the resulting position fit for XTE J1739-302. GRO J1744-28 (the bursting pulsar) and SLX 1732-304 (an x-ray burster in Terzan 1), were not active at this time.} \\end{figure}  On 12 August 1997 at 16:58 UT, the PCA scanned its 1\\dg\\ FWHM field of toward a weekly observing position near \\onee , approaching almost exactly along a line of constant right ascension.  The count rate, instead of rising smoothly from the background level as \\onee\\ entered the field of view, instead rose much higher just before the end of the scan, came back down, and leveled off at a rate over twice that expected for \\onee\\ after background subtraction (Figure 1).  This indicated a bright source positioned just before \\onee\\ in the scan direction, but still within the field of view of the PCA once it came to rest.  \\begin{figure}[t!] \\centerline{\\epsfysize=5.5in \\epsfbox{figure3.eps}} \\caption{ PCA count rates during the two scans over \\xj .  Top: the scan at nearly constant right ascension.  Bottom: the scan at constant Galactic latitude. The best-fit curve for each is made up of a sum of four contributions, each in the form of a collimator response times an intensity.  The four components are also shown and labeled as follows: A) \\xj\\ in its best-fit position; B) \\onee\\ ; C) Galactic Center diffuse emission (a different component in each plot - see text); D) a constant instrumental background. } \\end{figure}  Two days later, on 14 August 1997, another scan passed within 0.55\\dg\\ of \\xj , but no peak was seen.  This sets a 3$\\sigma$ flux upper limit of about 4\\% of the flux on 12 August at 16:58 UT.  The limit for 3 August can be derived from Figure 1: the rms variation from week to week in the observed flux from \\onee\\ (not counting the 12 August data point) is 8 counts/s.  Since the 3 August observation is, in fact, lower than average, the upper limit for \\xj\\ on this day is even stricter than the 3$\\sigma$ variation of 24 c/s, or 14\\% of the 12 August flux of \\xj .  Data from the Wide Field Camera on \\it Beppo SAX \\rm show no detection of the source on 6 September, with an upper limit of 20 mCrab from 2 to 10 keV, roughly 30\\% of the outburst flux. ",
        "conclusions": "Most bright Galactic x-ray transients are either black hole candidates or binaries consisting of a Be star and a neutron star (Be/NS).  We tentatively identify \\xj\\ as a Be/NS binary because its spectral shape is similar to that of these systems: a gradual steepening over the 2 to 25 keV range.  In this energy range, the black hole candidates show either two distinct components (a very soft thermal component and a power law) or else a single, very hard power law.  Our parameters for the cutoff-power-law model of the \\xj\\ spectrum fall within the ranges of the x-ray pulsars listed in \\markcite{Wh83}White et al. (1983), which include Be/NS binaries.  Our temperature for the exponential approximation to a bremsstrahlung spectrum (12.4 keV) is near the low end of the ranges of GS~0834-430 (12-18 keV; \\markcite{Wi97}Wilson et al. 1997) and A0525+262 (14 to 22 keV; \\markcite{Fi96}Finger, Wilson \\& Harmon 1996) measured with BATSE.  As noted above, spectral forms typical of black hole candidates were either poor fits or gave parameters far from the usual range.  With the OTTB model used in section 5, the luminosity from 20 to 50 keV $(1.05 \\times 10^{37}$ergs s$^{-1}$, about 1/4 of the 2 to 100 keV luminosity) is at the high end of the expected range for a ``giant'' outburst of a Be/NS binary (\\markcite{Bi97}Bildsten et al. 1997).  The short duration ($<$ 11 days) would be not be typical for a giant outburst (\\markcite{Bi97}Bildsten et al. 1997), although one giant outburst of A0535+262 was only 18 days long (\\markcite{Se90}Sembay et al. 1990).  With no detected pulsations and such a short sample of data, we can't distinguish between a normal and giant outburst by spin-up rate.  If we caught the source turning on, then at $< 2$ days it is not similar to any common class of transient, although the lightcurve is consistent with that recently observed from XTE J0421+560 = CI Cam (\\markcite{Sm98}Smith et al. 1998).  That bright transient showed a fast rise over hours followed by an exponential decay with a time constant on the order of a day (quick-look results provided by the ASM/\\it RXTE \\rm team).  Otherwise, it behaved very differently from \\xj , showing a bright Fe-K line and no time variability besides its smooth decay (\\markcite{Be98}Belloni et al. 1998).  What the events have in common is that they underscore the need for very frequent monitoring and very fast response in the study of Galactic x-ray transients.  If we are correct in the spectral identification of \\xj\\ as a Be/NS binary, the pulse period may simply have been too long ($>$ 300 s) to be seen in our short (1300 s) stretch of data, but the lack of detection could also be due to a very low pulsed fraction.  A pulsed fraction of only 4 to 6\\% rms was seen in X0331+53 (= V0332+53 = BQ Cam) along with flat-topped noise with a total rms of 25\\% from 0.01 to 1 Hz (\\markcite{Ma90}Makishima et al. 1990).  Our power spectrum is similar to that of the A0535+262 outburst of 1980 October with its pulsed component removed (\\markcite{Fr85}Frontera et al. 1985). Although that power spectrum had a broad peak around 0.05 Hz detected with more significance than the similar feature in our spectrum (see Figure 5), the authors interpreted it as due to imperfect subtraction of the pulsation, so the coincidence should be viewed cautiously. Like ours, their power spectrum showed significant low-frequency noise below 0.02 Hz and little power above 0.1 Hz.  Our best hope to learn more about this object is to catch a subsequent outburst with a longer exposure and multiple instruments.  If it is a Be/NS binary, its unusually brief and luminous outburst makes particularly interesting.  We encourage anyone with access to appropriate archival data to look for the August 1997 outburst and any earlier ones, particularly to search for pulsar periods longer than 300 s.  We also encourage anyone with new Galactic Center data to look at it promptly and call for target of opportunity observations from the various high-energy observatories during any subsequent outburst of this object.  The authors would like to thank the \\it RXTE \\rm Guest Observer Facility for their continuing assistance, and R. Hjellming for his prompt response in scheduling radio observations of the originally reported position.  This work was supported by NASA grants NAG5-3599 and NAG5-4110."
    },
    "9805/astro-ph9805050_arXiv.txt": {
        "abstract": "Certain problems in the field of stellar atmospheres, polarised radiative transfer and magnetic field diagnostics cannot be addressed by means of traditional sequential programming techniques because CPU times become prohibitive on even the fastest single processor machines when realistic physics and accurate numerical methods are employed. This contribution discusses the question of what kind of supercomputing approach is best suited for the modelling of Ap stars, pointing out the superiority of parallel computing with Ada95 over High Performance Fortran in all of the above-mentioned fields. ",
        "introduction": "\\label{intr}  In recent times, problems in the modelling of both nonmagnetic and magnetic stellar atmospheres have emerged that cannot be solved by traditional numerically intensive computing. Take as an example the calculation of line-blanketed LTE stellar atmospheres by means of opacity sampling as done in {\\sl ATLAS12}. Even on present-day fast single-processor machines realistic frequency step sizes lead to rather prohibitive CPU times (see Castelli, these proceedings). The same applies to the modelling of broadband linear (BBLP) and circular (BBCP) polarisation in sunspots, in the solar network and in magnetic stars, and to full Zeeman Doppler Imaging (ZDI). The usual remedies to this software crisis do not excel in imagination: employing coarse frequency grids, approximate formal solvers or Milne-Eddington atmospheres is a cheap expedient but does not attack the problem at its root.  Instead of waiting for the next increase in CPU clock rates to be able to run today's models on tomorrow's computers, it would be preferable to go parallel. Indeed, parallel architectures are readily available nowadays and languages with parallel constructs provide concurrent execution of program segments. Obviously there is no way around the restructuring and at least partial rewriting of existing programs, but isn't this preferable to physically doubtful approximations or to poor frequency sampling?  High Performance Fortran (HPF) would seem the obvious choice for the denizens of the Fortran universe, but does the data parallel paradigm of HPF really provide optimum parallelism in spectral line synthesis, be it LTE or NLTE, polarised or unpolarised? Would HPF really speed up stellar atmosphere calculations with a suitably modified version of {\\sl ATLAS}, and what about ZDI? Isn't it most likely that we have to rethink our approach in a far more radical way? ",
        "conclusions": ""
    },
    "9805/astro-ph9805099_arXiv.txt": {
        "abstract": "We present a dynamical study of the CL0023+0423 system at a redshift of $z = 0.84$.  This system consists of two components separated in velocity space by $\\sim 2900~{\\rm km~s^{-1}}$ and on the plane of the sky by $\\sim 0.23~h^{-1}~{\\rm Mpc}$. A kinematic analysis indicates that the two components are a poor cluster with a velocity dispersion of $415^{+102}_{-63}~{\\rm km~s^{-1}}$ and a mass of $\\sim 3 - 6 \\times 10^{14}~h^{-1}~{\\rm M_{\\odot}}$ and a less massive group with a velocity dispersion of $158^{+42}_{-33}~{\\rm km~s^{-1}}$ and a mass of $\\sim 10^{13}~h^{-1}~{\\rm M_{\\odot}}$ (Postman, Lubin \\& Oke 1998). The dynamics of galaxy groups at high redshift can provide important insights into the creation of present-day galaxy clusters. Therefore, we have performed a dynamical study on this system in order to determine whether the two groups are infalling. This analysis includes an analytic two-body calculation and N-body simulations.  The results of both studies indicate that the system is most likely not bound but simply a chance projection on the sky; however, within the observational uncertainties, there do exist bound solutions where the two galaxy groups are currently moving toward each other and will eventually merge into a larger system of galaxies.  We have run one-thousand N-body simulations with random initial conditions based on the observed parameters of the CL0023+0423 groups. A statistical analysis of these simulations indicates that there is an 20\\% chance that the two groups will merge.  If the CL0023+0423 system does merge, it will appear as a cluster on the sky, as well as in velocity space, within $1-2$ Gyrs. The cluster will evolve dynamically for more than 3 Gyrs, appearing during this time more similar to an open, irregular cluster. The final merged system has a velocity dispersion which is consistent with a local Abell richness class 1 cluster.  The morphological analysis of the galaxy populations of CL0023+0423 suggests that both groups are largely dominated by spiral galaxies. Early-type fractions are 33\\% or less (Lubin et al.\\ 1998).  These modest early-type fractions have implications for both cluster formation and group evolution. Studies of open clusters at $z = 0.31 - 0.54$ indicate that that they have early-type fractions between $45 - 80\\%$ (Dressler et al.\\ 1997; Andreon, Davoust \\& Helm 1997; Stanford, Eisenhardt \\& Dickinson 1997; Couch et al.\\ 1998).  If the CL0023+0423 system is the predecessor of such a cluster, the comparison may suggest that some fraction of early-type galaxies are formed between redshifts of $z \\sim 0.8$ and $z \\sim 0.3$; however, the morphological fractions are still highly uncertain.  In addition, the modest early-type fractions in both groups may be inconsistent with the strong correlation between velocity dispersion and early-type fraction observed in nearby groups of galaxies (Zabludoff \\& Mulchaey 1998). Both groups apparently have relatively low early-type populations, irrespective of their velocity dispersion.  If the groups of CL0023+0423 are typical of galaxy groups at high redshift, and if high-redshift groups are the progenitors of local groups, this result may also imply that some early-type formation is occurring at redshifts of $z \\simless 0.8$.  These results do not preclude the formation of early-type galaxies at very high redshift as many observations suggest (e.g.\\ Steidel et al.\\ 1996; Ellis et al.\\ 1997; Stanford, Eisenhardt \\& Dickinson 1997; Postman, Lubin \\& Oke 1998); however, the observations of the CL0023+0423 system may imply that a fraction of the early-type population is forming and/or undergoing significant evolution at redshifts of $z < 1$. ",
        "introduction": "In hierarchical clustering models, clusters of galaxies are formed by a progressive coalescence of an inhomogeneous system of subclusters (e.g.\\ White 1976; Evrard 1990; Lacey \\& Cole 1993,1994).  The early stages of these models provide a natural description of the formation of open, irregular clusters like Virgo or Hercules. These systems show no single, central condensation, though the galaxy surface density is at least five times as great as the surrounding field.  They often have high degrees of asymmetry and significant amounts of subclustering.  The progressive stages of evolution in cluster formation produces centrally-concentrated, compact clusters of galaxies. These clusters are dense, have a single, prominent concentration among the bright member galaxies, and typically display a high-degree of spherical symmetry (Abell 1958; Oemler 1975; Bahcall 1975; Dressler 1978; Dressler 1980a,b).  The remnants of hierarchical structure formation can be easily observed in the local universe. Substructure and more significant merger events are common characteristics of the optical and X-ray properties of nearby clusters of galaxies (e.g.\\ Geller \\& Beers 1982; Dressler \\& Shectman 1988; West \\& Bothun 1990; Jones \\& Forman 1992; Mohr, Fabricant \\& Geller 1993; Zabludoff \\& Zaritsky 1995).  These studies indicate that $30 - 40\\%$ of all clusters have substantial substructure where the subclump contains a significant fraction of the cluster galaxies (Dressler \\& Shectman 1987; West \\& Bothun 1990; Jones \\& Forman 1990). Simulations have shown that, when a cluster accretes smaller groups of galaxies, the effect of such a merger, such as optical or X-ray morphologies, can last for $\\sim 1$ Gyr (Evrard 1990). For a more significant merger of two equal-mass subclusters, it takes more than 4 Gyr after the first encounter until the collisionless dark matter (and galaxy) density contour shows a single peak.  The exact time depends on the density and velocity distribution of the dark matter.  A longer relaxation time is required when the dark matter is spatially extended and/or when the velocity distribution has a Gaussian shape (Nakamura, Hattori \\& Mineshige 1995).  It is worthwhile to note that hydrodynamic simulations indicate that the collisional, X-ray emitting intracluster gas can relax on timescales much shorter than the dark matter (see e.g.\\ Evrard 1990).  Spectral studies of rich clusters with significant substructure indicate that $\\sim 15\\%$ of the early-type cluster members have signs of current or on-going star formation (Caldwell \\& Rose 1997).  These observations may imply that merger activity can alter the star formation histories of the early-type members. Starbursts may be the result of shocks which are induced in the collision between the intracluster medium of the cluster and that of the subcluster. Zabludoff \\& Mulchaey (1998) find that a similar fraction of the early-type galaxies in groups of galaxies in the field have experienced star formation within the last $\\sim 2$ Gyr. Therefore, if some of the subclusters in these rich clusters are actually groups which have fallen into the cluster environment, the similarity between the star formation histories of the early-type galaxies in the subclusters and those in the galaxy groups indicate that the group environment may be the site of recent star formation. The modest intragroup medium and low velocity dispersions would argue that galaxy-galaxy collisions would be the dominant means to alter the star formation history of the early-type galaxies in these systems (Aarseth \\& Fall 1980; Barnes 1985; Merritt 1995; Zabludoff \\& Mulchaey 1998). The results suggest that both group and cluster environments should be examined as important sites for evolution in the star formation history of galaxies.  In order to understand the evolution of galaxies and their environments, it is essential to study groups and clusters of galaxies in their early stages of formation.  Unfortunately, because of the substantial background contamination rates at high redshift, finding systems which are in the process of forming is very difficult. However, we have found such a system as part of a high-redshift cluster survey.  At a redshift of $z = 0.84$, the CL0023+0423 system is part of an extensive photometric, spectroscopic, and morphological study of nine candidate clusters of galaxies at $z \\simgreat 0.6$ (Oke, Postman \\& Lubin 1998, hereafter Paper I).  First detected as a likely cluster by Gunn, Hoessel \\& Oke (1986), more detailed spectroscopic data reveal that this system is actually made up of two substructures separated closely in velocity, as well as position on the sky. The radial velocity histograms of the individual structures more closely resemble local groups or poor clusters of galaxies (Postman, Lubin \\& Oke 1998, hereafter Paper II). In addition, the morphological analysis of the galaxy population in this system indicates that these groups do not have a morphological composition which is typical of a massive, relaxed cluster. Their galaxy population is almost totally dominated by spiral galaxies, with early-type fractions that are 33\\% or less (Lubin et al.\\ 1998, hereafter Paper III). This study provides the first observational data on a possible group-group merger at high redshift.  Because of the intriguing nature of this system and its potential to reveal important information on the hierarchical formation of larger systems such as clusters of galaxies, we have decided to study in detail the dynamical state of this two group system. In this paper, we present the results of an analytic analysis of the system dynamics, as well as N-body simulations of the group-group system.  We address the possibility that this system is bound and will eventually merge, or that the system is simply a chance projection of two groups on the sky. A brief description of the observations are given in Sect.\\ 2. In Sects.\\ 3 and 4, we describe the analytic dynamical analysis and the N-body simulations, respectively. A discussion of the results and their implications for the formation and evolution of groups and clusters of galaxies is presented in Sect.\\ 5.  In the following analyses, we have assumed $q_{0} = 0.1$ and $H_{0} = 100~h~{\\rm km~s^{-1}~Mpc^{-1}}$. ",
        "conclusions": "In this paper, we have used an analytic analysis, as well as N-body simulations, to examine whether the observed group-group system of CL0023+0423 will eventually merge. Both studies have shown that this system is most likely unbound and simply a chance projection on the sky.  However, within the observational uncertainties, there do exist some system configurations where the two groups are bound and are in the process of merging. In the analytic analysis of the system dynamics, we have shown that there are bound-incoming solutions for this two body system if the radial separation of the two groups is on the order of $\\sim 0.23~{\\rm Mpc}$, the total mass of the system is larger by 1$\\sigma$ ($M \\sim 7 \\times 10^{14}~h^{-1}~{\\rm M_{\\odot}}$), and the radial velocity difference between the two groups is smaller by 1$\\sigma$ ($V_r \\sim 2700~{\\rm km~s^{-1}}$).  In addition, we have run 1000 N-body simulations with initial parameters that are randomly chosen to be within $\\pm 2\\sigma$ of the observed values.  We find that the two groups have successfully merged in 201 simulations.  Therefore, if the formal errors are an accurate measure of the uncertainty in the observational parameters, the group-group system of CL0023+0423 has an 20\\% chance of merging within a few Gyrs. This probability is, however, optimistic as the successful mergers depend strongly on certain parameters, specifically the mass of CL0023B and the radial velocity difference between the two groups. The N-body simulations indicate that, for a merger to occur, the system must have a mass which is on the high end and a radial velocity difference which is on the low end of the possible range in observed values. However, if the groups are not truly virialized, the mass will be overestimated by as much as a factor of 2 (Small et al.\\ 1998); therefore, the choice of a large mass would go in the opposite direction of the likely systematic bias.  If the two groups eventually merge, the resulting system has a characteristic velocity dispersion and harmonic radius which are typical of a richness class 1 or 2 cluster.  For the discussion of the implications for cluster formation, we will assume that the group-group system in CL0023+0423 will eventually merge. The simulations indicate that the merged systems will appear as a cluster of galaxies within $\\sim 1-2~{\\rm Gyr}$ (at $z \\simless 0.5$); therefore, the groups of galaxies in CL0023+0423 are likely to be the progenitors of some intermediate-redshift clusters.  As discussed in Sect.\\ 4, the resulting cluster is still evolving dynamically after 3 Gyrs or at redshifts of $z \\simless 0.1$. Therefore, over this time period, the system would appear more like an open cluster which are irregular, loose, and presumably dynamically young.  Several clusters belonging to this class have been studied at redshifts of $z = 0.31 - 0.54$.  High-resolution HST imagery of these clusters have been used to measure the morphological composition of the galaxy populations.  Because the number of spectroscopically confirmed cluster members is limited, the morphological fractions are determined statistically using background-corrected distributions of morphologies.  These studies imply elliptical fractions between $27-47\\%$ and total early-type (E + S0) fractions between $45-80\\%$ (Dressler et al.\\ 1997; Stanford, Eisenhardt \\& Dickinson 1997; Andreon, Davoust \\& Heim 1997; Couch et al.\\ 1998).  Classifiers of these galaxies estimate that the errors in a particular morphological bin is approximately $\\pm 20\\%$ (see Smail et al.\\ 1997; Andreon, Davoust \\& Heim 1997; Andreon 1998). Comparisons between independent studies confirm these uncertainties and may imply even larger ones; for example, take the open cluster CL0939+4713 at a redshift of $z = 0.41$. Andreon, Davoust \\& Heim (1997) find an elliptical fraction of 15\\%, while Dressler et al.\\ (1997) find 37\\% ($M_V < -20; h = 1, q_0 = 0.5$). These fractions are statistically different and are not the result of small number statistics (Andreon 1998).  The fractions for the total early-type population appear more consistent. For this cluster, independent classifiers find early-type fractions of 49, 56, and 60\\%, respectively (Andreon, Davoust \\& Heim 1997; Dressler et al.\\ 1997; Stanford, Eisenhardt \\& Dickinson 1997). However, as implied by the large discrepency in the elliptical fraction, there is contention about the ratio of elliptical to S0 galaxies in both open and compact clusters at these redshifts.  Dressler et al.\\ (1997) have examined 10 clusters at $z = 0.37 - 0.56$ and find that, compared to local clusters, the fraction of S0 galaxies is smaller by a factor of $\\sim 2-3$. However, other classifiers of the same clusters do not find this discrepency (Stanford, Eisenhardt \\& Dickinson 1997; Andreon, Davoust \\& Heim 1997). Therefore, it is fair to say that the exact morphological fractions in intermediate-redshift clusters are still not accurately known.  In Paper III, we have performed a similar morphological analysis on the two groups in CL0023+0423 using both high-resolution HST imagery and the spectral characteristics of the member galaxies.  The morphological classifications of the galaxies within the HST field-of-view ($\\sim 150^{''} \\times 150^{''}$) were made according to the Revised Hubble system of nearby galaxies (e.g.\\ Sandage 1961; Sandage \\& Bedke 1994).  The spectroscopic coverage of the HST field is limited with only 12 confirmed members. Seven are classified as spiral or irregular galaxies. The remaining five have been classified as either an elliptical or compact galaxy; none were classified as S0 galaxies.  We expect that these identifications are relatively robust and directly comparable to the classifications of nearby galaxies as we are morphologically classifying these high-redshift galaxies just slightly bluewards of the well-studied $B$ band and to roughly the same limiting surface brightness as local samples (see Paper III).  Of the five galaxies classified as ellipticals, three show old, red absorption spectra typical of present-epoch ellipticals. The other two have asymmetric disks and are identified as possible mergers (see Paper III).  Their spectra show strong star-forming features and are consequently bluer than the other ellipticals. Their properties are more consistent with those of blue compact galaxies found at redshifts of $z \\sim 0.1 - 0.7$ in previous HST observations (Koo et al.\\ 1994, 1995). All other galaxies in the HST field-of-view with strong \\ion{O}{2} emission (equivalent widths of greater than 10 \\AA) are classified as spirals or irregular/peculiar galaxies.  Therefore, based on both spectral and morphological evidence, we find an early-type fraction of 25\\% (3 out of 12) in the central region of the CL0023+0423 system.  If we try to improve the statistics by examining the spectral characteristics of the confirmed group members over the larger LRIS field-of-view ($\\sim 2^{'} \\times 8^{'}$; see Paper I), we find that only 17\\% (4 out of 24) of the confirmed members have old, red absorption spectra (ages of $\\simgreat 3$ Gyrs; see Paper II). Two-thirds of all group members have active star formation with \\ion{O}{2} equivalent widths which are typically much larger than 10 \\AA; the average \\ion{O}{2} equivalent width is 35 \\AA.  If we assume that all of these galaxies are spirals or irregulars (as is the case for those galaxies with HST morphologies; see above), the spectral results imply early-type fractions of 33\\% or less.  We have examined the effect of sample completeness on these percentages, i.e.\\ the possibility that some of the galaxies in this field which have spectra but no redshift measurement are actually system members with an absorption spectra that is consistent with a very faint, early-type galaxy.  Based on the numbers listed in Table 1 of Paper II, we present a statistical argument to estimate the number of such galaxies. We use the fraction of galaxies in this field that have an absorption spectrum and are also confirmed system members (25\\%), as well as the number of galaxies down to our limiting magnitude which have spectra but no redshift measurement (14).  If we assume that all 14 of these galaxies have faint absorption spectra, we expect that four (or less) are system members which have been missed. This implies that purely absorption spectra galaxies comprise less than 29\\% of the system population. Therefore, we conclude that our early-type fractions as estimated spectroscopically should not be significantly biased by incompleteness.  Because the number of confirmed group members with HST morphologies is small, we have also examined the background-subtracted distribution of galaxy morphologies in this system.  This study indicates that CL0023+0423 has a galaxy population which is more typical of the field.  The numbers are consistent with 100\\% of the galaxies in this two group system being normal spirals.  Of all the galaxies brighter than $M_V = -19.0 + 5~{\\rm log}~h$ in the central $\\sim 0.5~h^{-1}~{\\rm Mpc}$ of this system, early-type galaxies may comprise only $5^{+35}_{-5}\\%$ of the total population.  The formal uncertainties are large because of the small number statistics (see Paper III); however, there are consistent with the spectral results. If these groups are in the process of forming a cluster, a comparison between the morphological fractions in the CL0023+0423 system and those in open, intermediate-redshift clusters may indicate that a non-negligible fraction of early-type galaxies are formed between redshifts of $z \\sim 0.8$ and $z \\sim 0.3$.  These results are only suggestive because of the large uncertainties in the morphological fractions at both intermediate and high redshift.  If formation of early-type galaxies is occurring during this time period, we would expect that a significant fraction of these galaxies in open clusters would have spectral features which are characteristic of star formation activity within the last $\\sim 1$ Gyr.  Preliminary spectral studies indicate that early-type galaxies in intermediate-redshift clusters (both open and compact) show no signs of current or recent star formation; however, there is a non-negligible fraction which show post-starburst spectral features (Dressler et al.\\ 1994; Barger et al.\\ 1996; Oemler, Dressler \\& Butcher 1997; Poggianti 1997; Couch et al.\\ 1998).  Starbursts are thought to be the result of mergers or interactions between galaxies. Such events can drive substantial evolution in not only the stellar population, but also the morphological appearance of the galaxy (e.g.\\ Schweizer 1986; Hibbard et al.\\ 1994). The morphological signatures of a merger between two disk galaxies, such a long, bright tidal tail, would be visible for only $\\sim 1-2$ Gyrs.  After this time, the tidal debris becomes indistinguishable from the main merger remnant, a morphologically normal elliptical galaxy (Mihos 1995, 1998).  These timescales are consistent with the transformation of disk galaxies to spheroids between redshifts of $z \\sim 0.8$ and $z \\sim 0.5$.  In fact, the fraction of galaxies in intermediate-redshift clusters which exhibit signs of merging events, such as distorted morphologies or tidal tails, is as large as $\\sim$20\\% (e.g.  Lavery \\& Henry 1988; Lavery, Pierce \\& McClure 1992; Dressler et al.\\ 1994; Barger et al.\\ 1996; Oemler, Dressler \\& Butcher 1997; Couch et al.\\ 1998).  In the CL0023+0423 system, we also find that 41\\% (5 out of 12) of the confirmed cluster members with HST morphologies have either a ring structure, an asymmetric disk, or a double nuclei which may indicate an interaction or merger.  In addition, Koo et al.\\ (1995) suggest that blue compact galaxies (of which there are two in this system; see above) are galaxies which have undergone a recent, strong burst of star formation.  The galaxy luminosity will eventually fade by several magnitudes, resulting in a system with a surface brightness and velocity width which are typical of nearby low-luminosity spheroids (see Kormendy \\& Bender 1993). The wide range in redshift ($z \\sim 0.1 - 0.85$) over which these galaxies are found also implies that major star formation episodes have occurred in some spheroids over many Gyrs.  Because the group velocity dispersion is on the order of the internal velocity of its member galaxies, the kinematics of poor groups makes them a favorable environment for galaxy-galaxy encounters (Aarseth \\& Fall 1980; Barnes 1985; Merritt 1985). Therefore, it is conceivable that groups of galaxies like those of CL0023+0423 may be the sites of morphological evolution in galaxies.  Signatures of past starburst activity are not uncommon in local environments. For example, studies of nearby rich clusters with substantial substructure, as well as local groups of galaxies, indicate that $\\sim 15\\%$ of the early-type member galaxies have experienced recent or ongoing star formation (Caldwell \\& Rose 1997; Zabludoff \\& Mulchaey 1998). Field ellipticals also show remnants of past violent activity. Line strengths indicate that approximately 25\\% of all field ellipticals have produced 10\\% or more of their stars since redshifts of $z \\sim 0.5$ (Trager 1997).  These fractions are even higher for S0 galaxies where $25-50\\%$ show such activity (Fisher, Franx \\& Illingworth 1996; Trager 1998). These results also suggest that some early-type galaxies have experienced and are still experiencing considerable evolution since $z \\sim 1$ (see also Schweizer 1997).  If the results of the study of CL0023+0423 do support early-type formation at relatively recent epochs, it does not preclude in any way the formation of these galaxies at very high redshift or, subsequently, the existence of large populations of early-type galaxies in high-redshift clusters.  In fact, much data, including galaxy age and color scatter, support a very early epoch ($z > 5$) for the formation of early-type galaxies (e.g.\\ Dickinson 1995; Steidel et al.\\ 1996; Ellis et al.\\ 1997; Oke, Gunn, Hoessel 1997; Stanford, Eisenhardt \\& Dickinson 1997; Paper II). Additionally, already well-formed clusters at $z > 0.8$ appear to have fractions of early-type galaxies which are comparable to local clusters (Stanford, Eisenhardt \\& Dickinson 1997; Stanford et al.\\ 1997; Dickinson 1997; Lubin et al.\\ 1998). The relaxed, well-formed nature of these systems suggest that they formed at a much earlier epoch, allowing for perhaps enough time to create a significant population of early-type galaxies. The observations of the CL0023+0423 system may simply imply that a fraction of the early-type population is forming and/or undergoing significant activity at redshifts of $z < 1$.  On the other hand, the fact that a sizable portion of the early-type population in certain clusters are formed at redshifts of $z < 1$ may be difficult to reconcile with the tight color-magnitude relation found in clusters of galaxies between redshifts of $z = 0$ and $z \\sim 0.9$ (e.g.\\ Stanford, Eisenhardt \\& Dickinson 1995, 1997; Ellis et al.\\ 1997). The small intrinsic color scatter in early-type galaxies implies that the fractional differences in age between the stellar population in these galaxies are small (Bower, Lucey \\& Ellis 1992). Because the color scatter in the E/S0 populations remains roughly constant to $z \\sim 0.9$, this suggests that the last period of major star-formation took place at redshifts significantly greater than 1 (see also Arag\\'on-Salamanca et al.\\ 1993; Stanford, Eisenhardt \\& Dickinson 1995; Paper II). The majority of the clusters in these studies have been centrally-concentrated, compact clusters. Cluster concentration, which correlates well with degree of regularity, is measured by the compactness parameter $C \\equiv {\\rm log}(R_{60}/R_{20})$ where $R_{60}$ and $R_{20}$ refer to the radii containing 60\\% and 20\\% of the cluster population, respectively (Butcher \\& Oemler 1978).  It is conceivable that open, irregular clusters ($C < 0.35$) have systematically larger E/S0 color scatters which are the result of their more recent evolutionary history.  Five open clusters have been well-studied at intermediate redshifts, CL0303+17, CL0412-65, CL0939+47, CL1447+23, and CL1601+42.  The color scatter in their early-type populations are all consistent with the mean scatter of the entire cluster sample (Ellis et al.\\ 1997; Stanford, Eisenhardt \\& Dickinson 1997).  However, the relatively large uncertainties in the estimated scatters may still allow for a systematic difference between open and compact clusters. There are other apparent differences between these types of clusters; for example, the morphology-density relation which is characteristic of both open and compact clusters locally is qualitatively similar in compact clusters at intermediate redshift but completely absent in the loose, open clusters (Dressler et al.\\ 1997).  Even if the two CL0023+0423 groups are not in the process of merging, the observations of this system may have implications for the evolution of groups of galaxies.  CL0023A and CL0023B have similar kinematic characteristics to local groups.  Their velocity dispersions and estimated masses are typical of well-studied groups in the nearby universe (Hickson 1982; Ramella, Geller \\& Huchra 1989; Mulchaey et al.\\ 1996; Zabludoff \\& Mulchaey 1998). Locally, there exists a well-established correlation between the early-type fraction and the group velocity dispersion (Hickson, Kindl \\& Huchra 1988; Zabludoff \\& Mulchaey 1998).  As the velocity dispersion increases from 100 to $450~{\\rm km~s^{-1}}$, the early-type fraction rises from effectively zero to 55\\%.  These calculations include all member galaxies which are brighter than the absolute magnitude limit of $M_B \\sim -16$ to $-17 + 5~{\\rm log}~h$ and within a radius of $\\sim 0.6-0.8~h^{-1}~{\\rm Mpc}$.  Systems with larger velocity dispersions move into the more massive cluster regime.  Here, the relation between early-type fraction and velocity dispersion flattens out with clusters of galaxies typically having an early-type fraction of $\\sim 55 - 65\\%$ within these magnitude and radial limits (Dressler et al.\\ 1980a; Whitmore, Gilmore \\& Jones 1993).  Most clusters at $z \\simgreat 0.1$ are not sampled to such faint limits. If a brighter magnitude limit were adopted on the local group data ($M_B \\sim -18 + 5~{\\rm log}~h$), the statistics worsens though the early-type fraction does increase as the brightest galaxies in groups and clusters are normally early-types.  For the groups with the highest velocity dispersion, the early-type fraction approaches 80\\% or more, again consistent with the morphological composition of present-day clusters (Oemler 1974; Dressler et al.\\ 1980a; Whitmore, Gilmore \\& Jones 1993; Dressler et al.\\ 1997).  Based on the local correlation between early-type fraction and velocity dispersion described above, the modest early-type fractions in both CL0023+0423 groups may imply that the galaxy populations in groups of galaxies have evolved (see also Paper III).  We explore this possibility by examining the spectroscopically-confirmed group members. These data corresponds to an absolute magnitude limit of $M_B \\sim -19$ to $-20 + 5~{\\rm log}~h$ (see Paper II).  In CL0023A, four of the 7 galaxies are morphologically classified as spirals (see Paper III). Of the remaining three galaxies, one is classified as an elliptical. The other two are not within the HST field-of-view and, therefore, have no direct morphological information; however, they both have very high equivalent width \\ion{O}{2} emission (20 and 28.6 \\AA, respectively). Based on the spectroscopic and morphological data from the two cluster fields presented in Papers II and III, we find that 86\\% of all galaxies with such strong \\ion{O}{2} emission are classified as either a spiral or irregular/peculiar; the other 14\\% are classified as ellipticals though their photometric and spectral properties indicate that they are more similar to blue compact galaxies (see above).  This correlation between morphology and \\ion{O}{2} emission would suggest that the remaining two galaxies in CL0023A are most likely late-type galaxies.  Therefore, the elliptical fraction would be only 14\\% (1 out of 7).  Indeed, only 29\\% (2 out of 7) of the confirmed members have a typical K star absorption spectra which is old and red. These two galaxies are classified as an elliptical (MDS ID \\#17; see Sect.\\ 5.3.1 of Paper III) and a peculiar Sa (MDS ID \\#82; see Sect.\\ 5.3.2 of Paper III).  Based on the local relation, we would expect an early-type fraction of $10^{+20}_{-10}\\%$ for this low dispersion system of $158^{+42}_{-33}~{\\rm km~s^{-1}}$. The expected fraction should actually be higher, perhaps 33\\%, as the local data goes over two magnitudes fainter. In addition, with only 7 redshifts, the velocity dispersion of this group is likely biased low by a factor of $\\sim 1.5$ (Zabludoff \\& Mulchaey 1998).  This bias would also imply a predicted early-type fraction which is higher.  CL0023B has 17 confirmed members, of which 7 have direct morphological information. Of these seven, three are spirals, two are blue compact galaxies (classified as ellipticals), one is an elliptical, and one is classified as compact (see Paper III). Only 2 of these 7 galaxies have both spectroscopic and morphological properties which are characteristic of an elliptical galaxy.  Overall, 11 out of 17 confirmed group members have extremely strong \\ion{O}{2} emission (an average equivalent width of 43 \\AA). If we assume that all of these are late-type galaxies as is true for the vast majority of galaxies with direct morphological information (see above), this suggests an early-type fraction of 35\\% or less. In fact, only 12\\% (2 out of 17) of the confirmed group members have an old, red K star absorption spectrum.  These two galaxies are morphologically classified as an elliptical (MDS ID \\#20) and a compact galaxy (MDS ID \\#115; see Sect.\\ 5.3.1 of Paper III). For CL0023B with a velocity dispersion of $415^{+102}_{-63}~{\\rm km~s^{-1}}$, the local relation would predict an early-type fraction of $55^{+10}_{-20}\\%$.  For the brighter magnitude cut which is characteristic of our high-redshift data, this fraction would be 80\\% or more.  These observations suggest that there is no strong correlation between early-type fraction and velocity dispersion in the CL0023+0423 system. Consequently, the difference between our observations and that predicted by the local relation may imply early-type formation and, therefore, a change in the morphological composition of groups of galaxies between redshifts of $z \\sim 0.8$ and the present epoch. This evolution would suggest that, in order to create the correlation between velocity dispersion and early-type fraction that exists in local groups of galaxies, the new early-type galaxies will form only in the more massive systems.  These hypotheses are only valid if we assume that the groups of CL0023+0423 are typical of all groups at high redshift and that high-redshift groups are the progenitors of local galaxy groups.  It is also important to stress that the very limited number of confirmed members with direct morphology imply large statistical (and perhaps systematic) errors in both the velocity dispersion and early-type fractions of both components.  Therefore, we cannot say for certain that the properties of the CL0023+0423 groups are truly inconsistent with the early-type fraction versus velocity dispersion relation observed locally.  Indeed, even the nearby relation contains a large scatter with many local groups being significant outliers from the best-fit correlation (Zabludoff \\& Mulchaey 1998).  The observations of CL0023+0423 have provided us with the first detailed data on the optically-selected groups of galaxies and a potential group-group merger at high redshift. Though broad conclusions certainly cannot be made based on only two high-redshift groups, the results are suggestive, implying that there may be substantial evolution in galaxy morphology occurring at redshifts of $z < 1$.  \\vskip 0.5cm  We would like to thank the referee Ann Zabludoff for very useful comments on the text.  We also thank J.S. Mulchaey and S.C. Trager for helpful discussions and P. Brieu and L. Hernquist for expert advice on the N-body simulations.  The W.M. Keck Observatory is operated as a scientific partnership between the California Institute of Technology, the University of California, and the National Aeronautics and Space Administration.  It was made possible by generous financial support of the W. M. Keck Foundation.  Support for LML was provided by NASA through Hubble Fellowship grant HF-01095.01-97A awarded by the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., for NASA under contract NAS 5-26555.  This research was supported in part by {\\it HST} GO analysis funds provided through STScI grant GO-06000.01-94A.  \\newpage"
    },
    "9805/astro-ph9805266_arXiv.txt": {
        "abstract": "Two high galactic latitude B-type stars, \\pgdre\\ and \\pgsie , are analysed from low and medium resolution optical spectra and Str\\\"omgren photometry. A differential abundance analysis for He, C, N, O, Mg, Al, and Si reveals  that the He abundance is close to solar while metal underabundances relative to the  solar value of typically 1.3 and 1.2 dex for \\pgdre\\ and \\pgsie , respectively, are found. For both stars, carbon is even more depleted. The atmospheric parameters are consistent with  evolutionary tracks for stars evolving from the asymptotic giant branch  (AGB) with a stellar mass of 0.55~\\Msolar. The anomalous compositions are compatible with those  of other high galactic latitude post-AGB stars. Hence,  \\pgdre\\ and \\pgsie\\ are low mass post-AGB stars in an  evolutionary stage between those of A- and F-type supergiants of low mass  and central stars of planetary nebulae.  From kinematic data and  distances a population~II membership is probable for both stars. ",
        "introduction": "In a long-term project we have followed-up spectroscopically on  hot subdwarf candidates from the ``Palomar-Green Catalogue of Ultraviolet Excess Stellar Objects'' (Green et al., \\cite{gree86}). While most objects turned out to be hot subluminous stars (Moehler et al., \\cite{mohe90b}, Theissen et al., \\cite{thmo93}) or blue horizontal branch stars (Schmidt et al., \\cite{schm92}) a few candidates showed surface gravities too low for such  stars and more typical for giant stars. In addition these objects exhibit normal helium line strengths, again in contrast to the usually helium weak-lined hot subdwarfs.  The normality of the H/He spectrum might indicate that these stars are normal, i.e. massive giants in a strange place, very far from the  galactic disk. In this case they were either created in the disk and thrown out (cf. Leonard, \\cite{leon91}) or they were born in the halo  (which could be explained by the model of the ``galactic fountain'',  Dyson \\& Hartquist, \\cite{dyha83}). On the other hand, such stars could be post-AGB stars on their way from the asymptotic giant branch (AGB) to the white dwarf domain that spectroscopically mimic massive stars.  Since this part of the evolution proceeds very fast (10$^4$ years, Sch\\\"onberner, \\cite{scho83}), such objects should be rather rare. Indeed, not many have been found and analysed so far (Conlon et al., \\cite{codu91}, \\cite{conl93}).  Since a metallicity analysis offers a good possibility to distinguish between young, massive OB stars and post-AGB stars we decided to get higher resolution (about 1~\\AA) spectra of some of these stars. Here we report the results for two objects, \\pgdre\\ and \\pgsie . Lower resolution spectra of \\pgsie\\ have already been discussed by Conlon et al.  (\\cite{conl93}). In addition to these stars we observed spectra of well known and analysed main-sequence and sub-giant OB stars to allow direct comparisons between the spectra. In Section 2 we describe the observations and the reduction of the data. Sections 3 and 4 describe the determination of atmospheric  parameters and abundances. Section 5 gives the kinematic data of the stars,  while Section 6 gives a discussion of the results. ",
        "conclusions": "\\pgdre\\ and \\pgsie\\ are giant stars with normal helium abundances. Both stars share a characteristic abundance pattern with  \\bddre , the central star of a halo planetary nebula, and with  two high galactic latitude B-giants discussed by Conlon et al. (\\cite{codu91}), most notably the strong carbon deficiency. Depletions of N, O, Mg, Al, and Si by 1 dex or more in \\pgdre\\ and \\pgsie\\ point towards a low metallicity and, hence, membership of an old stellar population,  while they are not compatible with the values found for young B stars.  The positions of both stars in the (\\teff, log\\,g) diagram are consistent with evolutionary tracks starting from the AGB for masses of about 0.55~\\Msolar. We therefore conclude that both stars are in the post-AGB stage of their  evolution.  The similarity of the abundance patterns to that of \\bddre\\ suggests that  the successors of the B-type PAGB stars can be found amongst the halo CSPNe as already suggested by McCausland et al. (\\cite{mcco92}). Potential progenitors could be found among the optically bright post-AGB stars discussed by van Winckel (\\cite{vawi97}). However, van Winckel finds only one star, HD\\,107369, for which the abundance pattern comes close to that   of B-type PAGB stars. The CNO abundances of this star are  consistent with being the product of CN cycling. In the B-type PAGB stars, however, the C/N ratio can be explained by CN cycling, but not the N/O ratio.  Conlon et al. (\\cite{codu91}), McCausland el. (\\cite{mcco92}) and Napiwotzki et al. (\\cite{nahe94}) discussed various  scenarios to explain the abundance pattern of B-type PAGB stars  in terms of nuclear processing and  subsequent dredge-up, but no conclusion could be reached.  The high [O/Fe] ratio (+1.2$\\pm$0.5) observed in \\bddre\\ prompted  Napiwotzki et al. (\\cite{nahe94}) to suggest that its abundance pattern could be affected by gas-dust separation invoked by Bond (\\cite{bond91}) to explain the extreme [O/Fe]~$>$~4 ratio found in two peculiar PAGB stars of spectral type F (Lambert et al., \\cite{lahi88} and Waelkens et al., \\cite{wawi91}). Therefore, the determination of the iron abundance in the B-type PAGB  stars is a prerequisite for an interpretation of their abundance patterns.  According to the kinematic behaviour of \\pgsie\\ and the large distance  from the galactic plane for \\pgdre\\ both stars are probably Population~II  objects."
    },
    "9805/astro-ph9805116_arXiv.txt": {
        "abstract": "We report VLA $1\\sigma$ upper limits of 1.5\\% and 3\\% on the intrinsic circular and linear fractional polarizations, respectively, of the water vapor maser emission 0.2 pc from the central engine of NGC 4258. A corresponding 0.5\\% upper limit on any Zeeman-splitting-induced circular polarization translates to a 1$\\sigma$ upper limit on the parallel, or toroidal, component of the magnetic field of 300 mG. Assuming magnetic and thermal pressure balance in the disk, this magnetic field upper limit corresponds to a model-dependent estimate of the accretion rate through the molecular disk of $10^{-1.9}\\alpha$\\,\\smy\\ for the case where the magnetic field lies along the line of sight. ",
        "introduction": "\\label{s:mag.1}  Active galactic nuclei (AGN) are thought to be powered by accretion onto a central, supermassive black hole (e.g. Rees 1984) and it is probable that magnetic fields play an important role in this process.  They are often invoked as a potential source of viscosity and therefore dissipation and angular momentum transport in the accretion flow (c.f. Balbus \\& Hawley 1991). In addition, large-scale magnetic fields may be important in the energetics, collimation, and confinement of jets and broad line regions (Blandford \\& Payne 1982; Emmering, Blandford \\& Shlosman 1992).  Finally, the typical broad-band AGN spectrum is not adequately fit by a simple blackbody law, and many of the proposed non-thermal radiation processes require a substantial magnetic field.  Magnetic field strengths can be estimated from radio observations in a variety of ways. For example, if the source size can be estimated, then a determination of the synchrotron turnover frequency yields an estimate for the field strength. Alternatively, if the gas density can be estimated, then Faraday rotation measurements can also provide reasonably accurate field strengths. But the most direct method for estimating magnetic field strengths is via observations of the Zeeman splitting of spectral lines of various atomic and molecular species in regions of high density neutral gas.  Unfortunately, for gas in  thermal equilibrium it is impossible to apply this technique to AGN, as the interferometers needed to resolve the central engines of AGN have insufficient sensitivity to study the low surface brightness neutral gas emission where the Zeeman splitting occurs.  However, the extreme non-thermal equilibrium conditions characteristic of extragalactic masers offer an unusual opportunity to measure AGN magnetic fields. To date approximately 21 \\ho\\ megamasers and 50 OH megamasers have been detected in the nuclei of nearby AGN (e.g. Braatz, Wilson \\& Henkel 1997). They are often several orders of magnitude more luminous than Galactic masers, and many are bright enough to be studied with VLBI.  NGC 4258 is a mildly active Seyfert 2 galaxy with a 2--10 keV X-ray luminosity of $4\\times10^{40}$ \\ergs\\ and an obscuring column of $1.5\\times10^{23}$\\,cm$^{-2}$ (Makishima \\etal\\ 1994). Nuclear continuum and narrow line emission are seen in reflected, polarized optical light (Wilkes \\etal\\ 1995).  The galaxy harbors an \\ho\\ maser (Claussen, Heiligman, \\& Lo 1984), and VLBA observations reveal that the maser traces an extremely thin, slightly warped, nearly edge-on disk in the nucleus of the galaxy (Watson \\& Wallin 1994; Greenhill \\etal\\ 1995; Miyoshi \\etal\\ 1995; Moran \\etal\\ 1995; Herrnstein, Greenhill, \\& Moran 1996).  The masers are 0.13--0.26 pc from the center of mass of the disk, assuming a distance of 6.4 Mpc.  The nearly perfect Keplerian rotation curve of the masers requires a central mass of $3.5\\times10^{7}$ \\msun\\ within 0.13 pc. The velocity centroid of the disk is consistent with the optical systemic velocity of the galaxy and the rotation axis of the disk is well-aligned with a 100-pc-scale radio jet (Cecil, Wilson, \\& De Pree 1995). VLBA observations also reveal subparsec-scale jet emission along the disk axis (Herrnstein \\etal\\ 1997).  Deguchi, Nakai, \\& Barvainis (1995) report an upper limit on the linear polarization of about 20\\% for the highly Doppler-shifted masers and about 15\\% for those features near the galaxy's systemic velocity.  Here, we report an attempt to measure the Zeeman splitting of an isolated, strong maser feature in the sub-parsec-scale disk of NGC 4258 using VLA polarimetric observations. The observations are discussed in Section~\\ref{s:mag.2}, and the results are interpreted in Section~\\ref{s:mag.3}. Section~\\ref{s:mag.4} considers the implications of the measurements. ",
        "conclusions": "\\label{s:mag.4}  The upper limit on the magnetic field at 0.2 pc radius can be compared to a second upper limit relating to pressure balance in the disk. Moran \\etal\\ (1995) use the scatter in the vertical component of the systemic masers to place an upper limit of 0.0003\\,pc on the scale height, $H$, of the maser layer.  If the disk is supported by magnetic pressure and is in hydrostatic equilibrium then $H/r = v_{A}/v_{\\phi}$, where $r$ is the radius, $v_{A}$ is the Alfv$\\acute{\\mbox{e}}$n velocity ($v_{A}=B/\\sqrt{4\\pi\\rho}$), and $v_{\\phi}$ is the Keplerian rotation velocity. $\\rho$ is the density in the maser layer, and the Alfv$\\acute{\\mbox{e}}$n velocity in question relates to any magnetic field geometry capable of supplying pressure in the vertical direction: tangled, radial, and toroidal fields all suffice. We require $n \\la 10^{10}$ cm$^{-3}$ in order to support maser action and conclude that $B \\la 80$ mG in the maser layer at $r=0.2$\\,pc, the equality holding for a magnetically dominated disk characterized by the quoted upper limits for thickness and density. The 300 mG upper limit on \\bpar\\ provided by the Zeeman analysis does not preclude a magnetically supported disk. It does, however, provide a more robust upper limit on the field strength than do hydrostatic equilibrium considerations, since the masers could in principle be sampling a thin layer in a much thicker disk.  NGC 4258 is an exceptional laboratory for the study of AGN accretion processes. The nuclear maser reveals details about the kinematics and structure of the accretion disk on sub-parsec scales and permits the determination of the central mass with great precision. In addition to being constrained by these observations, efforts to construct self-consistent models of the accretion in NGC 4258 must necessarily address the fact that the central engine luminosity ($L$) is extremely sub-Eddington: ($L\\sim10^{42\\pm1}~\\ergs=10^{-3.6\\pm1}L_{E}$; Herrnstein \\etal\\ 1998). Neufeld \\& Maloney (1995) demonstrate that for an accretion rate, $\\mdot$, of approximately $10^{-4.1}\\alpha$\\,\\smy, the molecular disk becomes atomic throughout for radii greater than about 0.26 pc, thereby providing a physical explanation for the observed outer limit of the maser emission. Here, $\\alpha$ is the standard Shakura-Sunyaev parameterization of the kinematic viscosity. This accretion rate corresponds to a radiative efficiency of $10^{-0.6\\pm1}\\alpha^{-1}$ and suggests that NGC 4258 is powered by a highly efficient, but mass-starved, optically thick, geometrically thin accretion disk (Shakura \\& Sunyaev 1973).  Alternatively, Lasota \\etal\\ (1996) have proposed that NGC 4258 is powered by a central geometrically thick, optically thin advection-dominated accretion flow (ADAF; Narayan \\&  Yi 1995a\\&b). In this scenario, the vast majority of viscously dissipated energy is deposited into an extremely hot ($\\sim10^{12}$\\,K), largely non-radiative proton plasma, and the low luminosity of the system is due to low radiative efficiencies as opposed to mass starvation of the central black hole.  In the ADAF models the broad-band spectrum, from radio to hard X-ray, is the result of various nonthermal processes in a relatively cool ($\\sim10^{9.5}$\\,K) electron plasma.  Lasota \\etal\\ (1996) fit the optical and X-ray NGC 4258 emission with an $\\mdot\\simeq10^{-1.9}\\alpha$\\,\\smy\\ ADAF.  More recent calculations by Narayan (personal communication), suggest $\\mdot\\simeq10^{-1.7}\\alpha$\\,\\smy, corresponding to a radiative efficiency of $10^{-3\\pm1}\\alpha^{-1}$.  We note that recent high-resolution radio continuum observations suggest that if NGC 4258 harbors an ADAF, it must be truncated within about 100 Schwarzschild radii (Herrnstein \\etal\\ 1998).  \\begin{figure}[htb]          % \\plotone{apj-fig3.eps} \\caption {The shaded region shows the expected strength of the LOS-component of the equipartition magnetic field as a function of $\\mdot/\\alpha$ at a radius of 0.2 pc for a temperature of 600\\,K and for a range of field orientations. The accretion rates predicted by the thin-disk and ADAF models are shown (dashed vertical lines), as well as the upper limit on \\bpar\\ from the present experiment (solid horizontal line).  The dashed horizontal line shows the theoretical magnetic field sensitivity of a proposed ``broad-band'' Zeeman experiment.} \\label{fg:mag3} \\end{figure}  An observational determination of \\mdot\\ would be the most unambiguous way to discriminate between these two modes of accretion.  Unfortunately, \\mdot\\ is very difficult to measure. The most direct constraint is derived from the extreme precision of the maser Keplerian rotation curve, and requires $\\mdot \\la 10^{-1}\\alpha$\\,\\smy (Herrnstein \\etal\\ 1998). The present upper limit on \\bpar\\ provides an additional, albeit less direct, probe of \\mdot.  Following Neufeld \\& Maloney (1995), we treat the X-ray-heated outer disk in NGC 4258 as a thin isothermal disk in hydrostatic equilibrium, in which the standard Shakura-Sunyaev accretion solutions apply.  The midplane pressure, $p_{c}$, is then given by (Frank, King, \\& Raine 1992) \\bq p_{c}=\\frac{1}{3\\sqrt{2\\pi^{3}}}\\frac{GM\\mdot}{\\alpha c_{s} r^{3}}, \\label{eq:nm1} \\eq where $M$ is the central mass and $c_{s}$ is the isothermal sound speed. The magnetic field strength can be related to the accretion rate using equation~\\ref{eq:nm1} under the assumption of thermal and magnetic pressure balance in the disk ($\\vert B\\vert^{2}/8\\pi=p_{c}$).  One potential concern is that the very low ionization fraction in the molecular disk will significantly limit the coupling between the magnetic fields and the gas. However, we note that direct exposure to the central X-ray source (as a result of the warp), will generate a layer of atomic gas on top of the molecular disk with 1--2\\% fractional ionization, and that at $r=0.2$\\,pc the molecular layer is only about 20\\% of the total disk thickness (Neufeld, Maloney, \\& Conger 1994; Herrnstein, Greenhill, \\& Moran 1996).  The parallel component of the magnetic field is related to $B$ according to $\\bpar = uB$, where $u=\\cos(\\theta)$ and $\\theta$ is the angle between the LOS and the magnetic field at the position of the maser in question. A disk permeated by large-scale vertical or radial fields has $u\\simeq0$ along the midline, while a disk dominated by large-scale toroidal fields has $u\\simeq1$. {\\it Adopting a temperature of 600\\,K in order to accommodate the presence of masers in the disk, the present 300\\,mG upper limit on $\\vert\\bpar\\vert$ together with equation~\\ref{eq:nm1} implies $\\mdot\\la10^{-1.9}\\alpha u^{-2}$\\,\\smy.} Figure~3 shows the theoretical LOS magnetic field strength as a function of \\mdot\\ for $20^{\\circ}<\\theta<70^{\\circ}$ and illustrates that because of ambiguities in the magnetic field geometry, the present observations cannot distinguish between the ADAF and thin-disk models. We emphasize that this analysis implicitly assumes that a substantial component of the equipartition magnetic field is constant across the cross-section of the maser feature at 1306\\,\\kms, which is thought to have a linear scale of $\\sim10^{15}$\\,cm (Moran \\etal\\ 1995).  Furthermore, the theoretical field estimates are only as accurate as the assumption of pressure balance in the disk. It is at least in principle possible to construct either thermally or magnetically dominated disk models.  While the present VLA observations are not sufficiently sensitive to discriminate between the low and high \\mdot\\ models, it may eventually be possible to do this by extending the Zeeman S-curve formalism to search for coherent, ``broad-band'' Zeeman splitting across the systemic masers, which arise on the near side of the disk (see Figure~1).  In such a broad-band polarimetry experiment, one could search for a coherent frequency offset between the LCP and RCP spectra, by comparing $dI/d\\nu$ to the Stokes V spectrum across the entire 100 \\kms\\ of systemic emission. Because the systemic emission is 100 times broader and 10 times brighter than the feature at 1306 \\kms, the magnetic field sensitivity would be greatly enhanced.  Simulations of this experiment using actual VLBA spectra predict a 1$\\sigma$ uncertainty in \\bpar\\ of 30 mG, which in principle can discriminate between the low and high \\mdot\\ models, even accounting for ambiguities in the field configuration (dashed horizontal line in Figure~3). The broad-band Zeeman measurement would be sensitive to that component of the radial magnetic field that is constant across the systemic masers, which span about 4$^{\\circ}$ in disk azimuth and $10^{16}$ cm in radius."
    },
    "9805/astro-ph9805320_arXiv.txt": {
        "abstract": "The theory of radiatively-induced warps in accretion discs is applied to the discs of Be stars. It is found that these discs may develop warps in their inner regions, although once the warp amplitude is large enough then the interaction between the disc and fast radiatively-driven wind will determine its evolution. The warping is shown to be more important for later than earlier B stars. Although the interaction of the fast-wind with the disc will limit the amplitude of the warp, it cannot drive the warp radially outwards, and so the radial evolution of the warp depends on the dominant advective process within the disc.  Typical timescales associated with growing modes are shown to be short, of the order of days-weeks, although these are not likely to be the timescales inferred from observations of line-profile variations which are much longer, of the order of years. ",
        "introduction": "Be stars are accepted to have a dense, slowly expanding disc in their equatorial planes and a fast radiatively-driven wind over their polar regions (see Slettebak 1988 for a review). The emission line profiles generated in the disc have been monitored for many stars and are found to vary (e.g. Dachs 1987). Long-term variations in the ratio of the violet to red components of a line, $V/R$, are attributed to $m = 1$ density waves in a {\\em Keplerian} disc (Okazaki 1991, Papaloizou \\etal\\ 1992). Hanucshik (1996) also concludes that the disc is rotationally supported. Using the assumption that the disc is rotating at Keplerian velocities around the central star and is present throughout most of the star's life, Porter (1998) has recently showed that the radial velocity at the inner edge of the disc must be small ($v_r \\ltappeq 0.01\\kms$). The conclusion of these studies is that the disc material orbits many times before it travels a significant radial distance -- similar conditions to the material in an accretion disc.  Pringle (1996), Malony, Bagelman \\& Pringle (1996) and Armitage \\& Pringle (1997) have recently examined the possibility that a warp may be induced in an accretion disc due to the incident radiation field from the central object. They find that if the accretion disc is optically thick at the wavelength at which the disc emission is maximal then the disc may become unstable.  Here it is recognised that the physical situation examined by Pringle and co-workers and that of the Be star disc are similar: the Be star disc is being illuminated from both sides by the central B star; the Be star disc material orbits many times before moving a signifiant radial distance, and observational as well as theoretical studies have shown that at some radius the Be star discs are optically thick (e.g. Chokshi \\& Cohen 1987, Kastner \\& Mazzali 1989, Marlborough \\etal\\ 1997). Recently, two Be stars have been examined and have shown observational charactertistics of a warped disc (Hummel, 1998). In this paper it is attempted to apply instability criteria for radiatively-driven warping to the discs thought to exist around Be stars.  In \\S2 the disc structure is examined and the optically thick regions are identified. The criterion for the radiatively-driven warp instability to arise in the disc is then examined in \\S3. The evolution of a warp interacting with fast radiatively-driven wind is considered in \\S4. This is discussed in \\S5, and conclusions given in \\S6. ",
        "conclusions": "The theory of radiatively-induced warping of discs has been applied to Be stars and found to be important, especially for late B stars. The timescales associated with the growth of warps is short ${\\cal O}$(days), although these may not be the relevant timescales over which the modes are examined observationally. It is noted that warping has been observed in Be star discs already (Hummel 1998), and that the observationally derived precession timescales fall in the range in which the simple argument predicts.  A scenario has been presented in which radiatively-driven warps are created in the central parts of the disc and propagate outwards under the influence of the interaction of the disc and fast stellar wind."
    },
    "9805/hep-ph9805475_arXiv.txt": {
        "abstract": "We discuss collective effects like percolation, quark gluon plasma, or string fusion and their effect on the longitudinal development of high energy showers in air. It is shown that iron--air showers could develop more slowly than expected and produce the observed change in the slope of $X_{\\mbox{max}}$ at the highest energies. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805006_arXiv.txt": {
        "abstract": "We have reviewed the characteristics of the extended red emission (ERE) as observed in many dusty astronomical environments, in particular, the diffuse interstellar medium of the Galaxy. The spectral nature and the photon conversion efficiency of the ERE identify the underlying process as highly efficient photoluminescence by an abundant component of interstellar dust. We have compared the photoluminescence properties of a variety of carbon- and silicon-based materials proposed as sources for the ERE with the observationally established constraints. We found that silicon nanoparticles provide the best match to the spectrum and the efficiency requirement of the ERE. If present in interstellar space with an abundance sufficient to explain the intensity of the ERE, silicon nanoparticles will also contribute to the interstellar 9.7~$\\micron$ Si-O stretch feature in absorption, to the near- and mid-IR nonequilibrium thermal background radiation, and to the continuum extinction in the near- and far-UV. About 36\\% of the interstellar silicon depleted into the dust phase would be needed in the form of silicon nanoparticles, amounting to less than 5\\% of the interstellar dust mass. We propose that silicon nanoparticles form through the nucleation of SiO in oxygen-rich stellar mass outflows and that they represent an important small-grain component of the interstellar dust spectrum. ",
        "introduction": "The recent detection of extended red emission (ERE) in the diffuse interstellar medium (ISM) of the Galaxy by Gordon et al.\\ (1998) and the spectroscopic confirmation of its luminescence band nature by Szomoru \\& Guhathakurta (1998) fundamentally changed our perception of the ERE as an interstellar process in several ways. One, the ERE is no longer seen as a phenomenon limited to localized special environments, such as reflection and planetary nebulae. As a Galaxy-wide process, its characteristics must now be imposed as an additional observational constraint on models for interstellar dust in the diffuse ISM. Two, by correlating the ERE intensity with the \\ion{H}{1} column density in the diffuse ISM at intermediate and high Galactic latitudes, and thus with the dust column density, it has become possible to determine the efficiency with which UV/visible photons of the interstellar radiation field absorbed by interstellar dust are converted into ERE photons. An ERE photon conversion efficiency of ($10 \\pm 3$)\\% (Gordon et al.\\ 1998; Szomoru \\& Guhathakurta 1998) was derived by assuming the the ERE agent consumes {\\em all} photons absorbed by dust in the 91.2-550~nm wavelength range. To the extent that we know of the existence of other dust components which are not expected to contribute to the ERE, this assumption is false. The ERE agent, most likely, is not responsible for the absorption of all absorbed UV/visible photons. Thus, the derived efficiency is only a lower limit to the true, intrinsic, efficiency of the photoluminescence (PL) process seen as ERE. However, even if the intrinsic efficiency is as high as 50\\%, a reasonable upper limit for naturally occurring PL, the ERE agent still needs to absorb about 20\\% of all absorbed UV/visible photons in the diffuse ISM. This can only be, if the ERE agent consists of a cosmically abundant material which is capable of highly efficient PL. It appears that only carbon- or silicon-based materials fit these criteria. In this letter, we present the case for silicon nanoparticles being the component of the ISM responsible for the ERE. ",
        "conclusions": "We have examined the observed characteristics of the ERE, a photoluminescence phenomenon associated with interstellar dust and seen in a wide variety of astronomical environments as excess radiation in the 600-850~nm wavelength range. In the diffuse ISM, about 10\\% of absorbed UV/visible photons contribute to the ERE. The ERE carrier must be a cosmically abundant material exhibiting highly efficient photoluminescene properties. Silicon nanoparticles containing a few hundred silicon atoms each, surrounded by a SiO$_2$ shell, match the constraints regarding the spectrum and efficiency of the ERE.  Interstellar silicon abundances easily suffice to provide the needed material quantities, and the nucleation of SiO in oxygen-rich stellar outflows provides a likely source for their formation. We suggest that silicon nanoparticles are an abundant component of the interstellar dust spectrum."
    },
    "9805/astro-ph9805230_arXiv.txt": {
        "abstract": "We investigate the radial velocity and the magnetic field of the CP star $\\alpha^{2}$~CVn. The observed variation of the magnetic field is compared with that of our model. We search for a relation between the magnetic field and the distribution of the chemical elements. The period in the radial velocities is constant over a time interval of about 100 years. ",
        "introduction": "HD 112413 = $\\alpha^{2}$~CVn~is a CP star whose variability in line intensities and radial velocities (RVs), was detected already 100 years ago. In the earlier 50-ies $\\alpha^{2}$~CVn was one of the first stars in which a magnetic field was discovered. Photographic observations of the longitudinal magnetic field $B_{\\rm eff}$,~e.g. by Babcock \\& Burd (1952), Oetken et al. (1970), gave a very anharmonic field curve whereas the photoelectric ones by Borra and Landstreet (1977) show a more harmonic shape.\\\\ Michaud (1970) and Glagolevskij (1994) showed that the distribution of the chemical elements on the stellar surface of magnetic stars has to be connected with the magnetic field. Therefore, we search for a relation between the concentration of the chemical elements and the structure of the magnetic field assuming a concrete magnetic field model.\\\\ ",
        "conclusions": "Two findings seem to be important:\\\\ $-$~the rotation period, which is constant for nearly one hundred years, suggests that the abundance patches have remained at a fixed position\\\\ $-$~the Fe, Cr, Ti distributions show quite a unique relation to the magnetic field.\\\\"
    },
    "9805/astro-ph9805140_arXiv.txt": {
        "abstract": " ",
        "introduction": "The study of QSO clustering from large homogeneous surveys is likely to provide unique information on the nature of large-scale structure (LSS) in the Universe.  Even with relatively modest aperture telescopes, QSOs are readily detectable from the present cosmic epoch to a time when the Universe was less than 25\\% of its present age. This approximately eight billion year time interval provides a unique baseline over which to study the evolution of structure in the Universe with cosmic time.  QSO clustering also yields a direct measure of clustering on comoving scales $\\gg 100\\,$h$^{-1}$Mpc\\footnote{\\hangindent=1.37pc\\hangafter=1 Throughout this paper we use comoving distances, $r_{\\rm comoving} = (1+z)r_{\\rm proper}$, and define the Hubble constant as $H_0=100$h$\\,$km$\\,$ s$^{-1}$Mpc$^{-1}$}, providing key information (see figure 1) on LSS at scales between those provided by the COBE results ($>1000\\,$h$^{-1}$Mpc, Bennett {\\it et al.} 1994) and galaxy redshift surveys ($<100\\,$h$^{-1}$Mpc, Colless {\\it et al.}, this volume)  \\begin{figure} \\centering \\centerline{\\psfig{file=fig1.ps,width=12cm}} \\caption{The QSO power spectrum, P(k), based on a Zel'dovich approximation in a (2250h$^{-1}$Mpc)$^3$ box containing 192$^3$ particles with $z_{\\rm max}=2.2$.  A $\\tau$CDM model with $\\Omega_0=1$, $\\Gamma=0.2$ $\\sigma_8=0.67$ and QSO bias $b_{\\rm Q} =1.8$ at $z=0$ is assumed.  Errors on P(k) are based on the rms scatter from five simulations.  Note the 2dF QSO redshift survey comprises two such slices, effectively reducing errors by 1.4. Simulation taken from Croom et al.\\ (1998, in preparation).} \\end{figure}  To date, QSO clustering has remained largely unexploited for LSS studies.  This has largely been due to the lack of appropriate QSO catalogues.  Although over 10000 QSOs with measured redshifts are now known (V\\'eron \\& V\\'eron-Cetty 1997, see figure 2) less than $20\\,$\\% of these QSOs form part of homogeneous catalogues suitable for detailed clustering analysis.  The goal of the 2dF QSO redshift survey, currently underway at the Anglo-Australian Telescope (AAT), is to provide a new homogeneous catalogue of almost 30000 QSOs with which to carry out such an analysis.  The survey is almost two orders of magnitude larger than the previous largest homogeneous survey to this depth (Boyle {\\it et al.}\\ 1990).  In \\sect2 we review the existing results on QSO clustering and we describe the creation of the 2dF QSO redshift survey input catalogue in \\sect3. We report on the initial results from the survey in \\sect4 and discuss the future prospects for this and other QSO surveys in \\sect5.  \\begin{figure} \\centering \\centerline{\\psfig{file=fig2.ps,angle=-90,width=12cm}} \\caption{The all-sky distribution of the $\\sim 10000$ known QSOs from the catalogue of V\\'eron \\& V\\'eron-Cetty (1997).  The thirty UKST fields of the 2dF QSO redshift survey are also shown in their two declination strips.} \\end{figure} ",
        "conclusions": ""
    },
    "9805/astro-ph9805189_arXiv.txt": {
        "abstract": "Preliminary results from a study of the isotopic compositions of the elements Hg and Pt in a number of HgMn stars are presented. This work represents an improvement over previous studies thanks to the very high spectral resolution available (R = 118 000) and to the new information on wavelengths and atomic structure of \\HgII\\ and \\PtII. ",
        "introduction": "\\label{intr} Almost all HgMn stars exhibit a strong absorption feature at 3984~\\AA, which has been identified as a line of \\HgII. The wavelength at which this line is observed depends on the isotopic mix of Hg (White et al. 1976), which ranges from the terrestrial mix to nearly pure $^{204}\\!$Hg. Mercury is not the only very heavy element observed in HgMn stars. Lines of \\PtII\\ (Dworetsky \\& Vaughan 1973) and \\AuII\\ (Guthrie 1985) are also observed. Dworetsky \\& Vaughan (1973) studied the \\PtII\\ $\\lambda\\,4046$ line in a sample of nine HgMn stars. This line is the strongest Pt line at optical wavelengths, and in the nine stars studied it is shifted toward longer wavelengths by 0.04 to 0.09~\\AA, with respect to the centroid of the terrestrial platinum line. These shifts are interpreted as an isotopic effect. The corresponding anomalies are analogous to those found for Hg, in the sense that the heavier isotopes tend to dominate in cooler stars. Neither radiatively driven diffusion nor any other theory until now can account satisfactorily for the variations in the Hg and Pt isotope mix among the HgMn stars (Leckrone et al. 1993).  The main purpose of the work reported here was to provide additional observational constraints to guide the theorists in the understanding of the isotopic anomalies in HgMn stars, improving upon previous studies through the much better data quality obtainable now. Thanks to the availability of new laboratory measurements of isotope shifts in \\PtII\\ (Engleman 1989) it became possible to identify more definitely the \\PtII\\ isotopes.  \\begin{table}[t] \\small \\begin{center} \\caption{Isotopic compositions} \\label{t1} \\begin{tabular}{lrrrrrr} \\hline\\hline Star & {Terrestrial} & {$\\chi$~Lup} & {HR~7775} & {HR~1800} & {74 Aqr} & {HR 6520}\\\\ &{abundance}\\\\ \\hline\\hline $T_{\\rm eff}$ (K)  &&10680             &10830         &11070     &11880  &13250\\\\ $\\log g$     &&3.99              &4.11          &3.75       &4.03  &4.17\\\\ \\hline\\hline {[Hg]}              && +5.45  & +5.65        & +5.25    &        & +5.10\\\\ \\hline &\\multicolumn{6}{c}{Hg isotopic structure (\\%)}\\\\ \\hline 196 & 0.15 & 0.00 & 0.00 & 0.00 & 0.00 & 0.00 \\\\ 198 & 9.97 & 0.00 & 0.00 & 0.16 & 7.02 & 4.10 \\\\ 199a & 7.14 & 0.00 & 0.00 & 0.32 & 6.19 & 10.99 \\\\ 199b & 9.71 & 0.00 & 0.00 & 0.44 & 9.29 & 14.00 \\\\ 200 & 23.09 & 0.00 & 0.00 & 2.97 & 26.79 & 28.97 \\\\ 201a & 4.80 & 0.00 & 0.10 & 1.77 & 6.01 & 8.61 \\\\ 201b & 8.30 & 0.00 & 0.20 & 3.06 & 9.10 & 8.93 \\\\ 202 & 29.80 & 1.00 & 49.70 & 50.58 & 26.79 & 22.49 \\\\ 204 & 6.86 & 99.00 & 49.70 & 40.74 & 8.81 & 1.80\\\\ \\hline\\hline {[Pt]}                && +4.00           & +4.69         & +3.30\\\\ \\hline &\\multicolumn{4}{c}{Pt isotopic structure (\\%)}\\\\ \\hline 194 &32.90& 0.00& 0.00&  0.00 \\\\ 195b&18.78& 0.00& 7.50&  0.00 \\\\ 195a&13.15& 0.00&10.00&  0.00 \\\\ 196 &25.20&10.00&55.00&  0.00 \\\\ 198 & 7.19&90.00&27.50&100.00 \\\\ \\hline\\hline \\end{tabular} \\end{center} \\end{table} ",
        "conclusions": ""
    },
    "9805/astro-ph9805297_arXiv.txt": {
        "abstract": "We present an analytic description of the $r$-mode instability in newly-born neutron stars, using the approximation of uniform density. Our computation is consistently accurate to second order in the angular velocity of the star.  We obtain formulae for the growth-time of the instability due to gravitational-wave emission, for both current and mass multipole radiation and for the damping timescale, due to viscosity. The $l=m=2$ current-multipole radiation dominates the timescale of the instability. We estimate the deviation of the second order accurate results from the lowest order approximation and show that the uncertainty in the equation of state has only a small effect on the onset of the $r$-mode instability. The viscosity coefficients and the cooling process in newly-born neutron stars are, at present, uncertain and our analytic formaulae enables a quick check of such effects on the development of the instability. ",
        "introduction": "The recently discovered $r$-mode instability (\\cite{nanew}) in rotating neutron stars, has significant implications on the rotational evolution of a newly-born neutron star. The $r$-modes are unstable due to the Chandrasekhar-Friedman-Schutz (CFS) mechanism (\\cite{chandra}, Friedman \\& Schutz 1978) (see also Friedman \\& Morsink 1997). Andersson, Kokkotas \\& Schutz (1998) (see also Lindblom et al. 1998) find that the $r$-mode instability is responsible for slowing down a rapidly rotating, newly-born neutron star to rotation rates comparable to that of the initial period of the Crab pulsar ($\\sim$19 ms) or the recently discovered 16 ms X-ray pulsar in the supernova remnant N157B (Marshall et al. 1998) (with an estimated initial period of 6-9 ms). This is achieved by the emission of current-quadrupole gravitational waves, which reduce the angular momentum of the star. Additionally, as the initially rapidly rotating star spins down, an energy equivalent to roughly 1\\% of a solar mass is radiated in gravitational waves, which makes the process an interesting source of detectable gravitational waves (Owen et al. 1998).  In the present paper, we investigate the $r$-mode instability to 2nd order accuracy in the angular velocity of the star, in the approximation of uniform density (the actual density profile of realistic neutron stars is nearly uniform), where all results can be obtained analytically. \\footnote{The computation in \\cite{AKS98} is carried out in the 2nd order accurate, slow-rotation formalism for Newtonian stars by Saio (\\cite{saio}) while \\cite{LOM98}, arrive at the same conclusion, using a first-order, approximate, calculation.} While our analytic results provide an independent check of the numerical results in \\cite{AKS98} and Lindblom et al. 1998, our main objective is to present a simple set of equations, which enable one to obtain a qualitative insight into the mechanism of the $r$-mode instability and to quickly check the dependence of the instability on various important factors, such as the central density of the star, the different types of viscosity in neutron stars, the different possible cooling processes etc. Additionally, we expect that there is a number of issues related to the spinning mechanisms of pulsar such as accretion disc induced spin up, or the creation of millisecond pulsars due to accretion-induced collapse of a white dwarf (\\cite{AKS98}), for which one can use the simple relations provided here for a fast but still accurate evaluation of the various evolution scenarii. ",
        "conclusions": "Our analytical results for the onset of the $r$-mode instability in neutron stars agree well with numerically obtained results, for the same neutron star model. Using our analytic formulae, the uncertainty in the bulk viscosity can be easily explored  for different present and future estimates of the bulk viscosity coefficient. The shear viscosity for lower temperatures is also uncertain, since a high shear viscosity due to the mutual friction between superfluid vortices below the superfluid transition temperature could suppress the instability. If future investigations provide a definite answer on the effect of mutual friction in a superfluid, one can easily study the implication on the $r$-mode instability using the analytic formulae presented in this paper.  We would also like to point out that, although an $l=1$ dipole mode does not radiate in a non-rotating star, it does emit gravitational waves through the coupling to higher order terms, in rotating stars. According to (\\ref{tsv}), the shear viscosity for $l=1$ vanishes and thus cannot affect the emission of gravitational waves, this is not true for realistic equations of state but still the damping times are extremelly long. The $l=1$ $r$-mode will radiate through the coupling to spheroidal $l=2$ terms, i.e. it will generate mass quadrupole radiation. The frequency of this mode in the rotating frame is \\ $\\bar \\omega_r = \\varpi$,  while in the inertial frame the frequency is \\ $\\bar \\omega_i = (3/4) \\varpi^3$. According to the criterion for the onset of the CFS-instability, the $l=1$ $r$-modes are thus stable to the emission of gravitational waves, in contrast to the $l \\geq 2$ modes. Such stable oscillations, unaffected by shear viscosity at low temperatures, could be excited during a neutron star glitch. In analogy to the $l=1$ $r$-modes, an $l=1$ spheroidal mode, like the $f$ mode or the $p$-modes, will emit current quadrupole radiation and this case needs further study."
    },
    "9805/astro-ph9805068_arXiv.txt": {
        "abstract": "The observations in broadband linear polarization (BBLP) represent a powerful diagnostic tool for detecting the magnetic morphology of CP stars. In several cases, BBLP observations are not consistent with the prediction of the oblique rotator model with a simple dipolar field. Discrepancies between theoretical predictions and observations may partially be ascribed to the effect of desaturation due to the blending of spectral lines, which has not been taken into account so far in the models. However, there are strong evidences that in many cases the simple dipolar model is not sufficient to describe the magnetic configuration of CP stars. ",
        "introduction": "\\label{Sect_Techniques} The detection of magnetic fields in CP stars is based on the analysis of the Zeeman effect in the Stokes parameter profiles of spectral lines. Roughly speaking, the line intensity profile is sensitive to the mean magnetic field modulus, whereas the spectrum of the radiation observed in left and right circular polarization gives information on the mean longitudinal magnetic field. The linear polarization profile of a spectral line is sensitive to the transverse component of the magnetic field. Magnetic fields can be detected with several techniques, which have different constraints and provide different kinds of information.  From the relative wavelength shift of the line components split by the magnetic field, one can derive the \\textit{mean magnetic field modulus} $\\langle\\vert B \\vert\\rangle$, that is, the average over the visible stellar disk of the modulus of the magnetic field. So far, $\\langle \\vert B \\vert \\rangle$ has been measured in 42 CP stars. About a dozen of them are well monitored throughout the entire stellar rotation cycle (Mathys et al.\\ 1997).  The wavelength shift of the spectral lines between right and left circular polarization is proportional to the average over the stellar disk of the component of the magnetic field along the line of sight, $\\langle B_z \\rangle$. This quantity is called \\textit{mean longitudinal magnetic field}. An alternative method to derive the same quantity is based on the anaylsis of the wings of the H$_\\beta$ line (Borra \\& Landstreet 1980). Most of our knowledge of magnetic fields in CP stars relies on the observations of $\\langle B_z \\rangle$. Observations were performed e.g., by Babcock (1958), Borra \\& Landstreet (1980), Mathys (1991). About 150 stars have been monitored so far.  A quantitative analysis of the cross-over effect allows one to derive the average over the stellar disk of the component of the magnetic field along the line of sight times the distance between the point on the stellar surface and the plane defined by the line of sight and the stellar rotation axis (Mathys 1995a). This quantity, named \\textit{mean asymmetry of the longitudinal magnetic field}, $\\langle dB_z \\rangle$, has been detected in 44 stars. About a dozen of them were well monitored throughout the rotation cycle.  The second order moment of the profile $I$ about the line centre is proportional to the sum of the squares of the two previous quantities, $\\langle B^2 + B^2_z \\rangle$ (Mathys 1995b). This quantity, shortly called \\textit{mean quadratic magnetic field}, has been detected in about 40 stars, 11 of them well monitored in the rotation cycle.  The transfer of radiation in the stellar atmosphere is responsible for a differential saturation of the $\\sigma$ and $\\pi$ components of the Zeeman multiplet, which gives rise to a phenomenon of \\textit{broadband linear polarization}. This phenomenon has been known for a long time to be observable on sunspots (Leroy 1962), but only very recently this kind of observations have been systematically performed on CP stars (Leroy 1996). The observed BBLP is roughly proportional to some bilinear forms built up with the transverse components of the magnetic field, namely, $\\langle B^2_x - B^2_y \\rangle$ and $\\langle B_xB_y \\rangle$, where $xyz$ represents an orthogonal reference system with the $z$ axis oriented as the line of sight, and $x$ towards the north celestial pole (Landolfi et al.\\ 1993). BBLP has been detected in 55 CP stars, 16 of them are well monitored throughout the rotation cycle. ",
        "conclusions": ""
    },
    "9805/astro-ph9805318_arXiv.txt": {
        "abstract": "The results of a spectroscopic and photometric investigation of the central region of M~31 are presented. An analysis of absorption-index radial profiles involving magnesium, calcium, and iron lines has shown that the unresolved nucleus of M~31 is distinct by its increased metallicity; unexpectedly, among two nuclei of M~31, it is the faintest one located exactly in the dynamical center of the galaxy (and dynamically decoupled) which is chemically distinct. The Balmer absorption line $H\\beta$ has been included into the analysis to disentangle metallicity and age effects; an age difference by a factor 3 is detected between stellar populations of the nucleus and of the bulge, the nucleus being younger. The morphological analysis of CCD images has revealed the presence of a nuclear stellar-gaseous disk with a radius of some 100 pc, the gas component of which looks non-stationary, well inside the bulge of M~31. ",
        "introduction": "The stellar core of the nearby spiral galaxy \\object{M~31} is decoupled both dynamically and chemically from the bulge of the galaxy.  Discussion on the dynamical distinctness of the nucleus of \\object{M~31} began in 1988 when Dressler and Richstone (\\cite{dr88}) and Kormendy (\\cite{k88}) published kinematical data for stars in the center of \\object{M~31}: the rotation curve reached a sharp maximum at a radius of 1\\arcsec\\ followed by a drop practically toward zero velocity; the stellar velocity dispersion had also a prominent maximum in the center. The dynamical distinctness of the nucleus of \\object{M~31} was then explained by the presence of a supermassive black hole. Later the black hole hypothesis was confirmed when an image obtained with the HST revealed that a mass of $3 \\,\\ 10^7\\, M_{\\sun}$ attached to the dynamically decoupled nucleus is concentrated not in the bright star-like source P1 but in a fainter nucleus P2, $0\\farcs5$ from P1 (\\cite{letal93}). The true nucleus of \\object{M~31}, P2, is quite faint optical source, the mass-to-luminosity ratio is high; so the idea of the presence of a supermassive black hole in the center of \\object{M~31} is finally confirmed. However even proponents of the supermassive black hole hypothesis accepted the simultaneous presence of a dynamically decoupled stellar subsystem, namely, of a compact nuclear disk with a radius of 3\\arcsec--5\\arcsec\\ (10--17 pc). Kormendy (\\cite{k88}) found a zone of low stellar velocity dispersion in the radius range $1\\arcsec < r < 4\\arcsec$, that is, a cold nuclear stellar subsystem embedded into the bulge, and Tremaine (\\cite{tr}) had claimed from dynamical arguments that the \"double\" nucleus in \\object{M~31} can be stable during several billion years as a thick eccentric elliptical nuclear disk which contained the supermassive black hole at one of its foci. So the presence of a \\emph{stellar} dynamically decoupled nucleus -- or compact disk -- in the center of \\object{M~31} is also widely accepted.  Interestingly, the chemical distinctness of the nucleus of \\object{M~31} was noted some decades earlier than its dynamical distinctness, at the romantic epoch when one photographic spectrum of the galaxy had to be exposed during three nights. Joly and Andrillat (\\cite{ja73}) pointed out a change of the \\ion{Na}{i}, \\ion{Mg}{i}, CN, and \\ion{Ca}{ii}K equivalent widths between the nucleus and the bulge; if compared to galactic globular clusters, the magnesium and calcium abundance of the bulge is --0.84 dex, and the nucleus is more metal-rich by 0.6 dex in calcium and by 1.5 dex in magnesium. Surprisingly, the data of Joly and Andrillat (\\cite{ja73}) on the iron absorption lines show a prominent equivalent width gradient in the bulge, but the nucleus seems to share this gradient and does not look distinct. A CN break between the nucleus and the bulge of \\object{M~31} was also noted by McClure (\\cite{mc}) though not by spectroscopy but by narrow-band photometry: filters with passbands of 83--85 \\AA\\ were centered on the absorption line CN$\\lambda 4165$ and on the off-line continuum at $\\lambda 4255$. Spinrad and Liebert (\\cite{sl75}) found another photometric change, $\\Delta (U-V) = 0.13 \\pm 0.04$; but this colour change corresponds to a very modest metallicity break, not more than 0.2 dex. Morton and Andereck (\\cite{ma76}) have confirmed the result of Joly and Andrillat (\\cite{ja73}) on the equivalent width change for \\ion{Ca}{ii}K line -- $EW_{nuc}/EW_{bulge}=1.38 \\pm 0.11$, -- and Bica et al. (\\cite{basch90}) have also noted changes in the \\ion{Mg}{i} and \\ion{Ca}{ii} lines, but not so prominent as those in the paper of Joly and Andrillat (\\cite{ja73}): if calibrated into metallicity, they gives a metallicity difference of 0.2 dex between the nucleus and the bulge. Cohen (\\cite{cohen}) exposed a long-slit spectrum of \\object{M~31} along the east-west direction by using a linear digital detector; she found changes of the absorption lines NaD and Mgb and, by modelling them with stellar population synthesis, obtained a metallicity difference of four times (0.6 dex). In general, though the existence of metallicity difference between the nucleus and the bulge of \\object{M~31} has been established long ago, quantitative estimates range from 0.2 dex to an order of magnitude. Moreover, the studies of abundances of the stellar populations in the center of \\object{M~31} in the 60--70 was followed by a long pause, and investigations with modern CCDs have not been undertaken. The only exception is the work of Davidge (\\cite{dav97}); but he does not concern the nucleus being concentrated on gradients of absorption-line equivalent widths in the inner bulge. Another, even more recent work of Davidge et al. (\\cite{detal97}) treats the properties of the stellar populations in the nucleus of \\object{M~31} based completely on photometric data; their C-M diagrams for individual stars being constructed under very good seeing conditions ($FWHM \\approx 0\\farcs15$) in the narrow concentric rings with radii of $0 - 1\\farcs4$, $1\\farcs4 - 2\\farcs8$, and $2\\farcs8 - 4\\farcs0$ have allowed to detect a noticeable increase in mean stellar age with radius in the innermost part of the galaxy.  Our paper presents a spectral investigation of the stellar population properties in the nucleus and in the inner bulge of \\object{M~31}. In contrast to our precursors, we have now the possibility to compare both the mean metallicity and the mean age of stars and to determine in this way a sequence of star formation epochs. The work is undertaken within a large observational project on chemically distinct nuclei in galaxies and on possible relations between chemically and dynamically decoupled galactic nuclei. ",
        "conclusions": "Though \\object{M~31}, the most nearby spiral galaxy, has been studied for a long time and with unmatched spatial resolution, the true structure of its central region is still a puzzle. The large amount of accumulated observational data has led to reject many models, even regarded earlier as reasonable ones. Our work contributes to this process.  Two extraordinary features -- the strong increase of angular rotation velocity and of stellar velocity dispersion inside a radius of 1\\arcsec\\ and the shift of the bright point-like nucleus with respect to the center of the galaxy -- were initially explained in the frames of two equally successful models: a strong anisotropy (end-on tumbling minibar: \\cite{gerhard}) and an axisymmetric potential distribution with a supermassive black hole in the center (Kormendy \\cite{k88}). The former model looked preferable because of the isophote twist along the radius of \\object{M~31} and because of a noticeable line-of-sight velocity gradient along the photometric minor axis (Ciardullo et al. \\cite{ciar}) -- these observational facts had allowed Stark and Binney (\\cite{sb94}) to suggest the presence of a large-scale bar in \\object{M~31} extended up to 3\\arcmin\\ from the center. However Bacon et al. (\\cite{bemn94}) observed a central part, $r \\le 5\\arcsec$, of \\object{M~31} with the Integral Field Spectrograph TIGER and could compare the surface brightness map with the line-of-sight velocity field; they have found a coincidence of the dynamical and photometric major axes and in this way proved the circularity of stellar rotation inside a radius of 5\\arcsec. The nuclear region of \\object{M~31} has appeared to be axisymmetric. After that a model of Tremaine (\\cite{tr}) has become very popular: he has proposed an eccentric disk rotating in accordance with the Kepler's law and having a supermassive black hole in one of its foci, P2. In the frame of this model the bright nucleus P1 is an apocenter where stars decelerate and their orbits crowd, then forming a surface brightness excess. The model of Tremaine (\\cite{tr}) is often cited as the most realistic; however it has been dismissed two years ago. Gerssen et al. (\\cite{gkm95}) had observed \\object{M~31} with a long-slit spectrograph; they had oriented the slit along the minor axis of the nucleus ($P.A.=148\\degr$) and set its width as $1\\farcs25$, so that the putative disk of Tremaine falls completely within the slit. Due to angular momentum conservation, the luminosity-weighted velocity of a filled Keplerian orbit about a stationary object should be zero; but the analysis of the observations showed the presence of two distinct kinematical components at $r=0$! Indeed, only one model is not rejected: that of a giant stellar cluster falling to the center of \\object{M~31} under the attraction of a black hole (\\cite{emcomb97}); but such a configuration is unstable and very short-lived, so it implies that \\object{M~31} is observed during an unique evolutionary stage.  The results obtained in this work even worsen the situation. Earlier when we discovered a chemically distinct nucleus in galaxies more distant than \\object{M~31}, we supposed it to be a compact nuclear disk; in several cases where the chemically distinct nuclei appeared to be resolved, our supposition was confirmed by a coincidence of chemically, kinematically and photometrically distinct area radii (\\object{NGC~1052} -- \\cite{me95}, \\object{NGC~4621} -- \\cite{me97}; but note that they are elliptical galaxies). The case of \\object{M~31} has destroyed our presumption: the nuclear disk of \\object{M~31} detected by its photometric signature has a radius of 80--100 pc, while the chemically distinct nucleus is unresolved, being less than 3 pc. Moreover, it is not P1 which is chemically distinct, though only P1 is thought to be a giant stellar cluster, but P2 which is assumed to be a supermassive black hole the stellar content around which is not yet known. The central region with a radius of $2\\farcs3$ in Fig.~\\ref{diag} looks three times younger than the bulge; but P2 contributes only a small fraction into the luminosity of this region, namely, 30\\%\\ if to assume the photometric decomposition model C from the work of Bacon et al. (\\cite{bemn94}); so we must conclude that the stellar population of P2 is significantly younger than 5 billion years. Hence a relatively recent star formation burst exactly at the dynamical center of \\object{M~31} seems indubitable. It may not be the only and last one: dust spiral arms in the nuclear disk of \\object{M~31}, the local splitting of emission lines in several spots within a dozen arcseconds from the center seen in our data and also in the data of Ciardullo et al. (\\cite{ciar}), the bright nucleus P1 which cannot be on a stable orbit -- all these facts are evidences in favour of a continuous matter drift to the center of \\object{M~31}. Though less probable, there exists still another explanation of the observational facts reported by us: if in the proximity of the supermassive black hole located inside P2 the stars have some unusual structure, say, if their external atmospheres are removed by tidal effects, then we see their inner layers which are enriched by metals. In this case the nucleus P2 may look like a chemically decoupled nucleus too."
    },
    "9805/hep-ph9805288_arXiv.txt": {
        "abstract": " ",
        "introduction": "\\hspace*{1em} About thirty years ago it was pointed out that sufficiently energetic cosmic rays will lose energy scattering with the cosmic microwave background (CMB) or radio wavelength background radiation (RBR).  Greisen \\cite{Greisen} and Zatsepin and Kuzmin \\cite{ZK} (GZK) noticed that above threshold for pion production a cosmic-ray proton will lose energy by pion photoproduction. Sufficiently energetic nuclei also have a limited range before suffering photodissociation.  The mean-free-path for a proton with energy above pion production threshold is less than a dozen Mpc, and the mean--free--path for a nucleus above threshold for photodissociation is even shorter\\footnote{However, the original estimates of the infrared and far infrared background have been recently reexamined, leading to a possibility of a larger photodissociation mean free path \\cite{stecker}.}.  With the assumption that very high energy cosmic rays originate from sources more distant than this, Greisen and Zatsepin and Kuzmin predicted that there should be a maximum observed energy in the cosmic-ray spectrum. This maximum energy, known as the GZK cutoff, is usually taken to be around $5 \\times 10^{19}$eV.  Surprisingly, today there are more than 20 events observed with energy in excess of the GZK limit, and 8 events\\footnote{This number does not include events from southern hemisphere detectors.} above $10^{20}$eV \\cite{vr,hp,yak,flys,akeno}. The slope of the energy spectrum appears to change around $10^{19}$eV (in fact it appears to harden),  possibly indicating a new hard component in the cosmic-ray spectrum above the cutoff.  The most straightforward interpretation of the observation of cosmic rays above the GZK cutoff is that the sources of the cosmic rays must be ``local'' (within 50 Mpc or so).  Since cosmic rays above the GZK cutoff should be largely unaffected by intergalactic and galactic magnetic fields, there should be candidate cosmic accelerators associated with the arrival direction of the UHE cosmic rays.\\footnote{In this paper, {\\it ultra-high energy} (UHE) will refer to energies near or above the GZK cutoff, which we take as $5 \\times 10^{19}$eV.}  But there seems to be no unusual astrophysical sources such as active galactic nuclei or quasars {\\it within 50 Mpc} in the direction of the events above the GZK cutoff.  There are hints of unusual sources at cosmological distances (redshift of order unity) associated with the arrival directions of the highest energy cosmic rays \\cite{elbert,glennys-peter}, but there is no way a UHE nucleon or nucleus could reach us from such distances.  This suggests that new physics or astrophysics is required to understand the nature, production, composition, and propagation of such high-energy cosmic rays.  This issue has attracted a lot of attention, because even the most conservative explanations involve exciting new physics.  The first possible explanation is that the UHE cosmic rays actually do originate within 50 Mpc from us but we cannot see the accelerator. One might argue that the fact that there are events above the GZK cutoff is evidence that there must be galactic or extragalactic regions where the magnetic fields are far stronger than generally believed, so the arrival direction would not be correlated with the source direction \\cite{colgate}.  Or, perhaps the reason we do not observe any vestige of the cosmic accelerator is that the acceleration results from a transient phenomenon such as a cosmic defect \\cite{cosmicstring} or a gamma-ray burst in a relatively nearby galaxy \\cite{gamma-ray-burst}.  In these scenarios the source could be within 50 Mpc and a primary cosmic-ray hadron could reach us without appreciable energy loss.  Note however that the integrated flux reduction factor for a $3.2 \\times 10^{20}$ eV cosmic ray is of order 100 or more at 50 Mpc (c.f., Fig.\\ 2 of Ref.\\ \\cite{elbert}).  A second possibility is that the UHE cosmic rays are not nucleons, and originate from sources at cosmological distances without suffering energy loss while propagating through the radiation backgrounds. Stable non-hadronic primary candidates are electrons, photons, and neutrinos.  Electrons are not candidates since their energy loss during propagation is huge. Photons interact with the CMB and RBR and lose energy by $e^+e^-$ production, resulting in a propagation distance comparable to protons.  Furthermore, the highest energy cosmic-ray events are unlikely to be due to photons because photon-induced showers do not resemble the observed Fly's Eye shower \\cite{vasquez} and the muon content of other events is consistent with expectations for a hadronic rather than photon primary \\cite{akeno}.  Neutrinos could reach us from cosmological distance without energy loss, but it was shown that even allowing for new physics, the observed events could not be due to neutrino-induced showers \\cite{birdman}.  Even with the meager number of UHE events it seems very likely that the atmospheric shower was induced by a strongly interacting particle.\\footnote{Weiler has proposed a hybrid model \\cite{weiler-nu} where UHE neutrinos propagate from cosmological distances, then produce hadrons in collisions with massive neutrinos in the galactic halo.  This model seems to require several parameters to be finely tuned.}  A third possibility is that the primary is a hadron, as suggested by the properties of the observed air showers, but for some reason it has a longer path length than a normal nucleon.  There are two ways this might happen.  The first way is if the cross section for interaction with the CMB is smaller.  The second way is for the energy threshold for {\\it resonant} photoproduction interactions to be higher, so that the GZK bound is ``postponed\" to higher energy.  Farrar \\cite{glennysI} pointed out that the existence in certain supersymmetric models of a new stable hadron (e.g., $S^0$) with a mass of several GeV can realize this possibility.  Due to its greater mass and its large mass-splitting to its resonances, the threshold energy for resonance pion production is increased compared to a proton.  This increases its effective path length and postpones the GZK cutoff to higher energy.  A complete calculation of energy loss due to pion photoproduction, redshift, and $e^+e^-$ production was done by Chung, Farrar and Kolb\\cite{dan} (CFK) for $S^0$'s in the mass range around 2 GeV.  Their analysis demonstrated that the effective range through the CMB of a neutral primary as light as 2 GeV, with cross sections typical of a neutron but with a larger gap to the lowest resonance, would be fifteen times more than a nucleon.  Although their detailed analysis involved masses of around 2-3 GeV, they pointed out that one might entertain the possibility that the primary is {\\it much} more massive than the nucleon.  Indeed, extensions of the standard model often involve new colored particles with masses of order a TeV, which may be stable or long-lived.  Since the pion production threshold energy is directly proportional to the mass of the primary, the GZK cutoff energy trivially increases for a massive primary and, in addition, the fractional energy loss per collision is smaller for a more massive primary.  Thus, from purely kinematical considerations the path length of the primary increases rapidly as its mass increases.  If a cosmic accelerator produces primary protons with energies in excess of $10^{21}$ or so eV, then the accelerated proton could collide with a proton at rest near the accelerator and produce massive, high-energy particles with energy above the GZK cutoff: in the collision of a primary of energy $10^{21}$eV with a proton at rest, the invariant energy is $\\sqrt{s} \\sim10^3$TeV!  CFK coined the name ``uhecron\" to describe a new species of long-lived or stable, electrically neutral\\footnote{The reason CFK take the uhecron to be neutral is to avoid its interaction with galactic and extragalactic magnetic field and the consequent energy loss due to pair production and other mechanisms. This may not be an essential requirement, depending on the distance to the accelerator.} hadron which, due to its mass, can provide the answer to the cosmic-ray conundrum.  Such a primary of mass $M$ can have sufficient range through the cosmic backgrounds to reach us from sources at cosmological distances ($L\\sim 1\\, {\\rm Gpc}$) if its mass is more than a few GeV, and its lifetime is greater than $10^6{\\rm s}(M/3{\\rm GeV})(L/1{\\rm Gpc})$.  But, as CFK pointed out, the fact that the uhecron is hadronic does not guarantee that it will shower in the atmosphere like a nucleon. If the uhecron contains a heavy constituent of mass $M$, the fraction of its momentum which is carried by light degrees of freedom is approximately $ \\Lambda_{QCD}/M$.  It is energy from the light degrees of freedom that is released in a typical hadronic interaction.  Thus, if the uhecron contains a massive constituent, the energy spectrum of the particles produced in a uhecron-nucleon interaction will be very soft.  CFK argued that for a sufficiently massive uhecron it should be possible to differentiate between a shower induced by a uhecron and a shower induced by a proton.  The purpose of this paper is to determine the maximum possible mass of a primary consistent with present observations by comparing the shower of a massive uhecron with the development of the highest-energy cosmic-ray shower observed, the $E\\sim 3.2 \\times 10^{20}$eV event observed by the Fly's-Eye detector \\cite{flys}.  On the basis of the comparison, we show that the uhecron must be less massive than about 50 GeV in order to explain the highest energy event.  We also perform a more general comparison of proton-induced and uhecron-induced air showers to find means of discriminating between them.  Since the fluctuations in the development of hadronic showers is large, it is impossible to identify reliably the difference between a massive uhecron and a proton on a shower-by-shower basis.  But we show that with sufficient statistics a uhecron with mass larger than 10 GeV may be distinguishable from proton showers.  It is important again to restate explicitly the picture of the uhecron we will adopt for this study.  We will assume the uhecron consists of a single constituent which accounts for the bulk of the mass of the particle, surrounded by light hadronic degrees of freedom (gluons and/or light quarks).  This definition of a uhecron strictly does not encompass the archetypical uhecron, the $S^0$ \\cite{glennysI, dan} which is a bound state of light quarks and a relatively light gluino ($uds\\tilde{g}$).  The momentum of the $S^0$ is shared roughly equally by all constituents, so its shower properties are expected to resemble those of a nucleon.\\footnote{Furthermore, even if the $S^0$ momentum were carried by a single constituent, our results (see below) indicate that with a mass of order 2 GeV its shower is not distinguishable from a proton.}  However there are a number of interesting uhecron candidates for which the present definition is appropriate: \\begin{enumerate}  \\item Raby has proposed a class of low-energy SUSY models in which the gluino is the lightest supersymmetric particle \\cite{raby}. Presumably the lightest color-singlet state containing the gluino is the $R^0$ (a gluino--gluon color singlet).  For gluinos more massive than the QCD scale, the mass of the $R^0$ would be approximately the mass of the gluino, which could be anywhere from a few GeV to 100 GeV in his model.  \\item Extensions of the standard model often include new colored particles of mass in the 1 to 10 TeV range, often stable because of some accidental symmetry or a new conserved quantum number (see \\cite{dan,rabi} for examples).  Note that cosmological restrictions require them to be unstable over some mass ranges \\cite{nardi-roulet}.  \\item The uhecron may be a massive magnetic monopole (mass around $10^{10\\pm1}$GeV), accelerated by galactic magnetic fields to UHE energies \\cite{weilerpole}.  The magnetic monopole would have a small core, but could also be surrounded by a gluon cloud.  Since the monopole is accelerated by galactic fields, there need be no striking astrophysical sources in the direction of the incoming UHE cosmic rays.  Note, however, that it has been argued that their spectrum and arrival direction disagrees with observation \\cite{ubatuba}.  \\end{enumerate}  In the next section we describe the longitudinal development of air showers.  After a very brief description of the parameterization of the development of the air shower, we describe the procedure used in the simulation of the phenomenon.  We then discuss the modifications necessary to study the development of uhecron-induced showers.  In Section 3 we present the results of the uhecron simulations and contrast the development of uhecron showers and proton showers.  We next compare the uhecron results to the data for the highest-energy event observed and derive a limit on the mass of the uhecron.  We then discuss the kind of limit one might conceivably obtain with a larger data set.  Finally, before ending the section, we describe the ground-level particle content of a uhecron shower as might be observed with a ground array of detectors.  We then present a short concluding section. ",
        "conclusions": "\\label{sec:conc} \\hspace*{1em} We conclude from our simulations that a uhecron (a massive, long-lived and strongly interacting particle) that penetrates Earth's atmosphere can produce an air shower with characteristics qualitatively similar to those of a proton shower.  We find an approximate upper limit of 50 GeV on the mass of such a particle in order to have a shower compatible with the longitudinal development observed in the highest energy cosmic ray event \\cite{flys}.  If the cross section of the uhecron is lower than a ``normal\" hadronic cross section, this limit is even more stringent.  For instance it drops to 40 GeV for a cross section 1/10 of the pion-nucleon cross section.  We note that as the mass of the uhecron increases, the shower develops more slowly (larger average $X_{MAX}$) and the fluctuations in $X_{MAX}$ increase.  With a large number of events whose longitudinal shower development is measured, these differences could be sufficient to distinguish uhecron from proton primaries.  We estimate that about 30 events may be enough to discriminate between uhecron and proton, if the uhecron is more massive than 10 GeV. At lower uhecron mass the discrimination is difficult because the longitudinal development profiles become too similar.  In this case the ratio of muon over electromagnetic density at large distances from the core could be helpful with very high statistics.  We conjecture that having combined information on a single event, longitudinal development as well as lateral shower properties and the mu/em ratio, may give improved resolving power.  Such combined information would be available if the event were observed simultaneously with both fluorescence and ground level detectors, as in the Pierre Auger project \\cite{auger}.  It is encouraging that approved and proposed experimental efforts \\cite{auger,hires,teles} are likely to produce a substantial increase in the statistics available on ultrahigh energy cosmic rays.  We have shown that the detailed shower properties can allow candidate new-particle explanations for the highest energy events to be ruled out or conceivably confirmed.  Many possible new-particle explanations are already excluded.  In particular, the highest energy cosmic ray event is very unlikely to have been produced by a new superheavy ($M \\ga 100$ GeV) hadron containing a heavy colored particle or monopole.  \\vspace{24pt} We would like to thank Clem Pryke, Jim Cronin, Gustavo Burdman, Sergio Sciutto and Albert Stebbins for useful discussions. IA was supported in part by NSF Grant AST 94-20759 and the DOE through grant DE-FG02-90ER40606. The work of GRF was supported by NSF-PHY-94-23002.  The work of EWK was supported at Fermilab by the Department of Energy and by NASA under number NAG5-7092."
    },
    "9805/astro-ph9805362_arXiv.txt": {
        "abstract": "This review discusses the current status of our knowledge of emission by dust and molecular gas in high redshift radio galaxies, and the uncertainties in the derivation of physical parameters from these data. The similarity of far-infrared luminous, gas-rich high redshift radio galaxies and local ultraluminous infrared galaxies (ULIGs) is discussed. Given that local ULIGs rapidly convert most of their gas reservoir into stars, far-infrared luminous high-$z$ radio galaxies are likely undergoing immense bursts of star formation, possibly accounting for a large fraction of the final stellar populations in these systems. These results are discussed in the context of formation scenarios of massive galaxies. ",
        "introduction": "In recent years, evidence has been accumulating that radio galaxies at various\\footnote{\\hspace{-1.8em}{\\sl to appear in {\\it The most distant radio galaxies}, eds. H.J.A.\\ R\\\"ottgering, P.N.\\ Best, \\& M.D.\\ Lehnert, Kluwer Academic Publishers (Dordrecht)}} redshifts can be used to trace the formation and evolution of massive elliptical galaxies. At low redshift, it has long been known that that radio-loud active galactic nuclei (AGNs) are found predominantly in luminous ($L\\sim2-5\\,L_*$) elliptical galaxies, which are often the dominant cD galaxies of rich clusters (e.g., \\citebare{Matthewsetal64}). At $z\\sim1$, the host galaxies of well-studied 3CR sources are fully formed massive ellipticals \\cite{Bestetal98a}. High redshift ($z>2$) radio galaxies on the other hand, show a variety of morphologies, which is often characterized by the presence of a number of clumps (e.g., \\citebare{Mileyetal92}; \\citebare{Pentericcietal97}), totally unlike the more relaxed appearance of their lower $z$ counterparts. The properties of these high-$z$ radio galaxies likely reflect processes related directly to the buildup of the bulk of the stellar population. Hence the formation of present-day massive ellipticals may be observed directly in high-$z$ radio galaxies.  The purpose of this review is to address the implications of the recent detections of dust emission in a number of high-$z$ radio galaxies for our understanding of the role of these objects in the formation and early evolution of massive galaxies. After a description of the general observational and theoretical context (\\secref{sec.context}), the present observational situation in this rapidly developing field is reviewed (\\secref{sec.obs}). These results are then discussed in the context of what we know about far-IR luminous galaxies in the local universe (\\secref{sec.local}). Throughout this paper we adopt a value $H_0=75\\,\\kms\\,\\pun{Mpc}{-1}$ for the Hubble constant, and a deceleration parameter $q_0=0.1$, unless otherwise indicated. Where necessary, results from other publications are tacitly converted to the cosmology adopted here. ",
        "conclusions": ""
    },
    "9805/astro-ph9805224_arXiv.txt": {
        "abstract": "Highlights from the RXTE All Sky Monitor (ASM) during 1996 and 1997 are reviewed with particular attention to X-ray transients. The ASM has detected 117 sources. These include 12 recurrent transients and 10 new X-ray sources, some of which began their outbursts before the launch of RXTE. The majority of the outburst profiles are strikingly different from the classical form, with its fast rise and slow decay. Some of the light curves appear quasi-persistent, which suggests that they are associated with secular changes in the accretion rate from the donor star rather than with a cyclic instability in the accretion disk. The nature of the compact object is uncertain in many cases, but there is a likelihood that the majority of new sources are black hole systems, while the majority of recurrent transients contain neutron stars. There is a correlation between the gross shape of the energy spectrum, as seen in the ASM hardness ratio $HR2$, and the accretion subclass. These data, with important contributions from SAX, CGRO, and Granat, provide a deeper probe of the X-ray sky than has been available previously.  However, there are important systematic effects to be investigated before a meaningful evaluation of transient occurrence rates can be derived. ",
        "introduction": "The All Sky Monitor (ASM) on the {\\it Rossi X-ray Timing Explorer} (RXTE) has been regularly observing bright celestial X-ray sources since 1996 February 22. Detector problems had been encountered during the first days of operation (1996 January 5-12), but the instrument has remained stable under an observing plan restricted to low-background regions of the RXTE orbit (580 km altitude).  The ASM currently operates with 20 of the original 24 detector anodes, and the typical observation duty cycle is 40\\%, with the remainder of the time lost to the high-background regions of the orbit, spacecraft slews, and instrument rotation or rewinds. The net yield from the ASM exposures is about 5 celestial scans per day, excluding regions near the Sun. The ASM instrument calibration, construction of the data archive, the derivation of source intensities, and efforts to locate new X-ray sources are all carried out with integrated efforts of the PI team at M.I.T. and the RXTE Science Operations Center at Goddard Space Flight Center.  The ASM instrument consists of three scanning shadow cameras (SSC) attached to a rotating pedestal. Each camera contains a position-sensitive proportional counter, mounted below a wide-field collimator that restricts the field of view (FOV) to 6$^{o}$ x 90$^{o}$ FWHM and 12$^{o}$ x 110$^{o}$ FWZI. One camera (SSC3) points in the same direction as the ASM rotation axis. The other two SSCs are pointed perpendicular to SSC3. The latter cameras point toward a common direction, but the long axes of their collimators are tilted by +12$^{o}$ and -12$^{o}$, respectively, relative to the ASM rotation axis. The ASM can be rotated so that the co-pointing SSCs are aligned with the larger instruments of RXTE (i.e. the PCA and HEXTE). The top of each collimator is covered with an aluminum plate perforated by 6 parallel (and different) series of narrow, rectangular slits.  These slits function as a coded mask by casting a two-dimensional shadow pattern onto the position-sensitive anode wires in the detector. The histograms of accumulated counts from the anodes represent the superposition of shadows from each X-ray source in the collimator's field of view. Further information on this instrument is given by Levine et al. (1996).  The ASM raw data includes 3 types of data products that are tabulated and formatted for telemetry by the two ASM event analyzers in the RXTE Experiment Data System. In the current observing mode, position histograms are accumulated for 90 s ``dwells'' in which the cameras' FOVs are fixed on the sky. Each dwell is followed by a 6$^{o}$ instrument rotation to observe the adjacent patch of sky. The rotation plans for ASM dwell sequences are chosen to avoid having any portions of the Earth in the FOVs of SSCs 1 and 2. The position histograms are accumulated in three energy channels: 1.5--3.0, 3--5, and 5--12 keV. The second ASM data product consists of various measurements from each camera binned in time. These data are useful in studying bright X-ray pulsars, bursters, $\\gamma$-ray bursts, and several other categories of rapid variability. The ``good events'' from each camera are recorded for each energy channel in 0.125 s bins, while 6 different types of background measures are recorded in 1 s bins. Finally, 64-channel X-ray spectra from each camera are output every 64 s. These data provide a means of monitoring the detector gain, since we may integrate the spectra over long time scales to observe the 5.9 keV emission line from the weak $^{55}$Fe calibration sources mounted in each collimator. In addition, ASM spectra may be useful in investigations of spectral changes in very bright X-ray sources.  The ASM data archive is a public resource available for both planning purposes and scientific analysis. The source histories are available in FITS format from the RXTE guest observer facility at http://heasarc.gsfc.nasa.gov/docs/xte/xte\\_1st.html. The ASM archive is also available in ASCII table format from the RXTE web site at M.I.T., http://space.mit.edu/XTE/XTE.html.  The majority of scientific applications for ASM data can be organized into four categories: \\begin{itemize} \\item Locating and Monitoring X-ray Transients, \\item Measuring Intensity and Spectral Variations in Persistent X-ray Binaries, \\item Long Term Behavior of Bright AGNs, \\item Positions and X-ray Properties of $\\gamma$-Ray Bursts. \\end{itemize}  This paper briefly reviews the ASM results in the first category, with consideration of results obtained during 1996 and 1997. ",
        "conclusions": ""
    },
    "9805/astro-ph9805012_arXiv.txt": {
        "abstract": "It is likely that the reionization of the Universe did not occur homogeneously. Using a model that associates ionized patches with overdense regions, we find that the resulting cosmic microwave background (CMB) anisotropy power spectrum peaks at angular scales corresponding to the extent of the ionized regions, and has a width that reflects the correlations between them. There is considerable uncertainty in the amplitude. Neglecting inhomogeneous reionization in the determination of cosmological parameters from high resolution CMB maps may cause significant systematic error. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805080_arXiv.txt": {
        "abstract": "We employ Schwarzschild's method of orbit modeling to constrain the mass profiles of the central lens galaxies in Q0957+561 and PG 1115+080. We combine the measured central projected stellar velocity dispersions of these galaxies with the self-similar radial profiles of the rms velocity and of the Gauss-Hermite moment $h_4$ observed in nearby galaxies for $0 \\lsim R \\lsim 2 R_{\\rm eff}$.  For Q0957+561, we find a 16\\% uncertainty in the galaxy mass, and formal 2-$\\sigma$ limits on the Hubble constant of $H_0 = 61^{+13}_{-15}$ km s$^{-1}$ Mpc$^{-1}$.  For PG 1115+080, we find that none of the viable lens models can be ruled out, so that $H_0$ is not yet strongly constrained by this system. ",
        "introduction": "\\longpagen[8.52in] The gravitational lens system Q0957+561 (Walsh, Carswell, \\& Weymann 1979) has been modeled extensively in an effort to determine the Hubble constant $H_0$ from measurements of the time delay between its two primary images. Since the long-running dispute over the time delay measurement has been resolved in favor of the short delay (Schild \\& Thomson 1997; Haarsma {\\it et al}. 1997, 1998; Kundi\\'{c} {\\it et al}. 1997), the largest remaining uncertainty arises from the mass model.  The lens consists of a cluster with a large central elliptical galaxy G1.  The asymmetric radial positions of the images accurately constrain the parameterized radial mass distribution of G1 (Grogin \\& Narayan 1996, hereafter GN).  However, the presence of the cluster introduces a degeneracy in the overall mass normalization of G1 (Falco, Gorenstein, \\& Shapiro 1985; Gorenstein, Falco, \\& Shapiro 1988), and thus in the determination of the Hubble constant ($H_0 \\propto \\sigma_0^2$, where $\\sigma_0$ is a velocity dispersion characterizing the mass of G1).  Therefore, additional independent constraints are needed on the relative contributions of G1 and the cluster to the $6\\arcs$ image separation.  These can be obtained by inferring {\\it the mass of the cluster} from cluster dynamics (Garrett, Walsh, \\& Carswell 1992; Angonin-Willaime, Soucail, \\& Vanderriest 1994), from hot intracluster gas X-ray emission (Chartas {\\it et al}. 1995, 1998), or from the weak lensing of background galaxies (Dahle, Maddox, \\& Lilje 1994; Fischer {\\it et al}. 1997).  Alternatively, one can infer {\\it the mass of G1} from stellar dynamical measurements.  \\normalpage The quadruple lens PG 1115+080 (Weymann {\\it et al}. 1980) is the second system with a well-determined time delay (Schechter {\\it et al}. 1997; Barkana 1997b).  The projected mass of the primary lens galaxy G inside the ring of images can be precisely determined (unlike the case of Q0957+561, uncertainties in the mass distribution of nearby galaxies have only a minor effect on the models).  However, the geometry of the system does not permit distinguishing between different mass profiles.  This has important consequences for the Hubble constant, since $H_0$ can vary by 40\\%, depending on the mass model assumed (see Keeton \\& Kochanek 1997; Courbin {\\it et al}. 1997; Saha \\& Williams 1997; Impey {\\it et al}. 1998). Stellar dynamical measurements of G may be helpful for breaking the degeneracy in the mass model, and thus in $H_0$.  Falco {\\it et al}. (1997) measured the central projected stellar velocity dispersion of Q0957+561 G1 to be $\\hat{\\sigma}_{\\rm p} = 279 \\pm 12$ km s$^{-1}$, improving on an earlier measurement by Rhee (1991). Similarly, Tonry (1998) measured the central dispersion of PG 1115+080 G to be $\\hat{\\sigma}_{\\rm p} = 281 \\pm 25$ km s$^{-1}$.  However, the conversion of the measured $\\hat{\\sigma}_{\\rm p}$ to $\\sigma_0$ is subject to systematic uncertainties, which include the unknown anisotropy structure of the stellar orbits, the radial variation of the mass-to-light ratio, and the ellipticity of the galaxy.  Previous galaxy models have arrived at a relatively small uncertainty in this conversion by making arbitrary simplifying assumptions.  Kochanek (1993, 1994) considered a singular isothermal mass model with constant anisotropy.  For Q0957+561, GN used these dynamical models, but limited them to be nearly isotropic, leading to a 2\\% systematic uncertainty in $\\sigma_0^2$ (and thus in $H_0$).  Barkana (1997a) also assumed near-isotropic orbits, leading to an uncertainty in $H_0$ of 4\\%, but he noted that allowing for more anisotropy gives an uncertainty of 14\\%.  In fact, dynamical studies of nearby elliptical galaxies have clearly demonstrated that there is little basis for the assumption of isotropy, or even of constant anisotropy.  Binney \\& Mamon (1982) and Tonry (1983) first illustrated that an elliptical galaxy's surface brightness and projected stellar velocity dispersion profiles, $I(R)$ and $\\sigma_{\\rm p}(R)$, could not determine both its mass distribution and its anisotropy profile.  Richstone \\& Tremaine (1984) and Katz \\& Richstone (1985) used orbit modeling methods to demonstrate that the conversion from the dispersion profile $\\sigma_{\\rm p}(R)$ to a mass parameter $\\sigma_0^2$ can be uncertain by an order of magnitude.  Further theoretical studies (Dejonghe 1987; Merritt 1987; Merrifield \\& Kent 1990; Gerhard 1991; Dejonghe \\& Merritt 1992; Merritt \\& Saha 1993; Merritt 1993) demonstrated that {\\it complete} knowledge of the stellar line-of-sight velocity distribution (LOSVD) ${\\cal L}(v_{\\rm p},R)$ gives a unique solution for the two-integral distribution function (DF) $f(E,L)$ in a known spherical potential $\\Phi$, and may even strongly constrain an unknown $\\Phi$.  The general efficacy of {\\it incomplete} knowledge of ${\\cal L}$ is less clear, but important constraints on $f$ and $\\Phi$ could be further provided by large-radius measurements of higher-order velocity moments ({\\it e.g.}, the Gauss-Hermite moments $h_l$ --- van der Marel \\& Franx 1993; Gerhard 1993; see also Rix \\& White 1992; Zhao \\& Prada 1996).  Rix {\\it et al}. (1997) and Gerhard {\\it et al}. (1998) made dynamical fits to nearby galaxies with higher-order moments ($h_3,h_4$) measured to $\\sim 2.5 R_{\\rm eff}$; they determined the total mass to $\\sim$  10-15\\%, and ruled out a constant mass-to-light ratio with $>99\\%$ confidence.  For the galaxy NGC 2434, the solutions typically had nearly constant radial anisotropy $\\beta(r) \\equiv (1-v_\\theta^2/v_r^2) \\simeq 0.5$, while NGC 6703 showed an anisotropy rising from $\\beta \\sim 0.1$ at the center to $\\sim 0.4$ near $R_{\\rm eff}$.  {\\it Thus the assumption of $\\beta(r)=0$, or even of constant $\\beta(r)$, is certainly unwarranted and probably incorrect. }  While our current knowledge of elliptical galaxies does not permit us to make arbitrary assumptions about the anisotropy of the DF, we can impose the constraint that the unmeasured $\\sigma_{\\rm p}(R)$ and $h_4(R)$ profiles of these lens galaxies are similar to those of other galaxies (provided they are universally homologous).  In this study we rigorously consider the utility of the central $\\hat{\\sigma}_{\\rm p}$ measurement for determining $H_0$ from these two lens systems.  To ensure physically correct, robust results, we use a spherical orbit modeling method after Schwarzschild (1979), Richstone \\& Tremaine (1984), and Rix {\\it et al}. (1997) --- a fully general way to construct realistic models of a galaxy, given an assumed potential.  In \\S 2 we review the observational constraints on the lens galaxies, and introduce constraints on their LOSVD profiles by demonstrating that the profiles of nearby elliptical galaxies are self-similar.  We describe our modeling method and demonstrate it with a test-case problem in \\S 3.  In \\S 4, we report the range of model solutions for Q0957+561, and discuss the implications for $H_0$.  We examine the solutions for PG 1115+080 in \\S 5.  In \\S 6 we give our conclusions.  \\normalpage ",
        "conclusions": "Using very general orbit modeling methods, we have examined the uncertainty in the mass of the lens galaxies Q0957+561 G1 and PG 1115+080 G, given observations of their central projected stellar velocity dispersions $\\hat{\\sigma}_{\\rm p}$ (Falco {\\it et al.} 1997; Tonry 1998).  As many past studies have shown, such a measurement alone is inadequate to strongly constrain the galaxy's mass.  In order to put additional realistic {\\it a priori} constraints on the galaxy's properties, we have derived ``mean profiles'' of the rms projected velocity $v_{\\rm rms}(R)$ and the fourth-order Gauss-Hermite moment $h_4(R)$ from a large sample of nearby elliptical galaxies.  These mean profiles prove to be remarkably self-similar, even over a large range of galaxy types.  This universality is not too surprising, given the homology of early-type galaxies implied by the existence of the fundamental plane --- an even stronger implication if the central kinetic energy is considered instead of the velocity dispersion ({\\it e.g.}, Busarello {\\it et al}. 1997).  For Q0957+561 G1, given the best-fit SPLS mass model from Grogin \\& Narayan (1996), with only the surface brightness profile $I(R)$ and central $\\hat{\\sigma}_{\\rm p}$ as constraints, we find 1-$\\sigma$ limits on the mass dispersion $\\sigma_0$ of $295^{+143}_{-121}$ km s$^{-1}$.  The addition of the mean profile constraints reduces this permitted range to $280^{+19}_{-26}$ km s$^{-1}$.  In conjunction with the GN lens model constraints, this implies a cluster convergence of $\\kappa=0.25^{+0.14}_{-0.10}$, which is consistent with the constraints on $\\kappa$ from other independent studies. Using the time delay measurement of Kundi\\'{c} {\\it et al}. (1997), we find 1-$\\sigma$ limits on $H_0$ of $61^{+9}_{-11}$ km s$^{-1}$ Mpc$^{-1}$, and 2-$\\sigma$ limits of $61^{+13}_{-15}$ km s$^{-1}$ Mpc$^{-1}$.  {\\it Thus, current measurements of the lens system Q0957+561 do not constrain $H_0$ to better than 15\\%.} To obtain useful limits on $H_0$, we will need better constraints on the cluster convergence and/or better velocity profile measurements for the galaxy G1. We will also need substantial revision of the lens model --- the GN solutions fit the lens data poorly, inaccurately assume a spherical galaxy, and use an oversimplified Taylor expansion model of the cluster. Presumably a more accurate lens model will eventually provide a good fit to the data and a different value for $H_0$ (see Barkana {\\it et al}. 1998 for some improved models), in which case our dynamical model will still be illustrative of the systematic uncertainties expected.  We have also examined the lens galaxy PG 1115+080 G, which has a total mass that is relatively well-determined by the lensing constraints, but a radial mass distribution that is unconstrained (Impey {\\it et al}. 1998).  As with Q0957+561 G1, we model this galaxy by including constraints on $I(R)$ and the central $\\hat{\\sigma}_{\\rm p}$ along with the mean profile constraints. We find that these constraints are not sufficient to break the degeneracy between the different lens models, so that a large range of values is still permitted for $H_0$ (44-68 km s$^{-1}$ Mpc$^{-1}$). Further constraints on the mass distribution in this system are needed.  The stellar dynamics of gravitational lens systems show considerable promise for determining $H_0$, even if we have only central dispersions. Larger samples of dispersion measurements would be particularly valuable, since each lens system will have a different set of systematic uncertainties (it is encouraging to note that the results for $H_0$ from Q0957+561 and PG 1115+080 are so far consistent with each other). Of particular value would be results from isolated systems with lens geometries that probe the radial mass distribution, such as MG 1654+1346 (see Ellithorpe, Kochanek, \\& Hewitt 1996) and MG 1549+3047 (Leh\\'ar {\\it et al}. 1993, 1996).  The results from a large set of these systems could be combined using Bayesian methods to converge on a robust value for $H_0$ (see Press 1997), avoiding the correlated systematic uncertainties that plague ``distance ladder'' approaches. Furthermore, the independent measurement of galaxy properties from gravitational lensing and from stellar dynamics has an enormous potential for shedding light on the detailed structure of galaxies.  \\vskip 67pt  We thank Ralf Bender and Roberto Saglia for providing their data in tabular form; Marijn Franx, Ramesh Narayan, John Tonry, and the anonymous referee for helpful comments; and Roeland van der Marel for both.  We also thank Norm Grogin for many further demonstrations of his lens models.  C.S.K. is supported by NSF grant AST 94-01722 and NASA ATP grant NAG 5-4062.  \\newpage"
    },
    "9805/astro-ph9805049_arXiv.txt": {
        "abstract": "Before 1985, attempts to detect radio or X-ray emission from Magnetic CP (MCP) stars were either fruitless or ambiguous. However, more successful results have been obtained in the last dozen years: (i) Radio emission has now been detected from $\\sim 35$ MCP stars of the Helium-peculiar and Silicon-strong subclasses, with a functional dependence of radio luminosity $L_{R} \\propto T_{eff}^7 H_s^{1.2} P_{rot}^{-0.6}$, where $T_{eff}$ is the effective temperature, $H_s$ is the surface magnetic field strength, and $P_{rot}$ is the rotational period; rotational modulation of the radio emission has also been observed for several MCP stars. All of this evidence suggests that it is the MCP stars themselves, not close companions, that are responsible for the radio emission; (ii) The X-ray emission properties of MCP stars are however still poorly characterized: although a moderate number ($\\sim 18$) have by now been associated with X-ray sources, the lack of a clear correlation of this X-ray emission with other stellar parameters has made it difficult to rule out the close companion hypothesis. ",
        "introduction": "\\label{intr} Magnetic chemically peculiar stars (hereafter MCP stars) have been recognized as a distinct subclass of chemically peculiar (CP) stars since the 1960s. The strong (kiloGauss) and predominantly dipolar magnetic fields of these stars have been considered to be possible sites of high-energy phenomena, both thermal and non-thermal, for almost as long. The closed field regions of MCP stars can, due to the high magnetic pressure, clearly confine any ionized plasma that flows from the stellar surfaces, and interaction between such putative stellar winds and the strong magnetic fields might also result in the acceleration of ions to extremely large energies. An important theoretical paper which clearly presented much of this physics was published by Havnes \\& Goertz in 1984 (q.v.). It was quickly realized that, since the stars with the higher effective temperatures have the strongest radiation-driven winds, the B-type MCP stars (i.e., the helium-peculiar and silicon-strong subclasses) were much more likely to exhibit high-energy phenomena than the `classical' A-type MCP stars (i.e., the CrSrEu-strong subclass). Shore in a series of papers (Shore 1987; Shore {\\it et al.} 1987; Shore \\& Brown 1990) followed up on earlier work to propose that the He-pec stars had predominantly polar stellar winds as well as magnetospherically trapped plasma in the equatorial regions, and used this model to explain the variation with rotational phase of ultraviolet emission lines formed at temperatures $T_e \\sim 10^5$K. Earlier papers (e.g., Groote \\& Hunger 1982) had fit the hydrogen Balmer lines of the He-strong star $\\sigma$ Ori E with a model in which plasma with $T_e \\sim 10^4$K was confined in clouds in the equatorial regions.  In the present review, I will restrict my discussion to observations of MCP stars aimed at detecting plasma at temperatures $\\ga 10^6$K using X-ray telescopes, and to observations to detect high-energy, thermal or nonthermal electrons using radio telescopes. Thus, these studies can be regarded as complementing the aforementioned optical and UV studies that have helped to constrain the characteristics of the cooler components of the circumstellar plasma. I will briefly review the early studies in the radio and X-ray energy ranges and then discuss what has been newly learned in the last decade. ",
        "conclusions": "\\label{Conc} There is an essential difference between the radio and the X-ray properties of the MCP stars, with the former being well-established and the latter quite the opposite. Nonthermal radio emission is prevalent among the B-type MCP stars, and the basic emission mechanism is almost certainly gyrosynchrotron emission from mildly relativistic nonthermal electrons. Their radio emission properties are predictable from their basic properties (cf. eq. 1), and, in general terms, explainable using current models, although detailed models which fit the radio spectra and circular polarizarion as a function of aspect for specific stars are still pending. Their X-ray emission properties are still uncertain: the existence of significant amounts of plasma at $\\ga 10^6$K has still not been confirmed to be a common feature in MCP stars despite 20 years of observations. The presence of X-ray emission is not predictable from either their MCP type or their radio properties. In this sense, their status as a class of intrinsic X-ray emitters is, in my opinion, still not demonstrated. If the existing candidate X-ray emitters are intrinsic, then a revision of the Usov \\& Melrose (1992) model is clearly required, since, in its present form, it is underpredicting the X-ray luminosities by 7 or 8 orders of magnitude. Conclusive evidence for the intrinsic hypothesis might be (a) the discovery of a distinct X-ray spectral signature, such as elemental abundances showing a similar pattern to the underlying photospheric abundances, or (b) the finding of an X-ray modulation with a similar period to that of the magnetic light-curve, or (c) the acquisition of enough X-ray detections among MCP stars that a correlation with some other stellar properties may be found."
    },
    "9805/astro-ph9805339_arXiv.txt": {
        "abstract": "There is enormous potential to advance cosmology from statistical characterizations of cosmic microwave background sky maps.  The angular power spectrum of the microwave anisotropy is a particularly important statistic. Existing algorithms for computing the angular power spectrum of a pixelized map typically require $O(N^3)$ operations and $O(N^2)$ storage, where  $N$ is the number of independent pixels in the map. The {\\it MAP} and {\\it Planck} satellites will produce megapixel maps of the cosmic microwave background temperature at multiple frequencies; thus, existing algorithms are not computationally feasible. In this article, we introduce an algorithm that requires $O(N^2)$ operations and $O(N^{3/2})$ storage that can find the minimum variance power spectrum from sky map data roughly one million times faster than was previously possible.  This makes feasible an analysis that was hitherto intractable. ",
        "introduction": "The {\\it MAP} and {\\it Planck} satellites have the potential to yield enormous amounts of information about the physical conditions in the early universe (Bennett et al. 1995; http://map.gsfc.nasa.gov; Bersanelli et al. 1996; http://astro.estec.esa.nl/Planck). For example, with full sky coverage and an angular resolution of $0.21^{\\circ }$ at its highest frequency, {\\it MAP} should return accurate measurements of the cosmic microwave background (CMB) temperature anisotropy at roughly one million independent points on the sky. If the temperature fluctuations are consistent with inflationary models, these measurements will provide accurate determinations of most of the significant cosmological parameters (Spergel 1994; Knox 1995; Hinshaw, Bennett \\& Kogut 1995; Jungman et al. 1996; Zaldarriaga, Spergel \\& Seljak 1997; Bond, Efstathiou \\& Tegmark 1997). If they are not consistent with inflationary models, it will be even more exciting as we will need to rethink our ideas about the physics of the early universe.  There are a number of non-trivial numerical steps involved in comparing $10^{11}$ temperature differences with the predictions of a particular cosmological model. Several of the steps in the calculation are now clear. Wright, Hinshaw \\& Bennett (1996) found a rapid and exact algorithm for producing megapixel cosmic microwave background maps from differential data. Seljak \\& Zaldarriaga (1996) present a fast algorithm for computing the power spectrum for a given cosmological model. However, to date, we have lacked an efficient technique for computing the power spectrum from an observed sky map. The techniques that were used to extract the power spectrum from the {\\it COBE} data (G\\'orski 1994; G\\'orski et al. 1994; Bond 1995; Tegmark \\& Bunn 1995; G\\'orski et al. 1996; Hinshaw et al. 1996; Tegmark 1996; Bunn \\& White 1997) cannot easily be extended to the high resolution data since these techniques require $O(N^3)$ operations and O($N^2$) storage. For a single frequency of {\\it MAP} data, for example, this would take $>10^{19}$ operations, with $>10^{21}$ operations required for a combined analysis of polarization and temperature maps. All proposed techniques to date are thus wholly unequal to the task of analysing upcoming data sets (Borrill 1997).  In this paper, we present a fast, accurate method for extracting the power spectrum from realistic simulations of high resolution data at a single frequency. In \\S 2-5 we develop our numerical method for determining the power spectrum from data containing only cosmological signal and detector noise. In \\S 6, we show how the same approach can be used to directly estimate cosmological parameters from the maps.  In \\S 7, we apply the method to realistic simulations of {\\it MAP} data that include spatially varying noise and a galactic sky cut.  We show that our numerical method recovers minimum variance, unbiased estimates of the power spectrum and of cosmological parameters.  We summarize our results in \\S 8.  A subsequent paper will extend the techniques developed here to the analysis of multi-frequency data and polarization data. ",
        "conclusions": "We have developed an unbiased, minimum variance estimator for the power spectrum of temperature fluctuations in CMB maps. We have used it to estimate the power spectrum up to $l=1024$ from a 2 million pixel map that simulates single frequency {\\it MAP} data (94 GHz) with realistic instrument noise and sky cuts.  In contrast with existing algorithms, which require $O(N^{3})$ operations, our algorithm is $O(N^{2})$, and thus can be run overnight on existing workstations, rather than running for months at national supercomputing facilities. We anticipate further improvements in performance with more aggressive optimization (and parallelization) of the code.  In addition, we have estimated cosmological parameters from the recovered $c_l$ and find we can recover the input model with good precision. Thus, the pipeline from differential time-ordered data to cosmological parameters is now complete and well within present day computational capabilities.  Possibilities for future work include extensions to multi-frequency and polarization data, as well as the inclusion of Galactic foregrounds and systematic effects, such as striping, in our treatment of the data. The former only incur a linear increase of order a few in the operations count. The latter must be dealt with by inclusion of the appropriate new terms in the covariance matrix, as well as subsequent simulation of the systematic effects in the Monte-Carlo trace computations."
    },
    "9805/astro-ph9805175_arXiv.txt": {
        "abstract": "Recent ROSAT measurements show that the x-ray emission from isolated neutron stars is modulated at the stellar rotation period.  To interpret these measurements, one needs precise calculations of the heat transfer through the thin insulating envelopes of neutron stars. We present nearly analytic models of the thermal structure of the envelopes of ultramagnetized neutron stars.  Specifically, we examine the limit in which only the ground Landau level is filled.  We use the models to estimate the amplitude of modulation expected from non-uniformities in the surface temperatures of strongly magnetized neutron stars.  In addition, we estimate cooling rates for stars with fields $B \\sim 10^{15}-10^{16}$~G which are relevant to models that invoke ``magnetars'' to account for soft $\\gamma$-ray emission from some repeating sources. ",
        "introduction": "Since the launch of the ROSAT satellite, our knowledge of isolated neutron stars has expanded into new realms.  Before ROSAT, neutron stars were somewhat unique among astronomical objects.  Although they had been observed over a range of energies from radio to ultra-high-energy gamma rays, and had been evoked to power a variety of astrophysical objects from pulsars to soft X-ray repeaters and gamma-ray bursts, one could not argue unequivocally that a single photon from the surface of a neutron star had ever been detected.  For the first time, we have direct evidence for radiation from the surfaces of neutron stars.  More than a dozen such sources have been detected by ROSAT (\\eg\\ \\cite{Ogel95}), and more than ten have been fitted with spectra.  The spectra divide the objects into two classes: 1) objects with hard spectra whose X-ray emission is best attributed to the magnetosphere, and 2) neutron stars whose soft flux is well-described by a blackbody spectrum.  \\tabref{psrlist} lists several of those that fall into the second group with pertinent references. \\figcomment{\\bt \\caption{Several pulsars with observed surface blackbody emission} \\label{tab:psrlist} \\begin{tabular}{ll} Pulsar & References \\\\ \\hline PSR J0437-4715        & \\cite{Beck93} \\\\ PSR 0630+18 (Geminga) & \\cite{Halp92,Halp93} \\\\ & \\cite{Halp97} \\\\ PSR 0656+14           & \\cite{Finl92}, \\\\ & \\cite{Ande93,Grei96}, \\\\ & \\cite{Poss96} \\\\ PSR 0833-45 (Vela)    & \\cite{Ogel93b} \\\\ PSR 1055-52           & \\cite{Ogel93a,Grei96} \\\\ PSR 1929+10           & \\cite{Yanc94} \\end{tabular} \\et}  The X-ray spectra of these objects consist of a soft blackbody and hard power-law component.  Generally, the surface temperature inferred from modeling the blackbody is approximately what one would expect for a cooling neutron star at the characteristic age of the pulsar. If this interpretation is correct, the observations provide a direct probe of the structure of cooling neutron stars.  Furthermore, the observations show that the thermal radiation is modulated at the rotation period of the pulsars.  A strong magnetic field can modulate the thermal flux by causing the heat conduction in the outer layers of the star to be anisotropic.  Both to translate these observations into constraints on the structure of neutron stars and to understand modulation of the radiation, we must have a detailed understanding of the insulating layers of the neutron-star crust, \\ie the envelope.  We proceed in the spirit of \\jcite{Gudm82} and concentrate our analysis on the thin region, the envelope, which insulates the bulk of the neutron star.  The envelope is customarily defined to extend from zero density to $\\rho \\sim 10^{10}$~g/cm$^3$, and its thickness ($h_E$) is of the order of tens of meters, very small compared to the radius of the star, $R\\sim 10$ km.  By limiting the analysis, we focus on how various physical processes affect the thermal structure of the envelope and the relationship between the core temperature and the flux emitted at the surface.  An alternative point of view is to combine the envelope calculation with an estimate of the cooling rate due to neutrinos and the total heat capacity of the neutron star, yielding theoretical cooling curves (\\eg\\ \\cite{Tsur72,Nomo81,Glen80,VanR81})  Several authors have made much progress in understanding the properties of neutron star envelopes with and without magnetic fields. \\jcite{Gudm82} numerically calculate the thermal structure for unmagnetized envelopes, and \\jcite{Hern84b} present analytic models for the $B=0$ case.  \\jcite{Tsur79}, \\jcite{Glen80}, \\jcite{VanR81} and \\jcite{Urpi80} calculate the luminosity observed at infinity as a function of the core temperature for several magnetic field strengths less than $10^{14}$~G, including the zero-field case.  \\jcite{Hern85} calculates the thermal structure of envelopes for $B \\le\\ 10^{14}$~G for transport along the field, using the electron conductivities of \\jcite{Hern84a} which account for the quantization of electron energies in the magnetic field in a relativistic framework.  We will use these conductivities in the present work; therefore, \\jcite{Hern85} provides a natural benchmark.  \\jcite{VanR88} builds upon the \\jcite{Hern85} results by exploring various assumptions concerning the properties of the envelope at low densities and calculating profiles for many field strengths ($B<10^{14}$~G) and core temperatures.  Again, these calculations are limited to conduction along the field.  \\jcite{Scha90a}, using the electron conductivities calculated in \\jcite{Scha88}, calculates the thermal structure in two dimensions for $B \\lsim 10^{11}$~G.  Above a field strength of $10^{12}$~G, the calculations are not considered reliable.  Finally, \\jcite{Shib95} present the temperature distribution as a function of magnetic colatitude for $B=10^{12}$~G from a numerical solution to the two-dimensional thermal structure equation in a plane-parallel approximation.  The current work complements the previous ones by extending the results to stronger field strengths ($10^{14} \\rmmat{ G} \\le\\ B \\le\\ 10^{16}$~G) in a semi-analytical fashion.  We apply the approach of \\jcite{Hern84b} in the limit of a strongly magnetized envelope, and then justify and use the plane-parallel approximation to solve the two-dimensional structure equation.  We derive separable thermal structure equations in the high and low temperature limits for both liquid and solid material.  We calculate the thermal structure in terms of simple (although analytically intractable) integrals.  The plane-parallel approximation has the second important advantage that the detailed field configuration separates from the thermal structure problem.  Assuming that it is correct, we can synthesize the results for any field distribution $B(\\theta,\\phi)$ as long as $B$ is not too inhomogeneous on the scale of the envelope thickness (\\ie $|B/\\nabla B| \\gg\\ h_E$).  We find that the emission from a given surface element is a simple function of the location of the element.  Using this functional form, we derive light curves and time-dependent spectra including general relativistic effects.  Although we closely follow the formalism of \\jcite{Page95}, we calculate the two-dimensional thermal structure of the envelope and present results for several field strengths and fluxes. We internally verify and justify the geometric simplification used to translate our results into observables. ",
        "conclusions": "We have presented an analytic technique for calculating the thermal structure of ultramagnetized neutron star envelopes.  We use the exact thermal conductivities in an intense magnetic field of \\jcite{Sila80} and \\jcite{Hern84a} in the non-degenerate and degenerate regimes, respectively.  We make two simplifying approximations.  We assume that the interface between degenerate and non-degenerate material is abrupt.  \\jcite{Hern85} numerically calculated the thermal structure for $B=10^{14}$~G without this assumption.  Our agreement with this earlier result shows that an abrupt interface is a good approximation. Secondly, we use a standard simplification in the study of stellar atmospheres which is to use the radiative zero solution to fix the outer boundary condition (\\cite{Schw58}).  Because the equation for the thermal structure in the outermost layers is qualitatively similar to the relation for an unmagnetized envelope, we conclude that as in the unmagnetized case (\\cite{Hern84b}), this is also an accurate approximation.  A distinct approach is to treat the entire problem numerically, which allows us to estimate directly the possible errors that our simplifying assumptions introduce and alleviates the problem of how to treat the regions where more than one Landau level is filled. Numerical models for parallel and transverse conduction for $B=10^{12}-10^{14}$~G and $\\rho < 10^{10}$~g/cm$^{-3}$ are presented in \\jcite{Heyl98numens}.  We find that the relation between transmitted flux, core temperature and field strength may be approximated by a power law and that the effective temperature is proportional to $\\cos^{1/2} \\psi$ where $\\psi$ is the angle between the radial direction and the local direction of the magnetic field.  Using the geometric result, we calculate the observed spectra as a function of viewing angle including the effects of general relativity for dipole and uniform field configurations.  We extend the conclusions of previous work.  If the surface is assumed to radiate as a blackbody and neutron stars have radii within the currently accepted range, the anisotropic heat transport induced by a dipole field configuration is insufficient to produce the observed pulsed fractions even for ultramagnetized envelopes.  \\jcite{Pavl94} argue that in addition to the transmission of heat through the envelope, the emission at the surface is also strongly anisotropic.  Anisotropic emission can naturally produce large pulsed fractions even when the temperature on the surface is uniform (\\cite{Zavl95,Shib95}).  Additionally, the composition of the atmosphere may have a profound effect on the emergent radiation.  A magnetized iron atmosphere produces substantial limb darkening for $\\psi \\sim \\pi/2$, and the decrement is strongest at high energies (\\cite{Raja97}).  It is straightforward to graft these atmospheres onto the thermal envelopes calculated here to obtained the observed time-dependent spectra for a variety of realistic neutron star models. These effects along with anisotropic conduction may be sufficient to account for the large observed pulsed fractions."
    },
    "9805/astro-ph9805343_arXiv.txt": {
        "abstract": "A new numerical algorithm for calculating isothermal wind accretion flows has been developed and is applied here to the analysis of the hydrodynamics of two-dimensional plane symmetric accretion flows in wind-fed sources. Polar coordinates are used to ensure fine resolution near the object. It is found that a thin accretion column is formed which shows wave-like oscillations. Small accretion disks are formed temporarily around the object. Mass accretion rate and angular momentum accretion rate exhibit quasi-periodic oscillations. The amplitudes of the oscillations depend on the size of the inner boundary, the number of grid points and the method of calculation. For a smal\\-ler size of the inner boundary, finer grids and more accurate numerical schemes, the amplitudes of the oscillation become larger. ",
        "introduction": "Wind-fed accretion by a compact gravitating object is an important astrophysical phenomenon, which arises in many situations including massive X-ray binaries.  Wind accretion has been studied extensively both analytically and numerically. Hoyle \\& Lyttleton (\\cite{Hoy}) were the first to study an axisymmetric wind accretion onto a gravitating object immersed in a uniform flow. Assuming that the fluid particles collide at the axis and lose their tangential momentum, they obtained a mass accretion rate of, \\begin{equation} \\dot{M} = \\pi \\rho _ \\infty V _ \\infty R_{\\rm a} ^ 2 \\end{equation} where the accretion radius, $R_{\\rm a}$, is given by \\begin{equation} R_{\\rm a} = \\frac{2 GM }{V ^2 _\\infty}, \\end{equation} and $\\rho_\\infty$, $V _ \\infty$, $G$, $M$ are the density, velocity of the gas at infinity upstream, the gravitational constant and the mass of the object, respectively. In this estimate the pressure is neglected which means that the Mach number of the gas is infinitely large.  Bondi \\& Hoyle (\\cite{bon44}) have worked on the model in more detail and Bondi (\\cite{bon52}) obtained a formula for the case of spherical accretion. The latter case means that the Mach number at infinity is zero. He also proposed an interpolation formula for flows with finite Mach number.  The numerical computation of axisymmetric wind accretion was first attempted by Hunt (\\cite{Hun71,Hun79}) and general agreement between the accretion rates from analytical estimates and his numerical calculations was found. Shima et al. (\\cite{Shi}) and Koide et al. (\\cite{Koi}) calculated the same problem using a finer mesh, a modern algorithm and a supercomputer, and found accretion rates about a factor two larger than those obtained by Bondi.  If the flow at large distances is not uniform, angular momentum can be transferred to the compact object causing spin up or spin down. Ho et al. (\\cite{ho}) considered this problem based on non-axisymmetric simulations.  Two-dimensional (planar) numerical computations of an inhomogeneous flow were carried out and non-steady 'flip-flop'  motions were found by several authors (Matsuda et al. \\cite{Mat87}, Taam \\& Fryxell \\cite{Taa}, Fryxell \\& Taam \\cite{fry})  Furthermore it was found that non-steady motion occurs even in a homogeneous medium (Matsuda et al. \\cite{Mat91}). Livio et al. (\\cite{Liv}) suggested a possible cause of the instability. \\protect{Boffin} \\& Anzer (\\cite{bof}) calculated two--dimensional wind accretion using a smoothed particle method (SPH) and confirmed the unsteadiness. Benensohn et al. (\\cite{ben}) calculated uniform wind accretion of an adiabatic gas with polytropic index $\\gamma = \\frac{4}{3}$ and got 'flip-flop' nonsteady motions.  Three--dimensional simulations of wind accretion were performed and compared with two-dimensional cases by Sawada et al. (\\cite{Saw}) and Matsuda et al. (\\cite{Mat91, Mat92}). The three--dimen\\-sional calculations performed by Ishii et al. (\\cite{Ish}), Ruffert (\\cite{Rufa,Rufb,Rufc,Rufd}) and Ruffert \\& Arnett (\\cite{Rufe}) have sufficient spatial resolution to lead to non-steady oscillatory flows. Generally these oscillations have lower amplitudes than the corresponding two dimensional flows.  Since the real size of any astrophysical object is usually many orders of magnitude smaller than the accretion radius, it seems impossible to calculate the flow all the way to the surface. Thus the inner boundary of the computations is much larger than the real object. This is an artifact of the simulations.  But one finds that most of these calculations show non-steady behaviours of various degrees. The flows for which $\\gamma$ is close to units tend to exhibit the largest fluctuations.  In this present investigation we want to focus on the special case of isothermal flows. Wind accretion with $\\gamma$ close to unity has been studied by Matsuda et al (\\cite{Mat91}), Boffin \\& Anzer (\\cite{bof}), and Ruffert (1996). Since these flows were more violent than those of higher $\\gamma$ they seem of particular interest. Physically they can be thought of as approximations to configuration of low optical depth and efficient radiative losses resulting in a temperature distribution which is almost uniform.  In principle it would be desirable to include an energy equation which describes all gain and loss mechanisms of the flow. But this approach would complicate our calculations enormously. Therefore we shall use the approximation of isothermality.  The objective of this study is to calculate the wind accretion flows of an isothermal gas at various Mach numbers, $\\cal{M}$, and to study the unsteadiness of the accretion column and the appearance of accretion disks around compact objects using a fine mesh and high resolution scheme.  Here we present the results for two dimensional wind accretion of an isothermal gas by a new high resolution numerical scheme with fine numerical meshes. ",
        "conclusions": "Two dimensional planar wind accretion flows of a supersonic isothermal gas were  analyzed numerically. The computational efficiency was improved as compared to earlier calculations by using an accurate local time stepping and a new type of upwind scheme.  Comparing our new results with earlier calculations one realises that 2D flows with $\\gamma$ close to unity have so far not been studied very systematically. Matsuda et al. (1992) presented one case with $\\gamma = 1.05$ and one with $\\gamma = 1.005$. Their resolution near the central object was much coarser than our resolution. They had also obtained flip-flop configurations, but the instability took much longer to develop. Their amplitudes were smaller and the widths of the accretion columns considerably larger. Boffin and Anzer (1994) also calculated one model with $\\gamma = 1.1$ which has a narrow oscillating accretion column, but again a low amplitude. 3D flows with $\\gamma = 1.01$ have been systematically studied by Ruffert ( 1996). In these flows the accretion regions are always much wider and do not show the systematic oscillations of entire accretion region, although the accretion itself is non-steady. There differences between 3D and 2D flows are quite general and independent of the value of $\\gamma$.  We also have developed a numerical scheme which exactly conserves angular momentum (AMC scheme) and compared its performance to the earlier linear momentum conserving (LMC) scheme. We found that the results obtained with our LMC scheme depend very strongly on the resolution of the grid used, whereas the AMC scheme gives almost no difference for the two types of resolution considered here. Therefore we feel that the AMC calculations are much more reliable and all the results presented in the previous section are based on this scheme. The fact that our AMC and LMC results differ by large amounts suggests to us that many of the earlier calculations which basically conserve linear momentum should be taken with caution. This will be particularly important for the temporal behavior of the accretion of angular momentum $\\dot{J}$ (t).  Our calculations of supersonic flows show that in all cases the accretion of both mass and angular momentum is very erratic. It occurs in very short intervals and at extremely high rates. This behavior can be explained by the formation of Keplerian disks near the inner boundary. If the specific angular momentum of the infalling material is larger than that of the Kepler orbit at the innermost radius then this material cannot accrete.  If such high angular momentum material is flowing in long enough, a disk will form which blocks further accretion very efficiently. Material with very low angular momentum or with opposite rotation falling in during a subsequent phase interacts with the disk and can destroy it. This will lead to a burst of the accreted mass and angular momentum. After such a burst the process can be repeated and a new disk will form. Our calculations indicate that reversals of the disk rotation are quite common.  For the modeling of X-ray binaries fed by wind accretion the fluctuations of $\\dot{J}$ are of major importance. They can be \\- brought into relation with the observed spin-up and spin-down of these X-ray pulsars; see Anzer \\& B\\\"orner (\\cite{anz}). In their investigation they showed that the random fluctuations calculated by Ruffert (\\cite{Rufa}) for 3d models were by a factor $\\approx10$ too low in order to explain the pulse period variations observed in the source Vela X-1. However our new calculations give variations of $\\dot{J}$ which are substantially larger than those found by Ruffert. We have obtained typical values of RMS $(\\dot{J})$ of the order of 0.2 in our dimensionless units (see Table \\ref{case}). This result can also be formulated as: \\begin{eqnarray*} RMS (\\dot{J}) & = & 0.2~\\rho _\\infty ~{v^2_\\infty} ~ R_{\\rm a}^2  \\\\ & = & 0.1~ \\dot{M}_{HL}~{v_\\infty}~ R_{\\rm a}  \\\\ & \\approx & 0.1~{v_\\infty}~R_{\\rm a} \\end{eqnarray*} since $\\dot{M}$ is typically of the order unity. On the other hand Ruffert (\\cite{Rufd}) gives RMS(j)=0.01${v_\\infty} R_{\\rm a}$ for $\\gamma=1.01$ and RMS(j) = 0.03${v_\\infty}R_{\\rm a}$ for $\\gamma = 5/3$. Taking into account that RMS($\\dot {J}$)=RMS ($\\overline{\\dot{M}}$j) and $\\overline{\\dot{M}}$ $\\approx1$ we have RMS ($\\dot{J}$) = (0.01-0.03)${v_\\infty}R_{\\rm a}$. Therefore our values for the fluctuations area factor 3 to 10 larger than those of Ruffert's 3D calculations. Therefore on the basis of our calculations one might conclude that the observed period fluctuations could in principle, be caused by random fluctuations of $\\dot{J}$. But the amplitudes are only marginally large enough and any slight reduction of the efficency would rule out this interpretation. There is in particular the aspect that our calculations are two--dimensional whereas the real flows are three--dimensional and the difference in amplitudes between 2D and 3D flows could be sufficiently large to make the described interpretation invalid. To really answer this question requires full 3D computations, taking angular momentum conservation into account."
    },
    "9805/astro-ph9805033_arXiv.txt": {
        "abstract": "The search for chemically peculiar (CP) stars in open clusters using photoelectric photometry sampling the presence of the characteristic flux depression feature at 5200\\AA\\ via the $\\Delta a$-system (Maitzen 1976) has so far delivered data for objects usually no more distant than 1000 pc from the Sun. If one intends to study the presence of CP stars at larger distances from the Sun, classical photometry has to be replaced by CCD photometry.  For the first time, our investigation presents the results of CCD-photometry in the $\\Delta a$-system for a rich open cluster which is at a distance clearly beyond hitherto studied objects, Melotte 105 (2\\,kpc, log age = 8.5).  Comparison with published $uvby$-photometry yields the calibration of the colour index $g_{1}-y$ of our system, which is necessary for deriving the peculiarity index $\\Delta a$. For this we achieve an average accuracy of 0.007 mag.  Six objects with only marginally peculiar $\\Delta a$-values were found, but spectroscopic and additional photometric evidence is needed to substantiate their peculiarity. ",
        "introduction": "Chemically peculiar (CP) stars of the upper main sequence have been targets for astrophysical studies for exactly 100 years. Starting with the pioneering work of Maury (1897), most of this interval was devoted to the detection of peculiar features in their spectra and photometric behaviour, leading to the post World War II discovery of magnetic fields and the Oblique Rotator concept of slowly rotating stars with non-coincidence of magnetic and rotational axes. Since spectroscopic work required high resolution to accomplish magnetic field measurements, the determination of small Doppler shifts and the identification of the highly crowded narrow lines, investigations of CP-stars tended rather to the study of exotic individual behaviour than to the extraction of features common to this special group of main sequence stars.  Only 30 years ago a statistical study using the Str\\\"omgren system paved the way to a period of work aimed at shaping our concept towards questions on the origin of CP-stars, their evolution on the main sequence and their dependence on other parameters like metallicity, location in the galactic plane etc.  The prerequisite for investigating larger samples of CP-stars (including the generally fainter open cluster members) is unambiguous detection. Looking into catalogues of CP-stars, especially of the magnetic ones, it immediately becomes obvious that there are many discrepancies at classification dispersion. Even the short list of peculiar stars identified by Maury (1897) contains one object which is classified as erroneous entry in the catalogue of Renson (1991).  The reasons for discrepant peculiarity assessments are to be found in the differences of personal pattern recognition among different classifiers, differences of (mainly photographic) observing material (density of spectrograms, widening of spectra, dispersion, focussing), seeing conditions (for objective prism spectra), but also intrinsic variability of peculiar spectral features (e.g. silicon lines).  Photometry has shown a way out of this dilemma, especially through the discovery of characteristic broad band absorption features, the most suitable of them located around 520 nm. Two decades ago Maitzen (1976) introduced a 3 filter photometric system which samples the depth of this green flux depression by comparing the flux in the center (522 nm, $g_{2}$), with the adjacent regions (500, $g_{1}$, and 550 nm, $y$) using a band-width of 13 nm. The respective index was introduced as: \\begin{center} $a=g_{2}-(g_{1}+y)/2$ \\end{center} Since this quantity is slightly dependent on temperature (increasing towards lower temperatures), the intrinsic peculiarity index had to be defined as \\begin{center} $\\Delta a= a-a_o[b-y]$ \\end{center} i.e. the difference between the individual a-value and the a-value of non-peculiar stars of the same colour (the locus of $a_o$-values has been called {\\it normality line}).  Maitzen (1976) showed that virtually all peculiar stars with magnetic fields, the CP2-group according to the definition of Preston (1974), have positive $\\Delta a$ values larger than 0.012 mag up to 0.075 mag.  This result was corroborated by Maitzen \\& Vogt (1983) for a significantly larger sample. In a series of papers (first:  Maitzen \\& Hensberge 1981; last: Maitzen 1993) 38 open clusters have been surveyed for the presence of magnetic peculiar stars using $\\Delta$a-photometry with the conventional photomultiplier technique. The advantage of this approach is both a relatively high accuracy of the detection index $\\Delta$a (external scatter for normal stars 3-5 mmags) and its immediate availability, hence detection of CP2-stars right at the telescope (provided that the data acquisition system is able to display linear combinations of on-line magnitudes).  On the other hand the magnitude limit is about 12 for work with a 1m telescope. This limit corresponds to a vicinity of only about 1 kpc around the Sun. Even in this area we can reach in some cases only the hotter domain of CP2-stars.  If one considers to study the behaviour and appearance of peculiar stars over a significant range of galactocentric distances - speculating about the possible influence of different degrees of metallicity and/or galactic magnetic field strengths - then it is imperative to increase the actual magnitude limit by at least 2-3 magnitudes. Discarding the use of 4m telescopes as highly unrealistic for this purpose we are left with employing the CCD-technique. Maitzen et al. (1997) have prepared the ground by showing that for field stars in the range 8$<$V$<$10, $\\Delta a$ values obtained with CCD-photometry at a 60cm telescope exhibit essentially the same error level as conventional photoelectric photometry.  Aside from the basic advantage of CCD-photometry - i.e. simultaneous recording of many individual objects not only saving time but also increasing precision - the second benefit becomes more and more obvious when we observe clusters at larger distances with decreasing angular separations of their members, and this is the determination of stellar magnitudes by fitting of point spread functions.  The primary goal of the present study is to evaluate the accuracy performance of CCD-photometry, i.e. to which limiting magnitude the peculiarity index $a$ viz. $\\Delta a$ can be determined in an open cluster with a precision comparable to classical photoelectric photometry within reasonable amounts of observing time.  Here we present the results for our first open cluster, Melotte 105, measured in the $\\Delta a$ system with the Bochum 61cm telescope during a campaign in April/May 1995 at ESO La Silla, during which  3 dozens of other open clusters were observed.  Mel 105 is at a distance of 2000 pc in the 4th galactic quadrant and has a reddening of $E(B-V)=0.38$. Therefore, it is at a larger distance than all other open clusters surveyed so far by photoelectric photometry in $\\Delta a$ and a good starter for studying remote clusters in this system. ",
        "conclusions": ""
    },
    "9805/astro-ph9805205_arXiv.txt": {
        "abstract": "We performed a spectroscopic study of the SB2 Mercury-Manganese star $\\alpha$ And. Our measurements of the secondary's radial velocities result in improved orbital elements. The secondary shows abundances typical of the metallic-line stars: a Ca deficiency, small overabundances of the iron-peak elements, and 1.0 dex overabundances of Sr and Ba. ",
        "introduction": "\\vspace{-1mm} \\and\\, (HD\\, 358, HR\\, 15) is a bright, well known binary system with a HgMn primary. Pan et al. (1992) obtained its visual orbit from observations with the Mark III Stellar Interferometer. Later Tomkin et al. (1995) detected the weak secondary spectrum, measured the secondary's radial velocities near the nodes, and calculated the spectroscopic orbital elements of this SB2 system. Their derived mass ratio 0.42 $\\pm$ 0.02 differs from 0.48 of Pan et al. and gives too large a mass for the primary. No abundances of the secondary are published.  \\vspace{-3mm} ",
        "conclusions": ""
    },
    "9805/astro-ph9805211_arXiv.txt": {
        "abstract": "We present X-ray data of {\\rxj} obtained by {\\asc}. Although it shows a clear orbital intensity modulation with an amplitude of nearly 100\\% below 0.5~keV in {\\ro} data, the {\\asc} light curves are nearly flat except for a possible dip lasting about one-tenth of the orbital period. We discuss this within the model assumption of a stream-eclipsing geometry as derived from the ROSAT observations.  The {\\asc} X-ray spectrum can be represented by a two temperature optically thin thermal plasma emission model with temperatures of $\\sim$1\\,keV and $>$7\\,keV, suggesting postshock cooling as observed in {\\ex}. A remarkable feature of the spectrum is the strong iron {\\ka} emission line whose equivalent width is $\\sim$4\\,keV. To account for this, an iron abundance of greater than at least 1.3 times Solar is required. A combined spectral analysis of the {\\ro} PSPC and {\\asc} data indicates that the $N_{\\rm H}$-corrected flux ratio of the soft blackbody (0.1--2.4~keV) to the hard optically thin thermal plasma emission (2--10~keV) is as large as $\\sim 10^4$. ",
        "introduction": "A polar (or {\\am} type object; Cropper 1990) is an accreting binary composed of a mass-donating low mass secondary star and a magnetized white dwarf with a field strength of the order of 10-100\\,MG. Matter from the secondary accretes along the field lines onto a small region of the white dwarf close to the magnetic pole. Since the flow is highly supersonic, a standing shock is formed close to the white dwarf, and a hot plasma with a temperature of $10^8$~K is formed. From the postshock plasma, optical cyclotron emission and optically thin thermal plasma emission in X-rays have been observed. In addition to this, blackbody emission with a temperature of 10--40~eV is observed. Although the blackbody component is considered to be radiated from the surface of the white dwarf around the postshock plasma via reprocessing of the cyclotron and the hard X-ray radiation, its intensity is usually much larger than that of the cyclotron and hard X-ray radiation. This has become known as the so-called `soft excess problem' (Rothschild \\etal\\ 1981). Beuermann and Burwitz (1995) recently suggested that the amount of the soft excess is correlated with the strength of the magnetic field of the white dwarf.  It is known that the hard X-ray continuum spectrum of magnetic cataclysmic variables (mCVs) can, to a first approximation, be fitted by an optically thin thermal plasma emission spectrum with a single temperature undergoing photoelectric absorption represented by a single hydrogen column density. To represent the {\\exo} spectra of intermediate polars, Norton and Watson (1989) introduced the so-called `partial-covering absorber model' in which photoelectric absorption was represented by two column densities. Complex absorption was found also in polars with {\\gin} observations (Ishida and Fujimoto 1995).  It has also been expected that the hard X-ray emitting hot plasma is gradually cooled by cyclotron emission and bremsstrahlung (Aizu 1973, Frank, King and Lasota 1983, Imamura and Durisen 1983), and that the hard X-ray spectrum consists of multi-tem\\-pe\\-ra\\-ture emission components. Such a multi-temperature emission spectrum was first suggested by an {\\ein} observation of {\\ex} (Singh \\& Swank 1993), and later was established by an {\\asc} observation (Ishida, Mukai and Osborne 1994, Fujimoto and Ishida 1997). Note, however, that it is only for {\\ex} that the existence of multi-temperature emission is confirmed observationally.  {\\rxj} was discovered during the search for supersoft X-ray sources in the {\\ro} all-sky survey data (Greiner, Remillard and Motch 1995). Greiner, Remillard and Motch (1995, 1998) have analyzed all the {\\ro} data taken between 1990 September and 1993 Sep\\-tem\\-ber, and found a coherent period of 0.07847977(11)~d ( = 1.8835145$\\pm$0.000003~hr). The pulse profile in the band $<0.5$~keV is characterized by a deep intensity minimum, with basically no X-ray flux, lasting 0.1 orbital phase. The X-ray spectrum is characterized by  strong blackbody emission with a temperature of $20\\pm 15$~eV, with a clear excess emission above 1~keV, which has been approximated by thermal bremsstrahlung with a temperature of 20~keV. The absorption-corrected 0.1--2.4~keV fluxes of the blackbody and the thermal brems\\-strahlung components are $7 \\times 10^{-11}$~erg~cm$^{-2}$s$^{-1}$ and $8 \\times 10^{-13}$~erg~cm$^{-2}$s$^{-1}$, respectively, suggesting a huge soft excess of nearly 90 in the 0.1--2.4 keV band. Szkody {\\etal} (1995) carried out photometry, spectroscopy and polarimetry in the optical band, and found several characteristics of polars such as He{\\small II} emission line stronger than H$\\beta$ and circular polarization of 4\\%. All these properties strongly indicate that {\\rxj} is a polar.  In this paper, {\\asc} data of {\\rxj} taken in 1996 Sep-Oct are presented. In \\S~2 we describe how the {\\asc} observation was carried out. In \\S~3 light curve and spectral analysis are presented. We discuss these properties in \\S~4 in combination with the {\\ro} and {\\iue} data. In \\S~5 we summarize our results. ",
        "conclusions": "We presented X-ray data of {\\rxj} obtained by {\\asc}. From the {\\asc} light curves we find only marginal evidence for orbital intensity modulation which is seen in the {\\ro} light curve below 0.5~keV characterized by the sharp and deep minima. From this energy dependence, we conclude that the intensity modulation is caused mostly by photoelectric absorption in the pre-shock accretion column, and the accreting pole moves around on the hemisphere visible from the observer, consistent with the conclusions from Greiner, Remillard \\& Motch (1998). It is possible that the line of sight absorber is partly ionized or has a distribution in $N_H$ in the range \\lax 10$^{21}$~cm$^{-2}$.  The X-ray spectrum can be represented by a two temperature optically thin thermal plasma emission model with temperatures of $\\sim$~1~keV and $>$~7~keV. In analogy with {\\ex}, we deduce that {\\asc} observed the cooling of the postshock plasma, as indicated by the theory of the postshock accretion flow. A remarkable feature of the X-ray spectrum of \\rxj\\ is the strong iron {\\ka} emission line whose equivalent width is $\\sim 4$~keV. To account for this, an iron abundance greater than Solar by at least 1.3 times is required. From the combined analysis of the {\\ro} PSPC and {\\iue} spectra, the ratio of the bolometric luminosity of the soft component to the hard is revealed to be greater than 140."
    },
    "9805/astro-ph9805357_arXiv.txt": {
        "abstract": "{An expansion of a density field or particle distribution in basis functions which solve the Poisson equation both provides an easily parallelized n-body force algorithm and simplifies perturbation theories.  The expansion converges quickly and provides the highest computational advantage if the lowest-order potential-density pair in the basis looks like the unperturbed galaxy or stellar system.  Unfortunately, there are only a handful of such basis in the literature which limits this advantage.  This paper presents an algorithm for deriving these bases to match a wide variety of galaxy models.  The method is based on efficient numerical solution of the Sturm-Liouville equation and can be used for any geometry with a separable Laplacian.  Two cases are described in detail.  First for the spherical case, the lowest order basis function pair may be chosen to be exactly that of the underlying model.  The profile may be cuspy or have a core and truncated or of infinite extent.  Secondly, the method yields a three-dimensional cylindrical basis appropriate for studying galaxian disks.  In this case, the vertical and radial bases are coupled; the lowest order radial part of the basis function can be chosen to match the underlying profile only in the disk plane.  Practically, this basis is still a very good match to the overall disk profile and converges in a small number of terms. } ",
        "introduction": "\\label{sec:intro}  The basis function n-body force solver is optimal for studying the global response of galaxies to perturbations or stability (Earn \\& Sellwood 1995\\nocite{EaSe:95}).  This technique was developed for astrophysical problems by Clutton-Brock (1972, 1973\\nocite{Clut:72,Clut:73}), Kalnajs (1976\\nocite{Kaln:76a}), Fridman \\& Polyachenko (1984\\nocite{FrPo:84}) and more recently by Hernquist \\& Ostriker (1992\\nocite{HeOs:92}) who dubbed it the {\\it self-consistent field} (SCF) method.  Orthogonal function expansions are attractive Poisson equation solvers for two reasons: 1) the expansions can be chosen to filter the structure over an interesting range of scales and simultaneously suppress small-scale noise; and 2) the algorithm is computationally efficient, scaling linearly with the number of particles.  Mathematically, this entire class of algorithms relies on the general properties of the Sturm-Liouville equation (SLE) of which the Poisson equation is a particular case.  This same approach is common in perturbation theories and so facilitates direct comparison between n-body simulation and linear perturbation theory. In addition, this approach is straightforward to parallelize (e.g. Hernquist, Sigurdsson \\& Bryan 1995\\nocite{HeSB:95}); we find the algorithm scales linearly with the number of processors with low overhead.  If the basis set resembles the equilibrium galaxy, most of the computational work is concentrated on resolving the perturbation rather than the equilibrium.  This last point is also a disadvantage of this technique in applications to date.  If the equilibrium does not look like the basis set, the technique becomes less efficient and noisy because the expansion series must be sufficiently long to represent the equilibrium even without the perturbation.  This paper describes a general method based on a numerical construction of orthogonal bases which remedies this situation.  Solutions to the fundamental equation, the Sturm-Liouville equation, are well-understood and well-behaved.  A number of recently published algorithms take advantage of the special properties of this differential equation to yield high-accuracy solutions with low computational work.  Harnessing these developments to our needs leads to an algorithm for computing orthogonal bases whose lowest-order function matches any given {\\it any} regular equilibrium; spherical and three-dimensional cylindrical solutions are described in detail here.  The basic algorithm will be described in \\S\\ref{sec:SL}.  For the spherical case, the proposed algorithm is competitive in performance with evaluation by recursion relation used for the published bases cited above and has reproduced them with high accuracy as a check.  The cylindrical basis is a bit more cumbersome: one may rely on the same numerical solution to tailor the basis in the radial or vertical direction but not both simultaneously.  Here, I choose to derive the radial basis numerically.  The lowest-order radial basis functions then take the form $f(r)\\exp(\\pm ikz)$.\\footnote{Bases resulting the other choice has been explored by Earn (1996\\nocite{Earn:96}) using a different approach.}  These may then be adapted to the background by principal component analysis. Although more cumbersome to implement and more time consuming to execute than the spherical case, it is still fast relative to non-expansion-based solvers.  The details of the cylindrical basis are given in \\S\\ref{sec:PCA}. ",
        "conclusions": "This paper presents a numerical algorithm for constructing biorthogonal expansion bases for use in n-body force calculation and linear perturbation theory and explores its performance.  The major results of this investigation are as follows: \\begin{enumerate} \\item This algorithm removes one of the remaining limitations of the self-consistent field (SCF) method by providing basis sets tailored to any background galaxian profile. \\item The algorithm is applicable to any separable coordinate system. This paper details and benchmarks its implementation for spherical and three-dimensional cylindrical bases. \\item Sturm-Liouville equation solvers are publically available (e.g. see Pruess \\& Fulton 1993 for Fortran code) and a desired basis is readily obtained using equations (\\ref{eq:sle1})--(\\ref{eq:sle3}). \\item The main limitation of this method for n-body codes is the necessity to tabulate the basis functions rather than derive them from recursion relation on the fly (as in Clutton-Brock 1973 and Hernquist \\& Ostriker 1992).  On the other hand, this is largely a programming inconvenience; the algorithm is still easily parallelized and the table lookup is not a computational bottleneck. \\item For spherical expansions, the algorithm is conceptually equivalent to and computationally competitive with the published SCF expansions.  For three-dimensional cylindrical expansions, the coupling of the vertical and radial parts of the potential-density pairs requires an additional orthogonalization step.  This increases the computational overhead by up to 50\\% but does not effect scaling with particle number or parallelizability. \\item The use of these basis sets is not limited to n-body simulation. They are easily used in semi-analytic linear perturbation calculations and, moreover, facilitate the comparison between the n-body and perturbation theory. \\end{enumerate}"
    },
    "9805/astro-ph9805094_arXiv.txt": {
        "abstract": "We suggest that the bulge-to-disc ratios of spiral galaxies are primarily determined by the angular momenta of their host haloes predicted in current hierarchical clustering models for structure formation. Gas with low specific angular momentum becomes self-gravitating and presumably forms stars before it can settle into a rotationally supported disc. We assume this part of the gas in a dark halo to form a bulge, while the rest is assumed to settle into a rotationally supported disc. With these assumptions the predicted bulge-to-disc ratios in mass and in size, and other correlations between the bulge and disc components can match current observational results. This model predicts the existence of a population of low-surface-brightness galaxies which are bulgeless. The model also predicts that the bulge component has many properties in common with the disc component, because both form through similar processes. In particular, many bulges should be supported (at least partially) by rotation. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805057_arXiv.txt": {
        "abstract": "We describe a new method for the computation of opacity distribution functions (ODFs) useful to calculate one-dimensional model atmospheres in local thermal equilibrium (LTE). The new method is fast enough to be applied on current workstations and allows the computation of model atmospheres which deviate significantly from (scaled) solar chemical composition. It has reproduced existing ODFs and model atmospheres for solar abundances. Depending on the type of chemical peculiarity the ``individual'' model atmosphere may have a structure and surface fluxes similar to atmospheres based on (scaled) solar abundances or deviate in a way that cannot be reproduced by any of the conventional models. Examples are given to illustrate this behavior. The availability of models with ``individualized'' abundances is crucial for abundance analyses and Doppler imaging of extreme CP stars. ",
        "introduction": "Opacity distribution functions are used to describe line opacities $l_{\\nu}$ in LTE for a set of pairs of temperature T and gas pressure P assuming a fixed chemical composition and (in our case) a fixed value for the microturbulence $\\xi_t$: \\begin{equation}   \\label{kupka_eq1} l_{\\nu} = l_{\\nu}(T,P,{\\rm chem.~composition},\\xi_t). \\end{equation} The frequency dependence of the $l_{\\nu}$ is described by dividing the whole wavelength range relevant to the desired stellar atmosphere computations into 300 to 1200 channels. Each of these channels is devided into 10-12 subchannels which provide a statistical representation of line opacity (ranging from high to low values). The (radiative) flux integrated over a set of subchannels approximates the overall (radiative) flux for each particular channel. Thus, one can compute the total radiative flux throughout the model atmosphere, as well as surface fluxes and intensities. This technique was described in Strom and Kurucz (1966) and later used, for instance, by Gustafsson et al. (1975) and by Kurucz (1979). ",
        "conclusions": "We have presented a new method for the computation of model atmospheres for CP stars based on the ODF approach. The method was successfully compared with standard models. Application to various CP stars shows that model atmospheres for some abundance patterns ($\\lambda$ Boo stars or mildly peculiar roAp stars) can be computed using scaled solar abundances. For other patterns, systematic deviations occur which cannot be approximated by choosing a model atmosphere based on properly scaled solar abundances, making the computation of individual ODFs (or model atmospheres based on the OS technique) mandatory for applications such as abundance analyses and Doppler imaging. One prominent example are Si peculiar stars, essentially because Si is genuinely abundant and many of its absorption lines cluster in certain wavelength regions. More details on this work will be given in Piskunov \\& Kupka (1998)."
    },
    "9805/astro-ph9805261_arXiv.txt": {
        "abstract": "The results of numerical simulation of mass transfer in semidetached non-magnetic binaries are presented. We investigate the morphology of gaseous flows on the base of three-dimensional hydrodynamic calculations in interacting binaries of different types (cataclysmic variables and low-mass X-ray binaries).  We find that taking into account of a circumbinary envelope leads to significant changes in the stream-disc morphology. In particular, the obtained steady-state self-consistent solutions show an absence of impact between gas stream from the inner Lagrangian point $L_1$ and forming accretion disc. The stream deviates under the action of gas of circumbinary envelope, and does not cause the shock perturbation of the disc boundary (traditional `hotspot'). At the same time, the gas of circumbinary envelope interacts with the stream and causes the formation of an extended shock wave, located on the stream edge. We discuss the implication of this model without `hotspot' (but with a shock wave located outside the disc) for interpretation of observations. The comparison of synthetic light curves with observations proves the validity of the discussed hydrodynamic model without `hotspot'.  We also consider the influence of a circumbinary envelope on the mass transfer rate in semidetached binaries. The obtained features of flow structure in the vicinity of $L_1$ show that the gas of circumbinary envelope plays an important role in the flow dynamics, and that it leads to significant (in order of magnitude) increasing of the mass transfer rate. The most important contribution to this increase is due to stripping of mass-losing star atmosphere by interstellar gas flows.  The parameters of the formed accretion disc are also given in the paper.  We discuss the details of the obtained gaseous flows structure for different boundary conditions on the surface of mass-losing star, and show that the main features of this structure in semidetached binaries are the same for different cases.  The comparison of gaseous flows structure obtained in 2D and 3D approaches is presented. We discuss the common features of the flow structures and the possible reasons of revealed differences. ",
        "introduction": "The semidetached binaries belong to the class of interacting stars, where one component fills its critical surface that causes the mass transfer between components of the system.  In general, the form of critical surface may be complex (Kruszewski 1963) and special mathematical models are required to describe the process of mass transfer in such a system (see review of these models in Lubow 1993). However, in the standard treatment, semidetached binaries are considered under the assumption that orbits of components are circular and their rotation is synchronous with the orbital movement. In this case the critical surface can be identified with internal surface (Roche surface) in the restricted three-body problem and it is assumed, that the mass transfer between components of the system occurs through the vicinity of inner Lagrangian point $L_1$, where pressure gradient is not balanced by gravitational force.  The hydrodynamics of mass transfer through the inner Lagrangian point $L_1$ has been investigated by many authors. The detailed analysis of matter flow in the vicinity of $L_1$ was carried out by Lubow \\& Shu (1975). Using a perturbation method they evaluated main characteristics of the flow. In another approach, based on the analysis of Bernoulli integral, the stream parameters were specified as well and the dependence of the mass transfer rate upon the degree of Roche lobe overfilling was obtained (Paczy\\'nski \\& Sienkiewicz 1972; Savonije 1978).  For adequate description of the mass transfer process in the binary system besides determination of stream parameters it is also necessary to consider the further behavior of flowlines during movement of matter from $L_1$. It is the process of mass transfer produces the general flow structure and, accordingly, determines basic observation evidences, therefore the main attention was paid to study of this question. For the first time movement of particles leaving $L_1$ and moving in the gravitational field of binary system was considered by Warner \\& Peters (1972), Lubow \\& Shu (1975) and Flannery (1975). These results were obtained using a simplified ballistic approach for analysis of the gas movement without taking into account of hydrodynamic effects. To study the influence of the circumbinary envelope on gas movement and, accordingly, for a correct description of the flow, the solving of full system of hydrodynamic equations is required. This is possible only in the framework of rather complex mathematical models.  The use of numerical methods for investigation of hydrodynamics of mass transfer in semidetached binaries was limited by the computer power for a long time so 2D models were used for the analysis of the flow structure. Despite the restrictions of 2D approach, it allowed to consider some details of the flow structure correctly and to obtain a set of interesting results (see, e.g., Sawada, Matsuda \\& Hachisu 1986; Sawada et al. 1987; Taam, Fu \\& Fryxell 1991; Blondin, Richards \\& Malinowski 1995; Murray 1996). Last years the possibility of numerical hydrodynamic simulation of mass transfer in the framework of more realistic 3D models (Nagasawa, Matsuda \\& Kuwahara 1991; Hirose, Osaki \\& Minishige 1991; Molteni, Belvedere \\& Lanzafame 1991; Sawada \\& Matsuda 1992; Lanzafame, Belvedere \\& Molteni 1992, 1994; Belvedere, Lanzafame \\& Molteni 1993; Meglicki, Wickramasinghe \\& Bicknell 1993; Armitage \\& Livio 1996) appeared. In particular, these authors considered the formation of accreting disc in semidetached binaries (Nagasawa et al.  1991; Sawada \\& Matsuda 1992) and investigated the interaction of stream of matter leaving $L_1$ with the disc (Hirose et al.  1991; Armitage \\& Livio 1996).  Unfortunately, many 3D investigations were carried out during a rather small time-scales and this fact did not allow to consider the real flow morphology, accordingly, to evaluate the influence of forming circumbinary envelope on the flow structure. Some progress in the investigation of general flow structure in semidetached binaries was achieved in works Molteni et al. (1991), Lanzafame et al.  (1992, 1994), and Belvedere et al. (1993), where 3D numerical simulations were carried out on the sufficiently large time intervals. A set of interesting results was obtained in these works, however using of method SPH (Smoothed Particle Hydrodynamics) did not allow to consider the influence of the circumbinary envelope on the flow structure as far as the computational restrictions of SPH method did not permit to investigate flows with large density gradients, and, accordingly, the account of the influence of circumbinary envelope on the mass transfer was not quite correct. For the first time the morphology of gaseous flows in binaries was accurately considered by authors in Bisikalo et al.  (1997a,b).  In this work we present the results of 3D numerical study of the flow structure in semidetached non-magnetic binaries. TVD (Total Variation Diminishing) method of solving of hydrodynamic equations used in this paper has allowed to investigate the morphology of gaseous flows in the system and to consider the influence of forming circumbinary envelope, despite the presence of significant density gradients.  The numerical simulations of mass transfer in semidetached binaries have been conducted on large time intervals that allowed to consider the main features of flow structure in steady-state regime.  The earlier conclusions on the flow structure for low-mass X-ray binary (Bisikalo et al.  1997a,b) are generalized in present work for the wider class of objects.  The paper is structured as follows. In Section 2, the properties of used physical, mathematical and numerical models are described.  Section 3 contains the results of numerical simulations. In this Section the stream-disc interaction, comparison of synthetic light curves with observations, flow structure in the vicinity of $L_1$, influence of accepted boundary conditions, and comparison of the results obtained in 2D and 3D models are discussed. Our conclusions follow in Section 4. ",
        "conclusions": "This work deals with the results of 3D numerical simulation of the gaseous flows structure in semidetached binaries of various types. The analysis of results shows the significant influence of rarefied gas of circumbinary envelope on the flow patterns in these systems. The gas of circumbinary envelope interacts with the stream of matter and deflects it. This leads, in particular, to the shock-free (tangential) interaction between the stream and the outer edge of forming accretion disc, and, as the consequence, to the absence of `hotspot' in the disc.  At the same time it is shown, that the interaction of the gas of circumbinary envelope with the stream results in the formation of an extended shock wave located along the stream edge. The comparison of synthetic light curves with observations proves the validity of discussed hydrodynamic model without `hotspot'.  \\begin{figure} \\hbox{\\psfig{figure=fig8.eps,width=3.3in}} \\caption{Normalized density distributions along $Y$ axis passing through the accretor for 2D (solid line) and 3D (bold line) cases.  Accretor is placed at the point $Y=0$, and vertical dashed lines around this point correspond to the accretor size.} \\end{figure}  \\begin{figure} \\centerline{\\hbox{\\psfig{figure=fig9.eps,width=3.3in}}} \\caption{Schematic presentation of the main features of flow structure in semidetached binaries. Roche equipotentials are shown by dashed lines. Lagrangian points $L_1$ and $L_2$, location of the accretor, and quasi-elliptic accretion disc are marked. Shock wave located on the stream boundary is presented by bold line. Flows labelled by markers `$a$', `$b$', `$c$' and `$d$' are also shown. Flow `$d$' corresponds to matter leaving the system through the vicinity of outer Lagrangian point $L_2$. Flow `$c$' belongs to circumbinary envelope and interacts with the gas stream resulting in formation of the shock wave. Flows `$a$' and `$b$' belong to circumbinary envelope as well and strip off the atmosphere of mass-losing star increasing the mass transfer rate.} \\end{figure}  It should be mentioned that taking into account of circumbinary envelope also leads to a drastic change of the mass transfer parameters in the system. The calculated mass transfer rate increases in order of magnitude as compared with values got from the standard model. Moreover, the gas of circumbinary envelope changes the flow structure near the surface of mass-losing component that eventually influences the common structure of gas flows in the system and consequently affects the interpretation of observational data.  The qualitative similarity of the obtained solutions for the various types of semidetached systems permits to speak about some universal character of the considered hydrodynamic model for non-magnetic semidetached binaries. The main features of the obtained flow structure are summarized on Fig.~9, where gaseous flows in systems, location of the shock wave, as well as forming quasi-elliptical accretion disc are presented schematically.  It should be noted that the presented results are obtained for the steady-state case. For the non-stationary (transient) case when morphology of the flow is determined by external factors and is not self-consistent, more features of the flow may appear, in particular, arising of the zone of the disc-stream shock interaction  is possible. For example, if the disc was formed before the filling by the mass-losing star its Roche lobe than, after the beginning of mass exchange, the occurrence of `hotspot' would be possible.  Nevertheless, when the flow structure will become steady-state we will get the solution without `hotspot'. Therefore the lifetime of such formation is of importance. As the characteristic lifetime of `hotspot' it is naturally to take the interval, during which the quantity of matter, transferred to the system by the stream, will become comparable with the accretion disc mass, as far as after the total replacement of disc matter the solution will be self-consistent. For the mass transfer rate and accretion disc parameters typical for semidetached binaries it should be expected that on time intervals of the order of tens of the orbital period the solution will become steady-state. It means, that for the most part of time the gaseous flows structure is described by the model without `hotspot' as presented above.  Summarizing, we may conclude, that the correct consideration of circumbinary envelope in the numerical model of the mass transfer in semidetached binaries makes possible to find the new features of the flow structure and thus changes our view on the morphology of gaseous flows. We should note, that despite the similarity of observation evidences of the hypothetical `hotspot' and the shock wave obtained in calculations, the calculated morphology of matter flow in semidetached binaries differs from the standard model drastically, that, in turn, requires the revision of some standard concepts."
    },
    "9805/astro-ph9805111_arXiv.txt": {
        "abstract": "We present deep ultraviolet observations of a field containing the cluster Abell 2246 (z=0.225) which provide far-ultraviolet (FUV) images of some of the faintest galaxies yet observed in that bandpass.  Abell 2246 lies within the field of view of Ultraviolet Imaging Telescope (UIT) observations of the quasar HS1700+64, which accumulated over 7100 seconds of UIT FUV exposure time during the Astro-2 mission in March 1995.  For objects found on both the FUV and ground-based V-band images, we obtain FUV ($\\lambda \\sim$ 1520\\AA) photometry and V-band photometry, as well as mid-UV ($\\lambda \\sim$ 2490\\AA) photometry from UIT Astro-1 observations and ground-based I-band photometry. We find five objects in the images which are probably galaxies at the distance of Abell 2246, with FUV magnitudes (m(FUV)) between 18.6 and 19.6, and V magnitudes between 18.4 and 19.6. We find that their absolute FUV fluxes and colors imply strongly that they are luminous galaxies with significant current star formation, as well as some relatively recent, but not current, ($>$400 Myr ago) star formation. We interpret the colors of these five objects by comparing them with local objects, redshift-corrected template spectra and stellar population models, finding that they are plausibly matched by 10-Gyr-old population models with decaying star formation, with decay time constants in the range 3 Gyr $<\\tau<$ 5 Gyr, with an additional color component from a single burst of moderate ($\\sim$400-500 Myr) age. From derived FUV luminosities we compute current star formation rates.  We compare the UV properties of Abell 2246 with those of the Coma cluster, finding that Abell 2246 has significantly more recent star formation, consistent with the Butcher-Oemler phenomenon. ",
        "introduction": "UV images of local galaxies provide unique direct evidence about the morphology and quantitative nature of recent star formation, while UV images of objects at moderate redshifts (z up to $\\sim$0.3) extend the limits of direct knowledge of massive star formation back to more remote times.  The darkness of the UV sky (\\cite{oconnell}) further enhances observational prospects in this waveband. However, deep UV images of such objects are rare because of the dearth of instruments with sufficient resolution and sensitivity.  In this paper we discuss deep observations including photometry in the far UV ($\\sim$1520\\AA; FUV) and mid UV ($\\sim$2490\\AA; MUV) by the Ultraviolet Imaging Telescope (UIT) of a field which contains the cluster Abell 2246 (\\cite{abell}).  Abell 2246 is only lightly obscured by Galactic foreground reddening (E(B--V)=0.03) but has nevertheless attracted few published observations.  Cluster center, angular size, richness, and luminosity function data for Abell 2246 used here are from \\cite{hoessel1}, \\cite{hoessel2} and \\cite{struble}. With a distance of 900 Mpc and a look-back time of 1.2 Gyr (based on z=0.225, for q$_{0}$ = 0.1 and H$_{0}$ = 75 km~s$^{-1}$~Mpc$^{-1}$, adopted for this paper) galaxies in Abell 2246 are among the most distant and faintest yet observed in the far ultraviolet. They provide an important stepping stone to understanding observations of more distant objects in longer-wavelength bands. ",
        "conclusions": "Abell 2246, centered at RA=17$^{h}$ 00$^{m}$ 44$^{s}$, DEC=64$^{\\circ}$ 12.7 \\arcmin\\ (J2000.0)  with a diameter of 18\\arcmin, is completely contained within the extracted UIT sub-images centered on HS1700+64 as well as within the optical-band images. The objects 3,5,6,7, and 8 in Table~\\ref{datatab} are likely member galaxies of Abell 2246 from several lines of argument. First, they appear non-stellar on the V- and I-band images, with sizes corresponding to linear extents of a few tens of kpc at z=0.225. Second, their foreground- and redshift-corrected (FUV--V) colors, in the range -0.36 to +0.49, are similar to those of well-known local, bright star-forming galaxies (see Figure~\\ref{cmdiag} and Table~\\ref{abonly}, where derived FUV absolute magnitudes, UV colors, and star formation rates for these five galaxies are presented.). Third, their UV luminosities, if they are at the Abell 2246 distance, are approximately those of nearby UV-luminous galaxies.       \\begin{deluxetable}{lrrrr}  \\tablecolumns{5}  \\tablecaption{Intrinsic FUV Properties of Abell 2246 Galaxies \\tablenotemark{a} \\label{abonly}}  \\tablehead{ \\colhead{Source} & \\colhead{M(FUV)} & \\colhead{FUV--V} & \\colhead{MUV--V} & \\colhead{SFR} \\\\ \\colhead{\\#} & \\colhead{} & \\colhead{} & \\colhead{} & \\colhead{\\msol yr$^{-1}$} \\\\ }  \\startdata  3 & -21.20 & 0.28 & 0.49  & 2.9 \\nl 5 & -21.44 & -0.16 & 0.36 & 3.7 \\nl 6 & -20.43 & 0.49 &  0.75 & 1.4 \\nl 7 & -21.06 & 0.16 & 0.64  & 2.6 \\nl 8 & -20.62 & -0.36 & 0.13 & 1.7 \\nl  \\enddata  \\tablenotetext{a}{After correction for Galactic foreground (E(B--V)=0.03) and D.M.=39.8}  \\end{deluxetable}   Redshifts are necessary to confirm certainly that these five objects are not foreground galaxies.  However two points very much strengthen the case for this. From the UIT data compiled by \\cite{fanelli}, the vast majority of field candidates, with ``normal'' FUV fluxes (M(FUV)$>$-18.5) have significantly lower FUV surface brighnesses. Thus they cannot be placed in front of Abell 2246 so that both their diameters and magnitudes scale to the observed values. Nearly all elliptical galaxies (\\eg all in Virgo) as well as a large fraction of spirals (\\eg M81 and NGC 2903) are eliminated; high-luminosity objects which are allowed by this criterion must have DM$>$37.5, within a factor of 3 of the distance of Abell 2246.  The strongest argument against foreground interlopers, however, is the rarity of field galaxies, based on the very low frequency of {\\em any} extragalactic FUV sources seen in this (1520\\AA) bandpass in the complete UIT Astro-2 dataset. This bandpass excludes light longward of 1800\\AA\\ and is therefore not sensitive to stellar populations characterized by stars cooler than T$_{eff}$=9000K, as discussed below. Astro-2/UIT imaged 42 40-arcmin diameter extragalactic fields (that is, with $|b|>\\sim$30$^{\\circ}$ and excluding large foreground objects such as Galactic clusters and Local Group galaxies) with exposure times longer than 1000 seconds, with a conservative limiting magnitude of m(FUV)$\\sim$18.5.  This image set was searched at the positions of all known extragalactic objects from the catalogs of \\cite{devaucouleurs}, \\cite{paturel}, \\cite{huchra}, \\cite{nilson}, \\cite{veroncetty}, and the IRAS Point Source Catalog. After excluding each field's principal target and 9 known Coma Cluster members, a total of 16 objects were found and identified. Therefore, the rate of detection of serendipitous but catalogued FUV sources is about 0.4 objects per 40 arcmin UIT field, or 0.1 objects per (20-arcmin) Abell 2246 field, to m(FUV)=18.5.  We use the mid-UV ($\\sim$2000\\AA) luminosity function of \\cite{milliard} for background galaxies to scale by 1.2 magnitudes to our Abell 2246 field limit of m(FUV)=19.7, obtaining 0.6 possible foreground or background object per Abell 2246-sized field, compared with five detected objects.  Hence, the extreme rarity of serendipitous extragalactic FUV sources strongly supports identifying these objects as members of Abell 2246.  Figure \\ref{mosaic} shows $\\sim$1\\arcmin\\ subimages of the five sources in Table \\ref{abonly} identified here as galaxies at the distance of Abell 2246.  The subimages, in the V, MUV and FUV bands respectively, have the the same orientation as Figure \\ref{fullfield}.  The objects are clearly not point sources, although the amount of morphological information available from the images is limited.  \\begin{figure*}[hp] \\begin{center} \\epsfysize=6.0truein \\vglue 7.5truein \\end{center} \\caption{Subimages of the five sources 3,5,6,7 and 8 in Table \\ref{datatab} which we identify in this work as galaxies at the distance of Abell 2246, in the FUV (left) MUV (center) and V (right) bands respectively. Orientation is the same as in \\ref{fullfield}, and the angular scale is shown by a 10\\arcsec\\ bar. The distance and surface brightnesses of the galaxies limit morphological information in the images, although the objects are clearly not point sources. \\label{mosaic}} \\end{figure*}  Assuming a distance modulus of 39.8 mag (based on z=0.225; \\cite{struble}), we determine absolute FUV magnitudes (corrected for Galactic foreground extinction using E(B--V)=0.03 and the extinction law of \\cite{extinction}) in the range from --21.4 to --20.4 (Table~\\ref{abonly}). A (FUV--V), FUV color magnitude diagram for Abell 2246 galaxies and well-known local objects (\\cite{fanelli}) is shown in Figure~\\ref{cmdiag}. In this and subsequent plots all observations have been corrected for Galactic foreground extinction.  (Note that while the color error bars from Table~\\ref{abonly} would be a large fraction of the plot's color scale, the plot's entire range covers only the colors of stellar spectral types B9 through early A, showing the leverage of the FUV bandpass in this spectral range.) The effect of K corrections is shown by arrows, which display the changes in (FUV--V) color and FUV magnitude obtained by redshifting representative galaxy and starburst spectra. To compute these vectors we have used the template spectra of \\cite{kinney}, which provide spectral information across the wavelength range from 1200 to 10000 \\AA. (In this paper we use ``K correction'' to mean a transformation to rest-frame quantities, and ``redshift correction'' to mean its inverse, a transformation to observed quantities).  The vectors shown are the redshift corrections (z=0.225) of the Sc galaxy template, the mean redshift correction of the Sb and Sc galaxy templates, and the mean redshift correction of starburst templates of all internal reddenings, of \\cite{kinney}.  (These  redshift correction vectors are intended to show displacements only; their locations are arbitrary). The observed cluster galaxies evidently have color-magnitude loci which lie within the range of luminous blue local objects.  However, K corrections are significant.  For example, if all the Abell 2246 objects are Sc galaxies, some are intrinisically brighter in the FUV than M101, a luminosity class I Sc.  \\begin{figure} \\plotone{a2fig3.eps} \\caption{A (FUV--V),V color absolute-magnitude diagram for Abell 2246 (filled circles) and well-known local objects (open circles, labeled). All observations have been corrected for Galactic foreground extinction. (Note that while the color error bars from Table~\\protect\\ref{abonly} are a large fraction of the plot's color scale, the plot's entire range covers only the colors of stellar spectral types B9 through early A, showing the leverage of the FUV bandpass in this spectral range.) Without redshift corrections,  the observed cluster galaxies have FUV colors and magnitudes  which lie within the range covered by local objects.  Redshift-correction vectors for template spectra of Sa and Sb galaxies (averaged), starbursts with a range of internal reddenings (averaged), and Sc galaxies are also shown; appropriate  redshift corrections are evidently significant. The vectors are intended to show displacements only; their locations are arbitrary. \\label{cmdiag}} \\end{figure}   Figure~\\ref{colorcoloro} compares the (MUV--V, FUV--V) colors of Abell 2246 galaxies with the integrated colors of local galaxies (labelled).  The Abell 2246 objects, observed at z=0.225, have slightly redder FUV--V colors than the rest-frame values of the bluest star-forming systems that are plotted here.  For comparison, the colors of individual annuli from M33 and M74 ({\\em cf} Figure 7 in \\cite{cornett}) form a linear distribution that runs approximately from the locations of NGC1404 to NGC2993 on this plot, with colors from the outermost regions of these galaxies near those of NGC2993.  \\begin{figure} \\plotone{a2fig4.eps} \\caption{A comparison of the (MUV--V,FUV--V) colors for Abell 2246 galaxies (filled circles) with those derived from total magnitudes of observed local galaxies (open circles, labelled). All objects are corrected for foreground extinction; the potential effect of internal ``screen'' reddening for E(B--V)=0.1 for Galactic and SMC extinction laws is shown by arrows.  Abell 2246 galaxies observed at z=0.225 have slightly redder FUV--V colors values than the rest-frame colors of the bluest star-forming systems that are plotted here.  \\label{colorcoloro}} \\end{figure}   Figure~\\ref{colorcolort} compares the (MUV--V,FUV--V) colors with data from the individual template galaxy and population spectra of \\cite{kinney}, also shown as averages in Figure~\\ref{cmdiag}.  (This Figure has error bars appropriate for Figures~\\ref{colorcoloro},\\ref{colorcolort}, and \\ref{colorcolorm}).  The template spectra are constructed from composites of ground-based and IUE spectra of individual galaxies made through apertures matched to the $\\sim$ 10$\\times$20 \\arcsec\\ large IUE aperture. The colors plotted, from upper left to lower right, range from those of composite starburst galaxies with small (E(B--V)$<$0.1) to large (E(B--V)$\\sim$0.5) internal reddening as measured by Balmer decrements, through galaxies of increasingly earlier (``Sc, Sb, Sa, S0, E'') type, to those of a bulge (``B'') population.  Each template color is plotted as viewed in its rest frame (diamonds) and redshifted to z=0.225 (crosses). The template starburst and Sc colors as well as the observed galaxy colors are in good agreement with bluer examples from the same bands from disk annuli in the local spiral galaxies M74, M81, and M33 (\\cite{cornett}).  \\begin{figure} \\plotone{a2fig5.eps} \\caption{A comparison of the Abell 2246 data from Figure~\\ref{colorcoloro} with observation-based template galaxy spectra.  Diamonds, from upper left to lower right,  are the rest-frame colors of starbursts with small (E(B--V)$<$0.1) to large (E(B--V)$\\sim$0.5) internal reddening, through galaxies of increasingly earlier (``Sc, Sb, Sa, S0, E'') type, to those of a bulge (``B'') population.  Template colors are also plotted as redshifted to z=0.225 (crosses), connected to corresponding rest-frame template colors.  Error bars shown here are appropriate for all color-color diagrams in this paper. \\label{colorcolort}} \\end{figure}   We have also computed and redshifted the spectral energy distributions (SEDs) of selected stellar population models of \\cite{bruzual}, shown in Figure~\\ref{colorcolorm}. The lower pair of curves shows single-burst star-formation models (that is, which have no star formation after age $\\sim$ 0), having solar metallicity, an initial mass function (IMF) slope of --1.35, and a range of ages.  The evolutionary tracks of those models as viewed in their rest frame (diamonds) and with a redshift of z=0.225 (crosses) are shown, along with the colors of the observed cluster galaxies (filled circles).  The model starbursts shown range from age 0.1 Gyr (upper left of the diagram) to 0.6 Gyr (lower right) in $\\sim$0.1 Gyr increments.  Evidently the observed galaxies are $\\sim$1 mag bluer in (FUV--V) than redshifted single-burst models of the observed (MUV--V).  The upper pair of curves in Figure \\ref{colorcolorm} shows the effect of longer-term evolution.  Those curves display colors from a decaying continuous star-formation model of \\cite{bruzual}, with a star formation decay rate time constant of $\\tau$=1 Gyr, with age increasing from $\\sim$0 to 6 Gyr from upper left to lower right.  This model, which contains both younger and older stars than single-burst models, produces (FUV--V) colors which are up to $\\sim$0.7 mag bluer than observed galaxies in Abell 2246, for the observed (MUV--V).  \\begin{figure} \\plotone{a2fig6.eps} \\caption{\\small{The observed MUV--V, FUV--V color-color data from Figure~\\protect\\ref{colorcoloro} displayed with the rest-frame (diamonds) and redshifted (crosses) evolutionary tracks of selected stellar population models. All models have IMF slopes --1.35.  The lower pair of model curves represent aging single-burst star formation, with ages ranging from age .1 Gyr (in the upper left of the diagram) to 0.6 Gyr (lower right) in $\\sim$0.1 Gyr increments.  The upper pair of curves is a continuous-star-formation model, with a star formation rate which decreases exponentially with time constant $\\tau$=1 Gyr, and ages from $\\sim$0-6 Gyr in 1 Gyr increments.  The observed galaxies show strong evidence of recent ($<$400 Myr) star formation. Both rest-frame and redshifted model colors are evenly spaced with time in (FUV--V), showing a simple power-law increase to redder color values.  This reflects the exponential fall-off in numbers of the hottest stars (which dominate the FUV bandpass) caused by the decaying star formation rate; the strictly straight-line relationship of the redshifted model is caused by the bluer bandpass of observation which samples only extremely hot stars. The filled diamond symbols represent the rest-frame colors of models of 10-Gyr-old stellar populations with different SFR decay times, simulating a galaxy with a long history of star formation. After redshift correction, the best model for Abell 2246 galaxy colors evidently has star formation decay timescales of 3 Gyr$<\\tau<$5 Gyr. This results in models which are $\\sim$0.5 mag too blue in (FUV--V); however, the single-burst model loci show that contributions from a single recent burst can add MUV and V flux without significant FUV. \\label{colorcolorm}} } \\end{figure}   The models' behavior in the color-color plane of Figure~\\ref{colorcolorm} may be explained as follows. For any reasonable IMF, the UIT FUV bandpass is dominated by the hottest main sequence stars.  Redshifting the spectrum to z=0.225 accentuates this, shifting the effective bandpass of observation to shorter $\\lambda$ (1060-1460\\AA).  Therefore the evolution of FUV flux, and (FUV--V) color, tracks changes in the total flux of the hottest main sequence stars.  This is borne out in the ``Decaying SFR'' models: both the rest-frame and redshifted colors are evenly spaced with time in (FUV--V), showing a simple power-law decrease.  This behavior reflects the exponential fall-off in numbers of the hottest stars with decaying star formation rate.  The color decay is nearly linear in this log-log plot because radiation at shorter (K-corrected) wavelengths selects the hottest stars even more stringently than the uncorrected bandpass. Single-burst models become redder in (FUV--V) more rapidly than ``Decaying SFR'' models, because the upper end of the decaying main sequence becomes cooler and redder, not just less populous.  The bend in the evolutionary curve of the single-burst models, which strongly resembles the locus of the colors of stellar atmospheres of a wide range of temperatures in this plane, is caused by the rapid fall-off of sensitivity of the FUV bandpass to cooler stars: at about $T_{\\rm eff}=9000{\\rm K}$, the peak of those stars' UV continuum moves longward of the FUV long-wavelength cutoff at $\\sim$1800\\AA.  Stars much cooler than this contribute significantly to MUV flux, but not to FUV. The location of the observed objects at significantly bluer (FUV--V) than redshifted single-burst models implies the existence of star formation more recent than the burst's age ($\\sim$400-500 Myr) in the observed galaxies.  The effects of redshift in Figure~\\ref{colorcolorm} are of special interest. For SEDs with current star formation and therefore a significant hot-star component, correction to z=0.225 moves the models to bluer colors in both (FUV--V) and (MUV--V), {\\em upward} along the aging line.  In such SEDs both the FUV and MUV flux increase relative to V with redshift as more of the rising part of the hot-star continuum is shifted into those bandpasses.  This is opposite to the cases of the templates (Figure~\\ref{colorcolort}) and single-burst models, whose SEDs lack any contributions from the hottest stars.  The filled diamond symbols represent the rest-frame colors of models of 10-Gyr-old stellar populations with different SFR decay times (\\cite{cornett}).  They simulate a galaxy with a long history of star formation as well as strong current star formation, and may be used to constrain the time dependence of that star formation.  Since these model SEDs contain important contribution from the hottest stars, redshift corrections will move their colors upward and to the left along their locus, shortening their apparent SF decay timescale.  After this redshift correction, models consistent with Abell 2246 galaxy colors have star formation decay timescales of 3 Gyr$<\\tau<$5 Gyr.  The models are still $\\sim$0.5 mag too blue in (FUV--V); however, adding single-burst models with late A stars as their brightest component will add V and MUV flux, but not FUV.  From the above discussion it is evident that a combination of past star formation, as represented by single-burst models, and current star formation, as included in the models of \\cite{bruzual}, can produce the colors observed in Abell 2246.  The 10-Gyr-old decaying-SFR population with 3 Gyr$<\\tau<$5 Gyr overestimates the proportion of current star formation, with resulting (FUV--V) colors that are too blue.  Conversely, a single-burst population of age $\\sim$400-500 Myr is too red in (FUV--V), and implies that more recent star formation is required.  We conclude that the observed Abell 2246 galaxies a) are UV-luminous, with some as bright in the FUV as M101, depending on the assumed spectrum and resulting redshift correction. b) have FUV--V colors that show clear evidence of the flux of relatively young ($<$100 Myr) stellar populations, and are undergoing significant current star formation; but c) like observed late-type spiral galaxies, probably require flux contributed by stars of a range (up to $>$ 400 Myr) of ages.  This mixture of populations is like  that of the objects known as ``E+A'' or ``poststarburst'' galaxies identified by \\cite{dressler}.  Spectra of these objects are thought to show evidence of recent (10$^{8}$-10$^{9}$ yr preceding the rest-frame epoch), vigorous star formation which has since decreased dramatically. \\cite{liu} model optical-band spectrophotometry of eight such objects having 0.088$<$z$<$0.545.  Their models are constructed from a linear combination of two spectral components: a composite spectrum of either an E/S0 or Sbc galaxy template; and a starburst with an age in the range 10$^{8}$ to 2 $\\times10^{9}$ yr.  Figures~\\ref{colorcolort} and \\ref{colorcolorm} show that it is likely that such a combination would produce UV colors like those of Abell 2246 galaxies as well.  The red colors of all templates except the ``Sc'' example in Figure~\\ref{colorcolort}, compared to those of Abell 2246 galaxies, make it clear that a late-type underlying galaxy is required to model these FUV-bright objects.  The current FUV flux, a direct tracer of massive stars and therefore of recent star formation, permits estimating rates for formation of massive (M $>$ 5 \\msol) stars.  Following \\cite{lequeux}, we use stellar evolutionary models of \\cite{schaller}, \\cite{schaerer}, and \\cite{charbonnel} with the stellar atmosphere models of \\cite{kurucz}.  For comparison with \\cite{donas1}, our selected IMF is a two-component power law with a break at M=1.8 \\msol, with low-mass slope -0.6 and high-mass slope -2.5. For simplicity we use a model with constant star formation, which has constant UV flux after 100 Myr (Landsman, personal communication), so our integration times are chosen to be 300 Myr. This model allows us to approximate the star formation rate as a function of the FUV (1520\\AA) luminosity as  \\[ SFR = \\frac{L_{1520}}{4.46\\times 10^{39}} M_{\\odot} yr^{-1} \\]  This expression gives good agreement with other models having similar assumptions ({\\em cf} \\cite{donas1}), in view of the uncertainties present in the approach.  Star formation rates computed from this expression for Abell 2246 galaxies, ranging from 1.4 to 2.9 \\msol yr$^{-1}$, are given in  Table~\\ref{abonly}.  These values are lower limits for actual SFRs, since extinction will decrease the apparent $L_{1520}$ values.  The expected redshift correction for {\\em current} star formation rates at z=0.225, as discussed below, is small.  Only a few nearby clusters of galaxies have been observed in any ultraviolet bandpass.  \\cite{donas2} and \\cite{donas3} discuss $\\sim$20\\arcsec\\/ resolution images of a local benchmark, the Coma cluster, in a bandpass $\\sim$150\\AA\\/ wide centered at $\\sim$2000\\AA, deriving a UV luminosity function including 254 galaxies.  It is particularly straightforward to compare their observations with our MUV photometry of Abell 2246 galaxies, because the central wavelengths of the bandpasses in the clusters' rest frames are nearly equal (1955~\\AA\\/ for \\cite{donas2} observations of Coma {\\em vs} 2033~\\AA\\/ for our MUV bandpass at Abell 2246) and both  observations contained the entire clusters in their respective fields of view.  Assuming a distance modulus of 34.8 for Coma based on z = 0.023 (\\cite{thompson}) we use the MUV luminosity function of \\cite{donas2} to determine that Coma, at the distance of Abell 2246, would have $\\sim$12 UIT-detected galaxies given our limit of m(MUV)$\\sim$19.9.  A small richness correction derived by scaling the absolute numbers of Abell 2246 galaxies and Coma galaxies modifies the number of Coma galaxies seen at the Abell 2246 distance to be $\\sim$13-14. This result is consistent with the MUV data for Abell 2246: while only 5 Abell 2246 galaxies are {\\em uniquely} detected and identified because of the confusion due to multiple sources, many more are apparent in the image.  Large numbers of detectable MUV sources are present in the UIT MUV image of Abell 2246 because the MUV bandpass, sensitive to stars as late in spectral type as F, detects bright elliptical galaxies as well as those with current star formation.  The number of FUV-bright galaxies in Abell 2246 is apparently higher than in the Coma Cluster, however.  We may make that comparison using UIT Astro-2 FUV observations (1900 sec exposure time; detection limit m(FUV)$\\sim$19.0) of a field containing the central $\\sim$ 40\\arcmin\\ of the Coma cluster.  The two brightest galaxies in the UIT field, NGC4858 and IC4040, have UIT-determined m(FUV) of 14.65 and 14.70, respectively. Both are spirals and IRAS sources, and are among the six brightest Coma galaxies in the MUV according to \\cite{donas2}. These two galaxies, at the distance of Abell 2246, would have m(FUV) of 19.65 and 19.70 respectively--slightly fainter than the faintest observed Abell 2246 galaxy.  Taking into account the redshift correction to z=0.225 depends critically on stellar content.  SEDs with significant young stellar population ({\\em e.g.} upper curves in Figure~\\ref{colorcolorm}) have essentially unchanged FUV flux ($|\\Delta m(FUV)|<$0.1), while SEDs without the youngest stars become  0.8-1.0 magnitudes fainter at z=0.225. In either case, these two galaxies would be fainter than detected Abell 2246 galaxies.  We also compare estimates derived from MUV observations of the entire Coma cluster by {\\cite{donas2}. Coma's MUV luminosity function shows only 1 galaxy, NGC 4848 at m(MUV)=13.8, brighter than m(MUV)=14.6. (NGC 4848 and other brighter members were outside the UIT Coma field of view).  If we assume NGC4848 to have FUV flux and colors indicating significant recent star formation, then (FUV--MUV)$\\sim-0.5$, and the FUV redshift correction is small, resulting in m(FUV)=18.2--18.4, 0.3 mag brighter than any Abell 2246 galaxy.  If we assume NGC4848 to have the colors of normal spiral galaxies, then (FUV--MUV)$\\sim$0.0, and NGC4848 at Coma would have, after redshift correction, m(FUV) = 19.6-19.8, approximately equal to the faintest observed Abell 2246 galaxy.  A FUV measurement of NGC4848 would resolve this ambiguity; in either case NGC4848 is the only Coma galaxy that is possibly brighter in the FUV, and all others are certainly fainter, than Abell 2246 galaxies. It is noteworthy that the radial distribution within the cluster of identified Abell 2246 FUV sources is similar to Coma's distribution of MUV sources (\\cite{donas3}), with 3 of 5 Abell 2246 galaxies at a radius $\\sim$0.33 of the Abell radius (though the statistics are inadequate).  The rarity of FUV-bright Coma cluster galaxies is evidence that Abell 2246 has undergone significantly more recent star formation than Coma (which nonetheless shows, in its MUV characteristics, evidence of star formation within the past 1 Gyr; {\\em cf} \\cite{donas3}). This comparison is relevant to the Butcher-Oemler effect (\\cite{butcher1}; \\cite{butcher2}) wherein clusters are found to contain fractions of blue galaxies (as determined by restframe (B--V)) which increase with redshift. The FUV-bright galaxies in Abell 2246 directly show high rates of star formation $\\sim$1.2 Gyr ago at z=0.225, which are not present to the same extent at more recent times and smaller redshifts in Coma.  FUV data uniquely confirm the nature and scope of this star formation; additional FUV imaging of clusters such as Abell 2246 with higher sensitivity and finer resolution will be be vital in bridging the gap between the local neighborhood and the distant universe."
    },
    "9805/astro-ph9805327_arXiv.txt": {
        "abstract": "The decimetric radio continuum luminosity of a star-forming galaxy appears to be directly proportional to the rate of formation of supernovae in the galaxy. Since decimetric radiation does not suffer significant extinction and is not directive, radio luminosities may thus provide a particularly straightforward way to determine the current rate of star formation. Using a sample of over 700 local galaxies we confirm the utility of the radio luminosity as a measure of star formation rate by showing concordance with the rates predicted by U-band, \\halpha, and far infrared luminosites. We also show that there are systematic discrepancies between these various indicators, suggesting that the \\halpha~luminosity may underestimate the star formation rate by approximately an order of magnitude when the star formation rate is $\\gtrsim$ 20 M$_\\odot$ yr$^{-1}$. We use this calibration and the measured radio luminosities of sub-mJy radio sources to infer the star formation rate in approximately 60 star-forming galaxies at moderate ($z \\gtrsim 0.1$) redshifts, both as the actual rate and as the fraction of the existing mass of stars in the galaxy. For some of these objects the inferred current rate of star formation could increase the stellar mass in the galaxy by approximately 10\\% over an interval of $\\approx 30$ Myr. ",
        "introduction": "In the vast majority of extragalactic radio sources with 1.4 GHz flux density S$_{1.4} \\gtrsim 5$~mJy the radio emission appears to be due to the existence of a nuclear `monster' or `engine.'  These sources are members of the class of powerful radio galaxies, emitting a spectral power density $P_{1.4} \\gtrsim 10^{23}$~W~Hz$^{-1}$ (e.g. Condon \\markcite{Con:89}1989, especially his Fig. 12; Wall \\& Jackson \\markcite{Wal:97}1997).  Their optical hosts are intrinsically bright elliptical galaxies with optical luminosities of the order of $5 \\times 10^{37}$~W, often with red colours (e.g., Kron, Koo \\& Windhorst \\markcite{Kro:85}1985) and occasionally with high-excitation optical emission line spectra (e.g. Hine \\& Longair \\markcite{Hin:79}1979).  While the nuclear monster may stimulate some nuclear star formation, or vice versa, there is not a tight relationship between the radio power and either the number of stars or the current rate of star formation in the host galaxy.  Many of the remaining extragalactic radio sources with S$_{1.4} \\gtrsim 5$~mJy, as well as a large fraction of fainter sources, belong to a different population. These have optical counterparts that are spiral (if $V \\lesssim 18.5$) or blue peculiar (if $V \\gtrsim 18.5$) galaxies with a wide range of optical luminosities, often displaying evidence of ongoing star formation (Kron \\etal~\\markcite{Kro:85}1985; Benn \\etal~\\markcite{Ben:93}1993).  For the more luminous members of this population it is known that there is a tight correlation between the far-infrared (FIR) luminosity and the 1.4 GHz radio power (reviewed by Condon \\markcite{Con:92}1992).  Astrophysical interpretations of this correlation usually identify both the FIR and the radio emission as the consequence of ongoing massive star formation. As a corollary, the opportunity exists to determine the current rate of massive star formation from measurements of the FIR or radio luminosity.  The astrophysical significance of this opportunity is heightened by the fact that very deep radio surveys reveal large numbers of star-forming galaxies at redshifts well beyond $z \\approx 0.1$, at an epoch where there is growing evidence that some classes of galaxies experience star formation at a rate higher than in the local Universe. In this paper we explore the calibration of the relationship between radio luminosity and star formation rate using a sample of local star-forming galaxies with measured radio flux densities, and then apply the calibration to determine the current star formation rate in a sample of galaxies at $0.1 \\lesssim z \\lesssim 0.5$. ",
        "conclusions": "We consider firstly the `nearby' sample in which several different estimators of SFR are often available for a single galaxy.  Figure 1 indicates that SFR estimates based on the various indicators are in broad agreement with one another, but also points to the existence of systematic differences between different estimators of the SFR. There are also several objects in which at least one indicator is significantly discordant.  The tightness of the relationship between the star formation rates estimated from 1.4 GHz and 60 $\\mu$m data reflects the well-studied radio/FIR correlation.  Estimates of star-formation rates obtained from the two frequencies agree closely over more than four orders of magnitude. There is a tendency for objects with $SFR_{1.4} < 0.2$ M$_{\\odot}$ yr$^{-1}$ to have relatively high values of $SFR_{FIR}$, perhaps revealing some heating due to the general interstellar radiation field of older disk stars, without associated supernovae due to younger massive stars (see also Lonsdale \\& Helou \\markcite{Lon:87}1987).  A notable exception to the correlation is the very small number of galaxies in which the radio prediction is unusually high. These objects include NGC 4374 (M84; 3C 272.1) from Kennicutt \\& Kent (1983), IRAS 0421+040 from Eales \\etal~(1988), and Mrk 3 (4C+70.05) from Kennicutt (1992).  In each case there was clear prior evidence that the radio emission is not related to star formation, and the objects have been retained in our data set only to illustrate the effect of `radio-loud' galaxies.  The existence of these objects points to the potential problem that 1.4~GHz luminosities may be `contaminated' by radio emission arising from a nuclear monster, rather than star formation. In deducing the SFR from radio flux densities this contamination can be minimized by applying the following principles: (1) optical colours and optical spectral line ratios sucessfully delineate the class of objects with emission dominated by star formation (e.g. Benn \\etal~1993), (2) the fraction of galaxies which owe their radio emission to star formation rises quickly with decreasing flux density in the sub-mJy radio source population (e.g. Kron \\etal~1985; Wall \\& Jackson 1997), (3) any optical emission lines in classical radio galaxies with relatively low radio powers are relatively weak but lie in the high-excitation part of the relevant diagnostic diagram (e.g. Baum \\& Heckman \\markcite{Bau:89}1989), and (4) the bi-modal character of the radio-loud/radio-quiet dichotomy allows the use of the radio/optical flux density ratio to exclude the most powerful radio-loud objects (e.g, Sopp \\& Alexander \\markcite{Sop:91}1991, their Fig 1.)  Figure 1 shows that estimates of the SFR based on the widely-used \\halpha~luminosity indicator tend to be more scattered (relative to $SFR_{1.4}$) than the $SFR_{1.4}/SFR_{FIR}$ correlation. They also tend to lie about a factor of 10 above the trend defined by the 1.4 GHz/60 $\\mu$m estimates when SFR $\\lesssim$ 0.1 M$_\\odot$ yr$^{-1}$, and about a factor of 10 below the trend when SFR $\\gtrsim$ 20 M$_\\odot$ yr$^{-1}$.  The systematic deviations, which could have important implications for studies of the galaxy-to-galaxy variation of star formation rates, can be seen also in Figure 1 of Devereaux \\& Young \\markcite{Dev:90}(1990) and Figure 8 of Young \\etal~(1996), although these authors did not emphasise the discrepancy.  On the other hand, investigations spanning a smaller range of luminosities (e.g. Kennicutt \\markcite{Ken:83}1983b; Lonsdale \\& Helou 1987) have sometimes infered that the relationship is close to strict proportionality.  Devereaux \\& Young (1990) argue that the general correlation between \\halpha~and FIR results constitutes `strong evidence' for the view that the \\halpha~and the FIR luminosity are produced mainly by massive stars in galaxies with high FIR luminosity. Lonsdale \\& Helou (1987) argue on the other hand that a parallel correlation between the FIR and B-band luminosity implies that the FIR emission in most spiral galaxies is `dominated' by emission due to cool dust which is heated in part by low-mass stars. Even in this case, however, we would expect the FIR luminosity to be increasingly dominated by ionising radiation in the presence of vigorous star formation.  The systematic over-estimate of the SFR derived from \\halpha~relative to the 1.4~GHz estimates at {\\it low} values of SFR is related in part to the difficulty of determining the zero-point level for \\halpha~emission from H II regions in the presence of stellar \\halpha~emission or absorption. Furthermore, since the radio/FIR correlation can be `linearized' by allowing for the escape of cosmic rays from galaxies with weak star formation (Condon, Anderson \\& Helou \\markcite{Con:91a}1991), we must also entertain the possiblity that the SFR estimate based on $L_1.4$ is actually too low in these objects.  The deviation at high radio luminosities, as well as the increasing scatter in the correlation, could result from a relatively large amount of extinction in those objects undergoing the most vigorous star formation, or from the absorption by of Lyman photons by dust, or from an IMF which weights differently the high-mass stars mainly responsible for \\halpha~and the lower mass stars which dominate the supernova numbers (or to a combination of these factors). The fact that the estimates based on the edge-on sample of Lehnert and Heckmann (1996 - the open triangles in Figure 1) fall systematically low supports the view that extinction plays a particularly significant role. It is notable that the value of SFR derived from \\halpha~tends to be bounded above by a rate of $\\approx 25$ M$_{\\odot}$ yr$^{-1}$, corresponding to a luminosity of about $3 \\times 10^{35}$ W Hz$^{-1}$.  Estimates of the SFR based on U-band observations exhibit considerable scatter with respect to the radio estimates. Like the estimates based on \\halpha~they tend to be higher at low SFR and lower at high SFR. As noted above, even old stellar populations emit U-band light (e.g., Bruzual \\& Charlot \\markcite{Bru:93}1993, their Fig 1) providing a possible explanation of the trend observed at low values of SFR.  The trend for relatively low SFR values towards the right of the figure could reflect enhanced extinction in vigorous starbursts. As with \\halpha~there appears to be an upper limit to the SFR derived from U-band measurements.  The discrepancies highlighted here need to be further investigated, especially given the increasing use of broad-band far-UV luminosities to infer star formation rates in galaxies with very high redshifts (z $\\gtrsim$ 1; e.g.  Cowie \\etal~\\markcite{Cow:97}1997).  The points shown in Figure 2 correspond to galaxies selected by very deep surveys in the radio continuum. Of the star-formation indicators discussed above, only radio and \\halpha~luminosities are available at present for these samples.  Moreover, the \\halpha~values have been derived from noisy spectra which were obtained to determine redshifts rather than line fluxes, and hence they are expected to show significant scatter. The expedient of rejecting radio-selected objects that are red and/or show absorption line spectra has led to a sample which follows a trend that is essentially indistinguishable from the `local' sample. It is thus reasonable to conclude that (1) the `shortfall' in the \\halpha~estimates of the SFR for the distant sample is similar to that in the local population and is likely to arise for the same reasons, and (2) the radio luminosities of these types of galaxies provide estimates of their current star formation rates according to equation (1). A corollary of this is that several faint radio-selected galaxies have star formation rates comparable with those of the intrinsically luminous IRAS galaxies.  There is therefore considerable promise that deep radio surveys will reveal significant numbers of objects similar to the luminous IRAS galaxies, but lying at higher redshifts.  While the SFR of a galaxy is of interest in its own right, the ratio of the current SFR to the total number of stars that have been formed in a galaxy offers additional insight into the potency of the current rate of star formation.  We can estimate the total mass of stars formed in the faint radio galaxies, $M_{TOT}$, by using their $R$-band luminosities in the relation \\begin{equation} M_{TOT} = \\frac{L_R}{3.4 \\times 10^{11}~{\\rm W Hz}^{-1}} {\\rm M_{\\odot}}. \\end{equation} Here, we have used the approximation (deduced from Bruzual \\& Charlot 1993) that steady star formation, with a standard IMF and occurring over a period of more than 1 Gyr, will produce an $R$-band mass-to-light ratio of $R \\approx 5$~mag~M$_{\\odot}^{-1}$.  Since this approach fails to account for the contribution of any {\\it current} starburst to the $R$-band luminosity it will underestimate the fractional amplitude of very strong bursts.  \\placefigure{Fig3}  Figure 3 exhibits, for the sample shown in Figure 2, the relationship between the SFR (of massive stars) deduced from the radio luminosity and the total mass of stars deduced from equation (5). Lines are drawn to illustrate the locii of points for which the characteristic time $\\tau = M_{TOT} / SFR_{1.4}$ is constant. There is a tendency for galaxies that have already formed many stars to support a higher rate of current star formation. There is also a wide scatter in the ratio of SFR to total mass at any chosen size, not all of which is due to errors of observation. There are about a dozen galaxies with a stellar mass $M \\approx 10^{10}$ M$_{\\odot}$ and a star formation rate $SFR(M > 5 M_{\\odot}) \\geq 10$ M$_{\\odot}$ yr$^{-1}$. For such galaxies, the current burst of star formation is likely to increase the stellar mass by more than 10\\% if sustained for a time exceeding $10^7$~yr, implying an event of considerable significance in the development of the system. These objects are reminiscent of the IRAS galaxies that have high values of $L_{FIR}/L_B$ (e.g., Sanders \\& Mirabel \\markcite{San:96}1996, their Section 2.2)."
    },
    "9805/astro-ph9805039_arXiv.txt": {
        "abstract": "For the purpose of deriving the helium abundances in chemically peculiar stars, the importance of assuming a correct helium abundance has been investigated for determining the effective temperature and gravity of main sequence B-type stars, making full use of the present capability of reproducing their helium lines.  Even if the flux distribution of main sequence B-type stars appears to depend only on the effective temperature for any helium abundance, the effective temperature, gravity and  helium abundance have to be determined simultaneously by matching the Balmer line profiles. New MULTI NLTE calculations, performed adopting ATLAS9 model atmospheres and updated helium atomic parameters, reproduce most of the observed equivalent widths of neutral helium lines for main sequence B-type stars and they make us confident of the possibility to correctly derive the helium abundance in chemically peculiar stars.  An application of previous methods to the helium rich star HD\\,37017 shows that helium could be stratified in the magnetic pole regions, as expected in the framework of the diffusion theory in the presence of mass loss. ",
        "introduction": "\\vspace{-1mm} Chemically peculiar (CP) stars can present helium lines whose strength is different than in main sequence stars with equal Balmer lines and/or flux distribution. Usually, helium lines are weaker in the coolest CP stars and stronger in the hottest ones. As for other chemical elements, helium is supposed to be under- or overabundant in the photosphere because of diffusion processes. Computations by Vauclair et al. (1991) show that helium should also be stratified in the outer stellar layers.  All methods already known and used to determine stellar effective temperatures and gravities, which are necessary to derive helium abundances, are calibrated assuming a solar chemical composition. Thus, to correctly derive the peculiar helium abundance in CP stars, we have to understand \\begin{itemize} \\item the importance of helium abundance in determining the effective temperature and gravity. \\end{itemize}  A further fundamental step to derive the helium abundance of CP stars is the comprehension of \\begin{itemize} \\item the present capability of reproducing the helium lines of main sequence stars. \\end{itemize} ",
        "conclusions": "Since the stellar flux distribution of B-type stars depends only slightly on gravity and helium abundance, it can be used to determine the effective temperature of chemically peculiar stars. Anyway, since the Balmer lines are necessary to infer the gravity and are influenced by helium abundance, the effective temperature, gravity and helium abundance have to be determined simultaneously by matching their profiles.  The NLTE calculations which are available in the literature do not reproduce the observed equivalent width of the neutral helium lines for main sequence B-type stars. Before deriving the helium abundance, the reliability of the method used should be at least tested on main sequence B-type stars before it is applied to chemically peculiar stars.  Using the method described by Leone \\& Manfr\\`e (1997) to determine the stellar parameters from the Balmer line profiles, and using the NLTE calculations by Leone \\& Lanzafame (1998) of neutral helium lines, the helium abundance has been derived for the helium-rich star HD\\,37017 from a sample of lines which are mainly formed in different atmospheric layers. It appears that helium abundance decreases with optical depth in the magnetic pole regions while it remains constant with optical depth in the magnetic equatorial region. This result is consistent with the calculations by Vauclair et al. (1991), which foresee overabundant and stratified helium in the atmospheric regions close to the magnetic poles."
    },
    "9805/astro-ph9805333_arXiv.txt": {
        "abstract": "We report on the discovery of a variable radio source, \\VLAJ, associated with the proposed x-ray counterpart, 1SAX J0702.6+3850 of \\grb.  The source was monitored from one day after the burst to one month later at centimeter wavelengths (1.4, 4.9, 8.3 and 15 GHz; Very Large Array -- VLA), and in the 3-mm band (90 GHz; Owen Valley Radio Observatory -- OVRO).  Based on its variability, compactness and spectrum, we identify \\VLAJ\\ as the afterglow from \\grb.  We interpret the rapid flux density variations as interstellar scintillation and the sharp turnover in the radio spectrum below 13 GHz as arising from synchrotron self-absorption.  This suggests that the angular diameter of the fireball from \\grb\\ was of order a few microarcseconds in the first two weeks. The absence of a readily detectable afterglow in the optical, but clear detections in both the radio and infrared, can be understood as the result of extinction by dust. In this context it is interesting to note that half of all well-localized gamma-ray bursts have no optical counterpart. Based on our study of \\grb\\ we suggest that the progenitors of a sizable fraction of gamma-ray bursts are associated with regions of moderate to high gas density. ",
        "introduction": "The gamma-ray burst \\grb\\ was detected on 1998 March 29.1559 UT by the gamma-ray burst monitor on the BeppoSAX satellite and localized to an error circle of 3\\arcmin\\ radius by the Wide Field Camera (Frontera et al.\\ 1998). The burst lasted a total of 55 s and was bright -- a fluence of 5$\\times{10}^{-5}$ erg cm$^{-2}$ in the range 50--300 keV was measured by the Burst and Transient Experiment on board the Compton Gamma Ray Observatory satellite (Briggs et al.\\ 1998). Followup observations with the Narrow Field X-ray Instruments (NFI) on board BeppoSAX began 7 hrs after the burst.  in't Zand et al.\\ (1998) detected a previously uncataloged X-ray source 1SAX J0702.6+3850 (with an error circle of 1\\arcmin\\ radius) which was seen to fade by a factor of 3 over the 14 hr observation period. The burst position was further constrained by a timing annulus derived between the Ulysses and BeppoSAX satellites (Hurley et al.\\ 1998).  A power law decay in the X-ray flux, S$_x\\propto t^\\delta$ appears to be a universal signature of the X-ray afterglows; $\\delta$ ranges from $-1.1$ to $-1.5$ (e.g. Costa et al.\\ 1997, Piro et al.\\ 1998).  This phenomenon of long-lived post-burst emission -- afterglow -- is nicely explained by GRB models known as fireballs, in which relativistically expanding debris shells interact with a surrounding medium and accelerate a power law distribution of particles (M\\'esz\\'aros \\& Rees 1997). As the fireball decelerates the peak of the emitted radiation shifts in time to lower energies producing power law decays in the X-ray, optical and radio bands (e.g.  Waxman 1997a, Sari, Piran \\& Narayan 1998).  The successful detections at X-ray wavelengths have not been matched by equal numbers of afterglows at optical wavelengths. Only 4 of the $\\sim$10 GRBs with X-ray afterglows have detected optical afterglows (van Paradijs et al.\\ 1997, Bond et al.\\ 1997, Halpern et al.  1998, Groot et al.\\ 1998).  GRB\\,970828 is one notable non-detection; it exhibited a typical X-ray decay (Murakami et al.\\ 1998), yet no optical transient was associated with this burst to a limit 100 times lower than that of the optical afterglow from GRB\\,970508 (Odewahn et al. 1997).  Several authors have attempted to explain the diversity of afterglow behavior by invoking more complex fireball models (M\\'esz\\'aros, Rees, \\& Wijers 1998, Panaitescu, M\\'esz\\'aros, \\& Rees 1998), including such effects as non-isotropic expansion, non-uniform distributions of external media, and a time dependence in the radiative efficiency of the shock.  Paczy\\'nski (1998) suggested that foreground extinction could account for the absence of optical afterglows from GRBs. Indeed, modest extinction has been inferred for the optical afterglow from GRB\\,971214 (Reichart 1998, Halpern et al.  1998, Ramaprakash et al.\\ 1998). A logical extension of this extinction hypothesis is that, in some cases, the afterglow may be seen in the radio or infrared bands but not in the optical.  In this paper we report the detection of a variable radio source in the reduced error circle of \\grb\\ and argue that it is the radio afterglow from an optically obscured burst. ",
        "conclusions": "\\subsection{The Mystery of Optically Dim GRBs}  In \\S{1} we alluded to the fact that about half of the well-localized GRBs appear not to have any associated optical transient.  There are two solutions to this mystery (1) the afterglow decays rapidly in such GRBs or (2) Many GRBs suffer from extinction.  Photoelectric attenuation at keV energies decreases as $\\nu^{-3}$ whereas optical extinction $\\propto \\nu^q$ where $q$ is one or even larger.  Thus, a modest column density, say $10^{21}$ cm$^{-2}$, in the host has little affect on the SAX passband of 1--60 keV (Wide Field Camera and the Narrow Field Instruments) but can suppress the observed R-band flux by almost 5 mag.  Indeed the absence of detectable optical flux accompanying strong X-ray emission may simply be due to the strong redshift dependence of the suppression at observed optical wavelengths in comparison to the X--ray wavelengths, $\\tau_{\\rm opt}/\\tau_{\\rm X-ray} \\propto (1+z)^{3+q}$.  The standard afterglow model of M\\'esz\\'aros \\& Rees (1997) predicts that the peak flux density will be independent of frequency.  Our average 8.3 GHz flux density of 325 $\\mu$Jy from the period April 21--30 implies a peak R-band magnitude $\\sim$ 17, yet only 15 hours after the burst R was greater than 20 (Klose, Meusinger, \\& Lehmann 1998). Observations of very red colors for \\grb\\ show that, at least for this GRB, dust extinction is the cause of the dim optical emission (Djorgovski et al. 1998c, Palazzi et al.\\ 1998b).  Evidence for dust has been inferred in GRB\\,971214 (Reichart 1998, Halpern et al.\\ 1998, Ramaprakash et al.\\ 1998); however, the inferred column density, while uncertain, is rather modest, $A_R\\sim 1$ mag.  In the case of \\grb, the R-band extinction is higher, making the detection of the afterglow easier at IR wavelengths.  If dust is indeed the main cause of optically dim GRBs then we conclude that extinction is significant in at least half of the GRBs. At face value this is not compatible with the currently popular model of coalescing neutron stars (Narayan, Paczy\\'nski, \\& Piran 1992). Such ns-ns binaries are expected to live long enough ($10^9$ yr) and acquire sufficiently high velocities to remove them from their high gas density birthplaces. An alternative model for GRBs is the ``hypernova'' model (Woosley 1993, Paczy\\'nski 1998), in which very massive stars produce a ``dirty'' fireball with $\\sim$300 times the luminosity of a supernova.  Since such massive stars die young (age $\\sim10^6$ yr), a natural consequence of this model is that GRBs should be associated with high density, dusty regions actively undergoing star formation.  The large energies inferred for the $\\gamma$-ray burst and optical afterglow of GRB\\,971214 (Kulkarni et al.\\ 1998; Ramaprakash et al.\\ 1998) present a further challenge to the coalescing ns-ns models. The hypernova model is capable of producing a more energetic and long-lived afterglow via its ``dirty fireball'' than current ns-ns models which produce a ``clean fireball'' (Paczy\\'nski 1998).   \\subsection{Future Evolution of \\VLAJ}  Taking cues from the radio afterglow of GRB\\,970508 we expect that the current chromatic intensity variations will die and be replaced by broad-band lower modulation-index scintillation. Analyzing existing and future data in terms of diffractive and refractive scintillation will enable us to trace the angular evolution of the fireball (Frail \\etal\\ 1997, Waxman, Kulkarni, \\& Frail 1997). The character of the fluctuations from \\grb\\ are strikingly similar to GRB\\,970508 for which a size of 3 $\\mu$arcsec was inferred (Goodman 1997, Frail et al.\\ 1997). The angular size of \\grb\\ may be even smaller than this since it is at a lower Galactic latitude ($b=18.7^\\circ$ vs 26.7$^\\circ$) and the quenching of scattering scales inversely with sin($b$) (Goodman 1997).  When these flux modulations die down (due to the fireball expansion) the underlying spectrum is revealed. Between 8.3, 15 and 90 GHz the spectral slope $\\alpha$ is roughly consistent with the canonical 1/3 value expected for the low energy tail of the electron distribution (Fig.~2).  However, between 4.9 and 8.3 GHz a value of $\\alpha\\simeq{2}$ is suggested at late times (see Table 1). A similar slope was seen for GRB\\thinspace{970508} between 1.43 GHz and 4.86 GHz (Shepherd et al.\\ 1998), and was attributed to synchrotron self-absorption (Katz \\& Piran 1997). Again, as with GRB\\,970508, we determine from this an angular size for \\grb\\ of a few $\\mu$arcsec (Katz \\& Piran 1997). The higher self-absorption frequency inferred for \\grb\\ suggests that it is more compact than GRB\\thinspace{970508}, or in a denser environment, or that there exists an additional component of low energy electrons (Waxman 1997b) that increase the opacity at radio wavelengths.  In keeping with what has been observed from GRB\\,970508 and model predictions, the radio spectrum should evolve in time as the synchrotron peak shifts to lower energies.  Within a few months we will have sufficient data to carry out a comprehensive quantitative analysis of the radio observations with respect to fireball models."
    },
    "9805/astro-ph9805105_arXiv.txt": {
        "abstract": "We report the detection of the elusive carbon-chain radical C$_5$N in the dark cloud TMC1 and its tentative detection in the circumstellar envelope IRC+10216. C$_5$N appears to be two orders of magnitude less abundant than the related molecule HC$_5$N and much less abundant than expected from current gas phase chemistry models. In comparison the HC$_3$N to C$_3$N abundance ratio is of the order of 10, in reasonable agreement with model predictions.  We have also detected in IRC+10216 two lines arising from the C$_3$H radical in its excited $\\nu_4=1$ state. ",
        "introduction": "That long carbon chain radicals could be abundant and play a large role in interstellar chemistry was first recognized with the discoveries of C$_3$N in the circumstellar envelope IRC+10216 (Gu\\'elin \\& Thaddeus 1977) and in the dark cloud TMC1 (Friberg et al.\\ 1980). To date seven acetylenic chain radicals, C$_n$H, $n= 2-8$,  and five cyanopolyyne molecules, HC$_{2n}$CN, $n=1-5$, are identified in TMC1 and/or IRC+10216 (Bell et al.\\ 1997, Gu\\'elin et al.\\ 1997). Surprisingly, no cyanopolyyne radicals heavier than C$_3$N were so far detected, despite model predictions that at least C$_5$N should be abundant (Herbst  et al.\\ 1994).  The non-detection of C$_5$N was first blamed on a small dipole moment and on the lack of spectroscopic data. The  dipole moment $\\mu_0$ and the rotation constant $B_0$ of the  C-chains radicals depend critically on the nature of their electronic ground state. The two lowest states, $^2\\Pi$ and $^2\\Sigma$, are close in energy (Pauzat et al.\\ 1991). Pauzat and co-workers predicted from unrestricted Hartree-Fock calculations  that the $^2\\Pi$ state  of C$_5$N lay below the $^2\\Sigma$ state and that $\\mu_0$  was  very small. The line strengths scaling with $\\mu_0^2$,  it was no wonder that C$_5$N  escaped detection.  Things changed when Botschwina (1996) showed from more elaborate coupled cluster calculations that the  C$_5$N ground state was in fact $^2\\Sigma$ and that $\\mu_0$ was as large as 3.385~D, even larger than the dipole moment of C$_3$N. It then became clear that C$_5$N could be detected at least in the laboratory. Kasai et al.\\ (1997) succeeded to synthetize this species, to measure its microwave spectrum, and to derive  its rotational, fine and hyperfine constants, making at last possible a sensitive search for C$_5$N in space. In this Letter, we report the astronomical detection of this radical. ",
        "conclusions": ""
    },
    "9805/astro-ph9805275_arXiv.txt": {
        "abstract": "Hypersonic bipolar outflows are a ubiquitous phenomena associated with both young and highly evolved stars. Observations of Planetary Nebulae, the nebulae surrounding Luminous Blue Variables such as $\\eta$ Carinae, Wolf Rayet bubbles, the circumstellar environment of SN 1987A and Young Stellar Objects all revealed high velocity outflows with a wide range of shapes. In this paper I review the current state of our theoretical understanding of these outflows.  Beginning with Planetary Nebulae considerable progress has been made in understanding bipolar outflows as the result of stellar winds interacting with the circumstellar environment. In what has been called the \"Generalized Wind Blown Bubble\" (GWBB) scenario, a fast tenuous wind from the central star expands into a ambient medium with an aspherical (toroidal) density distribution.  Inertial gradients due to the gaseous torus quickly lead to an expanding prolate or bipolar shell of swept-up gas bounded by strong shock waves.  Numerical simulations of the GWBB scenario show a surprisingly rich variety of gasdynamical behavior, allowing models to recover many of the observed properties of stellar bipolar outflows including the development of collimated supersonic jets.  In this paper we review the physics behind the GWBB scenario in detail and consider its strengths and weakness.  Alternative models involving MHD processes are also examined.  Applications of these models to each of the principle classes of stellar bipolar outflow (YSO, PNe, LBV, SN87A) are then reviewed.  Outstanding issues in the study of bipolar outflows are considered as are those questions which arise when the outflows are viewed as a single class of phenomena occuring across the HR diagram. ",
        "introduction": "Bipolar outflows and highly collimated jets are nearly ubiquitous features associated with stellar mass loss.  From Young Stellar Objects (YSOs) to Luminous Blue Variables (LBVs) and Planetary Nebulae (PNe) - the stellar cradle to the grave - there exists clear evidence for collimated gaseous flows in the form of narrow high velocity streams or extended bipolar lobes (Fig 1.1).  In YSOs, LBVs and PNe these collimated hypersonic outflows are observed to transport prodigious amounts of energy and momentum from their central stars - enough to constitute a significant fraction of the budgets for the entire system. Thus outflows and jets are likely to play a significant role in the evolution of their parent stars. It is remarkable that such different objects, separated by billions of years of evolution and decades of solar mass, should drive phenomena so similar.  The similarity of jets and bipolar outflows across the H-R diagram must tell us something fundamental and quite general about the nature of stellar evolution as well as the interaction of stars with their environments.  The purpose of this paper is to review our current understanding of bipolar outflows.  As we will see there is a unique synergy between theory and observations in this field providing a window into fundamental processes such as shocks, instabilities, ionization dynamics and chemistry. These systems are, however, more than astrophysical laboratories.  The tension between an extensive multi-wavelength database and increasingly sophisticated theoretical tools allows bipolar outflows to act as an Archimedean lever yielding insights directly into the birth and death of stars.  Bipolar outflows consitute an exciting field of study with rapidly expanding frontiers.  The goal of this review is to act as a tutorial in bipolar outflow studies as well as focusing attention on particularly pressing unanswered questions.  \\subsection{Observational Background} Remarkable progress in understanding bipolar outflows has been achieved in the last two decades.  Observationally, the triad of morphologic, kinematic and spectroscopic studies have provided detailed portraits of bipolar outflows both individually and as a class of astrophysical object.  Morphological studies using both the HST and ground based instruments reveal bipolar outflows assuming a wide variety of large scale ($L \\approx 10^{17} ~cm$) {\\it global} configurations.  Spherical and elliptical outflows are observed in evolved systems such as PNe, Wolf-Rayet nebulae and LBVs (Schwarz \\ea 1992; Marstson \\ea 1994; Nota \\ea 1995).  True bipolar outflows, which appear as two opposing lobes joined at narrow waist (centered on the star) occur in both young and evolved systems.  The bipolar lobes exhibit different degrees of collimation ranging from wide figure {\\bf 8} shapes to long, narrow jets. Fig 1.2 shows a sample of PNe and underscores the extraordinary diversity of outflow shapes. The appearence of {\\it point symmetry}, where all features are reflected across a central point as in an {\\bf S} (NGC 5307, Fig 1.2 lower right), is a particularly intriguing feature to emerge from recent observations. Morphological studies also reveal an array of small scale features ($L \\le 10^{16} ~cm$) which go by a variety of suggestive labels: ansae, knots, arcs, and cometary globules (Balick \\ea 1994, O'Dell \\& Handron 1996).  Kinematical studies of bipolar outflows (using long slit echelle or Fabret-Perot observations) reveal flow patterns that simultaneously possess high degrees of symmetry and complexity (Bryce \\ea 1997).  The most important point for this review is the clear presence of {\\it globally axisymmetric expansion patterns} (Corradi \\& Schwartz 1992b, Lada \\& Fitch 1996).  The velocities of the bipolar lobes are larger than local sound speeds indicating that outflow physics will be dominated by processes inherent to hypersonic shock waves.  Spectroscopic studies of line emission provide detailed snapshots of the microphysical state in the outflows.  Ionization of the lobes can occur in different ways. Low mass young stars are too cool to produce significant ionizing flux so bipolar outflows associated with these stars can only be collisionally ionized by shocks. Outflows from evolved stars and those associated with high mass young stars are, however, often photoionized by strong stellar UV fluxes.  In addition to ions, most classes of bipolar outflow show some evidence for the presence of molecules and molecular chemistry (Latter \\ea 1995, Bachiller \\& Gutie\\'rrez 1997).  When spectroscopic studies of ionic and molecular transitions are coupled with shock emission/photoionization models, physical conditions in the outflows (density, temperature, ionization and molecular fractions) can be determined (Hartigan, Morse \\& Raymond 1993, Balick \\ea 1994, Dopita 1997).  \\subsection{Theoretical Background}  Rapid progress has also been made in the theory of bipolar outflows. Analytical and numerical studies have recovered many of the observed features of outflows through what we will call the Generalized Wind Blown Bubble (GWBB) paradigm.  In this scenario a fast wind from the central source expands into a strongly aspherical (toroidal) environment. The interaction of the wind and environment produces an expanding bubble bounded by strong shocks. The bubble's velocity is highest in the direction of lowest density. Thus it is the density gradient in the environment which establishes a preferred axis for the bipolar lobes. We will explore this mechanism in considerable detail in section 2.2 and 2.3.  Here we simply emphasize that this theoretical picture has been remarkably successful at explaining the properties of bipolar outflows in evolved stars.  It has also found some success in explaining the properties of bipolar outflows in YSOs but there the issues are more complex.  One aspect of YSOs which muddies the waters is the presence of both highly collimated jets and wider bipolar molecular outflows. The GWBB model provides a robust means of collimating jets (discussed in section 2.4 and 2.5) and may, therefore, be relevant to both the manifestations of outflows in YSOs.  The GWBB paradigm has been applied to almost all forms of bipolar outflows. Models which include both hydrodynamics and microphysics have been able to recover the global morphology, kinematics and ionization patterns in many PNe (Frank \\& Mellema 1994b) and SN87A (Martin \\& Arnett 1995). Strong correspondences also exist between GWBB models and the shapes and kinematics of WR nebulae (Garcia-Seguria \\& MacLow 1993), LBVs like $\\eta$ Carinia (Frank, Balick \\& Davidson 1995) and symbiotic stars like R Aquarii (Henny \\& Dyson 1992). In section 3 we will discuss the applications of the GWBB theory in more detail including its relevance to YSOs.  The GWBB scenario has enjoyed enough success in describing outflows from evolved stars that, in this regime at least, one can speak of a {\\it classical} GWBB model. Numerous variations on its basic theme have been already explored. In an observational science familiarity can breed contempt and the limits of the GWBB scenario are already being revealed. New observational aspects of evolved star bipolar outflows such as point-symmetry have been discovered which the GWBB scenario may have difficulty recovering.  Models invoking MHD processes have been proposed which may be capable of addressing these issues better than the GWBB model (Chevalier \\& Luo 1994). We will discuss these models in section 2.8.  MHD processes are also believed to produce YSO jets via accretion disks. These models may be relevant for PNe (Soker \\& Livio 1994) but may not be as useful for LBVs, WR bubbles and SN87A-type outflows (cf Washimi \\ea 1996).    \\subsection{Teleology} While bipolar outflows offer an unparalleled view into many important physical processes the ultimate purpose of their study is to understand stellar evolution.  The details of stellar birth and death are often hidden from view, shrouded behind dense dusty circumstellar veils.  Bipolar outflows and jets are, on the other hand, easily observed and have dynamical ages long enough to span important evolutionary transitions in the life of a star. Thus, bipolar outflows contain a ``fossil'' record of the central star's history. The simultaneous development of high resolution observational tools and detailed numerical models has allowed unprecedented contact between theory and reality in bipolar outflow studies.  Thus there exists the possibility of reading the history of an individual star or class of stars from the outflows.  This is the most exciting aspect of bipolar outflow studies however currently it remains in its infancy.  In section 4 we will consider what is required to build accurate stellar wind paleontological investigations.  A note on definitions is warranted before we venture any further.  In what follows we explore bipolar outflows primarily as a {\\it hydrodynamic} phenomena.  While we focus on jets and jet collimation we approach this issue in the context of the collimation of bipolar outflows as a whole.  I am, therefore, interested in systems in which both jets and bipolar outflows arise and examine the question: can both these outflow phenomena be driven by the same underlying hydrodynamic collimation process.  There are a number of excellent reviews of MHD collimation of jets (Pudritz 1991, Shu 1997). Here, however, the focus is first on bipolar outflows and second on jets.  Thus I will not address the relativistic beams associated with neutron stars and black holes. It is quite possible that the hydrodynamic mechanisms discussed in this review will have some relevance to those systems as well (Eulderink \\& Mellema 1994). ",
        "conclusions": "The studies reviewed in this paper demonstrate bipolar outflow formation is a common process associated with stellar evolution. Moreover, the similarieties between structural components of the outflows from low and high mass, evolved and young stars would seem to imply that similar processes shape them all. Such an implication is still open to debate however and it is useful to briefly consider what outflows at different ends of the evolution and mass spectrum have in common.  Bipolar outflows from evolved stars share many properties.  Wide-lobed bipolar Planetary Nebulae (PNe), Luminous Blue Variables (LBVs) nebulae and Supernova 1987A are quite similar in both their kinematics and shape.  Given what is already understood about mass loss from these stars the Generalized Wind Blown Bubble paradigm seems an approprate approach for describing the formation of bipolar outflows in many these stars. The existence of non-axisymmetric features in some bubbles however argues that this model requires either additional inputs (ionization effects, clumpy flows Hartquist \\& Dyson 1996, etc.) or a move to MHD scearios.  The relation between bipolar outflows in evolved stars and Young Stellar Objects is more complicated.  In YSOs the outflows are most likely associated with accretion disks. While low mass stars would not be able to produce any form of outflow without a strong magnetic field, high mass YSOs may not require magnetic fields to launch or collimate winds. The presence of long narrow jets {\\it and} bipolar outflows presents a twin challenge for theories which might link evolved and young stellar systems.  As we have seen GWBB models can produce both extended outflows and narrow jets but it is not clear if it can produce both simultaneously.  Recent studies of PNe and Proto-PNe revealing the presence of jets in evolved stars systems offer the hope of finding links betweeb evolved staroutflows and YSOs.  PNe jets often bear strinking resembelnce to those in YSO often appearing as smaller, shorter versions HH systems and displaying similar point-symmetric shapes.  For some of these systems hydrodynamic collimation models may work quite well (Dopita 1996). In others the Magnetized Wind Bubble scenario may be more attractive alternative.  The suggestion that accreation disks form in binary AGB systems (Livio \\& Soker 1994) provides yet another potential link between evolved and YSO bipolar outflows.  Along with physical processes different classes of bipolar outflows are connected by the tools and methodology used to study them.  Both observational diagnostics/techniques and numerical models can often be readily adapted between different classes of outflow. In some cases the types of questions that can be asked are identical.  For example the study of LBV nebula has led to questions about the geometry of the fast stellar wind which can be answered by examination of the mass and momentum distributions in the bipolar lobes. Such a procedure was used by Masson \\& Chernin 1992 in their examined of the WBB model of YSO molecular outflows.  \\noindent{\\bf Stellar Wind Palentology}  Given their ubiquity what do bipolar outflows teach us about stellar evolution?  One of the most compelling reasons for studying outflows is the stellar history encoded in their structure. The outflows trace the history of mass loss in the star. For systems where the succesive wind phases form extended circumstellar evelopes, the bipolar outflows can provide clues to evolutionary changes in $\\dot{M}$ and $V_w$.  For some systems the structure of the outflows can reveal properties of the pre-existing environment.  Thus the outflows can teach us about the history and physics of the stars themselves.  The high quality of observational data available combined with the variety of ``lookback'' timescales ($t_d$, equation \\ref{bas}) inherent to the outflows  ($t_d \\sim 10^2 -  10^5 ~y$) offers the possibilty that key transitions in an individual star's history might be recovered if we learn where and how to look. The use of collimated outflows as fossils can succeed if researchers can identify issues which provide a reasonably unique and unambiguous bridges between nebular stucture/dynamics and stellar evolution.  The development of  nebular palentological studies is still in its infancy. Researchers have, however, learned enough that one can point to examples of what such a investigation might look like. SN87A and $\\eta$ Carinae are both unique objects.  In each case the intense scrutiny the outflows have received has allowed their histories and the history of the central stars to be specfied in some detail.  There are so many PNe and YSOs that a similar critical mass of data has not accumulated around many of these objects as of yet.  The nebula surrounding SN87A is a particularly good example to consider. The properties of the central ring are known with enough accuracy that GWBB models can place strong constraints on the mass loss rates of both the Blue Supergiant and Red Supergaint winds. Assuming that the upper and lower rings are associated with the bipolar lobes then the geometry of Red Supergaint wind and the equator to pole density contrast can also be constrained.  This, in turn, leads to some limitation on the stellar rotation rate and the potential properties of a binary companion.  New studies of the extended circumstellar environment reveal a mild aspherical geometry on with lookback timescales of $t_d \\sim 10^5 ~y$ (Crotts \\& Heathcote 1997). It is possible that these data may provide futher evidence of evolution of rotation and binary interactions.  While many of the conclusions rely on the application of specfic models (GWBB, Wind Compressed Disks), as research progresses one would expect alternative models to be explored with as much detail.  The studies of $\\eta$ Carinae have also allowed the history of the enigmatic central star to be sketched out.  The expansion patterns of the lobes clearly link them to the outburst of 1843.  The attention focused on the equatorial ejecta however allows very different models for the history and geometry of the pre-outburst, outburst, and post-outburst winds to examined in detail.  It should also be possible to link the strutures which mottle the surface of the Homunculus' to instabilites dependant upon the the form of $\\dot{M}(t)$ and $V_w(t)$  With these examples it is not unresonable to expect that progress in bipolar outflows studies will someday allow the evolution of fundemental stellar properties to be read off the outflows.  The effect of mass loss, rotation, magnetic fields, and binary companions on stellar birth and death are all questions of fundemental importance. The rapid progress being made in bipolar outflow studies will produce insights not only into hydrodynamic and hydromagnetic phenomena in general but also provide direct links from the physics of these nebulae to the properties of the stellar sources.  Taken together these prospects will make the study of bipolar outflows an exciting field for years to come.  \\noindent{\\bf Acknowledgments} There are so many people who have contributed in to this review through interesting and helpful discussions that I do not have space here to acknowledge them all. I would however like to thank Bruce Balick, Garrelt Mellema, Vincent Icke, Mario Livio, Tom Jones, Dongsu Ryu, Jack Thomas, John Dyson, Mordecai-Mark Maclow, Guy Delamarter and Tom Gardner for their help. Support for this work was provided at the University of Rochester by NSF grant AST-9702484 and the Laboratory for Laser Energetics."
    },
    "9805/astro-ph9805043_arXiv.txt": {
        "abstract": "We report results of analysis of the ASCA observation of 1993, May 7th of the dipping LMXB source XBT\\thinspace 0748-676, and propose a new explanation of the spectral evolution in dipping in this source. The behavior of the source was very unusual in that, in the band 1 - 3 keV, dipping extended around most of the orbital cycle with almost no non-dip intensity evident and the depth of dipping reached 100\\%. At higher energies, eg 3 - 10 keV, the depth of dipping was less than 100\\%, and there were marked increases in hardness in dipping. We show that the non-dip and dip spectra in several intensity bands are well fitted using the same physical model that we have previously shown gives good explanations of several dipping sources, consisting of point source blackbody emission from the neutron star, plus extended Comptonised emission from the accretion disk corona (ADC), with progressive covering of the ADC during dipping. Best fit values of $\\rm {kT_{bb}}$ = 1.99$\\pm $0.16 keV and power law photon index $\\Gamma $ = 1.70$ \\pm $0.16 are found. The strong excess below 1 keV was well fitted by a Gaussian line at 0.65 keV. In dipping, good fits were obtained by allowing it to be covered by the same progressive covering factor as the extended continuum emission, providing strong evidence that the line originates in the ADC. Our approach of applying the two-component model, and explicitly including progressive covering of the Comptonised emission, differs radically from the ``absorbed plus unabsorbed'' approach previously used extensively for XBT\\thinspace 0748-676 and similar sources, in which the normalisation of the unabsorbed peak in dip spectra is allowed to decrease by a large factor in dipping. This decrease has often been attributed to the effects of electron scattering. By using our two-component model we show that the unabsorbed component is the uncovered fraction of the Comptonised emission, and in the band 1 - 10 keV, we do not need to invoke electron scattering to explain dipping. ",
        "introduction": "There are $\\sim $ 10 Low Mass X-ray Binaries (LMXB) that exhibit absorption dips in X-ray intensity at the orbital period, and it is generally accepted that these are due to absorption in the bulge in the outer accretion disk where the accretion flow from the Companion impacts (White and Swank 1982). XBT\\thinspace 0748-676 is an important member of the dipping class, discovered and observed extensively with {\\t Exosat}, exhibiting dipping, bursting and eclipsing (Parmar et al. 1986). Since then, it has undergone strong variations in brightness. The {\\it Exosat} observations showed that both dips and interdips were present, and that the depth and extent of dipping varied markedly from orbit to orbit. Eclipses showed that the inclination was high, and during these, about 4\\% of the non-dip emission in the band 2 - 6 keV remained, indicating the presence of an extended emission component. Using the eclipses as fiducial markers, Parmar et al. (1986, 1991) found a very accurate value for the orbital period and a decreasing period. Asai et al. (1992) and Corbet et al. (1994) showed evidence that the period is modulated sunusoidally, but this pattern was not continued when RXTE data was added (Hertz et al. 1997). However all determinations agree on the period to $\\sim $ 0.01~s, ie 13766.78~s.  \\medskip \\noindent The dipping sources do not, in general, have the spectral evolution expected for photoelectric absorption of a single emission component in the bulge in the outer accretion disk, ie a strong hardening of the spectrum. For example, X\\thinspace 1624-490 shows a softening of the spectrun (Church and Balucinska-Church 1995), and  X\\thinspace 1755-338 shows energy independence (White et al. 1984; Church and Balucinska-Church 1993). We have previously proposed a physical model for the dipping sources which explains the spectral evolution in these two sources (Church and Balucinska-Church, 1995). In this model, emission consists of two components: blackbody emission from the neutron star plus extended Comptonised emission from the accretion disk corona (ADC) represented as a power law at energies well below the Comptonisation break. X\\thinspace 1624-490 is also important because dipping reaches a ``saturated'' lower level which is strong evidence that two emission components are present (Church and Balucinska-Church 1995). The differences between the two sources above are mainly due to differences in blackbody temperature. In X\\thinspace 1624-490, $\\rm {kT_{bb}}$ is relatively high, and as dipping consists predominantly of absorption of the point-like blackbody emission, this leaves the residual spectrum softer. In X\\thinspace 1755-338, $\\rm {kT_{bb}}$ is smaller, with the peak emission at $\\sim $ 3 keV, so that removal of the blackbody leaves the spectrum neither harder nor softer in the band 1 - 10 keV.  \\begin{figure*}[t] \\leavevmode\\epsffile{fig1.eps} \\caption{ASCA GIS2 light curve for the complete 23~hr observation in the energy band 0.7 - 10.0 keV with 32~s binning. To show the quiescent source more clearly, the full heights of the bursts (B) have been suppressed. Eclipses are labelled (E).\\label{fig1}} \\end{figure*}  \\medskip \\noindent There is also an important group of dipping sources including XB\\thinspace 0748-676, XB\\thinspace 1916-053 and XB\\thinspace 1254-690 which have very similar spectral evolution in dipping, in that there is clearly part of the spectrum which is not absorbed. These sources also, do not fit a simple pattern of photoelectric absorption in the bulge in the outer accretion disk. Spectral evolution in dipping in these sources has been modelled using the ``absorbed plus unabsorbed'' approach (Parmar et al. 1986, Smale et al. 1988 and Courvoisier et al. 1986). In the case of XBT\\thinspace 0748-676, the non-dip spectrum was fitted with a cut-off power law model (Parmar et al. 1986). Dip data selected in intensity bands however, revealed a strong low energy excess, and the spectra could be fitted by dividing the non-dip model into two parts with the same form, one strongly absorbed and the other not absorbed. Good fits were obtained with the column density of the unabsorbed part fixed at the non-dip value and but with strongly decreasing normalisation as dipping progresses, whereas the absorbed component had increasing column density but with a normalisation remaining at about the non-dip value. The origin of the unabsorbed component has often in this group of sources been attributed to electron scattering producing an energy-independent decrease of intensity, although it is not clear that electron scattering should play an important role in the bulge in the outer accretion disk in competition with photoelectric absorption. In the case of XBT\\thinspace 0748-676, Parmar et al. (1986) concluded that the unabsorbed component was probably due to fast variations of column density which could mimic a soft excess.   \\medskip \\noindent In the case of XB\\thinspace 1916-053, we have shown that the ASCA spectra may be fitted well by our two-component model including progressive covering of the extended Comptonised emission (Church et al. 1997). The blackbody emission from the neutron star was covered rapidly. Very good fits were obtained in this way, with large increases in column density $\\rm {N_H}$ for the blackbody and smaller increases in $\\rm {N_H}$ for the Comptonised emission, which are averages across the absorber. The unabsorbed peak in the spectrum is the uncovered part of the power law component, and the increase in the partial covering fraction from zero to unity in deepest dipping is the behavior represented by the decreasing normalisation in ``absorbed + unabsorbed'' modelling. This is however radically different from that type of modelling. Firstly, we are explicitly using partial covering of extended emission in the model which has not previously been done. Secondly, the emission model for the source has two terms, and is the same model that we have shown gives good explanations of other dipping sources. Dipping can be modelled well by photoelectric absorption alone, without the need for electron scattering, and calculations from the cross sections showed that scattering is not expected to be important in the absorbing bulge in competition with absorption below 10 keV (Church et al. 1997). The main aim of the present work was to test using the ASCA data whether this approach can also successfully describe spectral evolution in dipping in XBT\\thinspace 0748-676. ",
        "conclusions": "The two-component model which we have previously proposed for the dipping sources has been shown to fit the source XBT\\thinspace 0748-676. In addition, we have shown that the spectral evolution in dipping can be explained in the same way as we did for XB\\thinspace 1916-053, by allowing the extended Comptonised component to be progressively covered by the absorber. In a separate analysis of the PV phase data for XBT\\thinspace 0748-676, Thomas et al. (1997) use the absorbed + unabsorbed approach for dip data, and find for non-dip emission that a single component generalised thermal model is `preferred from selected models'; however they do not report results using our two-component model, so their results should be seen in this light. Our spectral fitting results give a total absorbed flux of $\\rm {1.94\\cdot 10^{-10}}$ erg $\\rm {cm^{-2}}$ $\\rm {s^{-1}}$ in the band 1 - 10 keV, of which the blackbody contribution is 29.9\\%. The luminosity of the blackbody emission calculated from the unabsorbed flux in the band 0.01 to 50 keV, and assuming a distance of 10 kpc (Gottwald et al. 1986) is $\\rm {9.5\\cdot 10^{35}}$ erg $\\rm {s^{-1}}$, and from this, we can calculate the height of the emitting region assuming it is an equatorial strip around the star. This is found to be very small: 0.09 km, indicative of a very thin accretion disk, consistent with the fading of the source by a factor of about 30 from the {\\it Rosat} observation of 1991 Nov 11. In the case that the magnetic field strength of the neutron star were non-zero, the emitting area at the magnetic poles would still be very small: 5.8 $\\rm {km^2}$.  The duration of eclipse or dip ingress and egress can in principle be used to derive the size of the extended source region, ie the ADC. However, in the present observation the source was faint, much fainter than during the {\\it Exosat} observations, and we have only two dips that are well covered and 2 eclipses, compared with the 9 eclipses used by Parmar et al. (1986) to obtain the duration of eclipse ingress. The count rate of the source in {\\it RXTE} was also much higher than in {\\it ASCA} (Hertz et al. 1997). In these circumstances we can only estimate the size of the ADC. Using individual dips, we get an approximate duration of ingress $\\rm {\\Delta t}$ of about 110~s. In the present case, the angular size of the absorber is known to be greater than that of the source because dipping is 100\\% deep and so the duration of dip ingress is the time taken for the bulge in the outer disk to cross the diameter of the ADC. We estimate the radius of the disk as $\\rm {3.4\\cdot 10^{10}}$~cm, and the velocity of the bulge is $\\rm {2\\pi r_{disk}/P}$, where P is the orbital period. From these we derive the radius of the ADC to be $\\rm {8.5\\cdot 10^{8}}$ cm. From the 2 eclipses, we find that the duration of ingress is between 16~s and 64~s. Combining these with a binary separation of the stars of $\\rm {1\\cdot 10^{11}}$ cm, we get a radius of the ADC of between $\\rm{3.9\\cdot 10^8}$ cm and $\\rm {1.5\\cdot 10^9}$ cm. This can be compared with the results from {\\it Exosat} as follows. Fig.~3 of Parmar et al. (1986) shows the profile of 9 eclipses added by folding on the orbital period, having a steep transition to eclipse and also shoulders to the transition. The total duration of ingress can be seen to be about 30~s, equivalent to a $\\rm {r_{ADC}}$ of $\\rm {7.2\\cdot 10^8}$ cm. Our eclipse ingress times are consistent with their value. Thus our dip and eclipse data suggest a value of $\\rm {r_{ADC}}$ between $\\rm{3.9\\cdot 10^8}$ cm and $\\rm{1.5\\cdot 10^9}$ cm, consistent with an extended region above the inner part of the accretion disk.  The spectrum of XBT\\thinspace 0748-676 is clearly dominated by non-thermal emission which is identified with Comptonised emission of the ADC, and we can derive some information about the ADC from our results. Firstly, the value of the Comptonisation y-parameter is found to be $\\sim $ 1.5 from our power law index of 1.7 (eg Rybicki and Lightman 1979). There is no evidence (or expectation) for the Comptonisation break being less than 10 keV; however we can assume that the break energy is similar to that we have determined for the related source XB\\thinspace 1916-053 using the very broad band of 0.1 - 300 keV of BeppoSAX (Church et al. 1998), ie $\\sim $ 80 keV. From this value, the electron temperature $\\rm {kT_e}$ is $\\sim $ 30 keV, from which we derive values in XBT\\thinspace 0748-676 of the average optical depth and column density in the ADC of 2.5 and $\\rm {4\\cdot 10^{24}}$ $\\rm {cm^{-2}}$. We have also shown that the low energy feature can be fitted by a Gaussian line at 0.65 keV in non-dip and dip spectra, and in addition have shown that it originates in the ADC, since it can be fitted by the same progressive factor as the extended continuum emission. Future identification of this line, perhaps as O VIII L$\\alpha$ will, of course, provide direct information on conditions in the ADC. High energy data is clearly required for XBT\\thinspace 0748-676, together with higher resolution data below 1 keV to reveal the nature of the line emission.  The explanation of spectral evolution in XBT\\thinspace 0748-676 is radically different from previously modelling of the source using the absorbed + unabsorbed approach. Previously partial covering has not explicitly been used in modelling, and fitting our two-component model including partial covering of the extended component shows that the unabsorbed emission can be identified with uncovered non-dip emission. Gradual covering is represented in absorbed + unsabsorbed modelling by the decreasing normalisation often attributed to electron scattering. We do not need to invoke electron scattering in the absorber, so that dipping can be explained by photoelectric absorption alone. In the case of XB\\thinspace 1916-053 we presented calculations showing that electron scattering is not expected to be important in the bulge in the outer disk at energies below 10 keV (Church et al. 1997). In XBT\\thinspace 0748-676, the density increase in dipping is less, and so the effects of scattering will be even smaller."
    },
    "9805/astro-ph9805018_arXiv.txt": {
        "abstract": "Interstellar absorption in the galactic plane is highly variable from one direction to another. In this paper colour excesses and distances from a new  open cluster sample are used to investigate the spatial distribution of the interstellar extinction. An inverse method (Tarantola \\& Valette, 1982) is used to construct the extinction map in the galactic plane below $|b| < 10^{o}$. The $A_{v} (r,l) $ diagrams are  compared with those derived from individual stars (Arenou et al. 1992, Neckel \\& Klare 1980). An analytic expression for the interstellar extinction as a function of galactic longitude and distance in the solar neighborhood  is given. The comparison of the model predictions with Hipparcos observations in the 4-dimensional space of ($V$, $B-V$, $H_v$, $r$) shows that our extinction model provides a better fit to the data. However, a new and more detailed extinction model is still lacking. ",
        "introduction": "The light extinction caused  by the interstellar medium in the solar neighborhood alters the apparent brightness of the stars and thus their intrinsic properties (absolute magnitude, intrinsic colour, etc.). For this reason, it is crucial to know the effects of the interstellar medium on the radiation we receive from the stars, and to build models up which reasonably predict the amount of extinction in a given direction.   Bahcall \\& Soneira (1980)  have used Sandage's extinction law (1972) in their model of the Galaxy, and compared model predictions with star count observations in 17 fields. They found that,  at $\\mid$ b $\\mid$ $>$ 10$^{o}$, the  Sandage absorption law is a good approximation to observations. However, at low galactic latitudes, the absorptions are patchy and poorly  known. In the context of our study on stellar kinematics from Hipparcos observations and open cluster (Chen et al, 1998), we need a reliable extinction model at low galactic latitudes.  Neckel \\& Klare (1980)  derived extinctions and distances for more than 11000 stars and  investigated the spatial distribution of the interstellar extinction at $\\mid$ b $\\mid$  $<$ 7.6$^{o}$. Their results are shown in a series of diagrams. Using a larger sample (about 17000 stars) with MK spectral types and photoelectric photometry, Arenou et al. (1992) (hereafter AGG92) constructed an extinction model in which the sky was divided into 199 cells. In each cell, a quadratic relation was adopted. This extinction model will be the reference in this paper. Hakkila et al. (1997) have developed a computerized model of large-scale visual interstellar extinction. The code merges several published studies (FitzGerald 1968; Neckel \\& Klare 1980; Berdnikov \\& Pavlovskaya 1991; AGG92; etc.), and can be used for making corrections to individual observations and for correcting statistical sample. With extinctions determined from Str\\\"{o}mgren photometry and parallaxes from Hipparcos for a sample of 3799 stars, Vergely et al (1997) have constructed a three dimension  distribution of absorbing matter in the solar neighbourhood. However, all these authors use similar dataset (stars along with their corresponding colors,  magnitudes, and spectral/morphological classification types) so that the results are not statistically independent (Hakkila et al. 1997).   In this paper, we are going to use open clusters rather than individual stars to investigate the distribution of interstellar extinction, and compare our results with those  found in the literature. The reason for this investigation is twofold. First, the advent of CCDs with their sensitivity and linearity has allowed to obtain deep photometry of open clusters (Carraro \\& Ortolani, 1994, Kaluzny 1994). Second, distances and reddening for open clusters are more reliable than those from individual stars, simply because one is dealing with a group of stars and therefore the results are less sensitive to individual errors. Thus the open clusters system provides us with another way to obtain informations on A$_{v}$ as a function of the galactic longitude $l$ and distance $r$.   Pandey \\& Mahra (1987) firstly  used the open cluster catalogue of Janes \\& Adler (1982) to investigate the spatial distribution of interstellar extinction within 2 kpc. Their results are presented in graphical form. No analytical expression for the interstellar extinction has been presented. In this paper, we have used a new open cluster database and an inverse method (Tarantola \\& Valette, 1982) to derive an analytical expression of the interstellar extinction in the solar vicinity, and compared different extinction models with Hipparcos observations. In section 2 we present the data used in this paper. In section 3 we briefly explain the AGG92's model and compare it with our open cluster data. In section 4, we construct a new extinction model from open cluster system. In section 5 the predictions from AGG92's extinction model and our extinction model are compared with the Hipparcos data. The main conclusions are summarized in section 6. ",
        "conclusions": "At low galactic latitudes ($\\mid$ b $\\mid$ $<$ 10$^{o}$), the absorptions are patchy and not well known. In this paper colour excesses and distances of open clusters compiled by Mermilliod (1992) have been used to study the interstellar extinction. We have compared our extinction model with AGG92 and Hipparcos observations. The main results of this paper can be summarized as follows.  1. An inverse method (Tarantola \\& Valette, 1982)  was used to construct the extinction map in the galactic plane. We found the lowest absorption between $l$ = 210$^{o}$ to 240$^{o}$, and the highest absorption between $l$ = 20$^{o}$ to 40$^{o}$.   2. By using three different samples,  we found that AGG92's results are higher than the observations for small distances.  3. We have constructed an extinction law by combining Sandage extinction law with $\\mid$ b $\\mid$ $>$ 10$^{o}$ and our extinction law in the galactic plane.  4. AGG92 and our extinction models have been compared with a well selected sample from Hipparcos observations with Str\\\"{o}mgren photometry. We found that AGG92's model provides a slight overestimation of A$_{v}$.   5.  Galaxy model predictions with our and AGG92's extinction laws have been compared with the Hipparcos observations for the whole sky and at low galactic latitudes, respectively. The main characteristics in the observed distributions (for example, two peaks in $B-V$) are well predicted by both models. The $\\chi^2$ statistics, used to choose the best fit to Hipparcos observations, shows that our extinction model provides a better fit to data.  6. Interstellar absorption in the galactic plane is highly variable from one direction to another. We believe that a more detailed absorption law, which can represent the extinction at smaller scales, is needed for modeling the extinction  substructure in the galactic plane.   These results are very useful for us to  investigate the stellar kinematics in the solar neighborhood from Hipparcos and Tycho data in the near future. But a new and more detailed extinction model from Hipparcos observations is still missing."
    },
    "9805/astro-ph9805197_arXiv.txt": {
        "abstract": "Galaxies in the Las Campanas Redshift Survey are classified according to their spectra, and the resulting spectral types are analyzed to determine if local environment affects their properties.  We find that the luminosity function of early-type objects varies as a function of local density. Our results suggest that early-type galaxies (presumably ellipticals and S0's) are, on average, fainter when they are located in high-density regions of the Universe. The same effect may operate for some, but not all, late-type galaxies. We discuss the implications of this result for theories of galaxy formation and evolution. ",
        "introduction": "\\label{sect:intro}  As a snap-shot of a dynamic, evolving system, the present-day distribution of galaxies offers strong constraints for theories of structure formation. One of the best established of these, the density-morphology relation (Dressler~\\cite{Dre80}), quantifies the extent to which the mix of galaxy types is dependent on the local environment.  It is likely that the principal process responsible for this relation is the merging of galaxy pairs: Elliptical galaxies lie preferentially in high-density regions because, there, they had access to an abundant supply of objects with which to merge.  Any galaxy-formation scenario must also account for the observations of galaxies at earlier epochs (e.g., Ellis \\cite{Ell98}).  For example, a population of late-type galaxies observed at $z \\approx 1$ has no apparent counterpart among galaxies seen in the present epoch (Broadhurst, Ellis \\& Shanks~\\etal\\ \\cite{BroEtal88}). Conversely, some early-type (S0) objects are relatively abundant at present, but similar objects, or even obvious progenitors, are less prevalent at a redshift near unity (Dressler~\\etal\\ \\cite{DreEtal97}).  It may not be simple to connect galaxies of the distant past and those of the present epoch with a consistent model of cosmic structure formation.  The purpose of this note is to report an effect which adds to the growing list of requirements for a successful theory of galaxy formation. Specifically, in a study of a population of galaxies in the present epoch, we find that the luminosity function of some galaxy types depends on local density: Early-type galaxies tend to be fainter when they are located in dense environments.  This result is based on an analysis of galaxies in the Las Campanas Redshift Survey (Shectman~\\etal\\ \\cite{SheEtal96}), the parameters of which are briefly outlined here in \\S 2.  We use a spectral classification scheme and a simple partition of galaxies into low- and high-density subsamples to evaluate the environmental dependence of type-specific luminosity functions; details and specific results are given in \\S 3. We finish this note with possible interpretations of our results (\\S 4). ",
        "conclusions": "The density dependence of the faint-end slope of the type-specific luminosity functions, call it the $\\alpha$-density relation, suggests that on average galaxies of a given type are fainter in dense regions than in the field.  This behavior is qualitatively consistent with the observation that the dwarf-to-giant ratio of early-type galaxies is higher in groups than in the field (Ferguson \\& Sandage~\\cite{FerSan91}).  However, from a theoretical standpoint, this is perhaps an unexpected result.  Press-Schechter theory predicts a relative decrease in low-mass objects in overdense regions, as a result of a steeper decline of power with scale. One might assume that this prediction applies to galaxies of a single type and therein lies the unexpectedness of the $\\alpha$-density relation.  It may be that the Press-Schechter prediction is more suggestive of the change in the mix of galaxy types with density as opposed to the dependence of intrinsic properties of any one galaxy type.  Thus, to understand any sort of $\\alpha$-density relation, one may have to identify the processes which determine a galaxy's type.  Whatever these processes, the LCRS data suggest that they allow galaxies of a specific type to end up with less mass in dense regions than in the field.  Toy models which show this behavior may be easy to construct, but it may be more challenging to identify such an effect in more developed theories of structure formation or numerical simulations.  Without subscribing to specific toy models, we note a few processes which might affect the dependence of type-specific luminosity functions. The most promising is tidal stripping of an evolved galaxy or its progenitors in high-density regions. This would cause faint galaxies to become fainter, and bright galaxies to become brighter, assuming that the stripped material is transferred from one galaxy to another.  There is a hint that this might be occurring: $\\Mstar$ shifts slightly toward the bright end for galaxies in high-density regions. Alternatively, the gaseous component of the stripped material may be lost to the intergalactic or intracluster medium, and hence would not strongly affect the bright population of galaxies.  Another possible process that could effect the $\\alpha$-density relation is quenching of star formation in high-density regions. If one were to move a late-type galaxy from the field to a compact group or cluster, it is possible that a hot intergalactic (or intracluster) medium would produce sufficient $X$-radiation to reduce the star formation rate in the galaxy. This would cause the galaxy to be ``misclassified'' as an earlier type in our spectral scheme. Such a type migration would, to a crude approximation, steepen the faint-end slope of an early-type population simply by mixing early- and late-type luminosity functions.  One must be careful in associating the environmental dependence of the luminosity function with changes in the brightness of individual galaxies as predicted, for example, in the tidal stripping scenario. In particular, making individual galaxies fainter does not necessarily cause the slope of the luminosity function to rise toward the faint end. If the faint-end slope is steep to begin with, and if tidal stripping effects become more important with decreasing luminosity, then the faint-end slope could actually decrease, at least when measured over a finite range of luminosity. This might be a partial explanation for the fact that two of the late-type luminosity functions do not show a significant change in $\\alpha$.  While the effect we have identified in the LCRS data may be a useful constraint for theories of galaxy formation in its own right, we are missing a possibly crucial body of information, namely a list of galaxy morphologies.  The evidence is that morphology and spectral type are strongly correlated, but not perfectly so (Kennicutt \\cite{Ken92}; Connolly \\etal\\ \\cite{ConEtal95}; Zaritsky, Zabludoff \\& Willick \\cite{ZarZabWil95}). Differences in this correlation from high- to low-density regions might help to distinguish between the various scenarios that could produce the environmental effect reported here."
    },
    "9805/astro-ph9805368_arXiv.txt": {
        "abstract": " ",
        "introduction": "HII galaxies are dwarf galaxies in a bursting phase of star formation of low luminosity (mass), low heavy element abundance and low dust content where the triggering mechanism of the present episode of violent star formation is not so obvious \\cite{tt}.   HII galaxies are less clustered than normal bright galaxies and tend to populate regions of low galactic density \\cite{ims}\\cite{cm}\\cite{cmm}\\cite{pults}\\cite{rsm}\\cite{tt}\\cite{v}. They are not associated with giant galaxies, therefore, HII galaxies are not tidal debris of strongly interacting systems.   A possible and appealing alternative that other dwarfs or intergalactic HI clouds could be the triggerers \\cite{b}\\cite{m} was followed up by Taylor and collaborators \\cite{tbgs1}\\cite{tbgs2}\\cite{ta}.  They have used the VLA 21 cm maps to search for HI companions around HII galaxies.  The main VLA results can be summarized as follows: (i) 12 out of 21 HII galaxies have HI companions; (ii) 13 out of 17 Low Surface Brightness Galaxies (LSBG) from a control sample of quiescent galaxies do not have HI companions. As also pointed out by these authors, some questions remain intriguing from this: Why are these 9 out of the 21 HII galaxies with \\underline{no} companions violent forming stars now ('bursting')?  Why are these 4 out of the 17 LSBGs with companions {\\em not} 'bursting'?  Since most of the detected clouds in HI surveys have optical counterparts at faint levels (c.f. Hoffman and Zwaan at this meeting), we have carried out a further investigation of the galaxy environments of a unbiased sample of over 160 low redshift HII galaxies by cross-correlating their accurate position in the sky, also derived from the APM scans, to faint field galaxies ($15 < b_{\\rm J} < 20$) galaxies in the APM galaxy catalogue.  The HII galaxy sample used in this work is taken from the {\\em Spectrophotometric Catalogue of HII Galaxies} \\cite{tmmmc}. Our sample of faint field galaxies was selected from the APM Galaxy Survey, which is described in detail in \\cite{msegl}. The galaxy sample selected from the survey data at a magnitude limit of $b_{\\rm J} =20.5$ has a completeness $\\sim90$--$95\\%$, stellar contamination $\\sim 5\\%$ \\cite{mes}. ",
        "conclusions": ""
    },
    "9805/astro-ph9805124_arXiv.txt": {
        "abstract": "On January 19, 1993, a very bright bolide (peak magnitude $-23$) crossed the sky of Northern Italy, ending with an explosion approximately over the town of Lugo (Emilia Romagna, Italy).  The explosion (14 kton of energy) generated shock waves which were recorded by six local seismic stations.  A reanalysis of the available data leads us to the hypothesis that the meteoroid was a porous carbonaceous chondrite, somehow similar in constitution to the asteroid 253 Mathilde. ",
        "introduction": "The atmospheric interaction of large meteoroids provides our primary tool to characterize their population, physical and chemical properties, and dynamical evolution.  In turn, this can lead to a better understanding of the diverse populations of small bodies of the Solar System.  Currently, our knowledge is still quite limited, although, especially after the impact of comet D/Shoemaker--Levy~9 on Jupiter, the research efforts in this field have been intensified.  In particular, in 1994 the US Department of Defense made of public domain its records on energetic bolides over a time span of about twenty years (Tagliaferri et al.  \\cite{DOD}).  These data indicate that, from 1975 to 1992, there were 136 airbursts of energy greater than 1 kton, but the real number was probably at least 10 times higher, because the satellite system does not cover the entire Earth surface.  Both data and theories are required to assess the impact hazard and to understand the very bright bolides.  From this point of view, the Lugo bolide is a very interesting event, because the airburst was detected by several seismic stations.  The corresponding data allow us to characterize the meteoroid and to draw some tentative inferences about its nature and origin.  We have carried out a reanalysis of this event and we found that the data are most consistent with the hypothesis that the involved meteoroid was a porous carbonaceous chondrite, somehow similar to the asteroid 253 Mathilde. ",
        "conclusions": "The Lugo bolide has been reanalysed by taking into account only the data recorded by seismic stations.  We summarize the main inferred properties of the bolide in the following Table~\\ref{summary}.  \\begin{table}[h] \\centering \\caption{Summary on the properties of the Lugo bolide.} \\begin{tabular}{|l|r|} \\hline Apparition time (UT) & 1993 01 19 00:33:29 $\\pm 1$ s\\\\ \\hline Latitude of airburst$^{\\mathrm{a}}$ & $44.48\\degr \\pm 0.01\\degr$ N\\\\ \\hline Longitude of airburst$^{\\mathrm{a}}$ & $11.91\\degr \\pm 0.01\\degr$ E\\\\ \\hline Airburst height$^{\\mathrm{a}}$ & $30\\pm 3$ km\\\\ \\hline Explosion Energy & $14\\pm 2$ kton\\\\ \\hline Mass & $(4\\pm 1)\\cdot 10^{5}$ kg\\\\ \\hline Abs.  Visual Magnitude & $-22.7\\pm 0.5$\\\\ \\hline Velocity & $18\\pm 3$ km/s\\\\ \\hline Inclination$^{\\mathrm{b}}$ & $5.0\\degr \\pm 0.3\\degr$\\\\ \\hline Path azimuth$^{\\mathrm{a,c}}$ & $146.5\\degr \\pm 0.5\\degr$\\\\ \\hline \\end{tabular} \\label{summary} \\begin{list}{}{} \\item[$^{\\mathrm{a}}$] Calculated in Cevolani et al. (\\cite{LUGO2}).  \\item[$^{\\mathrm{b}}$] Over the horizon. \\item[$^{\\mathrm{c}}$] Clockwise from North. \\end{list} \\end{table}  We are now carrying out calculations on the orbit and the dynamical evolution this bolide, whose results will be available soon.  However, from the analysis described here it appears likely that the meteoroid was a porous carbonaceous chondrite, somehow similar in constitution to the asteroid 253 Mathilde.  The porosity would have increased the braking and as a consequence the airburst occurred at a height higher than for a compact carbonaceous chondrite object."
    },
    "9805/astro-ph9805312_arXiv.txt": {
        "abstract": "Until recently synthetic AGB models had not taken into account the break-down of the core mass-luminosity ($M_{\\rm c} - L$) relation due to the occurrence of envelope burning in the most massive ($M \\ga 3.5 M_{\\odot}$) and luminous ($M_{\\rm bol} \\la -6$) stars.  Marigo et al. (1998) made the first attempt to consistently include the related over-luminosity effect (i.e. above the $M_{\\rm c}-L$ relation) in synthetic TP-AGB calculations. The method couples complete envelope integrations with analytical prescriptions, these latter being presently updated with the highly detailed relations by Wagenhuber \\& Groenewegen (1998).  In this paper the reliability of the solution scheme is tested by comparison with the results of complete evolutionary calculations for a $7 M_{\\odot}$ AGB star undergoing envelope burning (e.g. Bl\\\"ocker \\& Sch\\\"onberner 1991).  Indeed, the method proves to be valid as it is able to reproduce with remarkable accuracy several evolutionary features of the $7 M_{\\odot}$ star (e.g. rate of brightening, luminosity evolution as a function of the core mass and envelope mass for different mass-loss prescriptions) as predicted by full AGB models.  Basing on the new solution method, we present extensive synthetic TP-AGB calculations for stars with initial masses of $3.5, \\, 4.0, \\, 4.5,$ and $5.0 \\, M_{\\odot}$, and three choices of the initial metallicity, i.e. $Z=0.019$, $Z=0.008$, and $Z=0.004$. Three values of the mixing-length parameter are used, i.e. $\\alpha=1.68, \\, 2.0, \\, 2.5$.  We investigate the dependence of envelope burning on such stellar parameters ($M$, $Z$, and $\\alpha$). The comparison between different cases gives hints on the interplay between envelope burning over-luminosity and mass loss, and related effects on TP-AGB lifetimes. ",
        "introduction": "\\label{intro} Synthetic AGB models may provide a powerful tool of investigation (Renzini \\& Voli 1981; Groenewegen \\& de Jong 1993; Marigo et al. 1996a). Evolutionary calculations can be easily carried out over the whole mass range of AGB stars and for various metallicities, and the results tested according to the adoption of various input prescriptions (e.g. analytical relations derived from full AGB models, mass-loss laws, parameters of the third dredge-up). In this way,  it is possible to explore the sensitiveness of theoretical predictions to different physical assumptions and, at the same time, to readily get an overall picture of aspects related both to single star evolution (e.g. location on the H-R diagram, maximum AGB luminosities, stellar lifetimes, initial mass-final mass relation, changes in the surface chemical composition, stellar yields) and to integrated properties of the stellar aggregates in which AGB stars are present (e.g. luminosity functions of oxygen- and carbon-rich AGB stars, contribution of evolved stars to the integrated light and to the chemical enrichment of the host galaxy).  Of course, the computational agility and flexibility typical of the synthetic approach are paid with a certain loss of details if compared to complete AGB  models. However, in order to get reliable results from synthetic analyses the degree of approximation in treating the physical processes must be good enough so that no essential feature is missed.  Actually, a clear point of inadequacy of synthetic AGB models has so far concerned the treatment of envelope burning in the most massive stars  ($M \\ga 3.5 M_{\\odot}$). In brief, the erroneous assumption common to AGB analyses performed either with the aid of convective envelope models (Scalo et al. 1975; Renzini \\& Voli 1981; Marigo et al. 1996a), or purely analytical prescriptions (Groenewegen \\& de Jong 1993) is that the quiescent surface luminosity of a TP-AGB star experiencing envelope burning still obeys the standard  $M_{\\rm c}-L$ relation, as in the case of lower mass stars ($M \\la 3.5 M_{\\odot}$).  On the contrary, over the years complete AGB calculations have clearly indicated that nuclear burning in the hottest convective envelope layers of the most massive TP-AGB stars may be a considerable energy source, making these stars depart significantly from the $M_{\\rm c}-L$ relation towards higher luminosities (Bl\\\"ocker \\& Sch\\\"onberner 1991 (hereinafter also BS91); Lattanzio 1992; Boothroyd \\& Sackmann 1992; Vassiliadis \\& Wood 1993; Bl\\\"ocker 1995 (hereinafter also B95); Wagenhuber 1996).  Indeed, the recent discovery of the over-luminosity produced by envelope burning above the $M_{\\rm c}-L$ relation has notably changed and improved our current understanding of the AGB phase, setting  important implications.  In principle, the luminosity increase of a TP-AGB star is not bounded by the classical AGB luminosity limit of $M_{\\rm bol} \\sim -7.1$, predicted by the Paczy\\'nski (1970) $M_{\\rm c}-L$ relation for a mass of the degenerate core equal to the Chandrasekhar critical value of $\\sim 1.4 M_{\\odot}$ (BS91). For this reason, the observation of AGB stars with luminosities close to $M_{\\rm bol} \\sim -7.1$ does not imply that their core masses are close to $1.4 M_{\\odot}$, a circumstance that has been until recently considered as a supporting indirect evidence for the occurrence of Type I$_{1/2}$ Supernovae, consequent to disruptive carbon ignition (see for example Wood et al. 1983).  Moreover, the maximum quiescent luminosity, $L_{\\rm max}$, attained by an AGB star with envelope burning does not coincide with the final luminosity at the so-called AGB tip, as in the case of lower mass stars following the core mass-luminosity relation. In higher mass stars the maximum of the luminosity occurs, in practice, at the onset of the superwind, and hence before the AGB tip is reached (when the envelope has been almost completely ejected and the core mass-luminosity relation is recovered).   It follows that at high luminosities, say $M_{\\rm bol} < - 6.5$, the core-mass luminosity relation cannot be longer employed in combination with the initial mass-final mass  ($M_{\\rm i}-M_{\\rm f}$) relation to estimate the age of a coeval system on the base of its brightest AGB stars (usually inferred through the steps: $L_{\\rm max} \\rightarrow M_{\\rm f} \\rightarrow M_{\\rm i} \\rightarrow$ age; see for instance the review by Iben \\& Renzini (1983)).  The over-luminosity effect is also expected to intensify the mass-loss process suffered by stars with envelope burning, so as to possibly anticipate the onset of the superwind regime. Then, this would result in a reduction of the TP-AGB lifetimes and hence of the remnant white dwarf masses. The latter consequence may concur to solve the long-standing problem related to the excess of white dwarfs more massive than $\\sim 0.7 M_{\\odot}$, as predicted by synthetic AGB models (see Bragaglia et al. (1995) for a punctual discussion of this point).  In this context, Marigo et al. (1998) first pointed out a possible solution scheme to overcome the limit of synthetic models. This paper aims at verifying the validity of the original treatment of envelope burning -- coupling the use of analytical relationships with complete envelope integrations -- by testing its capability of reproducing the results from full AGB calculations.  The general organisation of the paper is as follows. In Sect.~\\ref{method} the rationale underlying envelope integrations is briefly recalled, together with the main analytical prescriptions. In Sect.~\\ref{comp} the validity of the method is checked by comparison with full evolutionary calculations for a $7 M_{\\odot}$ star performed by BS91 and B95. Section~\\ref{synt} presents the results of synthetic TP-AGB evolutionary calculations for stars with initial masses in the range $3.5 M_{\\odot} \\le M \\le 5.0 M_{\\odot}$, and initial metallicities $Z=0.019, \\, Z=0.008,$ and $Z=0.004$. The sensitiveness of envelope burning to stellar mass, metallicity, and mixing-length parameter is discussed in Sect.~\\ref{overlres}. Some relevant quantities characterising the evolution of the model stars are tabulated in Table~\\ref{hbbtab}. Finally, Sect.~\\ref{end} contains some concluding remarks and illustrates the intents of future works. ",
        "conclusions": ""
    },
    "9805/astro-ph9805062_arXiv.txt": {
        "abstract": "We describe the early results of an investigation into the spectral characteristics of the hottest Am stars (A0 - A2) in an attempt to link the HgMn and Am classes of chemical peculiarity.  A limited sample of hot-Am stars was searched for the presence of lines from the very heavy elements platinum and mercury, as well as a search for lines from the rare-earth elements in HgMn stars. Our analysis of the strong optical platinum and mercury lines in the spectrum of the HgMn star HR 7775 has detected isotopic shifts that are different from those found at ultraviolet wavelengths and, in the case of mercury, vary with ionization stage. ",
        "introduction": "The creation of stellar groupings by spectral characteristics is an often necessary first step in being able to explore underlying physical processes. The descriminating, or classification, criteria are often chosen arbitrarily from available observational data and without regard for the physical mechanisms which create them. Thus, a single physical process may be responsible for the variation of spectral characteristics over a great range of temperature or luminosity. In some sense, the warm chemically peculiar stars may be overclassified.  Various subclasses exist which are loosely arranged by temperature or enhanced element, but they are now beginning to loose their identity and purpose as additional elements and spectral regions are explored.  We describe here the analysis procedure and early results of a project that has been undertaken to ascertain the extent of common spectral properties amongst warm chemically peculiar star groups. Our motivation is to test the theoretical hypothesis that diffusive element separation in the stellar atmosphere acts throughout the HgMn and Am star temperature range to create the abundance and isotopic anomalies observed (Michaud 1991). The concept that the Am and HgMn stars may be a related sequence can be traced back to Smith (1974) and perhaps even earlier, and has since been reiterated by others. However, an insufficient amount of evidence has been produced to either prove or disprove this assertion.  The well accepted, but perhaps arbitrary, boundary at spectral type B9.5/A0 separates the HgMn stars from the realm of the cooler, metallic-lined A (Am) stars. The classical Am stars are usually limited to spectral types later than A3. More recently the term {\\it hot-Am} star has found its way into the literature, representing those stars within the A0 - A3 subclasses whose spectra display a certain degree of Am star spectral characteristics, most notably an apparent deficiency of calcium and/or enhancements of iron-group element lines.  We are particularly interested in exploring the A0 - A2 stars for signs of an extension of the HgMn phenomenon to cooler temperatures as well as the B8 - A0 stars as an extension of the Am star phenomenon to hotter temperatures. ",
        "conclusions": ""
    },
    "9805/astro-ph9805254_arXiv.txt": {
        "abstract": "We extend our analysis of the three-dimensional response of a vertically polytropic disk to tidal forcing at Lindblad resonances by including the effects of a disk atmosphere.  The atmosphere is modeled as an isothermal layer that joins smoothly on to an underlying polytropic layer.  The launched wave progressively enters the atmosphere as it propagates away from the resonance.  The wave never propagates vertically, however, and the wave energy rises to a (finite) characteristic height in the atmosphere.  The increase of wave amplitude associated with this process of wave channeling is reduced by the effect of the atmosphere.  For waves of large azimuthal mode number $m$ generated by giant planets embedded in a disk, the increase in wave amplitude is still substantial enough to be likely to dissipate the wave energy by shocks for even modest optical depths ($\\tau\\ga10$) over a radial distance of a few times the disk thickness.  For low-$m$ waves generated in circumstellar disks in binary stars, the effects of wave channeling are less important and the level of wave nonlinearity increases by less than a factor of 10 in going from the disk edge to the disk center.  For circumbinary disks, the effects of wave channeling remain important, even for modest values of optical depth. ",
        "introduction": "Tidally generated waves can play an important role in the dynamics of gaseous disks and the tidal perturbers.  The Lindblad resonances (LRs) are important sites of wave generation in disks (Goldreich \\& Tremaine 1979).  Most previous studies have approximated the disk as two-dimensional (2D) by ignoring effects in the vertical direction (perpendicular to the orbit plane).  For disks whose structure and thermodynamic response are locally vertically isothermal, the 2D approximation accurately describes the wave that is generated at an LR, which is indeed a 2D sound wave.  In many cases of astrophysical interest, however, the disk is thermally stratified (optically thick) in the vertical direction.  Such cases include the planet-forming regions of protostellar disks (Bell et~al. 1997) and cataclysmic variable (CV) disks (La Dous 1994).  In an earlier paper (Lubow \\& Ogilvie 1998, hereafter Paper~I), we analyzed the linear response of a thin, vertically polytropic disk to tidal forcing at LRs.  We found that the dominant mode of excitation was the f mode of even symmetry.  Close to the resonance, this mode behaves compressibly and occupies the full vertical extent of the disk, like the 2D mode in an isothermal disk.  However, as the wave propagates away from the resonance through a radial distance of order $r_{\\rm L}/m$ (for resonance radius $r_{\\rm L}$ and azimuthal mode number $m$), it behaves incompressibly, like a surface gravity mode, and becomes confined close to the disk surface.  The degree of confinement increases with distance from the resonance.  As a consequence of this wave-channeling process, there is usually a strong increase in wave amplitude with radial distance from the LR on the scale $r_{\\rm L}/m$.  Although this analysis did not include damping or nonlinear effects explicitly, we estimated that shocks would sometimes develop and damp the wave.  Other damping mechanisms such as radiative damping (Cassen \\& Woolum 1996) or viscous damping may also be of importance.  Once the wave energy is channeled to a region close to the disk surface, the wave's properties are determined by the structure of the disk's outer layers.  In particular, the effects of a disk atmosphere can be important. Such effects are not well represented by a vertically polytropic disk.  In the case of a polytrope, there is a definite surface on which the density and temperature drop to zero. With an atmosphere, the disk density drops more gradually with height and the temperature is approximately constant. In fact, there is no actual disk surface in the isothermal case.  The density falls as a Gaussian with height above the mid-plane.  In the limit of very high (vertical) optical depth in the disk, the polytropic approximation should be adequate because the atmosphere contains an insignificant fraction of the disk's mass.  In the opposite limit of an optically thin disk, the disk structure is vertically isothermal and the wave launched at resonance behaves as a 2D sound wave.  The purpose of this paper is to determine to consequences of a finite, non-zero optical depth on the propagation of a wave launched from an LR. For disks of optical depth of about 100, which is thought to be typical in planet-forming regions of protostellar disks, nearly one per cent of the disk's mass resides in the atmosphere.  As a result of wave channeling, the wave energy could be contained within the atmosphere.  The purpose of this paper is to explore the consequences of a disk atmosphere on the propagational properties of the waves generated at LRs.  We aim to determine whether the wave in the disk atmosphere would propagate upwards into very low density regions where shock dissipation would occur.  More generally, we want to determine how much the wave amplitudes can be amplified by the wave-channeling process in the presence of a disk atmosphere as a function of disk optical depth. ",
        "conclusions": "We have analyzed the effects of a disk atmosphere on the propagation of a wave, the f mode, launched from a Lindblad resonance in a thermally stratified disk. The atmosphere was modeled as an isothermal layer that resides above a polytropic region that extends down to the mid-plane.  The polytropic region represents the (vertically) optically thick interior of the disk.  The effects of the atmosphere can be important because the f mode becomes progressively more confined to the disk surface with distance from resonance.  The wave energy is concentrated in the vertical direction as the wave propagates radially (see Figure 4). The peak of wave action density in the disk follows the trajectories shown in Figure 2 for various values of azimuthal wavenumber $m$ and disk optical depth $\\tau$.  The dominant amplification of a wave occurs in the thermally stratified region below the disk atmosphere through the process of wave channeling, as was anticipated in Paper I (see the discussion in \\S8.2). The disk atmosphere lessens the degree of wave amplification.  As was found in Paper I, the launched wave (f mode) behaves more like a surface gravity wave as it propagates radially away from the resonance.  In a purely vertically polytropic disk with no atmosphere, the vertical extent of the surface gravity wave decreases with increasing radial distance from resonance.  As a result, there is usually a substantial increase in wave amplitude on a radial scale of order $r_{\\rm L}/m$ from resonance, through a process termed wave channeling.  In the presence of an atmosphere, the effects of wave channeling are lessened (see Figure 3).  The increase in wave amplitude becomes limited once the wave enters the disk atmosphere. For disks of smaller optical depths, the thermal stratification is small and the wave enters the isothermal layer on a radial scale somewhat smaller than $r_{\\rm L}/m$.  Consequently, the wave behaves more like the 2D acoustic mode at most radii, in accordance with the expected limiting behavior of a vertically isothermal disk, in which the wave is 2D at all radii and its peak never rises above the mid-plane.  For disks around individual stars (circumstellar disks) in close binary star systems ($m = 2$), the increase in wave Mach number from resonance to disk center is typically less than a factor of 10, and somewhat less for lower optical depths (see Figure 3a).  For disks around binary star systems (circumbinary disks) ($m = 2$) the increase in wave Mach number from resonance is large over a scale of order the resonance radius (see Figure 3c).  For waves that are mildly nonlinear at resonance, nonlinear wave damping is likely to be important in such disks of modest vertical optical depth ($\\tau > 10$).  These effects are much stronger for circumbinary, rather than circumstellar disks.  For protostellar disks ($m \\sim 10 $, see Figure 3b and 4b), planet-forming regions are expected to be optically thick (Bell et~al. 1997).  Typical wave velocities near a giant planet are likely to be mildly supersonic, with Mach number around 0.2 (Goldreich \\& Tremaine 1979). In such a situation, wave channeling would likely lead to shocks for disks of moderate optical depths (of order 100) on a radial scale of a few times the disk thickness.  We emphasize that the detailed nonlinear outcome of the wave-channeling process remains uncertain.  While the wave action is concentrated in the polytropic layer, the motion is approximately incompressible, but this is no longer true once the wave migrates into the atmosphere, where shocks may limit the amplitude.  For this reason we have used the Mach number of the wave motion as a measure of nonlinearity.  The development of shocks or other nonlinear wave phenomena under these circumstances could be usefully investigated using numerical simulations."
    },
    "9805/astro-ph9805289_arXiv.txt": {
        "abstract": "A simple model for variability in relativistic plasma outflows is studied, in which nonthermal electrons are continuously and uniformly injected in the comoving frame over a time interval $\\Delta t$.  The evolution of the electron distribution is assumed to be dominated by synchrotron losses, and the energy- and time-dependence of the synchrotron and synchrotron self-Compton (SSC) fluxes are calculated for a power-law electron injection function with index $s = 2$.  The mean time of a flare or pulse measured at photon energy $E$ with respect to the onset of the injection event varies as $ E^{-1/2}$ and $ E^{-1/4}$ for synchrotron and SSC processes, respectively, until the time approaches the limiting intrinsic mean time $(1+z)\\Delta t/(2\\cal D)$, where $z$ is the redshift and $\\cal D$ is the Doppler factor.  This dependence is in accord with recent analyses of blazar and GRB emissions, and suggests a method to discriminate between external Compton and SSC models of high-energy gamma radiation from blazars and GRBs. The qualititative behavior of the X-ray spectral index/flux relation observed from BL Lac objects can be explained with this model.  This demonstrates that synchrotron losses are primarily responsible for the X-ray variability behavior and strengthens a new test for beaming from correlated hard X-ray/TeV observations. ",
        "introduction": "Space-based observatories such as {\\it ASCA}, {\\it RXTE}, {\\it Beppo-SAX} and the {\\it Compton Gamma Ray Observatory}, and ground-based air Cherenkov telescopes such as the {\\it Whipple Observatory} and {\\it HEGRA}, have gathered data from blazars and GRBs of sufficient quality to allow combined temporal and spectral analyses. A general trend is becoming apparent in blazar studies (see Wagner \\markcite{Wagner97}1997 and Shrader \\& Wehrle \\markcite{sw97}1997 for reviews): variability of optical and $> 100$ MeV emission in flat spectrum sources such as 3C 279 (e.g., Hartman et al. \\markcite{Hartman96}1996) and BL Lacertae (Bloom et al. \\markcite{Bloom97}1997), and of X-ray and TeV emission in the BL Lac objects Mrk 421 and Mrk 501 (e.g., Macomb et al. \\markcite{Macomb95}1995; Buckley et al. \\markcite{Buckley96}1996; Catanese et al. \\markcite{Catanese97}1997) appears to be temporally correlated, implying that the same population of electrons produces both the optical/X-ray and the 100 MeV - TeV emission. In the specific case of the 1994 May flare of Mrk 421, {\\it ASCA} observations (Takahashi et al. \\markcite{Takahashi96}1996) also show that the time lag of photons in the 0.5-2.0 keV range relative to 5 keV X-rays varies as $E({\\rm keV})^{-1/2} - 5^{-1/2}$.  Moreover, the Mrk 421 flare data follow a well-defined trajectory in a spectral index/flux display, and this behavior is also found in several other BL Lac objects (e.g., OJ 287, Idesawa et al. \\markcite{Idesawa97}1997; PKS 2155-304, Sembay et al. \\markcite{Sembay93}1993)  Recent analyses of GRB data show that the light curves tend to become narrower at higher energies (Fishman et al. \\markcite{Fishman92}1992), and that the decaying phases of GRB light curves are generally longer than the rising phases (Link, Epstein, \\& Priedhorsky \\markcite{Link93}1993).  Fenimore et al. (\\markcite{Fenimore95}1995) showed that the autocorrelation function of 45 bright BATSE GRBs, aligned at their peak fluxes, displays an energy dependent width $\\propto E^{-0.4}$. In the specific case of GRB 960720, which displays a single well-defined pulse, an energy-dependent duration $\\propto E^{-0.46}$ is found (Piro et al. \\markcite{Piro98}1998).  The radio/optical continuum of blazars is thought to be nonthermal synchrotron radiation, and blast wave models of GRBs also indicate that the hard X-ray/soft gamma-ray continuum from GRBs is produced by the same process (e.g.,  M\\'esz\\'aros, Rees, \\& Papathanassiou \\markcite{MRP94}1994; Tavani \\markcite{Tavani1996}1996; Waxman \\markcite{Waxman97}1997; Vietri \\markcite{Vietri97}1997).  The origin of the high energy gamma-ray emission in blazars is generally thought to originate from Compton processes, though it remains unclear whether internal synchrotron photons or photons produced outside the jet represent the dominant soft photon source.  The high-energy radiation from GRBs might also originate from Compton-scattering processes (see, e.g., Hurley et al. \\markcite{Hurley94}1994; M\\'esz\\'aros et al. \\markcite{MLR93}1993).  The likelihood that synchrotron emission from cooling nonthermal electrons can  explain some of the previously mentioned trends has been discussed in various approximations (e.g., Tashiro et al. \\markcite{Tashiro95}1995; Takahashi et al. \\markcite{Takahashi96}1996; Tavani \\markcite{Tavani1996}1996; Kazanas, Titarchuk, \\& Hua \\markcite{Kazanas98}1998; Kirk, Rieger, \\& Mastichiadis \\markcite{krm98}1998), but without a consideration of the associated SSC emission and a full treatment of the relativistic boost and energy normalization for the injected nonthermal electrons.  In this {\\it Letter}, we analyze the simplest possible model that retains the essential physics of nonthermal electron injection and cooling in relativistic plasma outflows. Even so, it goes a long way towards explaining the observed trends, and in addition suggests a fruitful avenue of research which could help to discriminate between models of high-energy blazar emission, to determine the origin of the high-energy radiation in GRBs, and to strengthen a new beaming test for blazar radiation (Catanese et al. \\markcite{Catanese97}1997). ",
        "conclusions": "Figure 1 shows the time-dependence of the synchrotron (Fig. 1a) and SSC (Fig. 1b) spectral fluxes for a variety of different observing energies, obtained by numerically solving equations (5) and (7), respectively.  Because $S\\propto \\eo^{-1/2}$ from uncooled electrons injected with $s=2$, the spectral flux is multiplied by $\\eo^{1/2}$ for clarity of presentation on a linear scale. In this calculation, we let $H = 0.1$ Gauss, $\\gamma_1 = 10$, $\\gamma_2 = 10^8$, $z=0.1$, $\\D = 10$, and $\\Delta t = 10^5$ s, as might be appropriate for a blazar flare.  We wish to determine the mean time $\\langle t(\\eo )\\rangle$ of the radiation observed at different energies, measured with respect to the onset of the injection event.  This is given by the expression $$\\langle t(\\eo )\\rangle =  {\\int _{-\\infty}^{\\infty} d\\to \\cdot \\to \\cdot S(\\eo;\\Omega_{\\rm obs},\\to )\\over \\int _{-\\infty}^{\\infty} d\\to\\cdot S(\\eo;\\Omega_{\\rm obs},\\to ) }  \\,.\\;\\eqno(8)$$ \\noindent Note that as defined here, a Heaviside (or boxcar) function has a mean time equal to one-half its temporal duration.  Equation (8) can be easily generalized for for higher moments of the temporal profile giving, for example, the FWHM duration of a flare.  After substituting equation (5) into equation (8), one obtains two cases depending on whether $T\\leq \\Delta t$ of $T > \\Delta t$.  When $\\gamma_2\\rightarrow \\infty$, the same result is found in both cases, namely $$\\langle t (\\eo )\\rangle _{\\rm syn} = {1+z\\over 2\\D} \\;\\big[ \\Delta t +\\nu_0^{-1}\\sqrt{{\\D\\epsilon_H\\over (1+z)\\eo}} \\;\\big] \\,.\\;\\eqno(9)$$  Equation (9) provides a convenient expression for fitting energy-dependent time-lag data, such as the Mrk 421 flare (Takahashi et al. \\markcite{Takahashi96}1996), or the energy-dependent GRB widths measured by Fenimore et al. (\\markcite{Fenimore95}1995) and Piro et al. (\\markcite{Piro98}1998), though the appropriate moment analysis should be used in more detailed treatments. Equation (9) indicates that the energy-dependence of the mean time varies more slowly than $E^{-0.5}$ when the light-crossing time $l/c$ or duration of the energization event are comparable to or longer than the comoving electron cooling time scale.  Fitting high quality data to equation (9) to determine the photon energy $\\bar\\eo$ where the two branches of the expression intersect, and using the shortest variability time scale  $\\delta\\to^{\\rm min} = (1+z)\\Delta t/\\D $ observed at $\\eo \\gg \\bar\\eo$, we derive the plasmoid magnetic field $$H({\\rm G}) \\cong 0.8\\;\\{ {(1+z)\\over \\D \\bar E({\\rm keV}) [\\delta \\to^{\\rm min}({\\rm hr})]^2 }\\}^{1/3}\\;.\\eqno(10)$$ Causality implies that $l\\lesssim c\\D \\delta \\to^{\\rm min}/(1+z)$. It will be important in future studies to understand how equations (9) and (10) are modified for a non-spherical plasmoid geometry and (especially for GRBs; see, e.g., Fenimore, Madras, \\& Nayakshin \\markcite{Fenimore96}1996) a blast wave geometry.   Expressions similar to equations (9) and (10) have been noted (e.g., Tashiro et al. \\markcite{Tashiro95}1995; Takahashi et al. \\markcite{Takahashi96}1996; Tavani \\markcite{Tavani96}1996; Buckley et al. \\markcite{Buckley98}1998; Catanese et al. \\markcite{Catanese97}1997; Kazanas et al. \\markcite{Kazanas98}1998), but this treatment provides a precise fitting function for analysis of data when $s= 2$ and yields a generalization for arbitrary values of $s$.  In the paper by Catanese et al. (\\markcite{Catanese97}1997), moreover, it is noted that correlated X-ray/TeV data imply an upper limit on $H$ because the electrons producing the highest energy synchrotron emission have Lorentz factors $\\gtrsim 2\\times 10^6 E_{\\rm C}({\\rm TeV}) (1+z)/\\D$, where $E_{\\rm C}({\\rm TeV})$ is the measured energy in TeV of the highest-energy gamma-rays. Synchrotron emission correlated with the TeV flux requires that the electrons radiate in a magnetic field at least as great as $H({\\rm G}) \\cong 11 \\epsilon_{\\rm obs, syn}\\D/[ E^2_{\\rm C}({\\rm TeV}) (1+z)]$, where $\\epsilon_{\\rm obs, syn}$ is the measured dimensionless energy of the highest energy synchrotron photons produced by the electrons which produce the TeV radiation. When compared with the value of $H$ inferred through equations (9) and (10), we obtain an expression for the Doppler factor, given by  $$\\D \\cong 1.7\\;{(1+z)  [E_{\\rm C}({\\rm TeV})]^{3/2}\\over \\bar \\epsilon_{\\rm obs}^{1/4}({\\rm keV}) (\\delta \\to^{\\rm min})^{1/2}(\\epsilon_{\\rm obs, syn})^{3/4}}\\;.\\eqno(11)$$ A lower limit to $\\D$ is obtained if the TeV flux does not exhibit a clear cutoff due to the high-energy cutoff in the the electron distribution function.  The inset to Fig. 1a shows the energy dependence of the mean time $\\langle t (E)\\rangle $ for the synchrotron and SSC processes.  As can be seen by examining Fig. 1b, the SSC emission decays more slowly than the synchrotron process, yielding an energy dependence $\\propto E^{-1/4}$ until the shortest mean time $(1+z)\\Delta t/2\\D $ is reached, which occurs at much higher energies than for the synchrotron process.  Thomson scattering of external photons, having an energy loss rate of the same form as the synchrotron energy loss rate, also gives $\\langle t(\\eo )\\rangle \\propto E^{-1/2}$ for $s=2$.  When more sensitive blazar $\\gamma$-ray observations become available with the upcoming $\\it INTEGRAL$ and $\\it GLAST$ missions, the energy-dependence of the mean time or duration of blazar flares can be used to determine whether SSC or ECS processes dominate in specific sources or between different source classes. This test might also be possible for bright TeV flares from BL Lac objects given the steadily improving sensitivity of air Cherenkov telescopes. More detailed studies need to be performed which take into account Klein-Nishina effects and general values of $s$.  Figure 2 shows the evolution of the synchrotron spectrum at different observing times, using the same parameters as in Figure 1.  At a fixed photon energy, the flux rises due to the accumulation of injected nonthermal electrons until cooling plays an important role in depleting the electron spectrum (cf. Dermer \\& Chiang \\markcite{Dermer98}1998).  The effects of cooling are seen earliest at the highest photon energies, and cause a softening of the spectrum.  Consequently, the spectral flux at a given photon energy displays a hard spectrum while its intensity is increasing, since the electrons that are radiating this emission have not yet felt the effects of cooling.  When synchrotron cooling becomes important, the flux begins to fall rapidly.  This behavior is shown by the solid curve (1) in the inset to Figure 2, which displays the clockwise evolution of the 2-10 keV spectral index as a function of a quantity proportional to the 2 keV spectral flux.  In reality, there will be some level of background which will be reached.  This is crudely modeled here by adding an underlying spectral component with spectral flux $S_{\\rm bg}(\\eo )=K_{\\rm bg}\\eo^{-0.5}$, where $K_{\\rm bg} = 1$  and $5\\times 10^4$ ergs cm$^{-2}$ for the dashed curve (2) and dotted curve (3), respectively.  The behavior illustrated in the inset to Fig. 2 is in qualitative agreement with evolutionary tracks of the spectral index/flux observed from some blazars, as noted in the Introduction (see also Kirk et al.\\ \\markcite{krm98}1998). Precise fitting of such tracks will require extending this model for general injection indices and for various forms for the background emission. The well-known hard-to-soft evolution of GRB pulses (e.g., Norris et al.\\ \\markcite{Norris86}1986) could be a manifestation of this effect, and the approach outlined here can also be used to analyze the fluence dependence of the peak of the GRB $\\nu F_\\nu$ spectrum (Liang \\& Kargatis \\markcite{Liang96}1996).  In summary, a very simple model has been presented for the observed nonthermal synchrotron and SSC emission emitted by cooling nonthermal electrons which are injected over a comoving time interval $\\Delta t$ into plasma with relativistic bulk motion.  In spite of the model's simplicity, empirical trends that have become better defined through recent combined temporal and spectral analyses of data from blazars and GRBs are qualitatively understood.  Straightforward generalizations, necessary for detailed fits to data, were indicated thoughout the analysis and will be treated in future studies."
    },
    "9805/astro-ph9805240_arXiv.txt": {
        "abstract": "I use recent photometric and stellar velocity dispersion measurements of the super-star-clusters (SSCs) NGC 1569A and NGC 1705-1 to determine their present-day luminosity/mass ($L_V/M$) ratios.  I then use the inferred $L_V/M$ ratios, together with population synthesis models of evolving star-clusters, to constrain the initial-mass-functions (IMFs) in these objects.  I find that $(L_V/M)_\\odot=28.9$ in 1569A, and $(L_V/M)_\\odot=126$ in 1705-1. It follows that in 1569A the IMF is steep with $\\alpha\\sim 2.5$ for $m^{-\\alpha}dm$ IMFs which extend to 0.1 $M_\\odot$. This implies that most of the stellar mass in 1569A is contained in low-mass ($< 1$ $M_\\odot$) stars. However, in 1705-1 the IMF is either flat, with $\\alpha\\lesssim 2$, or it is truncated at a lower mass-limit between 1 and 3 $M_\\odot$.   I compare the inferred IMFs with the mass functions (MFs) of Galactic globular clusters. It appears that 1569A has a sufficient reservoir of low-mass stars for it to plausibly evolve into an object similar to Galactic globular clusters. However, the apparent deficiency of low-mass stars in 1705-1 may make it difficult for this SSC to become a globular cluster. If low-mass stars do dominate the cluster mass in 1705-1, the large $L_V/M$ ratio in this SSC may be evidence that the most massive stars have formed close to the cluster cores. ",
        "introduction": "Recent optical and ultraviolet {\\it Hubble Space Telescope} ({\\it HST}) and near-IR speckle observations have revealed the widespread presence of luminous ($L_V\\gtrsim10^7$ $L_\\odot$) and compact ($r<2$ pc) ``super-star-clusters\" (SSCs) in a variety of star-forming galaxies \\markcite{Holtzman92} \\markcite{Whitmore95} \\markcite{Maoz96} (e.g. Holtzman et al. 1992; Whitmore \\& Schweizer 1995; Maoz et al. 1996; \\markcite{Tacconi96} Tacconi-Garman, Sternberg \\& Eckart 1996). The observations suggest that in starburst galaxies a large fraction of the OB stars are formed in compact SSCs, as opposed to more diffuse stellar associations.  The various photometric and spectroscopic observations are sensitive mainly to the massive stars ($\\gtrsim 3$ $M_\\odot$). Lower mass stars are generally not detectable directly. The {\\it total} stellar masses of the SSCs are usually estimated using models which predict the luminosity/mass ratios for an assumed set of parameters such as the cluster age and initial mass function (IMF).  In most models the luminosity is dominated by the most massive (and observed) stars in the system, while the mass is dominated by the more numerous lower mass (and unobserved) stars. Application of such methods has led to inferred cluster masses ranging from 10$^4$ to 10$^6$ $M_\\odot$ depending on the particular SSC and model employed.  The large inferred masses and the small observed radii have led to suggestions that the SSCs are young globular clusters.  NGC 1569 and NGC 1705 are two nearby dwarf galaxies. Each contains a prominent and well studied SSC, designated 1569A and 1705-1 \\markcite{Arp85}\\markcite{Melnick85} (Arp \\& Sandage 1985; Melnick, Moles \\& Terlevich 1985). \\markcite{Ho1996a}\\markcite{Ho1996b} Ho \\& Filippenko (1996a,b) carried out high-resolution (Keck) spectroscopy of 1569A and 1705-1, and were able to measure the stellar velocity dispersions in these SSCs. The stellar velocities together with the small cluster sizes indicated by the {\\it HST} observations \\markcite{O'Connell94}\\markcite{Meurer95}\\markcite{DeMarchi97} (O'Connell, Gallagher \\& Hunter 1994; Meurer et al. 1995; DeMarchi et al. 1997) imply cluster crossing times much shorter than the likely cluster ages. Ho \\& Filippenko concluded that the clusters are gravitationally bound, and that the implied virial masses are as large as $\\sim 10^5$ $M_\\odot$. Ho \\& Filippenko also argued that 1569A and 1705-1 might evolve into objects similar to Galactic globular clusters.  The independent estimates of the total cluster masses made possible by Ho \\& Filippenko's velocity dispersion measurements can be used to ``invert\" the usual analysis of the SSCs. In this paper I derive luminosity/mass ratios ($L_V/M$) for 1569A and 1705-1, based on the available observations.  I then compare the observed $L_V/M$ ratios with model predictions for a wide range of initial conditions.  My main goal is to constrain the IMFs in these clusters. I also address the question of whether the inferred IMFs are consistent with the observed mass-functions (MFs) of present-day globular clusters. ",
        "conclusions": "I have used the observed values of $L_V/M$ in 1569A and 1705-1 to constrain their IMFs. In 1569A the IMF must be steep, and a large fraction of the SSC mass is contained in low-mass ($m<1$ $M_\\odot$) stars. This cluster could plausibly evolve into a $2\\times 10^5$ $M_\\odot$ object resembling Galactic globular clusters. However, in 1705-1 the IMF must be flat or truncated at a lower-mass limit exceeding 1 $M_\\odot$. Most of the cluster mass in 1705-1 is contained in high-mass stars, implying that it may be difficult for this SSC to evolve into a globular cluster.  Several uncertainties could affect my analysis and conclusions. First, because $L_V/M$ is proportional to distance, and the distances to 1569A and 1705-1 are uncertain to within factors of $\\sim 1.5$ \\markcite{Meurer92}\\markcite{O'Connell94}(Meurer et al. 1992; O'Connell et al. 1994), the luminosity/mass ratios could be $1.5$ times smaller or larger than assumed here. Second, the observed luminosities have been corrected only for Galactic foreground extinction, and if there is any internal extinction the $L_V/M$ ratios could be much larger. The observations suggest that the internal extinction is very small \\markcite{Meurer92}\\markcite{Gonzalez97}(Meurer et al. 1992; Gonzalez-Delgado et al. 1997). However, I note that the initial extinction in these objects was probably very large. The fact that the SSCs are gravitationally bound implies that they must have formed with high ($\\sim 50\\%$) gas-to-star conversion efficiencies \\markcite{Mathieu84}\\markcite{Lada84}(Mathieu 1984; Lada et al. 1984). If the protostellar molecular cloud masses and sizes were approximately equal to the observed SSC masses and sizes, the initial molecular hydrogen ($H_2$) gas densities must have been as large as $\\sim 10^5$ cm$^{-3}$, with corresponding $H_2$ column densities equal to $\\sim 5\\times 10^{23}$ cm$^{-2}$.  For Galactic gas-to-dust ratios these column densities correspond to visual extinctions $A_V\\sim 100$.  So if the internal extinction is now very small any remaining gas must have been blown out very efficiently once the stars started forming.  This also implies that the star formation was ``instantaneous\" rather than ``continuous\" in these SSCs.  Third, the velocity dispersions measured by \\markcite{Ho96a}\\markcite{Ho96b}Ho \\& Filippenko (1996a,b) were based on observations of massive red supergiants (RSGs).  If these stars have attained equipartition, and are dynamically segregated in the cluster cores, the RSG velocities would be smaller than the mean stellar velocities, and the RSG virial masses would underestimate the total cluster masses. However, equipartition occurs on a cluster relaxation time $t_r\\approx (N/8{\\rm ln}N)t_c$, where $N$ is the number of stars in the cluster, and $t_c$ is the crossing time \\markcite{Bonnell98}(Bonnell \\& Davies 1998).  Adopting a mean stellar mass $<m>=1$ $M_\\odot$ (see Table 1), and the cluster masses and crossing times derived in \\S 3, it follows that $t_r \\gtrsim 100$ Myr.  It appears that the SSCs are too young for the RSGs to have undergone dynamical mass segregation. Alternatively, the massive stars may have formed preferentially closer to the proto-cluster cores \\markcite{Hillenbrand97}({\\it cf.} Hillenbrand 1997). The velocities could be independent of the stellar masses, but the observed half-light radii would provide lower limits to the true cluster sizes. The cluster masses would again be underestimated. If low-mass ($m<1$ $M_\\odot$) stars do dominate the total cluster mass in 1705-1 its large value of $L_V/M$ can be taken as evidence of mass segregation. I note that if the massive stars are spatially segregated the SSCs should appear more compact in the UV and near-IR since such light is produced by the most massive stars, whereas a significant fraction of the optical light is produced by intermediate mass stars (see \\S 2).  Finally, if significant populations of RSGs exist in both components of 1569A (see \\S 3) it is possible that part of the signal measured by Ho \\& Filippenko (1996a) is due to relative motion between the components rather than an intrinsic stellar velocity dispersion. The inferred $L_V/M$ ratio for 1569A would then be larger. Additional high-resolution spectroscopy is required to clarify this point.  Recent stellar census studies of nearby young clusters provide direct probes of the high-mass portions of the IMFs in star forming regions.  In Galactic and Magellanic OB associations \\markcite{Massey95}Massey et al. (1995) found that $\\alpha=2.3\\pm 0.3$ between 3 and 120 $M_\\odot$. Near-IR adaptive optics imaging \\markcite{Brandl96}(Brandl et al. 1996) and optical {\\it HST} imaging \\markcite{Massey98}(Massey \\& Hunter 1998) of R136, the ``core\" ($r<2$ pc) of the 30 Doradus star-forming region in the LMC, imply an IMF with $\\alpha\\sim 2.3$ from 120 $M_\\odot$ down to the confusion limit of about 2 $M_\\odot$, with no hint of a break near the observed low-mass limit.   In the Galactic cluster NGC 3603 \\markcite{Eisenhauer98} Eisenhauer et al. (1998) found that within the central parsec the MF is actually biased toward massive stars, with $\\alpha\\sim 1.7$ from the detection limit of $\\sim 1$ $M_\\odot$ to $\\sim 30$ $M_\\odot$, steepening to $\\alpha\\sim 2.7$ for higher masses. (This steepening may be due to the small cluster mass rather than to an intrinsic steepening of the IMF). \\markcite{Hillenbrand97}Hillenbrand (1997) found that the Orion Nebula Cluster is globally consistent with a Miller-Scalo IMF, but that within the central 0.3 pc the IMF is biased towards massive stars. It is unclear what these various observations imply about the IMFs in the much more massive systems 1569A and 1705-1. Nevertheless, Figs. 1-3 illustrate that the IMF in 1569A may be similar to the global MF in the Orion cluster, whereas the IMF in 1705-1 may be similar to the MF in NGC 3603.   Recent observations of many starburst galaxies indicate that a large, and perhaps dominant, fraction of the recent star formation has occured in massive SSCs \\markcite{O'Connell95}\\markcite{Maoz96}\\markcite{Tacconi96} (O'Connell et al. 1995; Maoz et al. 1996; Tacconi-Garman et al. 1996). If most of the SSCs have steep IMFs similar to 1569A, this would imply that dynamically significant populations of low-mass stars are formed togther with the observed high-mass stars. However, if most of the SSCs are similar to 1705-1, with flat or truncated IMFs, this would support the long-standing hypothesis \\markcite{Rieke80}\\markcite{Rieke93} (Rieke et al. 1980, 1993) that in starburst galaxies massive stars are formed preferentially. Additional dynamical measurements of SSC masses would be valuable."
    },
    "9805/astro-ph9805076_arXiv.txt": {
        "abstract": "A number of about 1500 spectroscopically classified CP2 stars in the southern sky ($\\delta =-90\\degr$ to $-12\\degr$) was extracted from the Michigan Catalogue (Vols. I\\,-\\,IV).  This sample was compared with the classification from Bidelman \\& Mc Connell (1973). We confirmed the spectral classification with the known photometric peculiarity indices in the Geneva system and in the 3-filter $\\Delta a$ system (Maitzen 1976). 10\\% of these stars show discordance between their respective types from the Michigan and Bidelman catalogues. Several objects were measured with a CCD in the $\\Delta a$ system in spring 1995. Eight stars are peculiar in $\\Delta a$. Furthermore, we have investigated the galactic distribution of all programme stars. We conclude that the distribution resembles the one of early-type stars, where the hotter (=\\,Silicon) stars are more concentrated towards the galactic plane than the cooler (=\\,Strontium) objects. ",
        "introduction": "About 30.000 stars with B, A and early F type classification were extracted from the Michigan catalogue of two dimensional spectral types for the HD stars (Nancy Houk et al., 1975-1988; hereafter MC). In the southern sky region from $\\delta_{1900} = - 90\\degr$ to $-12\\degr$, 1500 objects are classified as CP2 stars. The limiting magnitude of this survey is 11, the maximum of the distribution of magnitudes is between 8.5 and 10 for both subgroups.  74\\% of all extracted CP2 stars are members of the Si-subgroup. They are more concentrated towards the galactical plane than the cooler stars. In the galactic area from $b = -15\\degr$  to $b=+15\\degr$, there are 84\\% of all hotter silicon stars and 58\\% of all SrCrEu stars. As expected, this result is in accordance with the galactic distribution of main sequence stars. A tendency to clustering for the investigated stars was not found.  Bidelman \\& MacConnell (1973; hereafter BC) list 781 CP2 objects. Their classification is based on the same objective-prism plates as the sample of the Michigan catalogue. Most of the compared stars have the same or a similar spectral classification in MC and BC.  Photometric values are given for 981 stars in the Geneva system (Rufener 1988) as well as in the 3-filter $\\Delta a$ system (Maitzen \\& Vogt 1983).  430 out of 645 BC stars with known Geneva photometric indices, have a significant peculiarity.  75\\% of 339 BC-stars are peculiar in $\\Delta a$. Three quarters of this sample are members of the Si subgroup.  HD 134185 was classified as F0/2 V in MC (F2 V in the Geneva catalogue), but as Si in BC. Si may be a typographical error meaning Sr, but the photometric $\\Delta a$ value confirms non-peculiarity.  The spectral class of HD 91756 is given as Fm $\\delta $ Del in MC, (A0)  in the Geneva catalogue and Ap(SrCr) in BC. The Geneva peculiarity index $\\Delta (V1-G)$ shows no significant value for this star.  HD 110072 is identified as Ap(SrCr) in MC and BC but as K0 in the Geneva Catalogue and the Simbad database. The photometric values in the Geneva catalogue point to a late A-type star.  A total of 78 stars are identified as non-CP2 stars in MC but as Si, Sr or SrCrEu in BC. The spectral classification and photometric values in the Geneva catalogue are also different for these objects. Therefore, we started an observing run with $\\Delta a$ photometry for a sample of them. ",
        "conclusions": ""
    },
    "9805/astro-ph9805130_arXiv.txt": {
        "abstract": "We analyse synthetic galaxy spectra from the evolutionary models of Bruzual \\& Charlot and Fioc \\& Rocca-Volmerange using the method of Principal Component Analysis (PCA).  We explore synthetic spectra with different ages, star formation histories and metalicities, and identify the Principal Components (PCs) of variance in the spectra due to these different model parameters.  The PCA provides a more objective and informative alternative to diagnostics by individual spectral lines.  We discuss how the PCs can be used to estimate the input model parameters and explore the impact of noise in this inverse problem. We also discuss how changing the sampling of the ages and other model parameters affects the resulting PCs. Our first two synthetic PCs agree with a similar analysis on observed spectra obtained by Kennicutt and the 2dF redshift survey.  We conclude that with a good enough signal-to-noise ($S/N\\gg 10$) it is possible to derive age, star formation history and metallicity from observed galaxy spectra using PCA. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/hep-ph9805405_arXiv.txt": {
        "abstract": "The effects of radiatively decaying, long-lived particles on big-bang nucleosynthesis (BBN) are discussed. If high-energy photons are emitted after BBN, they may change the abundances of the light elements through photodissociation processes, which may result in a significant discrepancy between the BBN theory and observation.  We calculate the abundances of the light elements, including the effects of photodissociation induced by a radiatively decaying particle, but neglecting the hadronic branching ratio.  Using these calculated abundances, we derive a constraint on such particles by comparing our theoretical results with observations.  Taking into account the recent controversies regarding the observations of the light-element abundances, we derive constraints for various combinations of the measurements. We also discuss several models which predict such radiatively decaying particles, and we derive constraints on such models. ",
        "introduction": "\\label{sec:intro}  Big-bang nucleosynthesis (BBN) has been used to impose constraints on neutrinos and other hypothetical particles predicted by particle physics, because BBN is very sensitive to the thermal history of the early universe at temperatures $T \\lesssim 1$ MeV~\\cite{BBN-review}.  Weakly interacting, massive particles appear often in particle physics. In this paper, we consider particles which have masses of $\\sim O(100{\\rm ~GeV})$ and which interact with other particle only very weakly ({\\it e.g.}, through gravitation).  These particles have lifetimes so long that they decay after the BBN of the light elements (D, $^3$He, $^4$He, etc.), so they and their decay products may affect the thermal history of the universe.  In particular, if the long-lived particles decay into photons, then the emitted high-energy photons induce electromagnetic cascades and produce many soft photons.  If the energy of these photons exceeds the binding energies of the light nuclides, then photodissociation may profoundly alter the light element abundances.  Thus, we can impose constraints on the abundance and lifetime of a long-lived particle species, by considering the photodissociation processes induced by its decay.  There are many works on this subject, such as the constraints on massive neutrinos and gravitinos obtained by the comparison between the theoretical predictions and observations~\\cite{Lindley,Raddec,KM1,KM2,hadron}.\\footnote {As pointed out in Ref.~\\cite{hadrBR}, even if the parent particle decays only into photons, these photons will produce hadrons with a branching ratio of at least 1\\%.  However, since there is no data on some crucial cross sections involving $^7$Li and $^7$Be, we cannot include hadrodissociation in our statistical analysis.  Since we have neglected hadrodissociation, our constraints may be regarded as conservative bounds.}  A couple of years ago, Hata {\\it et al.}~\\cite{HSSTWBL} claimed that light-element observations seemed to conflict with the theoretical predictions of standard BBN.  Their point was that standard BBN predicts too much $^4$He, if the baryon number density is determined by the D abundance inferred from observations; equivalently, standard BBN predicts too much D, if the baryon number density is determined by the $^4$He observations.  Inspired by this ``crisis in BBN,'' many people re-examined standard and non-standard BBN by including systematic errors in the observations, or by introducing some non-standard properties of neutrinos~\\cite{xinu,mstau}.  In a previous paper~\\cite{PRLHKM}, we investigated the effect upon BBN of radiatively decaying, massive particles. These particles induce an electromagnetic cascade.  We found that in a certain parameter region, the photons in this cascade destroy only D, so that the predicted abundances of D, $^3$He, and $^4$He fit the observations.  However, since the ``BBN crisis'' was claimed, the situation concerning the observations of deuterium has changed.  The D abundances in highly red-shifted quasar absorption systems (QAS) have been observed. The abundance of D in high-$z$ QAS is considered to be the primordial value. Thanks to these direct new observations, we no longer need to use poorly understood models of chemical evolution to infer the primordial abundance from the material in solar neighborhood.  Moreover, there are also differing determinations of the primordial $^4$He abundance.  Hata {\\it et al.} used a relatively low $^4$He abundance (viz., $Y\\simeq 0.234$, where $Y$ is the primordial mass fraction of $^4$He)~\\cite{pagel,OliSkiSte}.  However, a higher $^4$He abundance ($Y \\simeq 0.244$) has also been reported~\\cite{IzoThuLip94,ThuIzo,IzoThuLip97}, and it has been noted that this higher observation alleviates the discrepancy with standard BBN theory~\\cite{KerSar}.  The typical errors in $^4$He observations are less than $\\simeq 0.005$, so we have discordant data for $^4$He.  Since we have discordant $^4$He abundances and new observations for D, the previous constraint on the radiative decay of long-lived particles must be revised.  In addition, the statistical analyses on radiatively decaying particles are insufficient in the previous works.  Therefore, in our present paper, we perform a better statistical analysis of long-lived, radiatively decaying particles, and of the resultant photodissociations, in order to constrain the abundances and lifetimes of long-lived particles.  In deriving the constraint, we use both high and low values of the $^4$He abundance, because it is premature to decide which data are correct. As a result, it will be shown that for low values of the $^4$He abundance, we have a poor agreement between the observations and the standard BBN theory. Moreover, we show in this case that a long-lived particle with appropriate abundance and lifetime can solve the discrepancy.  In the case of high $^4$He, standard BBN fits the observations, so we derive stringent constraints on the properties of long-lived particles.  In this paper, we also include the photodissociations of $^7$Li and $^6$Li for the first time. As we will show later, the destruction of $^7$Li does not dramatically affect the predicted D and $^4$He, in the region where the observed D and $^4$He values are best fit.  However, the $^6$Li produced by the destruction of $^7$Li can be two orders of magnitude more abundant than the standard BBN prediction of $^6$Li/H $\\sim O(10^{-12})$. We discuss the possibility that this process may be the origin of the $^6$Li which is observed in some low-metallicity halo stars.  In Sec.~\\ref{sec:SBBN}, we study how consistent the theoretically predicted abundances and observations are, in the case of standard BBN.  The radiative decay of long-lived particles is considered in Sec.~\\ref{sec:BBNX}, and the particle physics models which predict such long-lived particles are presented in Sec.~\\ref{sec:model}.  Finally, Sec.~\\ref{sec:summary} is devoted to discussion and the conclusion. ",
        "conclusions": "\\label{sec:summary}  We have discussed the photodissociation of light elements due to the radiative decay of a massive particle, and we have shown how we can constrain our model parameters from the observed light-element abundances.  We adopted both low and high $^4$He values in this paper, and we obtained constraints on the properties of the radiatively decaying particle in each case.  When we adopt the low $^4$He value, we find that a non-vanishing amount of such a long-lived, massive particle is preferred: $m_X Y_X \\gtrsim 10 ^ {-10} {\\rm GeV}$ and $10^4 {\\rm sec} \\lesssim \\tau_X \\lesssim 10^6 {\\rm sec}$. On the other hand, consistency with the observations imposes upper bounds on $m_X Y_X$ in each cases.  We have also studied the photodissociation of $^7$Li and $^6$Li in this paper. These processes do not affect the D, $^3$He, and $^4$He abundances, because $^7$Li and $^6$Li are many orders of magnitude less abundant than D,$^3$He, and $^4$He. When we examine the region of parameter space where the predicted abundances agree well with the observed $^7$Li and the low $^4$He observations, we find that the produced $^6$Li/H may be of order $10^{-12}$, which is two orders of magnitude larger than the prediction of SBBN (see Figs.~7 and~13). The predicted $^6$Li is consistent with the observed upper bound Eq.~(\\ref{Li6}) throughout the region of parameter space we are interested in. Although presently it is believed that the observed $^6$Li abundance is produced by spallation, our model suggests another origin: the observed $^6$Li may be produced by the photodissociation of $^7$Li.  We have also discussed candidates for our radiatively decaying particle. Our first example is the gravitino. In this case, we can constrain the reheating temperature after inflation, because it determines the abundance of the gravitino. We obtained the stringent bounds $T_R \\lesssim 10^8 {\\rm GeV} - 10^9 {\\rm GeV} $ for $100{\\rm ~GeV} \\lesssim  m_{3/2} \\lesssim 1{\\rm ~TeV}$. Our second example is the lightest neutralino which is heavier than the gravitino. When the neutralino is the lightest superparticle in the MSSM sector, it can decay into a photon and a gravitino. If we assume the lightest neutralino is pure bino, and its mass is about 100 GeV, then the relic number density of binos is related to the right-handed slepton mass, because they annihilate mainly through right-handed slepton exchange. For this case, we obtained an upper bound on the gravitino mass:  $m_{3/2} \\lesssim 400 {\\rm ~MeV} - 1 {\\rm ~GeV}$ for $100 {\\rm ~GeV} \\lesssim m_{\\tilde{l}_R} \\lesssim 1 {\\rm ~TeV}$.  Our third example is a modulus field. We obtained a severe constraint on its initial amplitude, $ \\phi_0 \\lesssim 10^8 {\\rm ~GeV} - 10^{9} {\\rm ~GeV}$ for $100{\\rm ~GeV} \\lesssim m_{3/2} \\lesssim 1{\\rm ~TeV}$. This bound is well below the Planck scale, so it suggests the need for a dilution mechanism, such as thermal inflation."
    },
    "9805/astro-ph9805183_arXiv.txt": {
        "abstract": "This paper presents a detailed description of the acquisition and processing of a large body of imaging data for three  fields in the globular cluster M4 taken with the  Wide Field and  Planetary Camera  2 aboard  the Hubble Space Telescope. Analysis with the ALLFRAME package yielded the deepest photometry yet obtained   for  this cluster.   The resulting   data-set  for 4708 stars (positions  and  calibrated photometry  in  V,  I, and,   in two  fields, U) spanning approximately six cluster core radii is available on the AAS CD-ROM (or email a request  to RAI).  The scientific analysis  is deferred to three companion  papers, which investigate the  significant white dwarf population discovered and the main sequence population. ",
        "introduction": "The refurbished Hubble Space Telescope (HST) made it possible, for the first time, to study two  stellar populations that  had remained almost completely unexplored from the  ground  even  in the   nearest  globular clusters:  the low-mass red dwarfs at the faint end of the main sequence (MS) and the white dwarfs (WDs).  To understand in detail the physics of  these objects, and in particular to constrain  theories of the cooling of  WDs and theories of the structure of faint MS  stars, we undertook a  study of the  nearest globular cluster,  M4.  The scientific results of  this study are  presented in three companion papers: the constraints placed on the  WD population are discussed in \\markcite{Richer95}    Richer  {\\it  et   al.}   (1995)   (paper~1)   and \\markcite{Richer97}  Richer  {\\it   et al.}    (1997)  (paper~2),  while the investigation of the MS population is presented in \\markcite{Fahl98} Fahlman {\\it et al.}  (1998) (paper~3).  The present paper details the observational data reductions performed for this study, and forms the basis from which the analysis of papers~1--3 was extracted. ",
        "conclusions": "We have  presented an  account of   the  reductions of  the  deepest stellar photometry yet derived  for the  globular  cluster M4.  The resulting  data, tabulated   in the CD-ROM accompanying this   contribution, show  that it is possible to obtain accurate photometry   with WFPC2 to $V=28$ in  relatively crowded  fields.  This  photometric survey  is  used in companion papers  to study the  main- and white   dwarf sequences with unprecedented  detail and accuracy."
    },
    "9805/astro-ph9805093_arXiv.txt": {
        "abstract": "The discovery of a $z_{abs}=0.0912$ damped Ly$\\alpha$ absorption-line system in the HST-FOS ultraviolet spectrum of the quasar OI$\\;$363 (0738+313) is reported.  This is the lowest redshift quasar damped Ly$\\alpha$ system known. Its neutral hydrogen column density is $N$(\\ion H 1)$=1.5 \\pm 0.2 \\times10^{21}$ atoms cm$^{-2}$, which easily exceeds the classical criterion for damped Ly$\\alpha$ of $N$(\\ion H 1)$\\ge2\\times10^{20}$ atoms cm$^{-2}$.  Remarkably, a $z_{abs}=0.2212$ damped system with $N$(\\ion H 1)$=7.9 \\pm 1.4 \\times10^{20}$ atoms cm$^{-2}$ has also been discovered in the same spectrum.  In the past, the standard paradigm for damped Ly$\\alpha$ systems has been that they arise in galactic or protogalactic \\ion{H}{1} disks with low impact parameters in luminous galaxies.  However, WIYN imaging of the OI$\\;$363 field shows that none of the galaxies visible in the vicinity of the quasar is a luminous gas-rich spiral with low impact parameter, either at $z= 0.0912$ or $z=0.2212$. Thus, these damped systems are among the clearest examples yet of cases that are inconsistent with the standard damped Ly$\\alpha$ $-$ HI-disk paradigm. ",
        "introduction": "While it is widely recognized that QSO damped Ly$\\alpha$ absorption-line systems are important probes of galaxy formation, their utility is hampered by their rarity, especially in the ultraviolet at low redshift. For example, only one damped Ly$\\alpha$ line was discovered in the {\\it Hubble Space Telescope} Quasar Absorption Line Key Project (Jannuzi et al. 1998) and, at the present time, only two damped systems have been confirmed from a survey of QSOs with the {\\it International Ultraviolet Explorer} (Lanzetta, Wolfe \\& Turnshek 1995).  In the ultraviolet ($\\lambda < 3220$ \\AA), Ly$\\alpha$ absorption has redshifts $z_{abs} < 1.65$. This corresponds to look-back times up to 77\\% of the age of the Universe for q$_0=0.5$.  Thus, it is particularly important that low-redshift examples of damped Ly$\\alpha$ systems be discovered and studied. This will eventually allow the evolutionary links between \\ion{H}{1} in nearby galaxies (Rao \\& Briggs 1993) and the high-redshift damped systems found in optical surveys (Wolfe et al. 1986; Lanzetta et al. 1991; Wolfe et al. 1995) to be explored.  We recently completed a highly successful HST Faint Object Spectrograph program to discover low-redshift ($z<1.65$) damped Ly$\\alpha$ lines in QSO spectra (Turnshek 1997; Rao \\& Turnshek 1998).  Coupled with earlier HST archival work (Rao, Turnshek \\& Briggs 1995), we have studied the ultraviolet Ly$\\alpha$ absorption line in 88 \\ion{Mg}{2} absorption-line systems. This targeted search has greatly increased the number of known low-redshift damped systems.  We currently have 12 damped systems in our sample, including two that were discovered serendipitously in spectral regions without available \\ion{Mg}{2} information.  In this $Letter$, we report our discovery of the lowest redshift damped Ly$\\alpha$ absorption-line system known ----- a $z_{abs}=0.0912$ system with neutral hydrogen column density $N$(\\ion H 1)$\\approx1.5\\times 10^{21}$ atoms cm$^{-2}$ in the spectrum of the quasar OI$\\;$363 (0738+313; $z_{em}=0.630$). This object has a known \\ion{Mg}{2} absorption-line system at $z_{abs}=0.2213$ (Boulade et al. 1987), but no evidence for an absorption system at $z_{abs}=0.0912$ had ever been reported prior to our survey. We obtained a HST-FOS G160L/BL spectrum of this object to determine the nature of the Ly$\\alpha$ line associated with the $z_{abs}=0.2213$ system and coincidentally discovered the damped system at $z_{abs}=0.0912$.  Remarkably, the slightly higher redshift system is also damped with $N$(\\ion H 1)$\\approx7.9\\times 10^{20}$ atoms cm$^{-2}$.  In \\S2 we describe our observations and data analysis. The observations include the HST-FOS spectrum of OI$\\;$363 and an R image of the field obtained in seeing of 0.55\\arcsec\\ with the WIYN 3.5-m Telescope on Kitt Peak. We find that none of the galaxies visible in the vicinity of the quasar are luminous gas-rich galaxies with low impact parameters, either at $z\\approx 0.0912$ or $z\\approx 0.2212$. In \\S3 we discuss these results. ",
        "conclusions": "The two damped Ly$\\alpha$ systems described in this $Letter$ provide an excellent opportunity to determine the nature of damped Ly$\\alpha$ galaxies. It is interesting to note that none of the objects listed in Table 1 conform to the standard paradigm that damped Ly$\\alpha$ galaxies are  $M_{R}^* \\approx -21$ spiral disks (Wolfe et al. 1986; Wolfe 1988) with small impact parameters ($b<20h^{-1}_{75}$ kpc, Steidel 1995), where we take $B-R=1$.  The usual convention is that this $M^*$  characterizes the  Schechter luminosity function of all galaxy types combined (see Fig. 2 in Rao \\& Briggs 1993).  However, the luminosity function of spiral galaxies is best described by a Gaussian function with a peak occurring at $M_R^*\\approx -18$ (see Fig. 1 in Rao \\& Briggs 1993).  Interestingly, the fact that brighter spirals have larger \\ion{H}{1} disks pushes the peak in the interception-probability distribution, which is the product of number density and \\ion{H}{1} cross-section, to higher luminosities, i.e., to $M_R \\approx -21$ (see Fig. 5).  If any one of the faint galaxies reported by Le Brun et al. within $\\approx$ 5\\arcsec\\ of the quasar is the absorber, then it would be an extremely faint dwarf galaxy at $z=0.0912$, with\\\\ $M_r$ $\\gtrsim-13.5$. On the other hand, the  low-surface-brightness feature, marked by a cross in Fig. 4, might be part of a galaxy that lies directly in front of the quasar (in this case for $z=0.09$, $M_R>-18$ is still likely).  Its existence needs to be confirmed. The candidate G1 is a likely absorber at $z=0.0912$, with absolute magnitude $M_R=-16.9$ ($M_B=-15.9$), $\\approx 4$ magnitudes fainter than the peak in the interception probability distribution.  It is a better candidate at $z=0.2212$, where it would be brighter and still at a reasonable impact parameter from the quasar. Galaxies G2 through G9 would be dwarfs at both redshifts and, given the high column densities of the two damped Ly$\\alpha$ lines,  make highly unlikely absorbers at the listed impact parameters. The galaxy G10 appears to be an early type galaxy and, though reasonably bright, is not expected to contain much gas, especially at large impact parameters.  The galaxy G11 is the only large spiral in the field (see the Fig. 4 caption); but its redshift is $z=0.06$ (\\S2.2). Thus, one could conclude that G1 is the only galaxy that is a reasonable candidate at either absorption redshift.  The absence of luminous spiral galaxy candidates at both damped absorption redshifts represents the most illustrative examples yet of inconsistencies with the standard HI-disk paradigm for damped Ly$\\alpha$ systems. Others have discussed similar findings in systems of higher redshift (Steidel et al. 1994; Steidel et al. 1997; Le Brun et al. 1997), but our upper limits on the luminosity of candidate damped Ly$\\alpha$ galaxies are about an order of magnitude lower than previous results.  Clearly, not all damped Ly$\\alpha$ absorbers are luminous spirals.  It has been known for some time that $\\Omega_{gas}$ at high-redshift is comparable to the present-day luminous (stellar) cosmological mass density (Rao \\& Briggs 1993).  Thus, the high-redshift damped systems are possibly the progenitors of all present-day galaxies, not just luminous spiral HI-disks.  The recent work of Khersonsky \\& Turnshek (1996), Pettini et al. (1997), Rao \\& Turnshek (1998), and references therein also reveal problems with this scenario.  In fact, the results of our recent survey for damped Ly$\\alpha$ absorption at $z_{abs}<1.65$ indicate that the cosmological mass density of neutral gas, $\\Omega_{gas} = 1.3\\Omega_{HI}$, at $z \\approx 1$ is also comparable to the luminous mass density at the current epoch (Rao \\& Turnshek 1998).  One solution might be the existence of  many unrecognized low-surface-brightness (LSB) galaxies or dwarf galaxies at low redshift. However, Briggs (1997a; b) and Zwaan et al. (1997) have concluded that the present-day population of LSB galaxies contributes little to the \\ion{H}{1} mass density at $z=0$.  They appear to be drawn from the same population as the ``normal'' high-surface-brightness (HSB) galaxies that are found in optical surveys. Moreover, Bothun et al. (1997) have shown that the average \\ion{H}{1} column density in LSB galaxies is lower than that in HSB galaxies.  Rao \\& Briggs (1993) used {\\it available data} to show that the local \\ion{H}{1} mass density and cross-section is dominated by the large spirals.  Thus, based on the present imaging results, we conclude that the nature of the damped Ly$\\alpha$ galaxies is not yet fully understood.  \\noindent {\\bf ACKNOWLEDGEMENTS.} We thank Ed Smith and Alex Storrs of the HST-FOS team for help and advice with FOS data reduction. We also thank Diane Harmer and Paul Smith for their assistance in obtaining the WIYN data, and Kathy Romer for kindly obtaining the spectrum of the $z=0.06$ galaxy (G11) in the field of the quasar OI$\\;$363 with the KPNO 4m.  This observation allowed us to exclude that galaxy as a possible site for the absorption.  We especially thank Eric Monier for his assistance with many aspects of this work.    \\clearpage"
    },
    "9805/hep-ph9805209_arXiv.txt": {
        "abstract": "We investigate nonthermal phase transitions that may occur after post-inflationary preheating in a simple model of a two-component scalar field with the effective potential $\\lambda (\\phi_i^2 - {\\rm v}^2 )^2/4$, where $\\phi_1$ is identified with the inflaton field. We use three-dimensional lattice simulations to investigate  the full nonlinear dynamics of the model. Fluctuations of the fields generated during and after preheating temporarily make the effective potential convex in the $\\phi_1$ direction. The subsequent  nonthermal phase transition with symmetry breaking leads to formation of cosmic strings even for ${\\rm v} \\gg 10^{16}$ GeV. This mechanism of string formation, in a {\\em modulated} (by the oscillating field $\\phi_{1}$) phase transition, is different from the usual Kibble mechanism. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805216_arXiv.txt": {
        "abstract": "We discuss the galaxy population of the rich cluster MS1621.5+2640 at z=0.4274, based on spectra and imaging in a field of size 9 arcmin by 23 arcmin ($\\sim$2 by 5 h$^{-1}$ Mpc). The sample comprises 277 galaxies, of which 112 are cluster members, 7 are `near-members', and 47 are field galaxies in the redshift range 0.37$\\le$z$\\le$0.50. The results are analyzed and compared with the z=0.2279 rich cluster Abell~2390.  MS1621.5+2640 has a higher blue fraction, a younger stellar population, and is a less evolved cluster. We do not find strong evidence of significant excess star formation compared with the field, although there is a small population of outlying near-members that is unusually blue and that may be affected by the cluster. There is a substantial population of red galaxies with significant H$\\delta$ absorption, which are not easily explained by any simple form of modeled star formation history. We detect two distinct populations of cluster galaxies: those where star formation ceased some time ago, and those with a gradual decrease over many Gyr. Our data suggests that the cluster formed by accretion from the field, with truncation of the star formation beginning at very large radii ($>$2 times the virial radius). The truncation process does not seem to be a sharp one though, in that lower-luminosity early-type galaxies in the inner core of the cluster are seen with significant H$\\delta$ absorption, indicating some star formation within the last 1-2 Gyr. Some combination of stripping of gas from the outer parts of the galaxy, together with gradual exhaustion of the gas in the inner parts would be consistent with our data. ",
        "introduction": "\\label{sec-intro}  Clusters at intermediate redshift allow us to study evolutionary processes in cluster structure, cluster growth by infall, and the evolution of cluster galaxies themselves, in morphology and stellar populations. Study of a well selected set of clusters at different redshifts provide snapshots of these processes at different epochs. The Canadian Network for Observational Cosmology (CNOC) survey of selected rich clusters (\\cite{yee96a}, \\cite{car96}) allows such studies and comparisons over the approximate redshift range 0.2$\\le$z$\\le$0.6.  The first cluster studied in detail using the CNOC database was Abell~2390 (\\cite{abr96}, \\cite{yee96b}), at redshift 0.2279. In those papers, the cluster is seen to have well-defined  and smooth radial gradients in galaxy color, morphology, and stellar population, showing that the cluster forms by galaxies continuously accreting from the field. This accretion process appears to involve the cessation of star-formation, and passive evolution, creating a gradient in time since last star formation with clustocentric distance. The overall blue fraction of galaxies  in the field around Abell~2390 is higher than in the zero redshift universe, both in the cluster and the nearby field, consistent with general aging of all galaxy populations.  One controversy in this area has been how severely galaxies are affected by accretion into a cluster: in particular, whether there is a population of merging or interacting infalling galaxies, whether infall disturbs the ISM enough to trigger a `starburst'\\footnote{One immediate cause for disagreement is in the definition of a `starburst'. Some authors seem to regard an Sc galaxy as a `starburst'. We will try to avoid some of this controversy by comparing our data with models with a simple star formation history which we will specify. This will allow others to decide whether any of the models are `starbursts' or not for themselves. } in an isolated galaxy, whether gas stripping may occur by the cluster IGM, and/or whether star-formation is halted on infall. We noted that we see little evidence of a population with excess star formation in Abell~2390, and advocated models of truncated star-formation to explain the observed radial gradients. Recently, it was pointed out to us that the issue of starbursting versus truncated star formation was carefully discussed in \\cite{new90}. Our conclusions here support and expand on their work, as well as complementing that in \\cite{abr96}.  Clearly it is of interest to see whether the same can be said of other clusters, particularly those at even higher redshift, where the fraction of blue star forming galaxies is known to be higher (\\cite{but84}).  This paper provides a comparison with Abell~2390 by studying the rich cluster MS1621.5+2640 at z=0.4274. To allow easy comparison, we tabulate some of the parameters of these two clusters in Table~\\ref{tbl-compare}. MS1621.5+2640 was chosen as the cluster in the CNOC database most likely to violate the passive truncated star formation model proposed for Abell~2390. We have data for a large number of members, and it was noted when the observations were being taken that line emission was common. As will be demonstrated, this impression was largely driven by the fact that most galaxies (field and cluster) at z=0.42 have higher star formation rates than those at z=0.23. We find that MS1621.5+2640 does not have an unusually large starforming population relative to the field.  A recent paper by \\cite{bar96} studies the galaxy population distributions in various observational parameter spaces, for the inner regions of three z=0.31 clusters.  They conclude that some 30\\% of the cluster galaxies have undergone a secondary burst of star formation in the previous 2 Gyr, and hence that infall causes a starburst. However, the difference between a galaxy with truncation following a period of rapid star formation, and one with simple truncation of star-formation is small in their key diagram (H$\\delta$ versus color, as discussed in \\cite{abr96} and \\cite{new90}), since the `active' stage is very brief, and the `post-starburst' stage (marked by strong H$\\delta$ absorption) is longer lived and indistinguishable from the `post-truncation' stage of a truncation model. In fact, they do not clearly test and evaluate the truncation model against a starburst model, and the number of active rapidly star forming galaxies is very small and may not be different from that of the field. \\cite{pog96} use the same data, but with independent models, coming to similar conclusions to \\cite{bar96}. We discuss this point in more detail in \\S~\\ref{sec-models}.  In \\cite{bal97}, we took a different approach to the above - looking at the [O~II] $\\lambda$3727 Equivalent Width (EW) alone, but summing up all the CNOC cluster data. Again we found no evidence for a population with excess star formation relative to the field at the same redshift. Cluster members were shown to have a lower star formation rate than the field regardless of their color. The present detailed study of a single well sampled cluster is meant to complement that more broad-brush method.  Considerable interesting work on galaxy clusters at a range of redshifts around 0.5 (overlapping the redshift of MS1621.5+2640) has been published recently by the ``MORPHS'' collaboration (see \\cite{dre97} and references therein). They have used WFPC-2 data from HST to morphologically classify the objects in these clusters, and interpret these results as showing (amongst other results) that the formation of ellipticals in clusters predates the formation of the rich cluster, and also that S0 galaxies are generated in large numbers only after the cluster virialization. It will be exciting to join our results with those in the ``MORPHS'' papers, but we will not be doing so in this paper. It is difficult to determine which of our current sample of galaxies are morphologically S0, even using the concentration index discussed in \\S~\\ref{sec-data}. The S0 class is a difficult one, involving objects with a wide range of bulge to disk ratio, and requires HST resolution to classify at these redshifts. We will be attempting to make this link using HST imaging of MS1621.5+2640 scheduled to occur over the second half of 1998.  The outline of this paper is as follows: We describe briefly the data and the line and morphology measurements in \\S~\\ref{sec-data}, and go on to summarize the selection effects and biasses in the data in \\S~\\ref{sec-bias}. Cluster membership criteria are explained in \\S~\\ref{sec-sample}. The changes in the various measured galaxy properties with location in the cluster are described in \\S~\\ref{sec-dist}, and ROSAT HRI X-ray observations are presented in \\S~\\ref{sec-xray}. Galaxy evolution models are described in \\S~\\ref{sec-models}, and are used to interpret the observed gradients in galaxy properties. Finally, we present our conclusions in \\S~\\ref{sec-discuss}. ",
        "conclusions": "\\label{sec-discuss}  We list below the main conclusions from the paper. \\begin{enumerate} \\item The blue population and general change in cluster membership begins at a radius of 100 arcsec (or approximately 350 h$^{-1}$ kpc). This is a larger linear radius than in Abell~2390 (approximately 150 h$^{-1}$ kpc).  Thus, the cluster core is considerably less compact in MS1621.5+2640. \\item There is an evolved subgroup some 3 arcmin to the E containing the true `BCG'. \\item The near-field galaxies may have enhanced star-formation. These are at very large radii (greater than 2 h$^{-1}$ Mpc), and the mechanism for this enhancement remains unclear. \\item A gradient is seen in the g-r colors of the red galaxies with projected radius, which we interpret as a gradient in the time since last star formation. This in turn fits well with models where the cluster formed by gradual accretion of field objects, which in turn then evolve (spectroscopically) into early-type galaxies. \\item The overall blue fraction is higher than at lower redshift, both in the field and the cluster. The outer cluster blue fraction matches that of the field, with the exception of the near-field (blue) galaxies. \\item We find no signs of extensive star-formation in the cluster compared with the field. Compared with the field, the average [O~II] is weaker in the cluster, as is H$\\delta$, and the galaxies are redder. This is consistent with simple truncation of star-formation as galaxies are accreted. \\item Detailed comparisons with GISSEL96 models show a good match in general except for an odd excess in H$\\delta$ index for all galaxies redder than g-r=1.2. Various attempts to explain this excess (resolutions differences, reddening, metallicity, truncated IMFs) were not successful, but certainly deserve further study. \\item At this redshift, the models allow us to distinguish two galaxy populations in the cluster: one best modeled with steadily decreasing star-formation, similar to that seen in spirals in the field, and another in which star-formation has terminated. \\end{enumerate}  To summarize further, and also indulge in some speculation, the above data, combined with that on Abell 2390, seems to be consistent with a fairly passive process of cluster formation. Galaxies are gradually accreted from the field, with no good evidence for a large burst of star formation as they are virialized into the cluster. The star formation in the galaxies begins shutting down at very large radii (well outside the virial radius), for reasons that remain unclear. However, there is evidence for recent (within the last 1-2 Gyr) star formation even in the inner 100 arcsec of the cluster. This star formation is mostly seen in the lower-luminosity early-type galaxies in the cluster core.  This suggests to us that the shut down process for star formation could well be some combination of stripping of gas from the outer parts of the galaxies as they fall in, followed by gradual exhaustion of the gas in the inner parts. This would allow the remnants of star formation to be still visible in galaxies seen in the core. One hundred arcsec at this redshifts is 350 h$^{-1}$ kpc, which would take 0.4 Gyr to cross, going at the cluster velocity dispersion speed of 850 km/s. The fact that the inner galaxies with stronger H$\\delta$ have lower luminosities could either imply that they have had different dynamical histories (i.e. the more luminous galaxies either formed there or fell in earlier), or that the lower luminosity galaxies were more gas rich, and hence more able to hang on to their star formation in the presence of stripping. One should also remember the general trend that more luminous galaxies even in the field tend to have formed their stars earlier than the lower luminosity ones."
    },
    "9805/astro-ph9805020_arXiv.txt": {
        "abstract": "We present new X-ray observations of the X-ray pulsar 1E 2259+586, obtained during March 1997, with the Rossi X-Ray Timing Explorer (RXTE). We have measured the pulse frequency derivative $\\dot \\nu = (-1.08 \\pm 0.04) \\times 10^{-14}$ Hz s$^{-1}$ from pulse arrival times obtained in a sequence of 5 observations spread over one month. This $\\dot\\nu$ is consistent with the long term spin-down trend. We also found that the observed X-ray luminosity is consistent with that measured at quiescent X-ray flux levels by previous missions. Our observations imply that 1E 2259+586 was spinning down steadily without exhibiting any stochastic torque noise fluctuations during the month covered by our observations. ",
        "introduction": "The X-ray pulsar 1E 2259+586 is located at the geometric center of the semi-circular shell of diffuse X-ray emission of the supernova remnant G109$-$1.0 (Gregory \\& Fahlman 1980).  Radio observations indicate that the age of G109$-$1.0 is about $10^4$ yr and the distance to it is 3.6--4.7 kpc (Hughes et al. 1984). The pulse frequency of the pulsar is 0.1433 Hz (pulse period 6.8 sec) and the average spin down rate since its discovery is $\\dot \\nu = -1 \\times 10 ^{-14}$ Hz s$^{-1}$.  If 1E 2259+586 is an isolated, spinning-down pulsar, this rate yields a rotational energy loss of 10$^{31}$ erg s$^{-1}$ which is much less then the intrinsic X-ray luminosity of $\\sim $10$^{35}$ erg s$^{-1}$ (Corbet 1995).  Koyama et al., (1987, 1989) claimed that the spin-down rate and unabsorbed X-ray luminosity can be understood if 1E 2259+586 is in a binary system with a neutron star magnetic field of 5$\\times 10 ^{11}$ Gauss and spinning close to the equilibrium period (Ghosh \\& Lamb 1979).  However, efforts to find orbitally induced Doppler shifts in the X-ray pulsations which would reveal the binary period have been unsuccessful.  The upper limits of $a_{x} /c \\sin i$ from previous missions such as Einstein, EXOSAT, Ginga and ROSAT have been very small.  For example, recent RXTE observations (Mereghetti et al., 1998) have given the lowest upper limit $a_{x}/c \\sin i < 30$ msec at the 99\\% confidence level.  According to these new upper limits, if 1E 2259+586 is in a binary system, its companion star must be either a white dwarf, or a helium-burning star with $M < 0.8 M_{\\circ}$ (Mereghetti et al. 1998). Alternatively, 1E 2259+586 could be an isolated star which is accreting from a disk and is formed by remnants of the common envelope evolution of a high-mass X-ray binary (van Paradijs et al. 1995, Ghosh et al. 1997).  Analyses of the ROSAT observations (Baykal \\& Swank 1996) have shown that the source is not steadily spinning-down.  A spin-up episode which is superposed on a long term spin-down trend was found. The deviations from the secular spin-down trend are also consistent with other accretion powered X-ray pulsars (Baykal \\& {\\\"O}gelman 1993), which are several orders of magnitude greater than those of radio pulsars. This result favors accretion onto a neutron star from a very low mass companion or from a residual disk. In this research note, we present the pulse timing results and X-ray spectrum of 1E 2259+586 from RXTE observations.  The main goal of these new RXTE observations is to resolve the pulse frequency derivative $(\\dot \\nu)$ during the observation time span and measure the source X-ray luminosity. These quantities enable us to further test the current accretion hypothesis for this source. ",
        "conclusions": ""
    },
    "9805/astro-ph9805166_arXiv.txt": {
        "abstract": "We present the first results of an ongoing project to investigate the present-day chemical abundances of the extreme outer parts of galactic disks, as probed by the emission line spectra of a  new sample of HII regions.   The galaxies studied here, NGC~628, NGC~1058 and NGC~6946, are all late-type spiral galaxies, characterized by larger than average HI-to-optical sizes. Our  deep H$\\alpha$ images have revealed the existence of recent massive star formation, traced by HII regions, out to, and beyond, two optical radii in these galaxies (defined by the B-band 25th magnitude isophote).  Optical spectra of these newly-discovered HII regions are used to investigate their densities, ionization parameters, extinctions and in particular their oxygen and nitrogen abundances.  Our measurements reveal gas-phase abundances of O/H~$\\sim$~10-15\\%~ of the solar value, and N/O~$\\sim$~20-25\\%~ of the solar value, at radii of 1.5--2~R$_{25}$.  Clear evidence also exists for diminished dust extinction (A$_V$~$\\sim$~0--0.2) at large radii.  The combination of our measurements of outer disk HII region abundances with those for inner disk HII regions published in the literature is a powerful probe of the shape of abundance gradients over unprecedented radial baselines.  The predictions of models of chemical evolution often diverge most strongly in the outer parts of galaxies.  Both the oxygen and the nitrogen-to-oxygen abundances generally decrease with increasing radius. Within the limits of the current dataset, the radial abundance variations are consistent with  single log-linear relationships, although the derived slopes can often differ considerably from those found if only inner disk HII regions are used to define the fit.  The small number of HII regions in our present sample, together with uncertainties in the calibrations of the empirical methods used here to determine abundances,  limit the ability to constrain both subtle changes in the radial gradient and intrinsic scatter at a fixed radius.  Nitrogen-to-oxygen ratios appear to be consistent with a combination of primary and secondary production of nitrogen.  Interestingly, both the mean level of enrichment and the ratio of N/O measured in extreme outer galactic disks are similar to those values measured in some high redshift damped Lyman-$\\alpha$ absorbers, suggesting that outer disks at the present epoch are relatively unevolved. ",
        "introduction": "Radial variations in the abundances of elements within galaxies are well-established as providing important constraints on models of disk galaxy formation and  evolution (eg. Pagel \\& Edmunds 1981; Vila-Costas \\& Edmunds 1992; Zaritsky, Kennicutt \\& Huchra 1994 (hereafter ZKH); Prantzos \\& Aubert 1995).  HII regions play a unique role in such studies because they yield relatively reliable elemental abundances from measurements of emission line intensities.  The early work of Searle (1971) and Shields (1974) used observations of HII regions in spirals to establish the existence of negative radial gradients  in the abundance of oxygen.  Subsequent work has shown that such gradients are a generic feature of disk galaxies, although the magnitude and the shape of the gradient, as well the characteristic abundance,  are observed to vary considerably from galaxy to galaxy (eg. Vila-Costas \\& Edmunds 1992; ZKH).  Unfortunately, most HII region abundance studies carried out to date have probed only the bright, easily-observed inner regions of galactic disks, lying at or within the classical optical radius, R$_{25}$ (defined by the B-band 25th magnitude isophote).  It is well known, however,  that disk galaxies have HI  disks which extend to typically $\\gtrsim$ 1.5--2~\\ropt (eg. Cayette et al 1994; Broeils 1994), and in some rare cases to $\\gtrsim$~3\\ropt (eg. van der Kruit \\& Shostak 1984).   These outer regions are characterized by low HI columns, high gas fractions and long dynamical timescales and thus provide an opportunity to study star formation and chemical evolution in rather unique physical environments.  Indeed, the outer regions of disks have physical properties which are reminiscent of those thought to exist during the early stages of galaxy formation, as well as those in giant low surface brightness galaxies, such as Malin 1 (eg. Pickering et al 1997).  Studying  the mean enrichment level, the shape of the abundance gradient and the amount of intrinsic scatter at fixed radius, in outer galactic disks  will therefore forward our understanding of a broad range of astrophysical objects.  Furthermore, since the predictions of models of chemical evolution often diverge most strongly in the outer parts of galaxies, abundance determinations extending as far in the outer disk as possible are needed to discriminate between competing theories.  Only a handful of outer disk HII regions have measured abundances and, of these, very few lie at galactocentric radii significantly beyond the optical radius (eg.  Garnett \\& Shields 1987; Garnett et al 1992; Garnett \\& Kennicutt 1994).  The bias of previous studies of HII regions to the inner disk  has certainly been due in part to the lack of known HII regions lying at large radii.  In the course of a large, deep, H$\\alpha$ imaging survey to search for and study star formation in the extreme outer regions of disk galaxies (Ferguson 1997), we have discovered numerous faint HII regions in the low gas surface density (N$_{HI}~\\lesssim$~few~$\\times~10^{20}$~cm$^{-2}$) outer limits  of several galaxies.  These HII regions are typically small (diameter~$\\sim$~150--500~pc) and of low luminosity (L$_{H\\alpha}$~$\\sim$~1--10~L$_{Orion}$), and do not clump into the giant complexes which populate the inner disks  of spiral galaxies. They are often observed to trace out narrow spiral arms, which are sometimes coincident with underlying HI arms and faint stellar arms ($\\mu_B \\sim$ 26--28 mag/$\\sq\\arcsec$).  A study of the ongoing and past  star formation in the outer disks of these galaxies and its implications for models of disk galaxy evolution will be presented elsewhere (Ferguson et al, 1998a,b).  We present here the first results from a long-slit spectroscopic study to determine the chemical abundances in the extreme outer parts of nearby disk galaxies, using the emission line spectra of our newly discovered HII regions.  This paper presents results for three late-type spiral galaxies, namely NGC~628, NGC~1058 and NGC~6946.   These galaxies have particularly extended disks of neutral hydrogen and are amongst our best examples of galaxies with extreme outer disk star formation.   A summary of their properties is presented in Table 1.  Chemical abundances for HII regions in the inner disks of NGC~628 and NGC~6946 have been derived by McCall, Rybski \\& Shields (1985: hereafter MRS). Chemical abundances have not been previously measured for HII regions in NGC~1058, and we present here a study of both inner and outer disk HII regions in that galaxy. ",
        "conclusions": "\\subsection{Constraining Outer Galactic Abundance Gradients}  The results presented here constitute the largest set of extreme outer disk abundances ever measured, and probe the chemical abundance content of these optically faint, previously unexplored regions of present day galactic disks.   We have found that the outermost HII regions studied have O/H abundances in the range  of 10--15\\% solar and  N/O abundances in the range of 20--25\\% solar.  Such low abundances have rarely been measured before in spiral disks (eg. Garnett \\& Kennicutt 1994).  We have shown that the outermost abundances play an important role in defining abundance gradients across the disks, and indeed often change the  nature of the gradient.  For example, in two of the galaxies studied here, the oxygen gradients are observed to steepen considerably when account is taken of the outermost HII regions.  Within the limits of the current dataset, the radial abundance gradients are consistent with single log-linear relationships, although hints of interesting behaviour can been seen at large radii in two of the galaxies.  There could be a flattening of the oxygen abundance beyond the edge of the optical disk in NGC~628 (although this result largely hinges on the metallicity of the outermost HII region), as well as a  steepening of the outer gradient in NGC~1058.  Unfortunately, given the relatively large errors in our abundance determinations, as well as the relatively small size of the current sample, it is not yet possible to assess the significance of these features.   Similarly, it is not yet possible to constrain the amount of intrinsic scatter in the abundances as a function of galactocentric radius, which is of interest for understanding the efficiency and timescales for elemental mixing  at large radii.  While the two outermost HII regions in NGC~1058 show differences of $\\gtrsim$ 0.2 dex in both O/H and N/O despite lying at very similar radii, a knowledge of the electron temperatures of these regions will be required before we can distinguish between real scatter and uncertainties inherent in the model calibrations.  Chemical evolution models predict a variety of different behaviours for galactic abundance gradients (eg. Prantzos \\& Aubert 1995, Molla et al 1996, Wyse \\& Silk 1989, Clarke 1989).   Of these models, some predict simple exponential declines while others produce steepenings or flattenings in the outer disk.    Our current dataset is not yet sufficient to discriminate between various chemical evolution models, but  we are continuing to obtain more and better data on these and other galaxies, spanning a range of physical environments and Hubble types.  \\subsection{Comparison to Other Galaxies}  It is of great interest to compare our measurements of outer disk abundances with the few published measurements in the literature.  High quality measurements exist for the Galaxy (Fich \\& Silkey 1991; Vilchez \\& Esteban 1996;  Rudolph et al 1997; Afflerbach et al 1997), M81 (Garnett \\& Shields 1987) and M101 (Garnett \\& Kennicutt 1994), but these measurements do not extend to the extreme galactocentric radii that we have studied here.  In case of Galaxy, outer disk HII regions have been  studied via optical and far-IR techniques out to 1.3$R_{edge}$, where  R$_{edge}$, the edge of the optical stellar disk, is taken to be 14~kpc (Ruphy et al 1997)\\footnote{It remains unclear as to the relationship between the optical edge, as defined by star counts, and the 25th B-magnitude isophote.}.  Both Fich \\& Silkey (1991) and Vilchez \\& Esteban (1996) found evidence for flat gradients in the nitrogen abundance beyond the solar circle; additionally, Vilchez \\& Esteban (1996) found evidence for only a mild outer gradient in oxygen abundance, with outer values of $\\sim$~20\\% solar, and outer disk values of N/O consistent with those measured in the solar neighbourhood.  More recently, Rudolph et al (1997) have re-examined outer abundance gradients using FIR lines for a sample of 5 HII regions.  When combined with the results of other studies, these authors do not see compelling evidence for a flattening of  the outer abundance gradients, and are able to fit the available data  with single log-linear relationships.  They caution, however, that their present sample is not yet sufficient to rule out the possible existence of a flattening  of the abundance gradient in the outer Galaxy.   They also find the variation of N/O throughout the Galactic disk to be consistent with a step function, with a mean log(N/O) of $-$0.50~$\\pm$~0.02 for R$<$6.2 kpc and, $-$0.83~$\\pm$~0.04 for R$>$6.2 kpc.  We note that HII regions have been discovered at much larger Galactocentric radii, out to $\\sim$ 28 kpc or 2R$_{edge}$ (de Geus et al 1993) however no abundance determinations have been made for these as of yet.  In M81, the outermost HII region (lying at $\\sim$ 1.3~R$_{25}$) has an O/H abundance of roughly 20\\% solar, consistent with an extrapolation of the inner gradient, whereas the N/O abundance is close to solar and consistent with there being no gradient across the disk (Garnett \\& Shields 1987). On the other hand,  the outermost HII region studied in M101 ($\\sim$ 1.1~R$_{25}$) has an O/H abundance of only 10\\% solar and an N/O abundance of 25\\% solar; both N/O and O/H are observed to decrease more or less smoothly with increasing galactocentric radius across the disk (Garnett \\& Kennicutt 1994).  Curiously, these abundances are as low as those observed in the extremeties of the galaxies studied here, despite that fact that we have probed the gas at significantly larger distances beyond the optical disk.  We find that the galaxies studied in the present work are more akin to M101 than the Galaxy or M81 since the outermost abundances of both O/H and N/O  are observed to decrease more or less smoothly across their disks.   The striking difference between the level of N/O enrichment seen in the outer disks of galaxies studied here (and M101)  and in the outer disks of the Galaxy and M81 is particularly puzzling and clearly warrants further study.  \\subsection{Implications for Understanding the Evolution of Galactic Disks}  Our outer disk measurements are largely consistent with those observed in  other low gas surface density objects, such as gas-rich dwarf irregulars (eg. Garnett 1990; Skillman et al 1989; Skillman et al 1997; Thuan et al 1995; Miller \\& Hodge 1996)\\footnote{Note however that the measured outer disk metallicities are still considerably in excess of the most metal poor, gas-rich objects known locally, eg. IZw18, UGC4483 and SBS 0335-052. These objects  have O/H abundances of only 2--3\\% solar (eg. Skillman \\& Kennicutt 1993, Skillman et al 1994, Izotov et al 1997), even although there is evidence for moderately long periods ($\\gtrsim$~10~Myr) of star formation in at least one case (eg.  Garnett et al 1997b).}  and some low surface brightness galaxies (McGaugh 1994).  It has often been suggested that outer galactic disks are $`$built' up through the accretion of gas, either smoothly (eg. Gunn \\& Gott 1972; Larson 1976) or as gas-rich low mass companions (eg. White \\& Rees 1987; Kauffman et al 1993; Kamphuis 1993; Zaritsky 1995).  There would appear to be no obvious argument against this hypothesis on the basis of chemical abundance content alone.  On the other hand, one might consider the scenario in which outer galactic disks evolve in relative isolation, with little  inflow or outflow.  The simple $`$closed box' model of Schmidt (1963) can be used to predict the mean metallicity expected for such regions, under the assumption of instantaneous recycling (appropriate for oxygen).   The closed box model can be represented by a simple relation (Searle \\& Sargent (1972)) $$\\mathrm Z=-p~ln~\\mu $$ where  p is the yield of the element in question and $\\mu$ is the gas fraction, defined as baryonic mass in gas to the total baryonic mass (stars + gas).  Our deep B-band surface photometry yields B-band surface brightnesses at two optical radii of $\\sim$~0.1~L$_{\\sun}$~pc$^{-2}$ in NGC~628 and NGC~6946 and $\\sim$~1.2~L$_{\\sun}$~pc$^{-2}$ in NGC~1058 (Ferguson et al 1998b). Under the assumption of M/L~$\\sim$~2, this leads to surface mass densities of 0.2 and 2.4~M$_{\\sun}$~pc$^{-2}$.  The B-band M/L ratio is sensitive to the star formation history of the outer disk, but is unlikely to be too much larger than the adopted value, which is found for the solar neighbourhood (Kuijken \\& Gilmore 1989).  Inspection of the HI maps of these galaxies reveals typical HI surface densities of $\\sim$~1~M$_{\\sun}$~pc$^{-2}$ (NGC~628) to $\\sim$~3~M$_{\\sun}$~pc$^{-2}$ (NGC~1058, NGC~6946) in the extreme outer disks (van der Kruit \\& Shostak 1984, Shostak \\& van der Kruit 1984, Kamphuis 1993).  Correcting for helium (a factor of 1.3), we derive gas fractions of $\\sim$~0.6 in NGC~1058 and $\\sim$~0.90 in NGC~628, NGC~6946.  Adopting a yield of 0.5~Z$_{\\sun}$, consistent with the observed mean metallicity of stars in the solar neighbourhood (Wyse \\& Gilmore 1995), the simple model then predicts  oxygen abundances of 6\\% solar for NGC~628, 25\\% solar for NGC~1058 and 2\\% solar for NGC~6946. These values can be compared to the observed metallicities of 10--15\\% solar.  Despite the many uncertainties involved, both NGC~628 and NGC~1058 have metallicities which lie within a factor of two of the closed--box model predictions, thus suggesting that the role of gas flows in the evolution of their outer disks is similar to that for the solar neighbourhood.  The outer disk of NGC~6946, on the other hand, does not appear to fit this picture. In the future, we plan to investigate the predictions of the simple model in more detail, including study of the radial variation in the derived effective yield.  \\subsection{Comparison to High Redshift Damped Lyman-$\\alpha$ Systems}  Much recent attention has been focused on uncovering the nature of the damped Ly$\\alpha$ systems (DLAs)  which cause absorption in the spectra of high redshift quasars.  While there remains much debate on this topic, follow-up imaging and spectroscopy of these systems lends support for the idea that at least some of them are young disk galaxies (eg. Wolfe 1988;  Briggs et al 1989;  Djorgovski et al 1996).  The distribution of impact parameters for these systems is not well-defined; however, in the particular case of Djorgovski et al (1996), an impact parameter of 18~kpc is inferred which clearly places the line-of-sight in the outer parts of what appears to be a large disk galaxy at z$=$3.15.   How do the chemical abundances measured in the DLAs compare to those measured in the extended parts of local disk galaxies?  There have been many recent efforts to measure the gas-phase chemical abundances in DLA systems (eg. Pettini et al 1994, 1995, 1997; Wolfe et al 1995; Lu et al 1996, 1998); studies have generally found them  to exhibit a wide range of low metallicities, ranging from less than 1/100th of the solar value to 1/10th solar, and showing a trend of increasing metallicity with decreasing redshift, albeit with considerable scatter at all redshifts (eg. Lu et al 1996, Pettini et al 1997).   Pettini et al (1997) calculate a column density-weighted mean metallicity of 1/13th solar for their sample of 34 DLAs (0.7$<$z$<$3.4), using the Zn lines which should be relatively unaffected by depletion onto dust.\\footnote{Note that the average HI column for this sample of DLA systems is roughly 10$^{21}$~cm$^{-2}$, which is several times higher than the typical HI column where our sample of outer HII regions reside (a few times 10$^{20}$~cm$^{-2}$).} This comparison suggests an intriguing  similarity in the mean chemical enrichment level of DLA systems  and present-day outer galactic disks, with DLA systems being only slightly more metal poor.  \\begin{figure} \\plotone{ferguson.fig13.ps} \\caption{Variation of log(N/O) vs 12+log(O/H) for our sample of HII regions, as well as the inner disk HII regions measured by MRS. Typical uncertainties in our measurements are indicated by the error bar in the lower left.  The location of the Sun in this diagram is indicated by a solid star.  We also plot values for the most metal poor object known in the local Universe, IZw18 (Skillman \\& Kennicutt 1993), as well as the N/Si and Si/H measurements for a sample  of high redshift DLA  systems from Lu et al (1998). (Note we have not shown the error bars for these data, but only the upper and lower limits). The dashed curves indicate the expectations for a pure secondary and primary+secondary origins for nitrogen from Vila-Costas \\& Edmunds (1993).  The extreme outer disk HII regions are consistent with a combination of both primary and secondary production of nitrogen.\\label{nooh}} \\end{figure}  Figure \\ref{nooh} shows a a plot of  nitrogen-to-oxygen abundance as a function of O/H  for our present sample of HII regions, as well as those inner disk regions from MRS.  We have indicated where IZw18, the most metal poor gas-rich object in the local Universe,  falls on this diagram (Skillman \\& Kennicutt 1993). Also shown are a set of  N/Si measurements (or upper limits) for a sample of 15 DLA systems with z$>$2  from Lu et al (1998). These authors argue that N/Si is equivalent to N/O, under the assumption that O/Si is approximately solar in DLA systems and that dust depletion is not significant.  This plot reveals that while many DLA systems in the Lu et al sample are considerably more metal poor than present-day outer galactic disks, others  are only slightly displaced from outer disks in terms of O/H and almost identical in terms of N/O.  The long dynamical timescales ($\\sim$10$^8$--10$^9$~years) which characterize the extreme outer disks imply that they evolve at a very slow rate but, nonetheless, we would expect them to have been more metal poor in the past than at the present epoch.  We therefore conclude that while, on average,   DLA systems tend to be slightly more metal poor than outer galactic disks at present, the two were very likely indistinguishable at large lookback times.   \\subsection{The Origin of Nitrogen}  A major unsolved question is  the importance of primary and secondary processes in the production of nitrogen.  Nitrogen is believed to be mostly a product of secondary nucleosynthesis, being produced via the CNO cycle, however it is also thought to have primary component which can be produced in the earlier helium burning stages (Renzini \\& Voli 1981).  It is thought that secondary nitrogen is produced by stars of a wide range of masses whereas primary nitrogen is produced by only intermediate mass stars (4~$\\lesssim$~M/M$_{\\sun}$~$\\lesssim$~8) (Renzini \\& Voli 1981, Matteucci 1986); there is evidence however that nitrogen produced in massive stars is also predominantly primary. As an example, Laird (1985) finds that [N/Fe] is constant for solar-neighbourhood stars with -1.8$\\le$[Fe/H]$\\le$0.5, implying that nitrogen and iron are produced in the same way, by the same stars.  Since it is well known that some iron is primary, and produced on short timescales by Type II supernovae from the core-collapse of massive stars (eg. Arnett 1995), then one expects that such stars must also contribute to nitrogen production.  The simple model for the chemical evolution of galaxies (ie.  closed box) predicts that N/O will be independent of O/H for a primary origin of nitrogen, and proportional to O/H for a secondary origin (see Vila-Costas \\& Edmunds 1993 for a discussion, including the complication of $`$delayed' primary production).  The timescales for nitrogen production differ in stars of varying mass. If nitrogen is predominantly produced by massive stars, then there should be no time delay between the release of the element with respect to oxygen,  and thus one expects the N/O ratio to exhibit only a small amount of scatter. If, on the other hand, most nitrogen comes from intermediate mass stars, then one expects a time delay of up to a few times 10$^8$~years between the release of N/O. This is expected to introduce a large scatter in the N/O ratio at low O/H, decreasing with increasing metallicity as the effects of the time delay become less and less important (eg. Garnett 1990; Pilyugin 1992).  In Figure \\ref{nooh}, we have overplotted  the expectations for primary and secondary production of nitrogen, as taken from Vila-Costas \\& Edmunds (1993).  At high metallicities, the expected linear trend between O/H and N/O is seen, indicative of the dominant role of secondary nitrogen production.  On the other hand, the extreme outer disk HII regions are seen to  populate the region bracketed by the pure secondary and primary + secondary curves, and thus appear consistent with a combination of  primary and secondary production of nitrogen. The spread in N/O at fixed O/H in the outer disk may very well reflect the timescale between $`$bursts' of star formation, produced perhaps by the passage of a spiral arm.   Further measurements, with improved accuracy, are needed to better understand these results."
    },
    "9805/astro-ph9805200.txt": {
        "abstract": "The High-Z Supernova Search is an international collaboration to discover and monitor type Ia supernovae (SN~Ia) at $z > 0.2$ with the aim of measuring cosmic deceleration and global curvature. Our collaboration has pursued a basic understanding of supernovae in the nearby Universe, discovering and observing a large sample of objects, and developing methods to measure accurate distances with SN~Ia. This paper describes the extension of this program to $z \\geq 0.2$, outlining our search techniques and follow-up program. We have devised high-throughput filters which provide accurate two-color restframe $B$ and $V$ light curves of SN~Ia, enabling us to produce precise, extinction-corrected luminosity distances in the range $0.25 < z < 0.55$. Sources of systematic error from K-corrections, extinction, selection effects, and evolution are investigated, and their effects estimated.  We present photometric and spectral observations of SN~1995K, our program's first supernova, and use the data to obtain a precise measurement of the luminosity distance to the $z=0.479$ host galaxy. This object, when combined with a nearby sample of SN, yields an estimate for the matter density of the Universe of $\\Omega_M = -0.2^{+1.0}_{-0.8}$ if $\\Omega_\\Lambda = 0$. For a spatially flat universe composed of normal matter and a cosmological constant, we find $\\Omega_M = 0.4^{+0.5}_{-0.4}$, $\\Omega_\\Lambda = 0.6^{+0.4}_{-0.5}$. We demonstrate that with a sample of $\\sim 30$ objects, we should be able to determine relative luminosity distances over the range $0 < z< 0.5$ with sufficient precision to measure $\\Omega_M$ with an uncertainty of $\\pm 0.2$. ",
        "introduction": "Measuring the cosmological parameters which describe the global properties of the Universe has been a fundamental quest in astronomy ever since Robertson \\markcite{R36}(1936) and Walker \\markcite{W36}(1936) formulated the metric for a homogeneous and isotropic Universe. By observing how a standard candle dims as a function of redshift, usually shown as a Hubble diagram, the effects of curvature and cosmic deceleration can be observed and quantified (\\markcite{S61}Sandage 1961). Early luminosity distance investigations (\\markcite{HMS56}Humason, Mayall, \\& Sandage 1956; \\markcite{B57}Baum 1957; Minkowski 1960\\markcite{M60}) used brightest cluster galaxies as standard candles and measured galaxy brightnesses in the range $0.01 < z < 0.5$. Attempts to trace luminosity distances versus redshifts with these galaxies at $z>0.1$ changed emphasis when it was realized both from theory (\\markcite{T72}Tinsley 1972) and observation (e.g., \\markcite{OGH96}Oke, Gunn, \\& Hoessel 1996) that the effects of galaxy evolution are much larger than the differences due to cosmology.  Although many other methods for measuring the global curvature and cosmic deceleration exist (e.g., Peebles \\markcite{Peebles93} 1993), none has yet delivered a definitive result.  For example, measuring the number of galaxies as a function of magnitude maps out the volume of space as a function of redshift, and can be used to gauge the global geometry. Attempts to use this method \\markcite{S84}(Shanks et al. 1984) have been hampered by galaxy evolution and merging, although some of the uncertainty may be eliminated by moving to the infrared (\\markcite{yp95}Yoshi \\& Peterson 1995). Another test examines the angular size of a standard rod as a function of redshift. Kellerman \\markcite{K93}(1993) resolved a large sample of compact radio sources out to $z\\approx3$ using Very Long Baseline Interferometry. Although the angular sizes increase at $z> 1$ as expected for a non-empty Universe, evolutionary effects are hard to quantify. Stepanas \\& Saha \\markcite{SS95}(1995) have also shown that the unknown intrinsic distribution of source sizes makes it difficult to obtain a statistically significant measurement of cosmological parameters. Guerra \\& Daly (1998)\\markcite{GD98} have also used extended radio galaxies as standard rods and show that the results are consistent with a low density universe.  Type Ia supernovae (SN~Ia) have long been considered promising tools for measuring extragalactic luminosity distances, but only recent searches, the resulting sets of light curves and spectra, and new methods of analysis (\\markcite {P93}Phillips 1993 [P93]; \\markcite{H96a}Hamuy et al. 1995, 1996a-d [H95,H96a-d]; \\markcite{RPK95} Riess, Press, \\& Kirshner 1995, 1996a [RPK95, RPK96]) have quantified the nature, power, and limitations of SN~Ia as distance indicators. SN~Ia offer high intrinsic luminosity [$M_B\\approx-19.4$~mag (\\markcite{Setal97}Saha et al. 1997)] and as individual stars, may not be subject to the same evolutionary effects which plague galaxies (although this must be demonstrated). Even before this breakthrough in understanding SN~Ia, two searches for distant SN~Ia were initiated (\\markcite{NN89}N\\o rgaard-Nielsen et al. 1989; \\markcite{Pe95} Perlmutter et al. 1995) to measure cosmological parameters. These searches demonstrate that it is possible to find SN at $z > 0.3$ with large-format CCD detectors, and give hope that a significant sample of SN~Ia can be gathered in just a few years. Perlmutter et al. (1997, \\markcite{Pe98}1998) have already presented observations of 8 objects at $z > 0.35$. A  sample of $\\sim 30$ objects, if carefully measured and shrewdly analyzed,  will provide a statistically interesting measurement of  global cosmological parameters.  Recently, much effort has been put into examining how to use the power spectrum of fluctuations in the cosmic microwave background (CMB) to measure cosmological parameters (see\\markcite{Hu96} Hu 1996 for a review). Future satellite missions, such as the Microwave Anisotropy Probe and Planck Missions, will measure  temperature fluctuations across the sky on scales substantially less than a degree, determining the power spectrum of fluctuations out to a multipole expansion of $\\ell >500$. Model fits to these observations promise to provide simultaneous measurement of ten cosmological parameters. However, because the fits are degenerate for certain combinations of interesting parameters such as $\\Omega_M$ and $\\Omega_\\Lambda$, the CMB observations will need to be combined with other observational data, such as those from high-redshift supernovae (\\markcite{ZSS97}Zaldarriaga, Spergel, \\& Seljak 1997), to determine $\\Omega_M$ and $\\Omega_\\Lambda$ individually.  This paper reports on the High-Z SN Search, a coordinated program to discover, spectroscopically classify, and measure photometrically in at least two filters a distant set of SN~Ia. This enterprise aims to measure the deceleration parameter, $q_0$, with an uncertainty smaller than $0.1$, and will place strong limits on combinations of cosmological parameters such as $\\Omega_M - \\Omega_\\Lambda$. The CMB observations provide a nearly orthogonal set of parameters, so we will be able to separate the effects of any exotic forms of matter-energy in the Universe from normal matter.  We will be able to discern whether the Universe is open, closed, or has nearly zero global curvature. Preliminary results from our program were reported by Schmidt et al. (\\markcite{S96}1996), Schmidt \\markcite{S97}(1997), and Leibundgut \\& Spyromilio (1997)\\markcite{LS97}. In addition, we have confirmed the predicted time-dilation of redshifted objects using SN~Ia light curves (Leibundgut et al. 1996\\markcite{L96}) and spectra (Riess et al. 1997 \\markcite{retal97}), and presented observations of 3 objects observed with {\\it HST} (Garnavich et al. 1998).  In \\S 2 we describe how the expansion, deceleration, and curvature of the Universe are related to luminosity distances, and in \\S 3 we discuss measuring distances with SN~Ia at $z<0.1$. Our supernova search and observational follow-up program are outlined in \\S 4. In \\S 5 we present the techniques and limitations of using SN~Ia to measure accurate luminosity distances at $z>0.2$. Observations of the first SN discovered by this program, SN~1995K at z=0.479, are presented  in \\S 6, with the techniques discussed in \\S 3-5 applied to this object. We summarize the High-Z SN Search to date in \\S 7 and use the results for SN~1995K to estimate the precision with which we will be able to constrain cosmological parameters. ",
        "conclusions": ""
    },
    "9805/astro-ph9805172_arXiv.txt": {
        "abstract": "We have detected HI  21cm line absorption by the Warm Neutral Medium (WNM) using the Westerbork Synthesis Radio Telescope. The absorption was detected toward Cygnus A at LSR velocities of --40 and --70 km s$^{-1}$. These two velocity ranges were previously identified as being relatively free of cold absorbing clouds. The measured  optical depth for the WNM along the line of sight to Cygnus A is 8.9$\\pm$1.9 $\\times$10$^{-4}$ at --70 km s$^{-1}$, and 8.5$\\pm$2.0 $\\times$10$^{-4}$ at --40 km s$^{-1}$, with corresponding spin temperatures of 6000$\\pm$1700 K and 4800$\\pm$1600 K, respectively. The volume filling factor for the WNM appears to be fairly high (f $\\approx$ 0.4).  {\\sl Subject headings: ISM: structure and atoms -- Galaxy: fundamental parameters -- Radio Lines: ISM} ",
        "introduction": "The CNM-free LSR velocity ranges toward Cygnus A are from --40 to --36 km s$^{-1}$, and --63 to --78 km s$^{-1}$, corresponding to inter-arm velocities,  between the Orion (0 km s$^{-1}$), Perseus (--50  km s$^{-1}$), and Outer (-85 km s$^{-1}$) spiral arms (Mebold and Hills 1975). We have searched for absorption by the WNM at these inter-arm velocities using the  WSRT. Using an interferometer mitigates problems of confusion by stray radiation and in-beam emission. Observations were made on May 24 and 25, 1996, July 15 and 16, 1996, June 28 and 29, 1997, and August 22 and 23, 1997, for 12 hours each day. The total bandwidth was 1.25 MHz with 255 spectral channels after Hanning smoothing, corresponding to 1 km s$^{-1}$ velocity resolution, centered at an LSR velocity of --45 km s$^{-1}$.  Absolute flux calibration and initial phase calibration were performed using 3C 286, and residual phase errors were corrected through self-calibration using a continuum data set made from line-free channels.  An accurate bandpass was determined using frequency switching to `high' and `low' frequencies by $\\pm$ 1.25 MHz on Cygnus A itself. The cycle time for bandpass calibration was two hours, with a 50$\\%$ duty cycle, and the bandpass for each cycle was the average of the high and low frequency settings. An estimate of the accuracy of the bandpass calibration process was obtained by applying the `high' frequency bandpass solutions to the `low' frequency spectra. The residual fractional deviations were generally $\\le$ 1.7$\\times$10$^{-4}$. We adopt this value as a conservative estimate of optical depth errors introduced by the bandpass calibration process.   The visibility data from each day's observation were reduced independently using the Astronomical Image Processing System (AIPS). The continuum emission was subtracted using the AIPS task UVLIN, which performs a linear fit to the line-free channels for each visibility, and the data from the eight days were then combined. Spectral line image `cubes' were constructed from the combined data, and the point response function of the array deconvolved, using the AIPS task IMAGR. The Gaussian restoring CLEAN beam  had a FWHM = 30$''$$\\times$15$''$ with a major axis oriented north-south.  The continuum image of Cygnus A made from line-free channels is shown in Figure 1. The observed total flux density at 1.4 GHz for Cygnus A was within 2$\\%$ of the expected value of 1575 Jy (Baars et al. 1977). The source is characterized by two `hot spots' of emission at the source extremities, separated by 2$'$, with a low frequency `bridge' of emission connecting the two hot spots. In the following section we analyze spectra taken at the the Northwest (NW) and Southeast (SE) hot spots. The peak surface brightness is 346 Jy beam$^{-1}$ for the NW hot spot and 398 Jy beam$^{-1}$ for the SE hot spot.  The HI absorption spectra toward the SE and NW peaks of Cygnus A are shown in Figure 2. The spectra show deep, narrow lines due to the CNM, clustered around the spiral arm velocities (0, --50,  and --85 km s$^{-1}$), with optical depths approaching 0.5. There are significant variations in optical depth for these narrow lines between the two lines of sight separated by 2$'$. Small scale structure in the CNM is a well studied phenomenon (Dieter et al. 1976, Diamond et al. 1989, Deshpande et al. 1992, Frail et al. 1994, Davis, Diamond, and Goss 1996), and a detailed analysis of the CNM structure toward Cygnus A will be presented elsewhere. The root mean square (rms) noise per channel in these spectra is 0.1 mJy channel$^{-1}$, corresponding to an rms optical depth ($\\sigma$) of about 2.5$\\times$10$^{-4}$ per channel.  Note the lack of deep, narrow absorption features in the inter-arm velocity ranges from --40 to --36 km s$^{-1}$ (region `B' in the spectra), and --63 to --78 km s$^{-1}$ (region `A'), as first pointed out by Mebold and Hills (1975). Figures 3 \\& 4 show these same spectra with an expanded flux and optical depth scale in order to investigate possible absorption in these two CNM-free velocity ranges.  The observed (continuum subtracted) surface brightnesses in regions A and B appear to be significantly below zero in all the spectra. The mean optical depth values in regions A and B are listed in Table 1. The errors are a (quadratic) sum of the sensitivity error (ie. the rms in off-line channels after averaging over the specified number of channels), and the estimated error due to bandpass  calibration.   The optical depth in velocity region B is marginally higher  toward the SE source relative to the NW source (1.8$\\sigma$). However the SE source spectrum shows a narrow absorption line by the CNM at --30 km s$^{-1}$ which is not seen toward the NW source. The wing of this line might perturb the measured value in region B. Similarly, the spectrum of the SE component in velocity region A shows what might be a narrow, weak absorption component at --69 km s$^{-1}$. In the analysis below we use the optical depths in regions A and B averaged over both SE and NW spectra (row three in Table 1), with errors as dictated by each measurement.  We are confident that the finite optical depths observed at the CNM-free velocities are reliable for a number of reasons. First, the observed optical depths  are significantly higher than the estimated accuracy of the bandpass calibration. Second, we have analyzed spectra from data taken from the four epochs separately (May and July 1996, and June and August 1997), and obtain consistent results for each epoch. And third, we have fit Gaussian models to the strong absorption lines nearest in velocity to the CNM-free regions, and we find that the finite optical depths seen in these regions are not due to the wings of the strong lines.  We should emphasize that we cannot rule-out a model in which the finite optical depth seen at the CNM-free velocities toward Cygnus A are due to a `continuum' of low optical depth cold clouds. If this were the case, the optical depths quoted in Table 1 would be strictly upper limits to the true optical depth of the WNM, and the spin temperature values quoted below then become lower limits. ",
        "conclusions": ""
    },
    "9805/astro-ph9805034_arXiv.txt": {
        "abstract": "In a previous paper we investigated means for constraining the mass distribution of cluster galaxies by weak lensing. We concluded that a comprehensive method should treat the lensing effects of individual cluster galaxies and those resulting from a general cluster component simultaneously. To this end we now develop a non-parametric maximum likelihood cluster reconstruction algorithm that can implicitly take into account the presence of cluster galaxies. The method includes an entropy-like regularization prescription and directly uses the ellipticities of individual source galaxy images as observables rather than relying on an averaged ellipticity field. The mass distribution of cluster galaxies is described by parametrized models. For each set of galaxy parameters the cluster reconstruction algorithm allows to determine the best representation of the global underlying cluster component that is consistent with the presence of the cluster galaxies and the observed image ellipticities of background galaxies. Tests with simulations yielded convincing and robust results.  We applied the method to a WFPC2 image of the cluster Cl0939+4713 and obtained a detection of the lensing effects of luminous elliptical cluster galaxies. We consider this application as a successful test of our technique. However, the small size of the image we analysed does not yet allow to draw strong conclusions regarding the mass distribution of cluster galaxies. ",
        "introduction": "Weak gravitational lensing was suggested by Tyson et~al. (1984) as a means to obtain information on the mass distribution of galaxies at large radial distances. The first detection of weak distortions of the images of background galaxies generated by the tidal gravitational field of intervening galaxies was reported by Brainerd, Blandford \\& Smail (1996). Griffiths et~al. (1996) studied the galaxy-galaxy lensing effects in the {\\it Medium Deep Survey}\\/, and Dell'Antonio \\& Tyson (1996) and Hudson et~al. (1997) analysed the {\\it Hubble Deep Field}\\/. Natarajan et~al. (1997) detected the weak lensing effects induced by cluster galaxies in a mosaic of HST-images of the cluster AC114.  In this paper we continue our discussion of methods to constrain the mass distribution of cluster galaxies from weak lensing presented in Geiger \\& Schneider (1998, hereafter referred to as Paper~I). For the maximum likelihood method described in that paper we first obtained a reconstruction of the cluster mass distribution using `conventional' techniques that rely on the inversion of an integral equation. Then we added parametrized models for the mass distribution of galaxies to the cluster component and applied empirical mass subtraction prescriptions to keep the total mass of the system constant. This approach proved to be workable in the outskirts of clusters where the surface mass density is low. In the non-linear lensing regime in the cluster centre, however, the result turned out to be sensitive to the details of the mass reconstruction and subtraction procedures. For values of the galaxy model parameters that imply a very large fraction of the total mass attributed to galaxies, conceptional problems arise regardless of the (non-)linearity of the lensing effects or the distance from the cluster centre.  In order to overcome these problems we decided to extend the maximum likelihood approach to the problem of determining the mass distribution of the cluster component. This allows to include prior information on the presence of the cluster galaxies directly into the cluster reconstruction procedure. We follow the work of Bartelmann et~al. (1996) and use the two-dimensional gravitational potential as the fundamental physical quantity describing the properties of the cluster lens. Likelihood methods for cluster reconstruction from weak lensing were also described by Squires \\& Kaiser (1996) and Bridle et~al. (1998). These methods use a grid of averaged image ellipticities of background galaxies as observational input. However, the averaging procedure already destroys the information on galaxy scales that we are interested in and therefore we are obliged to use the ellipticities of individual source galaxy images directly as observables. Such an approach for cluster reconstruction was investigated in detail by Seitz, Schneider \\& Bartelmann (1998), and some elements of the technical implementation of our method are similar to their study.  In Section~\\ref{meth} we describe the combined method for cluster lens reconstruction and galaxy-galaxy lensing that we have developed. The reliability of the method is confirmed by an application to simulated observations presented in Section~\\ref{appsim}. In Section~\\ref{appreal} we use our techniques to analyse a WFPC2 image of the cluster Cl0939+4713, which yields a detection of the lensing effects induced by elliptical cluster galaxies. Section~\\ref{disc} concludes the paper with a discussion of the results. ",
        "conclusions": "\\label{disc} Continuing our discussion of Paper~I we investigated methods to constrain the mass distribution of cluster galaxies by weak lensing. In Paper~I we concluded that a comprehensive method should simultaneously take into account the lensing effects of individual cluster galaxies as well as those of a general cluster component. To this end we developed a regularized cluster lens reconstruction algorithm that directly uses the image ellipticities of individual background galaxies as observables, and that allows to explicitly include the presence of cluster galaxies in the analysis. We parametrize the mass distribution of cluster galaxies by truncated isothermal spheres and apply simple scaling laws to relate the halo properties to those of $L_{\\star}$-galaxies according to the galaxy luminosity. For each value of the galaxy parameters -- velocity dispersion $\\sigma_{\\star}$ and cut-off radius $s_{\\star}$ of an $L_{\\star}$-galaxy -- the method allows to determine the best representation for the underlying cluster mass component that is consistent with the observed image ellipticities. We tested the method on simulations and achieved convincing and robust results. If the dark matter distribution around cluster galaxies is rather extended, it is not straightforward to make a clear-cut distinction between mass belonging to a global cluster component or to galaxy haloes. An important result from these simulations is that the distance between the grid points of the cluster component determines the scale for a somewhat artificial separation of the mass located in galaxies and the mass attributed to the global cluster component.  We applied our method to a WFPC2 image of the galaxy cluster Cl0939+4713. Our reconstruction of the cluster mass distribution is consistent with the result of SKSS who used the same object catalogue but a different reconstruction technique. However, the robustness of the result should finally be verified by using observations with a larger field of view. This would also allow to break the mass sheet degeneracy that prevents us from deriving absolute mass estimates here. In this respect another factor of uncertainty is the unknown redshift distribution of the source galaxies.  The galaxy-galaxy lensing analysis provided a significant detection (approximately at the 99\\% to 99.9\\% confidence level) of the lensing effects induced by luminous elliptical cluster galaxies. In fact it had already been speculated by Dressler et~al. (1994b) that some of the arclets in this WFPC2 image `appear to be background sources lensed by individual galaxies rather than the overall cluster potential'. The likelihood function attains its maximum at rather low values for the cut-off radius parameter. However, the small number of lens galaxies included in our study does not allow to derive strong limits for the parameters describing the galaxy haloes. Changing the assumptions that we were obliged to make regarding the mass sheet degeneracy and the redshift distribution of the sources leads to a rescaling of the galaxy-galaxy lensing result. For this reason we do not quote galaxy mass-to-light ratios or values for the fraction of the total mass bound to galaxies in this paper. Our result is consistent with the result of Natarajan et~al. (1997) who applied the methods of Natarajan \\& Kneib (1997) to analyse a mosaic of WFPC2 images of the cluster AC114. However, significant differences of their approach compared to ours make a quantitative comparison of the results difficult.  We consider the application to Cl0939+4713 described in this paper as a successful test of our maximum likelihood technique for galaxy-galaxy lensing in clusters of galaxies. The problems we encountered are not inherent to our method and mainly resulted from the small size of the image we analysed. Ideal observations for this project should include deep exposures with high image quality and large fields of view, which would make it feasible to test possible variations of the halo properties with the density of the environment. Another important ingredient should be supplementary imaging in different colours in order to assure a reliable selection of cluster galaxies and potentially for deriving photometric redshift estimates for the galaxies used as sources. The benefit of photometric redshift information for galaxy-galaxy lensing studies (in the field) was emphasized by Schneider \\& Rix (1997) and has been demonstrated by Hudson et~al. (1997) in their analysis of the {\\it Hubble Deep Field}\\/. Making use of available redshift information for individual source galaxies also allows us to reduce the noise in cluster reconstructions. [In our implementation the approximation (\\ref{papp}) for the probability density distribution of image ellipticities can then be replaced by the simpler expression (\\ref{pappsimp}).] This was also recognized by Seljak (1997) in a similar context.  The analysis of the kind of observational data sets envisaged above, which are most likely to be available within the next few years, will provide the opportunity to compare the lensing constraints to recent numerical studies of the properties of galaxy haloes within clusters (Tormen, Diaferio \\& Syer 1997, Ghigna et~al. 1998). In addition, a comparison and combination of the galaxy-galaxy lensing results with morphological studies of the galaxy population in these clusters (e.g. Oemler, Dressler \\& Butcher 1997) could help to obtain observational evidence for the physical processes [such as the `galaxy harassment' picture suggested by Moore et~al. (1996)] that might be responsible for the evolution of cluster galaxies."
    },
    "9805/astro-ph9805202_arXiv.txt": {
        "abstract": "An accurate fitting formula is reported for the two-point correlation function $\\xi_{hh}(r; M)$ of the dark matter halos in hierarchical clustering models. It is valid for the linearly clustering regime, and its accuracy is about 10\\% in $\\xi_{hh}(r; M)$ for the halos with mass $M>(10^{-2}\\sim 10^{-3})M_{\\star}$ where $M_{\\star}$ is the characteristic non-linear mass. The result is found on the basis of a careful analysis for a large set of scale-free simulations with $256^3$ particles. The fitting formula has a weak explicit dependence on the index $n$ of the initial power spectrum, but can be equally well applied to the cold dark matter (CDM) cosmological models if the effective index $n_{eff}$ of the CDM power spectrum at the scale of the halo mass replaces the index $n$. The formula agrees with the analytical formula of Mo \\& White (MW96) for massive halos with $M>M_{\\star}$, but the MW96 formula significantly underpredicts $\\xi_{hh}(r; M)$ for the less massive halos. The difference between the fitting and the analytical formulae amounts to a factor $\\gs 2$ in $\\xi_{hh}(r; M)$ for $M=0.01 M_{\\star}$. One of the most interesting applications of this fitting formula would be the clustering of galaxies since the majority of halos hosting galaxies satisfies $M\\ll M_{\\star}$. ",
        "introduction": "The two-point correlation functions of halos are studied both for scale-free models and for representative CDM models of hierarchical clustering. In the scale-free models, a power-law power spectrum $P(k)\\propto k^n$ is used for the initial density fluctuation and the universe is assumed to be Einstein--de Sitter, $\\Omega=1$. Four models with $n=-0.5$, $-1.0$, $-1.5$, and $-2.0$ are studied. Because these models are conceptually simple and exhibit interesting scaling properties, it is relatively easy to understand how physical properties depend on the shape of the power spectrum and, perhaps more importantly, to distinguish the physical effects from numerical artifacts, since the latter should not in general obey the scaling relations. For this reason, we will extensively use the scaling property that the bias parameter $b$ depends only on the halo mass $M$ scaled by $M_{\\star}$, i.e. $M/M_{\\star}$ for each $n$. This scaling property manifests itself when the bias parameter for $M/M_{\\star}$ is plotted for different evolution times.  Each of our simulations for $n\\ge -1.5$ is evolved for 1000 time steps with a total of seven outputs, and that for $n=-2.0$ is evolved for 1362 steps with eight outputs. The output time interval is chosen so that $M_{\\star}$ at each successive output is increased by a factor 2.5, and the $M_{\\star}$ values (in units of the particle mass) at the first output are 74 , 59, 35, and 13 for $n=-0.5$, $-1.0$, $-1.5$, and $-2.0$ respectively. Note that fixing the $M_{\\star}$ values is equivalent to fixing the normalization for the power spectra.  The time step and the integration variables are taken similarly to Efstathiou et al. (\\cite{efwd88}). In this Letter we will rely on these scale-free models to understand how the halo-halo correlation depends on the shape of the power spectrum. Then we will examine whether the result obtained from the scale-free models can be applied to CDM models, since CDM models, at least variants thereof, are believed to be close to reality.   Three CDM models are very typical: one is the (ever) standard CDM model (SCDM), one is an open model with $\\Omega_0=0.3$ and with a vanishing cosmological constant $\\lambda_0$ (OCDM), and the other is a flat lower density model with $\\Omega_0=0.3$ and $\\lambda_0=0.7$ (LCDM). These CDM models are completely fixed with regard to the DM clustering if the initial density power spectrum is fixed. For our simulations, the linear CDM power spectrum of Bardeen et al. (\\cite{bbks86}) for the primordial Harrison-Zel'dovich spectrum is used for the initial condition, which is fixed by the shape parameter $\\Gamma=\\Omega_0 h$ and the amplitude $\\sigma_8$ (the rms top-hat density fluctuation on radius $8\\mpc$). The values of ($\\Gamma$, $\\sigma_8$) are (0.5, 0.62) for SCDM, (0.25, 1) for OCDM, and (0.20, 1) for LCDM.  Each model is simulated with our vectorized $\\pppm$ code on the Fujitsu VPP300/16R supercomputer at the National Astronomical Observatory of Japan. Each simulation is performed with $256^3$ ($\\approx 17$ million) particles and with good force resolution $\\eta\\approx 1/2000 L$ (where $L$ is the simulation box size). To properly understand the effect of the cosmic variance, three to four independent realizations are generated for each simulation of one model. Furthermore, two different box sizes, $100\\mpc$ and $300\\mpc$, are adopted for each CDM model. Further details about the code and the simulations will be given in a forthcoming paper (Jing, in preparation), where many clustering statistics of the dark matter will also be presented. The CDM simulations with box size $100\\mpc$ were used by Jing \\& Suto (\\cite{js98}) to study the constraints on cosmological models of the high concentration of the Lyman Break galaxies at redshift $z\\approx 3$ discovered by Steidel et al. (1998). ",
        "conclusions": "We have determined the correlation function of the DM halos for the hitherto largest set of high-resolution N-body simulations. The excellent scaling exhibited in the bias parameter $b(M/M_{\\star})$ at the different evolutionary outputs for the scale-free models assures that the results in Figures~\\ref{fig2} and \\ref{fig3} are physical, not contaminated by numerical artifacts. The simulation results are in good agreement with the formula of Mo \\& White (1996) for massive halos $M\\gs M_{\\star}$. However, for less massive halos the simulation results are significantly higher. The MW96 formula was found in good agreement with the results of $100^3$-particle scale-free simulations by MW96 and with the results of $128^3$-particle CDM simulations of box size $\\sim 300\\mpc$ by Mo et al.(1996). However, their tests were limited to halos with $M\\gs M_{\\star}$ because of the relatively poorer mass resolutions. The results found here therefore do not contradict with, but in fact support and extend the previous N-body tests.  The fitting formula (\\ref{fitting}) we found for $b(M)$ is accurate for halo masses $M/M_{\\star}> 10^{-2}\\sim 10^{-3}$ with only about 5\\% error. The formula can be applied both to the scale-free models and to the CDM models. In the latter case, the index $n$ in equation (\\ref{fitting}) should be replaced with the effective power spectrum index $n_{eff}$ (Eq.~\\ref{neff}) and the original definition of $\\nu=\\delta_c/\\sigma(M)$ is used for $\\nu$ . This fitting formula could have many important applications for the studies of the large-scale structures. One of them would be to predict and to interpret the clustering of late type galaxies and dwarf galaxies both in real observations and in analytical modeling of galaxy formation, since these galaxies are believed to have formed recently ($z\\ls 1$; cf Mo et al. \\cite{mmw98}) with halo masses much less than $M_{\\star}$.  At present we do not know the {\\it exact} reasons which cause the MW96 formula to fail at the small halo masses $M/M_{\\star}\\ll 1$. In the derivation of MW96, two main assumptions are 1) the halo formation is determined by the peak height through the extended Press-Schechter formalism; 2) the mapping of the halo clustering pattern from the Lagrangian space to the Eulerian one is local and linear (with the spherical collapse model). The first assumption could break down more seriously for low peak-height halos, because the tidal force plays a more important role in their formation. Though, as Katz, Quinn \\& Gelb (\\cite{kqg93}; see also Katz et al. \\cite{kqbg94}) pointed out, the peak-height is not the sole parameter even for the formation of high-peak halos. The validity of the local linear mapping was recently questioned by Catelan et al.(\\cite{cmp98}) in a different context. Unfortunately their result is not directly applicable to this Letter.  In relation to the gentle rise observed for the bias parameter at the small $M/M_{\\star}$ (Fig.~\\ref{fig2}), we have visually inspected the spatial distribution for halos with $M/M_{\\star}\\approx 3\\times 10^{-3}$ ($b\\approx 0.8$), $M/M_{\\star}\\approx 10^{-1}$ ($b\\approx 0.6$), and $M/M_{\\star}\\approx 1 $ ($b\\approx 1$) in one late output of the $n=-0.5$ model. The small and large halos appear to delineate filamentary structures more closely than the median-mass halos, consistent with the measured $b$. This however might hint that either or both of the two assumptions are violated for the small halos, since the small halos are otherwise expected to be more preferentially located in low density regions. Obviously it would be very interesting to find out why the MW96 formula fails. In a future paper, we will examine this question more closely."
    },
    "9805/astro-ph9805278_arXiv.txt": {
        "abstract": "A large number of microlensing events have been observed in the direction of the Galactic bulge, with a measured optical depth in the range $2 - 3\\times 10^{-6}$. It has been shown that most of these events are due to bulge stars being lensed by other bulge stars or by foreground disk stars. Among the stars observed in the bulge fields, there should also be disk stars located behind the bulge; here, we consider their effect on the microlensing rates. The optical depth of background disk stars is much higher than that of typical bulge stars, reaching $10^{-5}$ at $6\\kpc$ behind the bulge. Thus, although background disk stars are a very small fraction of the stars in Baade's window, we find that $\\sim 5$ to $10\\%$ of the optical depth should be due to disk stars more than $3\\kpc$ behind the bulge. This fraction is sensitive to the luminosity function of disk stars at large scale-height, to the magnitude cutoff of the survey, and to the amplification bias effect causing large numbers of ``blended'' events. We consider also the effect of a warp and flare in the disk at large distances behind the bulge; this could increase the optical depth from the background disk to $\\sim 20 \\%$ of the total. Events on background disk stars should on average be longer than other events and could be distinguished also by measuring the proper motion or distance of the stars that have been microlensed. The number of these events could be an interesting probe to the structure and stellar population of the far-side of the Galactic disk. ",
        "introduction": "Microlensing has emerged over the last few years as a new field in astrophysics capable of probing the nature of the dark matter, the structure of our Galaxy, and the mass distribution of stars (see Paczy\\'nski 1996 for a review). So far, over 100 events have been detected in the bulge of our Galaxy, near Baade's window (Alcock \\etal 1997a), and over a dozen in the LMC (Udalski \\etal 1993,1994; Alcock \\etal 1997b). In both locations, the observed optical depth is larger than expected if we assume that the known stars are the only objects causing the microlensing events, although there remain substantial systematic errors in the theoretical calculations due to uncertainties in the spatial distribution of the stars, as well as statistical errors in the observational results.  The estimates of the microlensing optical depth to Baade's window assumed first that most of the events would be caused by disk stars lensing sources in the bulge. After the first few events started to suggest a higher than expected value of the optical depth, it was pointed out by Kiraga \\& Paczy\\'nski (1994) that lenses in the bulge should actually be more important. For a source in the bulge, an event due to a lensing star in the bulge should typically have a much smaller Einstein radius, $\\theta_E = [( 4GM/c^2) (D_{ls}/D_l D_s)]^{1/2}$, compared to an event due to a disk lens, because the distance ratio $D_{ls}/D_s$ is very small, given the small size of the bulge compared to its distance from us. However, the smaller Einstein radii are more than compensated for by the much larger abundance of bulge stars in the observed fields. The triaxiality of the bulge turns out to help as a way to increase the optical depth, due to the larger separation that is allowed between the lens and the source if the long axis is at a small angle relative to the line of sight. Assuming the maximum mass for the bulge that is consistent with dynamical measurements, an optical depth as high as $\\sim 2\\times 10^{-6}$ is possible (Zhao, Spergel, \\& Rich 1995). This is still below the observed value, but consistent with it given the errorbars ($\\tau=2.4 \\pm 0.5 \\times 10^{-6}$ for all events and $\\tau=3.9^{+1.8}_{-1.2} \\times 10^{-6}$ for clump giant events alone, in Alcock \\etal 1997a; and $\\tau=3.3 \\pm 1.2 \\times 10^{-6}$ in Udalski \\etal 1994).  In this paper, we shall analyze the contribution to the total optical depth of sources that are at larger distances than the bulge. These sources should generally belong to the disk of the Milky Way behind the bulge. Some of them could also be members of the recently discovered Sagittarius dwarf (Ibata \\etal 1994), which was shown to extend to fields close to Baade's window from a study of RR Lyrae variables (Alard 1996). Any such sources should have a much higher optical depth than the very numerous bulge sources, since they can be lensed by any bulge star with a large value of the distance ratio.  Mollerach \\& Roulet (1996) discussed the contribution to the optical depth from stars in the disk and estimated their event durations. In their models, they considered only sources within $3 \\kpc$ of the center of the Galaxy. In this paper we shall specifically address the contribution to the optical depth due to sources at larger distances. We also study the effect of a perturbation on the shape of the disk behind the bulge to the optical depth. We find that in general, the events on background disk stars can only account for a small fraction of the total optical depth toward the bulge, and therefore cannot increase the total optical depth significantly above the predictions obtained by assuming that all sources are bulge stars (in practice, some sources are also foreground disk stars, and this reduces the mean optical depth). Nevertheless, the fraction of background disk events is still large enough to be detectable and may be of interest for studies of Galactic structure. ",
        "conclusions": "This paper has discussed the contribution of background disk sources to the total optical depth of microlensing events observed in Baade's window. This contribution is appreciable and needs to be taken into account in comparisons between theoretical predictions and the observations. About $15 \\%$ of the optical depth in Baade's window is due to background disk stars; about half of this is contributed by stars at a distance greater than $3\\kpc$ behind the bulge. This fraction is sensitive to the luminosity function of the source stars and the apparent magnitude limit of the microlensing survey (and therefore, to the ``blending effects''). The possibility that the far-side of the disk is strongly warped and flared has also been considered, and we showed that this could significantly increase the number of background disk stars in Baade's window. Our results may be taken as upper limits on the contribution of disk perturbations to the optical depth, since we have chosen the most favorable parameters in our models (e.g., the angle of the line of nodes for the warp and flare model was chosen to be $20\\degree$, and a large mass of the Sagittarius dwarf was assumed).  Future microlensing surveys towards the bulge could provide an interesting probe to the structure and the stellar population of the Galactic disk behind the bulge. We have focused in this paper on predicting the contribution to the microlensing event rate on Baade's window, at a Galactic latitude $b\\simeq -4\\degree$, because the present surveys are being done near this area of low extinction. However, in lines of sight closer to the Galactic plane, the density of disk stars should decrease more slowly with distance, and the contribution of background disk stars should therefore be much higher; an example has been presented in Table 5. Extinction should of course be very high at lower Galactic latitude, so microlensing surveys in this area would probably need to be carried out in the infrared. This may become more feasible as the size of CCD cameras sensitive in the infrared increases.  We would like to thank the referee, Dr. HongSheng Zhao, for his comments and suggestions that have improved the content and the presentation of this paper.  \\newpage"
    },
    "9805/astro-ph9805365_arXiv.txt": {
        "abstract": "We report the detection of Cepheids and a new distance to the spiral galaxy NGC 2541, based on data obtained with the Wide Field and Planetary Camera 2 on board the Hubble Space Telescope (HST). A total of 25 exposures (divided into 13 epochs) are obtained using the F555W filter (transformed to Johnson V), and nine exposures (divided into five epochs) using the F814W filter (transformed to Cousins I). Photometric reduction of the data is performed using two independent packages, DoPHOT  and DAOPHOT II/ALLFRAME, which give very good agreement in the measured magnitudes. A total of 34 bona fide Cepheids, with periods ranging from 12 to over 60 days, are identified based on both sets of photometry.  By fitting V and I period-luminosity relations, apparent distance moduli are derived assuming a Large Magellanic Cloud distance modulus and mean color excess of 18.50 +/- 0.10 mag and E(B-V) = 0.10 mag respectively. Adopting A(V)/E(V-I)=2.45, we  obtain a true distance modulus to NGC 2541 of 30.47 +/- 0.11 (random) +/- 0.12 (systematic) mag (D = 12.4 +/- 0.6 (random) +/- 0.7 (systematic) Mpc), and a total (Galactic plus internal) mean color excess E(B-V) = 0.08 +/- 0.05 (internal error) mag. ",
        "introduction": "This paper presents, analyzes and discusses Hubble Space Telescope (HST) $V$ and $I$ observations of the Sa(s)cd (RC3, de Vaucouleurs et al. 1991) galaxy NGC 2541. These were obtained over a 47 day interval for the purpose of discovering and measuring Cepheid variables, and then using their period$-$luminosity (PL) relation to determine a Cepheid distance to their host system.  The observations are part of the HST Key Project on the Extragalactic Distance Scale (Freedman et al.  1994a, 1994b, 1994c; Kennicutt, Freedman \\& Mould 1995). Cepheid distances measured for the 18 Key Project galaxies, all within a redshift of approximately 1500 km s$^{-1}$, will provide an accurate absolute calibration for a number of secondary distance indicators, such as the Tully$-$Fisher relation, the planetary nebula luminosity function, the surface brightness fluctuation method, the expanding photosphere method for Type II supernovae, the globular cluster luminosity function and the type Ia supernova standard candle. The ultimate goal of the Key Project is to employ these distance indicators to determine the Hubble constant, {\\it H$_0$}, to within 10\\% accuracy.  NGC 2541, with an inclination of 58\\deg~(Bottinelli et al. 1985b), was selected as part of the Key Project because of its potential as a Tully$-$Fisher calibrator. Given previous estimates of its distance (e.g. $\\sim 8$ Mpc, Bottinelli et al. 1986) NGC 2541 was considered to be a relatively easy target for the detection and measurement of Cepheid variables using HST. In addition, the galaxy has the same metallicity as the LMC (which we use as calibrator for the zero points of the Cepheid PL relation), therefore the derivation of its Cepheid distance bypasses any complications associated with a dependence of the PL relation on metal abundance (e.g. Kennicutt et al.  1998).  The galaxy (RA = 08h14m40s, Dec = +49\\deg03\\min44\\sec~ at equinox 2000) belongs to the NGC 2841 group (de Vaucouleurs 1975), which occupies a 15\\deg $\\times$ 7\\deg~ region of the sky near the border of Ursa Major and Lynx.  Besides the loose triplet formed by NGC 2541, NGC 2500 and NGC 2552, the group is comprised of four additional large spirals and, possibly, several dwarf systems. All of the members have low systemic velocities, in the range 420 to 750 \\kms; in particular, the 21 cm velocity for NGC 2541 is $556 \\pm 4$ \\kms (RC3, de Vaucouleurs et al. 1991).  NGC 2541 is the eleventh Key Project galaxy for which a distance has been determined; published results include M81 (Freedman et al. 1994a), M100 (Freedman et al.  1994c, Ferrarese et al. 1996), M101 (Kelson et al. 1996, 1997), NGC 925 (Silbermann et al. 1996), NGC 3351 (Graham et al. 1997), NGC 3621 (Rawson et al. 1997), NGC 7331 (Hughes et al. 1998), NGC 2090 (Phelps et al. 1998), NGC 1365 (Silbermann et al. 1998) and NGC 4414 (Turner et al. 1998).  These measurements further expand the existing database of Cepheid distances, measured by a number of collaborations using both HST (NGC 4603, Zepf et al. 1997; M96, Tanvir et al. 1995; NGC 4639, Saha et al. 1997; NGC 4536, Saha et al. 1996a;  NGC 4496A, Saha et al. 1996b; IC 4182, Saha et al.  1994; NGC 5253, Saha et al. 1995) and ground based facilities (SMC and LMC, Welch et al. 1987; M31, Freedman and Madore 1990; M33, Freedman et al.  1991; IC 1613, Freedman 1988; NGC 2366, Tolstoy et al. 1995; GR 8, Tolstoy et al. 1995; NGC 6822, McAlary et al. 1983; IC10, Saha et al. 1996; WLM, Madore \\& Freedman 1991; NGC3109, Capaccioli et al.  1992; as well as for more distant galaxies, including NGC 2403 and the M81 group, Freedman \\& Madore 1988; NGC 300, Freedman et al. 1992; Sextans A, Madore \\& Freedman 1991; Sextans B, Sakai, Madore \\& Freedman 1997; Leo I, Hodge \\& Wright 1978; Leo A, Hoessel et al.  1994; Pegasus, Hoessel et al. 1990; M101, Alves \\& Cook 1995; and the Virgo galaxy NGC 4571, Pierce et al. 1994).  This paper is organized as follows:  \\S 2 describes the data and the preliminary reduction. \\S 3 and \\S 4 deal with the photometric reduction and the variable star search. The $V$ and $I$ PL relations and the apparent distance moduli are discussed in \\S 5, while \\S 6 deals with the extinction and the true distance modulus.  Discussion and conclusions can be found in \\S 7. ",
        "conclusions": "A few distance estimates to NGC 2541 are available in the literature and are summarized in Table 11.  Based on the RC2 $B_0^T$ magnitude and 21 cm data from Bottinelli et al.  (1992),  Bottinelli et al.  (1984, 1985b, 1986) calculated a distance modulus to NGC 2541 via the B$-$band Tully Fisher relation (calibrated by Bottinelli et al. 1983).  The derived distance modulus is 29.41 $\\pm$ 0.18 mag (Bottinelli et al. 1984). Based on the same data, but slightly different calibration, this was later revised to 29.53 $\\pm$ 0.26 mag (Bottinelli et al. 1985b) and 29.82 $\\pm$ 0.19 mag (Bottinelli et al.  1986). This distance modulus disagrees significantly with the Cepheid distance derived in this paper. However, we notice that the distance moduli given by Bottinelli et al. (1986) for NGC 2500 and NGC 2841, likely members of the same group, are 30.44 $\\pm$ 0.91 mag and 30.54 $\\pm$ 0.17 mag respectively, close to our value for NGC 2541.  NGC 2541 forms a triplet with NGC 2500 and NGC 2552. Their heliocentric velocities are 556 \\kms, 516 \\kms~and 519 \\kms~ respectively (RC3), and the galaxies are within a few degrees of each other, which strongly suggests physical association. It seems unlikely that NGC 2541 is a foreground object as the Bottinelli et al.  distance would imply. Note that NGC 2841 itself is only at a heliocentric velocity of 631 \\kms, although it is $\\sim 11$ degrees away.  An independent IR Tully$-$Fisher distance to NGC 2541 was kindly provided by Stephane Courteau (private communication). Based on the Willick et al. (1997) Tully$-$Fisher calibration, the distance modulus to NGC 2541 is 30.86 $\\pm$ 0.3 mag.  Using the inverse Tully$-$Fisher relation instead (Willick et al. 1997), a consistent result is obtained:  30.78 $\\pm$ 0.3 mag.  Within the quoted errors, these values are in agreement with the Cepheid distance modulus found in this paper, and strengthen the suspicion that the data used by Bottinelli et al. (1984, 1985b, 1986) are in error.  Unfortunately, there are no early$-$type galaxies in the vicinity of NGC 2541 that can be used to derive an independent distance to the group using the surface brightness fluctuation or planetary nebulae luminosity function methods. We note that the distance moduli derived by Bottinelli et al. (1985a) using the method of `sosie' galaxies, and by de Vaucouleurs (1975), based on a variety of secondary distance indicators (HII regions, brightest stars, luminosity class), are likely to suffer from much larger uncertainties than quoted in Table 11.  The following is a brief summary of the results presented in this paper. HST/WFPC2 images of the spiral galaxy NGC 2541, distributed over 13 $V$ and five $I$ epochs within a 47 day window, allowed the discovery of 34 Cepheid variables and 23 additional variable stars, none of which previously known.  Photometry for all the stars in the field was performed independently using a variant of the DoPHOT program (Schechter et al.  1993, Saha et al. 1994) and the DAOPHOT II/ALLFRAME package (Stetson 1994); very good agreement (within 0.05 mag) is found between the two sets of photometry.  A subsample of 27 Cepheids with well defined periods between 12 and 47 days, was chosen to fit $V$ and $I$ PL relations and derive $V$ and $I$ apparent distance moduli by assuming a Large Magellanic Cloud distance modulus and mean color excess of $\\mu_{LMC} = 18.50 \\pm 0.10$ mag and $E(B-V) = 0.10$ mag respectively.  Assuming $R_V = A(V)/E(B-V) = 3.3$, we derive $A(V)/E(V-I) = 2.45$ based on the extinction laws by Cardelli, Clayton \\& Mathis  (1989), Stanek (1996) and Dean, Warren \\& Cousins (1978). This leads to a true distance modulus to NGC 2541 of $\\mu_0 = 30.47 \\pm$ 0.11 (random) $\\pm$ 0.12 (systematic) mag($ = 12.4 \\pm 0.6$ (random) $\\pm$ 0.7 (systematic) Mpc, and a total (Galactic plus internal) mean color excess $E(B-V) = 0.08 \\pm 0.05$ (internal error) mag."
    },
    "9805/astro-ph9805153_arXiv.txt": {
        "abstract": "We have obtained long-slit spectra of \\gal\\ in the J, H, K, and N bands, broadband images in the J, H, and K$_s$ bands, narrowband images centered at the wavelengths of \\bg\\ and \\h2(1,0)S(1), and imaging spectroscopy centered on \\ne2(12.8\\micron).  We have subtracted a composite stellar spectrum from the galaxy spectrum to measure faint emission lines which otherwise would be buried in the complicated continuum structure.  We use these data and data from the literature in a comprehensive re-assessment of the starburst in this galaxy. We confirm that the \\fe2\\ emission is predominantly excited by supernova explosions and show that the rate of these events can be derived from the strength of the infrared \\fe2\\ lines. Although the \\h2\\ emission superficially resembles a thermally excited spectrum, most of the \\h2\\ infrared luminosity is excited by fluorescence in low density gas. We confirm the presence of a bar and also show that this galaxy has a circumnuclear ring. The relation of these features to the gaseous bar seen in CO is in agreement with the general theoretical picture of how gas can be concentrated into galaxy centers by bars. We derive a strong upper limit of $\\sim$ 37,000K for the stars exciting the emission lines. We use velocity-resolved infrared spectra to determine the mass in the starburst region. Most of this mass appears to be locked up in the old, pre-existing stellar population.  Using these constraints and others to build an evolutionary synthesis model, we find that the IMF originally derived to fit the starburst in M~82 also accounts for the properties of \\gal; this IMF is similar to a modified Salpeter IMF.  The models indicate that rapid massive star formation has been ongoing for 20-30 million years in \\gal---that is, it is in a late phase of its starburst. Its optical spectrum has characteristics of a transitional HII/weak-\\oa\\ LINER. We model the emission line spectrum expected from a late phase starburst and demonstrate that it reproduces these characteristics. ",
        "introduction": "A significant fraction of nearby galaxies have strong, centrally concentrated infrared sources indicative of rapid star formation in their central few hundred parsecs (Rieke and Lebofsky 1979; Devereux, Becklin, \\& Scoville 1987).  These so-called starburst galaxies make a dramatic contribution to the star formation activity in the local universe; Heckman (1997) calculates that just four galaxies (of which \\gal\\ is one) are responsible for 25\\% of the high-mass star formation within 10~Mpc.  Local starburst galaxies serve as a model for galaxy formation and evolution at higher redshifts.  An understanding of how these events occur and evolve will require study over the full range of scale and age.  At the same time, very detailed modeling is required of the nearest starbursts such as \\gal\\ to provide templates for models of less accessible systems.  \\gal\\ has been observed extensively in all accessible spectral regions (e.g., radio: Ulvestad \\& Antonucci 1997; molecular line emission: Mauersberger et al. 1996; far infrared: Carral et al. 1994; mid-IR: Keto et al.  1994, Telesco, Dressel, \\& Wolstencroft 1993; near-IR: Sams et al. 1994 ; optical: Watson et al. 1996; X-ray: Ptak et al. 1997; and gamma ray: Paglione et al. 1996). Additional references are given in those cited above.  \\gal\\ was one of the first galaxies to be studied with an evolutionary starburst model (Rieke et al. 1980). The subsequent observations taken together have substantially expanded on our knowledge of the starburst phenomenon from the inputs to that model, demonstrating the presence of supernovae remnants and HII regions, young ultracompact star clusters, a hot superwind, and a host of other starburst phenomena.  Young starbursts, including \\gal, are characterized by high levels of interstellar extinction that can make properties deduced from optical measurements unrepresentative (e.g., Rieke \\etal\\ 1980; Puxley 1991).  Consequently, infrared measurements play a critical role in defining nebular properties, morphologies, rotation curves, and other parameters. We have added to the extensive infrared observations of this galaxy with the deep NIR images covering a wide ($\\sim7^\\prime$) field, narrowband images in \\bg\\ and \\h2, near- and mid-infrared spectra, and a high-resolution spectral map of \\ne2.  Since the first models of M~82 and \\gal\\ (Rieke et al. 1980), evolutionary synthesis modeling has become an important component of starburst study (e.g., Doyon, Joseph, \\& Wright 1994; Leitherer \\& Heckman 1995 and references therein; Schinnerer \\etal\\ 1997).  We interpret the new infrared measurements with the other properties of \\gal\\ through a modern synthesis model (Rieke \\etal\\ 1993, hereafter RLRT93). In this way, the behavior of \\gal\\ can be related consistently to that of other galaxies we have modelled as part of a comprehensive study of nearby starbursts: NGC~6946 (Engelbracht \\etal\\ 1996, hereafter ERRL96) and M~82 (McLeod \\etal\\ (1993); RLRT93); a number of blue dwarf galaxies (Vanzi \\& Rieke 1997), and a sample of interacting galaxies (Vanzi et al. 1998). We demonstrate that the \\gal\\ starburst is at a relatively late stage of development, during which the high rate of supernova explosions leads to behavior that places it as a transition object between pure starbursts and weak-\\oa\\ LINERs.  This paper is organized as follows: In \\S\\ref{sec:obs} we describe the observations and the processing we performed on the data.  We then use our data and results from the literature to derive physical properties of the starburst in \\S\\ref{sec:parms}.  In \\S\\ref{sec:models} we use the properties derived in the previous section to constrain evolutionary synthesis models of the starburst.  We use the models to explore burst parameters such as age, intensity, IMF, and star formation history.  \\S\\ref{sec:other} discusses other results obtained from our data or implied by the starburst modeling results.  In \\S\\ref{sec:conclude} we summarize the paper. ",
        "conclusions": "\\label{sec:conclude}  We have presented high-quality spectra of the nuclear region of \\gal\\ in the J, H, and K bands. We have constructed a composite stellar spectrum and subtracted it from that of \\gal\\ to obtain accurate measurements of faint emission lines free from the spectral structure due to stellar absorptions. We have also presented J, H, and K$_s$-band images, narrow-band images in the lines of \\bg\\ and \\h2(1,0)S(1), spectra at 10.5\\micron, and a high-resolution spectral map at 12.8\\micron.  We have used these data and data from the literature to perform a detailed study of the interstellar medium and nuclear starburst in \\gal.  All of the observed properties, from the radio through the X-ray, can be explained plausibly by a starburst. Specifically, we find:  \\noindent 1.) The density in the \\fe2-emitting region is $\\sim 5 \\times 10^3$ cm$^{-3}$, similar to that observed for \\fe2-emitting regions in supernova remnants. The strength of the \\fe2\\ emission is consistent with the supernova rate estimated from radio observations and predicted by starburst models.  Together, these results confirm that the \\fe2\\ is produced predominantly in the supernovae in the starburst.  \\noindent 2.) Roughly two thirds of the infrared \\h2\\ luminosity is due to UV fluorescence. It is likely that significant fluorescent-excited components of \\h2\\ emission in other starburst galaxies have been missed because the distribution of emission over many relatively faint lines makes detection difficult in spectra of modest resolution and signal to noise.  If the Lyman continuum is produced by Orion-like HII regions, they would also come within a factor of $\\sim 2$ of producing the fluorescent \\h2.  \\noindent 3.) From the MIR fine-structure lines of [\\ion{S}{4}] and \\ne2, we find that the stars photoionizing the gas in \\gal\\ have T$_{eff} < 37000$~K, with $<15$\\% of the ionizing flux coming from stars with T~$>40000$~K.  We also use an upper limit to HeI(1.7)/Br10 to place an upper limit on the stellar temperature of $\\sim37,000$~K.  These values are consistent with the temperature determination of $34,500 \\pm 1,000$K derived from far infrared fine structure lines by Carral et al. (1994).  \\noindent 4.) The optical line ratios in \\gal\\ indicate it is a HII/weak-\\oa\\ LINER and they are fitted well by the hot-star ($T \\sim 45,000$K) photoionization models proposed for this type of active galaxy by Filippenko \\& Terlevich (1992).  However, the stellar temperature determinations are inconsistent with the hot star model. The line ratios can also be fitted by a combination of a photoionization model for metal rich HII regions excited by stars of $\\sim$ 38,000K, and the shock excited spectrum of supernova explosions occurring at the rate measured in the starburst. These components fit \\gal\\ with virtually no adjustment of free parameters. We suggest that many weak-\\oa\\ LINERs can be explained in the same manner, rather than by the presence of a mini-AGN or by hot stars.  \\noindent 5.) We have used the infrared data to derive extinction-independent estimates of the primary boundary conditions for starburst models of \\gal: mass, \\nlyc, maximum stellar temperature, bolometric luminosity, and CO absorption depth.  \\noindent 6.) We find that IMF8, proposed by RLRT93 as the best-fitting for M~82, also provides a good fit to the properties of \\gal\\ and suggests a starburst mass of $6 \\times 10^7$~\\msun. IMF3, which was assembled by RLRT93 to represent various representations of the local IMF, is not satisfactory although it cannot be ruled out as absolutely as for extreme galaxies like M~82 and NGC 1614.  IMF8 is very similar to the modified Salpeter IMF that other workers have recently found gives a good fit in starburst galaxy models. Due to the large variations or uncertainties in IMFs above 1~\\msun, it is no longer clear whether IMF3 is a meaningful representation of the local stellar population to be compared with those in starbursts.  \\noindent 7.) The model giving the best fit assumes that the star formation is extended over 20 to 30 million years. A portion of the supernovae is associated with stars formed near the beginning of the burst, while the ionizing flux arises from the most recently formed stars. The rate of star formation must have declined rapidly during the last roughly 5 million years, to account for the absence of very hot stars. This model is in agreement with our suggestion that a HII/weak-\\oa\\ LINER can arise from a late phase of a starburst."
    },
    "9805/astro-ph9805223_arXiv.txt": {
        "abstract": "The existence of low-mass sterile neutrinos is suggested by the current status of solar and atmospheric neutrinos together with the LSND experiment. In typical four-flavor scenarios, neutrinos would contribute to a cosmic hot dark matter component {\\em and\\/} to an increased radiation content at the epoch of matter-radiation equality. These effects leave their imprint in sky maps of the cosmic microwave background radiation (CMBR) and may thus be detectable with the precision measurements of the upcoming MAP and PLANCK missions. ",
        "introduction": "Neutrino oscillations are currently indicated by the solar~\\cite{solar} and atmospheric~\\cite{atmospheric} neutrino anomalies and by the LSND experiment~\\cite{LSND}.  Taken together, these three bits of evidence are too much of a good thing in that they are incompatible with a three-flavor mixing scheme among $\\nu_e$, $\\nu_\\mu$ and $\\nu_\\tau$. Apart from the obvious possibility that some of these preliminary indications may be unrelated to neutrino oscillations, one intriguing speculation is that there is a fourth low-mass neutrino, $\\nu_s$, which mixes with the standard flavors~\\cite{fourflavor}.  It would have to be sterile with regard to the electroweak interactions and thus is undetectable in any direct search experiment.  The mixing of $\\nu_s$ with standard flavors allows for its thermal production in the early universe, and even though it will typically not attain full equilibrium there will be a cosmic background of sterile neutrinos. The standard Big Bang nucleosynthesis (BBN) constraint on the cosmic radiation density thus provides nontrivial limits on the masses and mixing angles of a four-neutrino scenario consisting of $\\nu_e$, $\\nu_\\mu$, $\\nu_\\tau$ and $\\nu_s$ \\cite{BBN1,Viking,BBN2,sarkar}.  Likewise, if neutrinos are Dirac particles and thus have right-handed components and if they have anomalous magnetic dipole moments, a cosmic abundance of the sterile states can be produced by magnetically induced spin precessions and by electromagnetic spin-flip scatterings~\\cite{dipole}.  However, the most spectacular cosmological consequence of sterile neutrinos is their impact on the large-scale structure of the universe, and notably on the temperature variations of the cosmic microwave background radiation (CMBR).  The anticipated sky maps of the future MAP and PLANCK~\\cite{MAP+PLANCK} satellite missions have already received advance praise as the ``Cosmic Rosetta Stone''~\\cite{BTW} because of the wealth of cosmological precision information they are expected to reveal~\\cite{CMBreview,Jungman,Bond,HETW}.  In the previous discourse on sterile neutrinos it has been curiously overlooked that a successful deciphering of the CMBR hieroglyphs could well make or break the hypothesis of this elusive particle's existence. Even if its signature in real CMBR sky maps may not be unambiguously visible, the hypothesis of sterile neutrinos introduces two additional degrees of freedom into the game of cosmological parameter estimation, viz.\\ a hot dark matter component and additional radiation in the form of neutrinos. ",
        "conclusions": "Low-mass sterile neutrinos are a generic possibility, and indeed required if all current empirical indications for neutrino oscillations are correct.  This would imply a cosmological hot dark matter component in the form of massive neutrinos, and nonstandard contributions to the radiation density at the epoch of matter-radiation equality.  In contrast with previous discussions, both effects would simultaneously occur and would leave their imprint in the large-scale matter distribution as well as in the CMBR temperature sky~maps.  In a four-flavor scenario, the neutrino mass- and mixing scheme can be rather complicated, allowing for involved oscillation phenomena in the early universe because of the possibility of resonant effects. It is thus premature to attempt a complete discussion of all possible cases. However, if one takes the current empirical situation with regard to neutrino parameters seriously at all, then nonstandard neutrino properties will have a large impact on the cosmological observables to be extracted from precision CMBR experiments and galaxy surveys.  In some scenarios, the sterile-neutrino imprint will stick out very clearly, in others it may not be possible to disentangle it from other effects. The most difficult-to-detect scenario is where atmospheric neutrinos oscillate from $\\nu_\\mu$ to $\\nu_\\tau$ and solar neutrinos from $\\nu_e$ to $\\nu_s$ with the small mixing angle MSW solution or the vacuum solution.  Even if the signature of sterile neutrinos cannot be unambiguously seen in the CMBR sky maps and galaxy surveys, they still affect the interpretation of these cosmological precision observables. Therefore, the current experimental effort to pin down the neutrino mass spectrum and mixing angles is inseparably interwoven with a precision interpretation of the forthcoming CMBR~sky~maps."
    },
    "9805/astro-ph9805015_arXiv.txt": {
        "abstract": "A database of radio-quiet quasars\\footnote{Table~I is available in electronic form at the CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or via http://cdsweb.u-strasbg.fr/Abstract.html} detected with ROSAT is presented containing 846 quasars seen in the All-Sky Survey and/or in pointed PSPC observations. About $\\sim $ 70\\%  of the objects have been detected in X-rays for the first time. We present the soft X-ray fluxes and spectra, if available. Using an optically selected subsample compiled from this database we study  the broad band properties of radio-quiet quasars with high statistical significance.  We confirm that radio-quiet quasars have in general steeper soft X-ray spectra (\\mGammag = $2.58 \\pm 0.05$ for  $z < 0.5$) than radio-loud objects, with $\\Delta\\Gamma \\sim 0.4 $ and $\\Delta \\Gamma \\sim 0.3$, compared to the  flat- and steep-spectrum radio quasars (Brinkmann \\eta 1997), respectively. The spectral differences persist to high redshifts with $\\Delta\\Gamma \\sim 0.6$ at $z > 2$ compared to flat-spectrum radio-loud quasars. A spectral flattening with redshift is confirmed for the radio-quiet objects up to $z \\sim 2$, beyond which the spectral slopes seem to be independent of redshift, similar to that found for radio-loud quasars. The spectral slopes of the ROSAT radio-quiet quasars at $z > 2.5$ ($\\Gamma \\sim 2.23^{+0.16}_{-0.19}$) are consistent, within the errors, with those found for nearby quasars in the medium energy band (2--10~keV). This implies that X-ray spectral evolution is not important in radio-quiet quasars. We show that there is, in a statistical sense, little or no excess absorption for most of the radio-quiet objects at $z > 2$, in contrast to their radio-loud counterparts.  By dividing the sample into narrow redshift bins,  the existence of a correlation between the X-ray luminosity and the luminosity at $2500\\AA$, i.e., $l_x \\sim l_o^{e}$, is confirmed. Individual objects show a large scatter from this correlation and the slope $e$ takes values in the range $0.75 \\la e \\la 1.17$, depending on the mathematical method used to analyze the data.  The X-ray loudness \\alpox appears to be independent of {\\it z}, but regression analyses indicate a slight increase  of \\alpox  with optical luminosity. However, this  behavior is, very likely, not caused by physical properties inherent to the quasars but is the result of the intrinsic dispersion of the luminosities and the flux limits in both the optical and X-ray observations.  Finally, we find a small fraction of sources with a substantially larger value of \\alpoxe, objects which appear to be  relatively ``X-ray quiet'' compared to the bulk of the other quasars.  \\par ",
        "introduction": "Convincing evidence for a correlation between the X-ray and optical luminosity of quasars, for a surplus of X-ray emission in radio-loud quasars compared to radio-quiet quasars at the same optical luminosity, and for differences in the soft X-ray ($\\sim$ 0.1--3.5~keV) spectral properties in the two classes of objects was provided by {\\it Einstein} IPC observations (c.f. Zamorani \\eta 1981, Kriss \\& Canizares 1985, Avni \\& Tananbaum 1986, Wilkes \\& Elvis 1987, Canizares \\& White 1989, Wilkes \\eta 1994). Although the samples used were considerably larger than those of previous studies they did not allow an analysis of the X-ray properties of individual subgroups of quasars with sufficient statistical significance. This was especially true for  quasars at higher redshifts, where the results were ambiguous because of the small number of detections in X-rays.  The ROSAT observatory (Tr\\\"umper 1983) provides for the first time the opportunity to study a very large number of quasars in the X-ray domain. More than $\\sim$ 30000 AGN are expected to be detected in the \\RAS (RASS) of which the majority are not yet optically identified. A correlation of source lists from the \\RAS and from pointed observations with existing catalogues yielded more than 1500 detections of previously known quasars with many of them seen in X-rays for the first time. For the radio-quiet quasars this fraction amounts to about 70\\%.  In a previous paper (Brinkmann \\eta 1997, hereafter paper~I), we have presented the data for 574 radio-loud quasars\\footnote {We use $\\log~R = 1$ as dividing line between radio-loud and radio-quiet objects, where R is the K-corrected ratio of radio to optical flux R=f(5GHz)/f(2500\\AA). Objects with  $\\log~R > 1 $ are called radio-loud (Stocke \\eta 1992, paper~I). We used $\\alpha_{radio}$ = 0.5 and $\\alpha_{opt}$ = 0.5 for the K-corrections ($S \\propto \\nu^{-\\alpha}$).} and 80 quasars detected by ROSAT which show radio emission but either qualify as ``radio-quiet'' or have no radio-loudness available. Here we present the second part of the ROSAT quasar database---the radio-quiet objects. We give the  X-ray flux densities and estimates of the soft X-ray spectral indices, if available, for objects from the RASS (Voges 1992) and pointed observations (ROSAT-SRC, Voges \\eta 1995). We present data  of detected quasars only, although we will use  upper limits for the non-detections in some places in the statistical analyses.  In \\S~2 we present the data and give details of the derivation of the relevant parameters, and we discuss in \\S~3 the role of possible X-ray detection biases of the sample. The soft X-ray spectra of the radio-quiet quasars are analyzed in \\S~4. We then present the relationship between the X-ray and optical luminosity in \\S~5. Finally, the conclusions are given in \\S~6. ",
        "conclusions": "We have presented the X-ray spectra and flux densities for 846 radio-quiet quasars detected by ROSAT in both the All-Sky Survey and in pointed observations. This is the largest sample of X-ray detected radio-quiet quasars published so far, with many of the objects ($\\sim$ 70\\%) being detected in X-rays for the first time. We studied the broad band properties of the population by using a sub-sample of optically selected, radio-quiet quasars from this compilation, which comprises 644 detections. The large number of objects enables us to investigate their properties in a wide range of parameter space with high statistical significance. Our major results are the following:  1. For radio-quiet quasars we found a systematically lower and much faster decreasing X-ray detection probability in the RASS towards faint optical magnitudes and high redshifts than for radio-loud objects (paper~I). The on average higher X-ray luminosity of radio-loud quasars can account for these differences qualitatively.  2. A large fraction of the quasars, observed more than once in pointed observations, shows moderate flux variations, mostly by less than a factor of two. Extreme flux variations (by a factor of six) were found for only one radio-quiet object (PG~0844+349).  3. We confirm that radio-quiet quasars have in general steeper soft X-ray spectra (\\mGammag = $2.58 \\pm 0.05$ for  $z < 0.5$) than radio-loud objects ($\\Delta\\Gamma \\sim 0.4 $ and $\\Delta\\Gamma \\sim 0.3$, compared to flat- and steep-spectrum radio quasars, respectively) and that this spectral difference persists at $z > 2$ with $\\Delta\\Gamma \\sim 0.6$ compared to the flat-spectrum quasars (paper~I). It is also confirmed that the soft X-ray spectra in the ROSAT band are systematically steeper than those in the harder $Einstein$ IPC band by $\\Delta\\Gamma \\sim 0.5$.  4. A spectral flattening with redshift is also confirmed for radio-quiet objects, which can be described by a linear relation of the form $\\mbox{\\mGammag}= 2.63(\\pm 0.04) - 0.20(\\pm 0.04) \\times z$, for $z \\la 2.0$. As for the class of radio-loud quasars, the spectral slopes appear to be independent of redshift beyond $z \\sim 2$. A decrease of the dispersion of the distribution of spectral slopes with redshift (from $\\sigma=0.38^{+0.05}_{-0.04}$ for $z < 0.5$ to $\\sigma=0.26^{+0.20}_{-0.12}$ for $z > 2.5$) is also inferred, though the significance is not high. These results are consistent with a composite X-ray spectral model including an additional steep soft X-ray component.  5. The spectral slope of $\\Gamma \\sim 2.23^{+0.16}_{-0.19}$ found for the ROSAT radio-quiet quasars at high redshifts ($z > 2.5$), is consistent, within the errors, with those found for nearby quasars in the medium energy X-ray band (2--10~keV) of $\\Gamma \\sim 2.0$. This implies that X-ray spectral evolution is not important in radio-quiet quasars. The steep spectra of radio-quiet quasars seen at $z > 2.5$ strengthen the claims that they cannot be the dominant contributors to the cosmic X-ray background.  6. The data are, in a statistical sense, consistent with no or only little excess absorption at all redshifts for radio-quiet quasars, though the significance of the result beyond $z \\sim 2$ is not high.  7. By dividing the sample into narrow redshift bins ($\\Delta z = 0.1$) for $z < 1$, the correlation between the X-ray luminosity and the luminosity at $2500\\AA$ is explicitly tested and proved to be an intrinsic property of the quasar population, though a large scatter is present in the correlation. The slopes of the regression line, obtained from  different  mathematical methods, cover a wide range, $0.7 \\la e \\la 1.2$. The existence of an intrinsic non-zero dispersion in the correlation between X-ray and optical luminosities suggest the existence of an additional ``variable'', not accounted for in the analysis. One possibility would be a non-uniform, orientation dependent attenuation of the fluxes in the different energy bands.  8. In accordance with previous studies, the X-ray loudness \\alpox seems to be independent of redshift but to correlate with $\\log~l_o$. The \\alpox -- log~\\lo dependence is complex: the increase of \\alpox with \\lo is significant only for $\\log~l_o \\ga 30.5$ and  the low-\\lo  radio-quiet quasars show a slight increase of  \\malpox with $z$ up to $z \\la 0.5$. However,  this correlation might be primarily caused by luminosity selection biases and by the intrinsic dispersion of the data, as already discussed in paper~I for radio loud quasars. A detailed study of these effects by Monte Carlo simulations will be presented elsewhere (Yuan \\eta 1997).  9. Highly reduced X-ray emission, by more than a factor of $\\sim 30$ compared to the bulk of the objects, is found in a few objects in the high-\\alpox tail of the \\alpox distribution. There are eight  such ``X-ray quiet'' objects below $z < 0.5$, six detected by ROSAT and two upper limits, having $\\alpha_{ox} > 2$. The reason for the low X-ray emission  of these objects is unknown; we note, however, that X-ray variability might play a role. For example, the highly variable object PG 0844+349 is classified as ``X-ray quiet'' only in its low state.  \\vskip 0.4cm"
    },
    "9805/astro-ph9805259_arXiv.txt": {
        "abstract": "The average age difference between pairs of star clusters in the Large Magellanic Clouds increases with their separation as the $\\sim0.35$ power. This suggests that star formation is hierarchical in space and time.  Small regions form stars quickly and large regions, which often contain the small regions, form stars over a longer period.   A similar result found previously for Cepheid variables is statistically less certain than the cluster result. ",
        "introduction": "A previous study of the positions and ages of Cepheid variables in the Large Magellanic Cloud (Elmegreen \\& Efremov 1996) found that Cepheids closer to each other were also more likely to have the same age, as if star formation proceeds in small regions faster than in large regions.  Because Cepheids are relatively old ($>27$ Myr in the LMC sample, with most around 100 Myr) and could have drifted slightly from their point of origin, and because Cepheid ages are rather uncertain, we repeat this LMC study here using star clusters. The Bica et al. (1996) catalog of clusters in the LMC contains 590 members with accurate positions and UBV colors, and these colors have been converted to age by Girardi et al. (1995). Thus we can use the Bica et al. clusters to study the hierarchical properties of star formation in the LMC. ",
        "conclusions": "The positions and ages of clusters in the LMC suggest that there is a correlation between the duration of star formation in a region and the size of the region.  This correlation may be the result of star formation inside cloud complexes that always live for several turbulent crossing times.  The turbulent crossing time in a region of star formation may be estimated from the size-linewidth relation for the clouds that form stars. A compilation of the size-linewidth relations for molecular clouds in the Milky Way is shown in figure \\ref{fig:sl}. The size $S$ is the FWHM of clouds and clumps in various surveys, and the linewidth $c$ is the Gaussian dispersion. The average relation for all the surveys is \\begin{equation} c(km\\;s^{-1})\\sim 0.7S(pc)^{0.5}. \\label{eq:sl} \\end{equation} The ratio of $S$ to $c$ is shown on the bottom of the figure.  Half of this ratio gives the turbulent crossing time, \\begin{equation} t_{crossing}(Myr)\\sim{ {0.5S(pc)} \\over {c(km\\;s^{-1})} } \\approx 0.7 S(pc)^{0.5}. \\label{eq:mcs}\\end{equation} Because of the size-linewidth relation, the turbulent crossing time increases with cloud size approximately as the square root. This slope is comparable to, although slightly larger than, the $\\Delta t-S$ relation for clusters, suggesting that the LMC cluster relation results in part from turbulence. For separations of 10 pc and 100 pc, the average age difference between clusters in the LMC is 3.2, and 2.1 times the crossing time of galactic molecular clouds on the same scale. This suggests that star formation always proceeds with a time scale of $\\sim2.5$ crossing times in the progenitor cloud.  \\begin{figure} \\vspace{3.6in} \\special{psfile=lmcfg7.ps hoffset=-180 voffset=-250} \\caption{Size and linewidth data from molecular cloud surveys. Symbols are for giant molecular cloud surveys = {\\it dots}: Solomon et al. (1987), {\\it open circles}: Dame et al. (1986); quiescent clouds = {\\it filled triangles}: Falgarone et al. (1992), {\\it open triangles}: Williams et al. (1994; the Maddalena-Thaddeus cloud), {\\it inverted open triangles}: Lemme et al. (1995; L1498), {\\it inverted filled triangles}: Loren (1989; Ophiuchus), OB associations = {\\it crosses}: Williams et al. (1994; Rosette), {\\it plus signs}: Stutzki \\& G\\\"usten (1990; M17).} \\label{fig:sl} \\end{figure}   The size-linewidth relation for clouds in the LMC is not well known, but a recent study suggests that the linewidths in the L48 clumps are slightly larger for the clump size than the linewidths in galactic GMCs  (Kutner et al. 1998).  This could be a region of unusually high pressure and not normal for star formation in the LMC, but if this result is representative, then $t_{crossing}(S)$ would be slightly lower than for Galactic clouds, and the ratio of the star formation time to the crossing time slightly larger than $\\sim2.5$.  The $\\Delta t-S$ relation for clusters is not the result of OB association expansion or stellar drift at a fixed initial velocity because then the slope of the correlation would be 1 instead of $\\sim0.3$. It has to result from stellar or gas kinematics with velocity dispersions that scale with a fractional power of the size of the region.  There are several important implications for star formation of the cluster $\\Delta t-S$ relation:  1.  On all scales over which the correlation exists, small regions come and go while large regions continue to form stars.  This implies there might be some recycling of small-scale star-forming material during the lifetime of the larger region. Then metal enrichment from supernovae can be greater in the most active clouds, and more variable from cloud to cloud than previously thought (Elmegreen 1998), and total cloud ages can be longer than previously determined from cloud disruption times following only one epoch of star formation (Leisawitz et al. 1989).  2.  Larger star-forming regions have both larger velocity dispersions and larger average ages than smaller regions.  This trend is similar to that found for OB subgroups and whole OB associations and may contribute to the impression that OB subgroups expand into OB associations. There may not be this much expansion, however. Instead, there could be a difference in the sizes and velocity dispersions of the two types of regions from birth. OB subgroups are born small and they may stay moderately small during the formation time of the other subgroups.  All of the subgroups together define the association, which is a composite of clumpy subparts.  3.  The largest regions of star formation in giant spiral galaxies, regions like Gould's Belt and other Star Complexes measuring 300 pc to 1 kpc, take so long to form (30 Myr) that they are not particularly bright on average.  They are also not unified in appearance by a single bright HII region or concentration of O stars, because the general population of O stars that formed there has already aged off the main sequence.  Instead, they are visible primarily as concentrations of Cepheid variables and other supergiant stars, which is how they were originally discovered (Efremov 1979, 1989).  Most of the O stars are visible only in smaller concentrations, which appear as multiple cores inside the star complexes. Thus the largest regions of star formation in giant spiral galaxies are often overlooked, especially in H$\\alpha$ or UV studies.  This situation changes in smaller galaxies, where the star formation length and time scales are generally shorter than in large galaxies (Elmegreen et al. 1996). In small galaxies, the largest regions of star formation can form so quickly that there are still many OB stars, and then they appear very bright, like 30 Dor (see review in Elmegreen \\& Efremov 1998).  4.  Regions of star formation that are defined by HII regions in H$\\alpha$ images of other galaxies tend to be concentrations of O-type stars, and therefore have ages of around 10 million years. These regions are the classical OB associations.  Because of the $\\Delta t-S$ correlation, they have a characteristic size that corresponds to their age. For a measured characteristic size of OB associations equal to $\\sim80$ pc (Lucke \\& Hodge 1970; Efremov, Ivanov, \\& Nikolov 1987), equation \\ref{eq:tspc} confirms that their duration of star formation is $\\sim14$ Myr. The identification of these regions is entirely based on the presence of O-stars and bright emission nebulae, and is therefore only a selection of one particular scale out of a continuum of scales for the star-formation process.  This was implicitly the case in Efremov et al.  (1987) and Battinelli et al.  (1996), where stars in M31 were selected to be brightest in U or B to detect the O-associations. However, OB associations are not representative of the star formation process in general; they are only one level in a continuous hierarchy of self-similar processes that extends from parsec to kiloparsec scales (Elmegreen \\& Efremov 1996, 1998).  \\begin{figure} \\vspace{3.1in} \\special{psfile=lmcfg8.ps hoffset=-20 voffset=-50} \\caption{Schematic diagram showing the size dependence of the duration of star formation in various regions.} \\label{fig:hier} \\end{figure}   Figure \\ref{fig:hier} shows a schematic diagram of the sizes and durations of star formation in regions that are commonly discussed.  The $\\Delta t-S$ relation here is based on the square root approximation to the power in the size-linewidth relation (cf. Eq. \\ref{eq:sl}) and not on the slightly shallower power-law dependence for clusters written in equation (\\ref{eq:tspc}). The square root dependence gives a more sensible result over the whole range of star-forming scales. The figure considers small clumps in which individual stars form on $10^5$ year time scales, T-Tauri stellar associations, OB subgroups and associations, and star complexes, in which stars typically form on 1, 3, 10 and 30 Myr time scales, respectively. The figure also includes spiral arm segments (not density waves) that may extend for $\\sim10$ kpc.  The ages of spiral arm segments come from the pitch angles ($\\sim 15^\\circ$) and the rate of shear in the galaxy.  We discussed previously how such star-forming spiral arms, which are clearly distinct from density waves by their short lengths and lack of dustlanes, are a natural extension of the star formation process to scales larger than the disk thickness (Elmegreen \\& Efremov 1996).  5. Star formation may be intimately connected with turbulence because of the similarity between the $\\Delta t-S$ relation and the $S/c-S$ relation for molecular clouds. If true, then turbulence would contribute to the rate of star formation in a region. The formation of cloud structure by turbulent processes is well recognized (e.g., Langer et al. 1995; Falgarone \\& Phillips 1996)."
    },
    "9805/astro-ph9805290_arXiv.txt": {
        "abstract": "The colours and magnitudes of early-type galaxies in galaxy clusters are strongly correlated. The existence of such a correlation has been used to infer that early-type galaxies must be old passively evolving systems. Given the dominance of early-type galaxies in the cores of rich clusters, this view sits uncomfortably with the increasing fraction of blue galaxies found in clusters at intermediate redshifts, and with the late formation of galaxies favoured by CDM-type cosmologies. In this paper, we make a detailed investigation of these issues and examine the role that the colour-magnitude relation can play in constraining the formation history of galaxies currently found in the cores of rich clusters. We start by considering the colour evolution of galaxies after star formation ceases. We show that the scatter of the colour-magnitude relation places a strong constraint on the spread in age that is allowed for the bulk of the stellar population. In the extreme case that the stars are formed in a single event, the spread in age cannot be more than 4 Gyr. Although the bulk of stars must be formed in a short period, continuing formation of stars in a fraction of the galaxies is not so strongly constrained. We examine a model in which star formation occurs over an extended period of time in most galaxies with star formation being truncated randomly. This model is consistent with the formation of stars in a few systems until look-back times of $\\sim 5\\Gyr$. An extension of this type of star formation history allows us to reconcile the small present-day scatter of the colour-magnitude relation with the observed blue galaxy fractions of intermediate redshift galaxy clusters.  In addition to setting a limit on the variations in luminosity-weighted age between the stellar populations of cluster galaxies, the colour-magnitude relation can also be used to constrain the degree of merging between pre-existing stellar systems. This test relies on the slope of the colour-magnitude relation: mergers between galaxies of unequal mass tend to reduce the slope of the relation and to increase its scatter. We show that random mergers between galaxies very rapidly removes any well-defined colour-magnitude correlation. This model is not physically motivated, however, and we prefer to examine the merger process using a self-consistent merger tree. In such a model there are two effects. Firstly, massive galaxies preferentially merge with systems of similar mass. Secondly, the rate of mass growth is considerably smaller than for the random merger case. As a result of both of these effects, the colour-magnitude correlation persists through a larger number of merger steps.  The passive evolution of galaxy colours and their averaging in dissipationless mergers provide opposing constraints on the formation of cluster galaxies in a hierarchical model. At the level of current constraints, a compromise solution appears possible. The bulk of the stellar population must have formed before $z=1$, but cannot have formed in mass units much less than about half the mass of a present-day $L_*$ galaxy. In this case, the galaxies are on average old enough that stellar population evolution is weak, yet formed recently enough that mass growth due to mergers is small. ",
        "introduction": "The existence of a strong correlation between the colour and luminosity of early-type galaxies in rich clusters is well established (eg., Sandage \\& Visvanathan, 1978, Larson, Tinsley \\& Caldwell, 1980). This colour-magnitude relation (CMR) implies a physical link between the stellar populations of galaxies and their total stellar mass. The best explanation appears to be a correlation between the metal abundance of the stellar population and total galaxy mass (eg., Arimoto \\& Yoshii, 1987, Bower, Lucey \\& Ellis, 1992b, BLE), although explanations based on the age of galaxies and on their stellar initial mass function (IMF) have also been proposed (eg., Trager et al., 1993, Tantalo et al., 1997). Metal abundance variations are widely accepted as the more likely explanation because observations of the CMR in high redshift clusters show a uniform shift towards bluer colours for galaxies of all magnitudes while maintaining small scatter about the mean relation (Ellis et al., 1997, Kodama \\& Arimoto 1997, Stanford, Eisenhardt \\& Dickinson 1997, Kodama et al.\\ 1998). As we discuss below, the tightness of the CMR leaves little room for additional variations in galaxy age.  In this paper, we explore the constraints that can be placed on the formation of galaxies by the existence of the CMR. Through out this paper, we will assume that the metal abundance and age of a galaxy are uncorrelated --- this is the case for both the super-novae wind models of Arimoto \\& Yoshii (1987) and for self-regulated hierarchical galaxy formation models (White \\& Frenk, 1991, Kauffmann et al., 1993, Cole et al., 1994, and subsequent papers). Ferras et al.\\ (1998) provide an extended discussion of the formal limits on correlated variations. We focus on the scatter of galaxies about the mean correlation and on the slope of the correlation relative to the observed scatter. For galaxies in clusters the rms residuals are remarkably small, less than 0.05~mag rms in $U-V$ for E and S0 galaxies in the Coma cluster. This implies a marked homogeneity in the galaxies' formation that provides an important constraint for theoretical models. For example, if the general trend of increasingly blue colours at fainter magnitudes were due to age, the galaxies would be required to range in `age' (ie., the look-back time to the epoch at which the bulk of the stars were formed) by $\\sim 7$ Gyr across the first 3 mag of the relation (with the larger galaxies being formed at earlier times), while being coordinated so that galaxies of the same magnitude `formed' with an rms dispersion substantially less than this. Other models for the origin of the relation are similarly constrained.  Under the assumption that the CMR is driven by age, BLE used this argument to suggest that most galaxies must have formed the bulk of their stars at redshifts greater than unity. The CMR data also limit the possible contribution of bursts of star formation after the formation of the main stellar component and the degree to which present-day galaxies can have been built-up by the merger of stellar (as opposed to gaseous) systems.  Taken in isolation, the tightness of the colour-magnitude relation suggests that most galaxies in clusters are old systems, but this picture appears to conflict with observations of intermediate redshift clusters. Photometric studies (eg. Butcher \\& Oemler, 1978, 1984, Couch \\& Newell, 1984) have suggested very significant changes to the clusters' mix of galaxy types over look-back times of 3-5 Gyr. Probably, this change is associated with the transformation of Spiral galaxies into S0 types (Dressler et al., 1997, Couch et al., 1997). This picture is reinforced by spectroscopic observations showing that a large fraction of the galaxies in these clusters have only just completed their star formation cycle (eg., Dressler \\& Gunn, 1983, Couch \\& Sharples, 1987, Barger et al., 1996). Can this evidence of star formation activity at relatively modest look-back times be compatible with the homogeneity of the present-day galaxy populations? To answer this question, we must carefully examine the scatter that would be introduced by a population of galaxies that were forming stars in the recent past. In BLE's data for the Coma cluster, for example, a small number of galaxies do not fit on the ridge-line of the CMR. It is important that we correctly account for the treatment of out-lying galaxies when comparing the present-day and intermediate redshift observations, and that we allow for the full range of morphological types seen in present-day clusters.  The well-defined slope of the CMR implies that the star formation process is linked to the final mass of the system: if most of the star formation were to occur in systems of low mass, a physical link between the initial systems and their final mass would be required. In the absence of such a process, the CMR sets a strong constraint on the number of pieces that can be combined to form the final system.  The approach outlined by BLE does not rely on a specific model for galaxy formation in order to derive constraints from the observational data. The short-coming of this approach is that it takes no account of the inevitable gravitational evolution of the dark matter haloes in which the galaxies are contained. For example, although the early-type galaxies on which this data is based currently reside in rich clusters, at $z=2$ this cluster scale over-density would not yet have collapsed; the typical dark matter halo been would have had a mass close to that of a small galaxy group, and the finally galaxy would probably not be recognisable as a single unit. White \\& Frenk (1991), Kauffmann et al.\\ (1993), Cole et al.\\ (1994) and subsequent papers, have therefore pursued an alternative approach. They take as a starting point the extended Press-Schechter theory (Press \\& Schechter, 1974, Bower, 1991, Bond et al., 1991, Lacey \\& Cole 1993). This allows Monte Carlo simulation of the halo merger tree corresponding to a present-day halo of a given mass. Simple prescriptions for the cooling of gas, formation of stars and the feedback of the energy released in supernovae, and the merging and accretion of galaxies can then be fed into the realistic gravitational framework. The rudimentary parameters controlling these processes can then be adjusted until a realistic galaxy population is created. Using these procedures, Kauffmann (1996) and Kauffmann \\& Charlot (1997, KC97) have argued that an elliptical galaxy colour-magnitude relation can be produced that accurately mimics the observed one, both in its slope and in scatter about the mean relation. This occurs despite the fact that star formation continues until relatively late epochs in such hierarchical models (eg., 50\\% of stars form at $z<1$ in the models of Baugh et al., 1997 [BCFL]), and that considerable merging occurs between stellar components.  This paper re-examines the constraints which can be derived from the CMR in the light of these models. Following BLE, we limit the galaxy formation process in a model independent manner, while at the same time allowing for realistic growth of the gravitational potential. Our aim is to synthesize the model independent approach of BLE with the all-inclusive modeling of KC97 and BCFL. We consider a simplified scheme in which the bulk of star formation occurs at relatively high redshift giving rise to a well defined CMR for proto-galactic sub-units. Merging of these increases the mass of the galaxy causing the initial relationship to evolve increasing in scatter and decreasing in slope. Random merging of the sub-lumps creates a relationship that is very flat and has considerable scatter. The key issue we address is the extent to which merging in a hierarchical universe leads to a predominance of equal mass mergers and thus to a CMR that maintains it slope and keeps an acceptably small scatter. To do this we use the galaxy merging trees from BCFL but avoid using a detailed model for the galaxy formation process in order to isolate the effect of galaxy mergers. We do not attempt to assign morphological classifications to the model galaxies. Instead, we compare the sample with a morphologically complete sample of galaxies in the core of the Coma cluster (Terlevich et al., 1998).  In outline, the structure of this paper is as follows. The first part of the paper is concerned with the conclusions that can be drawn regarding the stellar populations of cluster galaxies. In \\S2, we revisit the scatter analysis of BLE. We consider a range of possible star formation histories ranging from continuous almost uniform star formation to star formation that declines strongly through the lifetime of the galaxy. We also consider the effects of the clipping algorithm used to determine the scatter from the observational data. \\S3 investigates whether the range of star formation histories considered allow a consistent paradigm that explains both the narrowness of the CMR at low redshift and the evidence of star formation activity in intermediate redshift clusters. In \\S4 we investigate the effects of mergers on the evolution of the CMR. In particular, we determine whether there is a conflict between the old ages implied for the bulk of the stars and the degree of merging implied by hierarchical gravitational evolution.  Finally, in \\S5, we present a discussion of the results of our study. We show that the small scatter of galaxies in the cores of rich clusters remains a challenging constraint for realistic galaxy formation models. However, by weighting the rate at which star formation is truncated in the galaxies towards large look-back times, and incorporation of the effects of the colour-magnitude fitting process, it is possible to explain both the increasing blue galaxy populations of intermediate clusters and the uniformity of the present-day galaxy populations. Furthermore, we show that while random merging quickly removes any initial slope of the colour-magnitude relation, inclusion of the dynamical biases inherent in a hierarchical model sustains the ratio of the scatter and slope of the colour-magnitude relation at a value close to that observed. A summary of our investigation is given in \\S6. ",
        "conclusions": "The narrowness of the colour--magnitude relation in cluster galaxies imposes a definite degree of homogeneity on their stellar populations. BLE applied this argument in the context of single burst models to conclude that the majority of star formation occured at look-back times greater than $\\sim 10\\Gyr$. In this paper, we have explored a greater variety of possible star formation histories. We find that the bulk of stars must still have old ages but that star formation continuing until relatively recent epochs can be accommodated within the observed CMR scatter in local clusters. This conclusion agrees with that of van Dokkum et al., 1998, who used analytical models for the evolution of galaxies luminosities to explore a wider range of possible star formation histories. As a result, the rising fraction of blue galaxies in clusters out to $z\\sim0.5$, as observed in the Butcher-Oemler effect, does not conflict with the homogeneity of the majority of cluster galaxies. Importantly, these constraints on the star formation histories of cluster galaxies can be naturally reconciled with hierarchical models of galaxy formation since the cluster galaxy population is assembled at substantially higher redshifts than galaxies drawn at random from the universe as a whole.  The CMR can also be used to limit the degree of merging that can occur between galaxies after the stellar population has been formed. Random mergers quickly flatten the CMR leading to a high ratio between the scatter and slope. With random merging, this point is reached quickly, so that the bulk of stars in these galaxies must be formed at redshifts less than 1. However, merging between galaxies is not well modelled by a random process since the dynamical friction time scale is strongly mass dependent. We have incorporated the effect of correlated mergers using galaxy merger trees taken from BCFL. This substantially lengthens the survival time of a recognisable CMR. Even at $z=2$, the ratio of the scatter to slope is kept small, lying just within acceptable bounds. As a result, it is possible to construct a model for the formation of cluster galaxies in which the bulk of the stars are formed at high redshifts (resulting in little differential colour evolution at the present-day) and yet undergo sufficiently little merging that the colour-magnitude relation is still well-defined at the present-day. Nevertheless, the factor by which the mass of a typical galaxy grows between the formation of the bulk of the stars and its present-day mass is small, less than a factor of 2--3 regardless of the merger tree considered.  This conclusion agrees with the study of KC97, who found that in their specific galaxy formation model the CMR of early-type galaxies was well defined and showed scatter comparable to the observational measurements. Our approach has been complementary: a tightly defined colour-magnitude relation at low redshift is not a property related to a single specific model, but will be met by any model in which most star formation occurs at early epochs ($z>1$) in cluster galaxies and in which subsequent mass growth is not too large. We have avoided distinguishing galaxies on the basis of their morphology by using robust algorithm to measure the scatter of the mean relation with all morphological types present. This results in a conclusion that is as model independent as possible."
    },
    "9805/astro-ph9805129_arXiv.txt": {
        "abstract": "We report on kinematic observations of $H\\alpha $ emission from four late-type galaxies of the Hickson Compact Group 16 (H16 a,b,c and d) obtained with a scanning Fabry-Perot interferometer and samplings of 16 $km$ $s^{-1}$ and 1$\\arcsec$.  The velocity fields show kinematic peculiarities for three of the four galaxies:  H16b, c and d. Misalignments between the kinematic and photometric axes of gas and stellar components (H16b,c,d),  double gas systems (H16c) and severe warping of the kinematic major axis (H16b and c) were some of the peculiarities detected.  We conclude that major merger events have taken place in at least two of the galaxies of the group, H16c and d, based on their significant kinematic peculiarities, their double nuclei and high infrared luminosities.  Their H$\\alpha$ gas content is strongly spatially concentrated -- H16d contains a peculiar bar-like structure confined to the inner $\\sim$ 1 h$^{-1}$ kpc region.  These observations are in agreement with predictions of simulations, namely that the gas flows towards the galaxy nucleus during mergers, forms bars and fuels the central activity.  Galaxy H16b, an Sb galaxy, also presents some of the kinematic evidences for past accretion events.  Its gas content, however, is very sparse, limiting our ability to find other kinematic merging indicators, if they are present.  We find that the merger remnants in the compact group HCG 16 have significantly {\\it smoother} optical profiles than isolated mergers, i.e., they show an amorphous morphology and no signs of tidal tails. Tidal arms and tails formed during the mergers may have been stripped by the group potential (\\cite{bar92}) or alternatively they may have never been formed.  The velocity field of the galaxy H16a shows grand-design isovelocity lines with no signs of disturbances inside a radius of $\\sim$ R$_{25}$.  This result is contrary to expectations given that the galaxy has a high infrared luminosity, central activity, tidal tails at large radii and it is embedded in a common group envelope observed in HI and x-rays. The normality of the velocity field suggests that this galaxy may be a fairly recent acquisition of the group.  Our observations suggest that HCG 16 may be a young compact group in formation through the merging of close-by objects in a dense environment. ",
        "introduction": "Nearby compact groups may provide one of the best laboratories for studying the effects of on-going collisions on the structure and dynamics of galaxies.  One of the most dense compact groups known is HGC 16 (H16), or Arp 318, originally catalogued by \\cite{hic82} as an association of four late-type galaxies at V $\\sim$ 4000 $km s^{-1}$ with a small velocity dispersion of 123 $km s^{-1}$ and a median galaxy-galaxy projected separation of 44 h$^{-1}$ kpc (\\cite{hic92}, $h$ is the dimensionless Hubble constant $H_o$/100 km sec$^{-1}$ Mpc$^{-1}$).  \\cite{rib96} found three other bright galaxies near the group, in the same redshift range, within a median radius of 0.197 h$^{-1}$ Mpc. All seven galaxies were found to have nuclear emission lines, making H16 the highest concentration of starburst and active nuclei galaxies in the nearby universe.  ROSAT observations of the group (\\cite{pon96}) revealed an intragroup medium with a total extent of $\\sim$ 400 h$^{-1}$ kpc, a temperature of 0.3 KeV and density of N$_H$ = 2.05 $\\times$ 10$^{20}$ cm$^{-1}$.  Such density and temperature are typical of compact groups of galaxies, which were shown to fit well on the extension of the correlations between the X-ray luminosity and both the gas temperature and velocity dispersion of clusters of galaxies (although with a steeper slope than for clusters (\\cite{pon96}).  The presence of diffuse X-ray emission around H16 is a strong evidence that this is a genuine group.  Compact groups that are true high-density systems, as H16 seems to be, provide excellent environments in which to study the properties of merger galaxies.  Such a study for a compact group galaxy is important not only to investigate the process of galaxy formation through mergers but it may also provide important information on the dynamical evolutionary stage of the compact group.  In this paper we investigate the dynamical state of the Hickson compact group 16 through the study of the detailed 2-D kinematics of four of its member galaxies:  H16a (NGC 835, Mrk 1022), H16b (NGC 833), H16c (NGC 838, Mrk 1021) and H16d (NGC 839).  The four galaxies are optically luminous systems, with M$_B$ ranging from --20.3 $+$ 5 log $h$ to --19.5 $+$ 5 log $h$. H16a, c and d are strong radio and infrared sources (\\cite{all96}, \\cite{men95}). \\cite{bos96} report on CO observations for three of the four galaxies and calculate large masses of molecular gas.  Analysis of their central spectra shows that H16a and d are LINERs, H16b is a Seyfert galaxy and H16c has a starburst nucleus (\\cite{coz98}). Rubin, Hunter \\& Ford (1991, RHF91) obtained rotation curves (RCs) for the four galaxies and found they have ``abnormal'' shapes, a common feature among strongly interacting and merging galaxies.  H$\\alpha$ emission-line observations of H16 a,b,c,d obtained with a scanning Fabry-Perot are used here to derive the spatial distribution and kinematics of their warm gas content.  The warm gas is a good tracer of the total gas mass (that is gas in all phases) in a galaxy (\\cite{sch93}) and it should also be a good tracer of the potential of the galaxy.  The details of the kinematics of the emission-line velocity field can be used as diagnostics of the gaseous merger dynamical state.  The organization of this article is as follows:  Section 2 gives details of the observations and data reduction.  In Section 3, we present the results for the photometry, the internal kinematics and the various mass determinations of the galaxies.  Section 4 contains the discussion and our conclusions. ",
        "conclusions": "Each of the four HGC 16 galaxies shows various evidences for interactions. This is clearly a very high density group. This gives an excellent opportunity to correlate the different galaxy properties suggested to be induced by galaxy-galaxy interactions. Via comparison with the recent simulations of interacting galaxies, we can also attempt to characterize the type of interaction each of the HCG 16 galaxies has undergone and to trace the interaction history of the group. Before discussing these principal results of the study, we address three other issues: the previous claims of unusual rotation curves for these galaxies, how these galaxies fit onto the Tully-Fisher and Fundamental Plane relations and the presence of dwarf galaxies in the tidal tails of H16a.  \\subsection  {Falling and/or sinusoidal rotation curves?}  RHF91 presented evidence that a large fraction of the compact-group spirals have falling rotation curves, unlike field spirals. If falling rotation curves are common in compact groups, this would suggest that the individual dark matter halos are absent most probably because they were stripped by interactions.  In particular both H16c and H16d were reported to have sinusoidal RCs, from measurements of the kinematics of the gas obtained through long-slit spectra taken along the photometric major axis of the galaxies.  Having the 2-D VFs of the galaxies we can check on the reality of the sinusoidal shapes of the RCs.  Fig. 10 shows two possible RCs for H16c (described in Section 3.2.3). Fig.  10b was built with the same parameters used by RHF91, i.e., PA=78$\\degr$ and the center coinciding with the optical/continuum center.  This curve clearly has a sinusoidal shape, in agreement with the curve presented in RHF91.  The rotation curve of H16c plotted using the kinematical center, inclination and position angle (Fig. 10a and see \\S3.3 for how these quantities are defined and derived) has a fairly normal shape typical of a disk in differential rotation.  The sinusoidal rotation curve derived through long-slit spectroscopy (Fig. 10b) along the photometric major axis of H16c was caused by a misalignment between the gas and stellar components and the presence of a second kinematic component.  There are a number of merger-candidates in the literature previously found to have sinusoidal rotation curves. It would be interesting to review these cases with a full 2-D velocity field analysis (e.g. \\cite{mih98}, \\cite{hib96}).  In the case of H16b and H16d it is not possible to make their rotation curves look normal by changing the center or the position angle along which the curve is plotted. H16b has the peculiarity of having one side (the northeast) with a much lower velocity amplitude than the other side, while for H16d the velocity curve reaches a maximum and drops on both sides.  \\subsection{ The Tully-Fisher and the fundamental-plane relations}  Dynamical processes may have worked very efficiently in HCG 16, as exemplified by the high frequency of interaction indicators found for all member galaxies.  It is of interest to investigate if the ``damage'' due to interactions and merging was large enough to affect the position of the galaxies in the Tully-Fisher or fundamental-plane relations.  In order to check if H16a, H16c and H16d follow the TF relation we have used the plots in Figs. 6, 10 and 12 to estimate the shape and amplitude of the rotational velocity curves and we have then compared these with ``template curves'' based on the TF relation for normal galaxies (RHF91, Fig. 5).  No attempt is made to check if H16b follows the TF relation since the two sides of its rotation curve do not coincide.  H16a has a normal velocity amplitude for its luminosity, following the TF relation.  For the main disk component of H16c the measured V$_{max}$ is also within the expected value, for a galaxy of this luminosity (see Fig. 5 of RHF91 and table 2).  A similar exercise, for H16d shows that V$_{max}$ is $\\leq 0.5$ of the value expected for a normal galaxy of this luminosity, if the inclination given by the velocity field (inclination = 47$\\degr$) or R image (inclination=57$\\degr$) is used.  There is a chance, however that the flattening of the velocity curve for H16d is due to dust. Alternatively there could be problems in obtaining dynamical information about gravity from the measured gas properties in the inner 5 $\\arcsec$ of an interacting galaxy. Dynamical information from the stars would be much more reliable in this case.  The surface-brightness of H16d follows an r$^{1/4}$-profile, typical of early-type galaxies. It is, therefore, of interest to check if it follows the r$_{eff}$ vs. $\\mu_{eff}$ relation, a projection of the fundamental-plane relation, observed for bulges and elliptical galaxies.  With the values of r$_{reff}$ = 10.5 and $\\mu_{eff}$ = 21.1 mag arcsec$^{-2}$, we find that H16d follows the fundamental plane relations of normal field galaxies (\\cite{kor89}).  \\subsection{Dwarf galaxies in tidal debris ? }  The analysis of R images of Hickson compact groups by \\cite{hun96} yielded a sample of 47 candidate dwarf galaxies that may be associated with tidal arms and tails in interacting and merging galaxies.  If the majority of these dwarf-galaxy candidates are confirmed as being gravitationally bound stellar systems, these authors estimated that a significant fraction (perhaps as much as one-half) of the dwarf population in compact groups is created in mergers occurred in the giant parent galaxies.  Three dwarf candidates are suggested by \\cite{hun96} to sit on the eastern tail of H16a.  Two of the candidates were observed in our fields.  We checked if emission lines are associated to the candidates.  In both cases, the emission-line map has a S/N twice higher on the corresponding pixels then in the direct neighbourhood of these pixels. We therefore confirm these may be star-forming knots at the redshift of the group. However, no conclusive information can be obtained from the examination of the velocity field, the line-width and the equivalent-line-width maps, due to insufficient S/N. We conclude that the ``dwarf candidates'' are at the same redshift of the group but we cannot exclude the possibility that they are material connected to the main galaxies.  Longer exposure and higher spatial resolution are necessary to confirm the nature of the emission, if these are galaxies formed in interactions or just part of the main galaxy which will fall back in a few crossing times.  It is interesting to note that these star-forming knots are found exactly in the galaxy for which we show that no merging has occurred (H16a).  \\placetable{tbl-3}  \\subsection{A census of the interaction-related properties for the H16 galaxies}  There are a number of ``interaction indicators'' that have been suggested over the years. These have been used to attempt to discern the different types of interactions (primarily tidal interactions vs. merging) that can occur between galaxies and are generally based on models of colliding systems.  Table 3 lists eight interaction/merging indicators and whether or not each is seen in the four galaxies of H16.  Close tidal encounters which do {\\it not} lead to merging can strongly affect the interstellar medium of the colliders, can trigger strong bursts of star formation on a timescale of a few 10$^8$ years but can only mildly affect the {\\it kinematics} of the central velocity fields of the colliding galaxies. In a merger, on the other hand, even if the two nuclei have not yet mixed completely, the regular kinematic structure of the disks disappear and the velocity fields may display strong signatures of the collision (e.g.  the Antennae galaxies, \\cite{amr92}, N7252, \\cite{hib96}).  We, therefore, use the indicator ``highly disturbed velocity field'' as one of our definite evidences for a merger.  Central double nuclei and/or double kinematic gas components are other definite indicators of mergers. Double nuclei are present, for example, in the well known mergers, the Antennae galaxy (\\cite{amr92}), N520 (\\cite{hib96}) and IRAS 23128-5919 (\\cite{mih98}) and double kinematic gas components are present in IRAS 23128-5919, the Cartwheel galaxy (\\cite{amr98}), HCG 31 and several other obvious merger candidates.  We list as interaction indicators for collisions which do not necessarily lead to merging the following properties:  kinematic warping, gaseous and stellar major axes misalignments, tidal tails, high infrared luminosity and central activity.  Although the presence of tidal tails and plumes are common in mergers and they are often taken as evidence for a major accretion event, we group it with the indicators of mild interaction since they can also be formed in interactions which do not lead to mergers. Kinematic and photometric major axes misalignment and kinematic warping are properties that are also usually associated with galaxy collisions.  In a sample of 75 normal spiral galaxies studied by \\cite{sch93}, for which velocity fields were available, only a very small fraction of the galaxies showed kinematic warping and/or small gaseous/stellar major axes misalignments. Another indicator listed in Table 3 is ``high infrared luminosities''. Certainly not all strongly interacting galaxies have high IR luminosities but most IR loud galaxies have disturbed morphologies.  Finally, ``central activity'' is taken as an interaction indicator. Although there is not a one-to-one correlation between interactions and central activity, there are several lines of evidence that strongly suggest that interactions may drive nuclear inflows and fuel central activity in interacting systems.  \\subsection {The interaction history of the H16 galaxies}  The data presented in this paper support the existence of a regular gaseous rotating disk in the center of galaxies H16a, b, c and d.  H16c may contain two disks. Based on the interaction indicators for each galaxy listed in Table 3 and in the context of the various models of galaxy interactions (\\cite{bar89}, \\cite{bar92}, \\cite{bar96}) we suggest the following histories for the H16 galaxies.  {\\bf H16a:} This galaxy does not present any of the indicators common to merging galaxies, listed in Table 3. It has a normal mass distribution inside a radius of 30 arcsec and it does not contain any multiple nuclei or gas components.  At large radii, however, the galaxy looks morphologically disturbed and shows tidal tails to the east and south.  Although numerical models of compact groups suggest that galaxies with extended tidal tails are excellent candidate mergers (Fig. 1 of \\cite{bar89}), the grand-design isovelocity lines of H16a with no signs of disturbances inside a radius of $\\sim$ R$_{25}$ strongly suggest that this galaxy has not yet suffered a major merger.  An interaction between galaxies H16a and H16b is very probable.  The difference in systemic velocities between these two galaxies is less than 200 km s$^{-1}$ and their isophotes overlap. The normality of the velocity field of this galaxy compared with the peculiarities observed for the velocity fields of the other three very kinematically disturbed companions, leads us to the conclusion this may be a fairly new member to the heart of the group.  {\\bf H16b:} Although H16b is classified as an Sb galaxy, it has a velocity field which is very peculiar, unlike that observed for any other normal spiral galaxy (e.g. \\cite{sch93}).  Strong warp of the kinematic major axis, misalignment between the optical and kinematic axes and a very slow rise in the velocities in the inner regions of the galaxies are obvious peculiarities related to past or ongoing interactions.  In addition, the blue and red sides of its velocity curve do not match in shape nor in amplitude and its visual morphology is asymmetric, although smooth and ``well behaved'' (if compared with the morphologies of other merger remnants).  The close proximity of H16a and the overlapping isophotes with this galaxy suggest a clear ongoing interaction. H16b has little ionized gas compared to the other members of the group studied here (Table 2). Also the mass of molecular gas is four times smaller than the average value for the other members.  It is the one galaxy from the four studied here that is not detected at radio or infrared wavelengths. Although H16b does not show a double nucleus or a double gas system, which would be a clear sign of an ongoing merger, the very disturbed velocity field calls for a scenario where strong interactions or accretion events have occurred in the past, which have, however, left the disky morphology of the galaxy intact.  A gas-poor accretion event is possible, given the lack of gas in H16b. The scarcity of gas and consequent low S/N of the data did not allow detection of a double gas component, if it was present.  In addition to having suffered accretion events H16b may be involved in an on-going interaction with H16a, which may result in a major merger.  Unlike the two ongoing mergers in H16 (H16c and H16d, see below) the warm gas content of H16b is not concentrated to the center of the galaxy.  {\\bf H16c:} This galaxy is clearly an ongoing starburst merging system.  It presents seven indicators (all but ``tidal tails'') for strong interactions and mergers, according to Table 3.  The VF of H16c shows the existence of a second velocity component to the southwest of the galaxy, closely spaced in velocity and position to the main component.  The observed kinematic warp  and misalignment between gas and stars are strong evidences for external accretion of the central ionized gas in H16c. Its infrared luminosity is comparable to that of other mergers (log L$_{IR}$ = 11.6 L$\\odot$). The morphology of the galaxy as a whole is amorphous.  No disks or spiral arms or tails are seen down to a surface brightness limit of R $=$ 25 mag arcsec$^{-2}$. The shapes of the continuum and monochromatic images of H16c (Fig. 4) indicate that the emission-line isophotes are significantly flatter than the isophotes of the underlying stellar component, typical of early-type galaxies. However, the light profile of the galaxy does not follow an r$^{1/4}$ law as is common for relaxed mergers.  The kinematics of H16c resembles that of the merger remnant N6240 (\\cite{bla91}). The overall optical morphology of H16c, however, is much less peculiar than that of N6240 or of any other ongoing merger remnant from the Toomre and Toomre's sequence.  It is possible that pre-existing tidal arms or tails were quickly ripped apart by the dynamical forces within the compact group, a situation which is unlikely to happen in less dense environments (\\cite{bar92}). Alternatively tidal tails may have never been formed.  The HI map of the group (Williams, private communication) shows a strong connection between galaxies H16c and H16d. It also shows that the centers of the HI contours are displaced from the optical centers of H16c and H16d by almost one optical radius, in the direction of the line that joins the centers of these two galaxies.  {\\bf H16d:} This galaxy may also be a merger remnant.  It presents five of the eight indicators of interactions and mergers, from Table 3. Double kinematic gas component, kinematic warping and tidal tails are not detected for this galaxy.  The first two indicators are not observed probably due to the small range in radii where the gas kinematics can be measured. The lack of tidal tails is common to all mergers in group H16 and may be a common feature of mergers in compact groups in general.  H16d has a double nucleus and a peculiar velocity field.  It has a high infrared luminosity (log L$_{IR}$ = 11.7 L$\\odot$), similar to that of other mergers.  Outside a radius of 8 arcsec the light profiles of the galaxy (in the R and I bands) follow an r$^{1/4}$-profile with an r$_{eff}$ of 10.5 arcsec.  In the inner parts the galaxy profiles flatten, as is typical of the mass profiles of modeled merger remnants (eg. \\cite{brn92}).  H16d has a peculiar bar-like structure in the central 10$\\arcsec$ which coincides with the peaked ionized gas, molecular gas and dust components.  The rotation curve of H16d peaks and falls on both sides within the inner 2 h$^{-1}$ kpc region. We do not have kinematical information on the second nucleus, which lies 7 arcsec to the east of the main nucleus. The second nucleus is only seen in the NIR due to obscuration from dust in the optical.  The H$\\alpha$ emission in H16d (and in H16c also) is strongly spatially concentrated, in qualitative agreement with the results of simulations of gaseous collisions, in which the gas is driven towards the nucleus of the galaxies (e.g. \\cite{bar96}).  Our measurements support the models of dynamical evolution of compact group galaxies (\\cite{brn92}) and should impose strong constraints to the detailed simulations of the gas kinematics during group merging. \\\\"
    },
    "9805/hep-ph9805300_arXiv.txt": {
        "abstract": " ",
        "introduction": "Supersymmetric (SUSY) theories possess a unique and efficient mechanism for the generation of a cosmological baryon asymmetry through the decay of scalar fields along nearly $F$- and $D$-flat directions of the scalar potential, using baryon number violation arising in GUTS \\cite{ad}, or other sources of baryon or lepton number violation such as neutrino mass effects \\cite{cdo1,cdo2,my}.  In its original form, applied to SUSY-GUTS, this Affleck-Dine (AD) mechanism  not only produces a baryon asymmetry but also entropy from the decaying fields and results in a net baryon-to-entropy ratio, $n_B/s \\sim$ O(1) \\cite{ad,lin}. In fact, one of the main problems associated with this mechanism in the GUT case is the dilution of the asymmetry down to acceptable levels of order $10^{-11}$--$10^{-10}$ \\cite{eno}. In the context of GUTS after inflation, the AD scenario can be shown \\cite{eeno} to produce an asymmetry of the desired magnitude. During inflation, scalar fields with masses, $m < H$, are driven  by quantum fluctuations to large vacuum values ( but less than $M_P$) which become the source of the scalar field   oscillations once inflation is over: however, inflation supplies  an additional source of entropy from inflaton decays.  Recently, it has been argued that the simple picture of driving scalar fields to large vacuum values along flat directions during inflation is dramatically altered in the context of supergravity \\cite{drt}.  During inflation, the Universe is dominated by the vacuum energy density, $V \\sim H^2 M_P^2$. The presence of a nonvanishing and positive vacuum energy density indicates that supersymmetry is broken and soft masses of order of $\\sim H$ are generated \\cite{all}. If such masses are generated along otherwise flat directions, the AD mechanism would be inoperative. While this is true for minimal supergravity, it was shown \\cite{gmo} that in no-scale supergravity \\cite{ns}, or more generally in any supergravity theory with a Heisenberg symmetry of the kinetic function \\cite{heis}, such corrections are absent at the tree level.  It was also shown \\cite{gmo} that, at the one-loop level, corrections to the scalar potential are generated along flat directions and, for all directions which do not involve stops, these corrections to the mass squared are negative of magnitude $\\sim 10^{-2} H^2$.  This is large enough to insure that the $vev$'s along flat directions do indeed run to $\\phi_0 \\sim M_P$, resulting once more in a baryon asymmetry of order unity.  In this letter, we will systematically consider possible effects which may contribute to diluting a baryon asymmetry $n_B/s \\sim$ O(1) from the AD mechanism.  We will consider in turn the effects of dilution by an inflaton and/or a massive moduli field, the effects of higher dimensional operators, and the effects of an intermediate scale.  We find that the dilution by an inflaton or moduli is insufficient in generic supergravity models, but may be sufficient in the context of certain string-derived models. In the absence of these effects the presence of the first nonrenormalizable contribution to the superpotential (of dimension 4), suppressed by the GUT scale, can lead to the desired asymmetry.  We also consider baryogenesis in the context of the minimal supersymmetric standard model, with only the baryon number violation, and nonperturbative corrections to the superpotential, implied by quantum gravity effects at the Planck scale. We find that these effects are sufficient to generate the observed baryon asymmetry of the universe (BAU) in the case of the minimal supersymmetric standard model with standard 4-dimensional (super)gravity. We also consider the possibility that the asymmetry is diluted by sphaleron effects at the electroweak scale. ",
        "conclusions": "We have considered several possibilities for controlling the potentially large baryon asymmetry produced by the Affleck-Dine mechanism in supergravity models with a Heisenberg symmetry of the K\\\"ahler potential.  Moduli, which normally produce a problematic amount of entropy were shown in fact to be insufficient in diluting the baryon asymmetry in models with $R$-parity conservation.  In the absence of $R$-parity, the untwisted sector of orbifold compactifications may provide a candidate for the desired dilution. Perhaps the simplest and most natural possibility for the reasonable baryon asymmetry invokes only gravitational interactions for both the baryon number violation as well as the higher dimensional operators necessary for controlling the sfermion $vev$s along flat directions.  In a GUT, we showed that it is necessary to have mutually exclusive GUT-flat and GUT-non-flat directions.  GUT-flat directions do not provide ample suppression of the baryon asymmetry.  Finally, we noted that the mechanism suggested by KRS using lepton mass effects via sphaleron interactions could dilute an asymmetry by a factor of $10^{-6}$.  Before concluding we note that string theory predicts a relation between the values of the gauge coupling $g$ at unification and the scale of unification, $\\Lambda_G \\sim g M_P$, that is  not satisfied in fits of the data to the MSSM. It has recently been realized that in the strongly coupled limit of heterotic superstring compactifications (``M-theory\"), string unification may in fact be effectively 5-dimensional \\cite{w,p}, with the extra dimension ``turning on\" at $O(10^{14})$ GeV for the gravitational modes of the theory.  In this scenario the two $E_8$'s of the heterotic string live on two ten-dimensional surfaces that are separated by this fifth (or eleventh) dimension $r$. This changes the dimensional analysis that relates the Plank scale to the string tension, and modifies the above relation in such a way that consistency with MSSM RGE's can be achieved for the choice $r\\sim 10^4$. One might expect that $r^{-1}\\sim 10^{-4}$ would represent an upper limit for values of $vev$'s that may be reliably considered in the low energy effective theory. However, just as the MSSM gauge coupling RGE's are left intact, because the observed sector is contained in a (weakly coupled) $E_8$ that does not see the extra dimension, the AD flat directions that are charged under the MSSM gauge group will also be unaffected. Unless effective interactions at the scale of the fifth dimension can be generated along the flat directions in such a way so as to preserve the running of the gauge couplings, these effects will not be capable of reducing the baryon asymmetry.        \\vskip 0.8truecm \\noindent {\\bf Acknowledgments} \\vskip 0.4truecm This work was supported in part by the Director, Office of Energy Research, Office of High Energy and Nuclear Physics, Division of High Energy Physics of the U.S. Department of Energy under Contracts DE-FG02-94ER-40823 and DE-AC03-76SF00098, by NSF grants AST-91-20005 and PHY-95-14797 and by the Alfred P. Sloan Foundation.  \\newpage"
    },
    "9805/astro-ph9805284_arXiv.txt": {
        "abstract": "The stellar population of the Pegasus dwarf irregular galaxy is investigated in images taken in the F439W (B), F555W (V), and F814W (I) bands with the Wide Field Planetary Camera 2 (WFPC2) on the {\\it Hubble Space Telescope}. With WFPC2 the Pegasus dwarf is highly resolved into individual stars to limiting magnitudes of about 25.5 in B and V and 25 in I.  These and ground-based data are combined to produce color-magnitude diagrams which show the complex nature of the stellar population in this small galaxy. A young (age $<$ 0.5~ Gyr) main sequence stellar component is present and clustered in two centrally-located clumps, while older stars form a more extended disk or halo. The colors of the main sequence require a relatively large extinction of A$_V =$ 0.47 mag. The mean color of the well-populated red giant branch is relatively blue, consistent with a moderate metallicity young, or older, metal-poor stellar population. The red giant branch also has significant width in color, implying a range of stellar ages and/or metallicities. A small number of extended asymptotic giant branch stars are found beyond the red giant branch tip. Near the faint limits of our data is a populous red clump superimposed on the red giant branch. Efforts to fit self-consistent stellar population models based on the Geneva stellar evolution tracks yield a revised distance of 760 kpc. Quantitative fits to the stellar population are explored as a means to constrain the star formation history. The numbers of main sequence and core helium-burning blue loop stars require that the star formation rate was higher in the recent past, by a factor of 3-4 about 1~Gyr ago. Unique results cannot be obtained for the star formation history over longer time baselines without better information on stellar metallicities and deeper photometry. The youngest model consistent with the data contains stars with constant metallicity of Z$=$0.001 which mainly formed 2-4 Gyr ago. If stellar metallicity declines with increasing stellar age, then older ages are allowed of up to $\\approx$8~Gyr. However, even at its peak of star forming activity, the intermediate-age dominated model for the Pegasus dwarf most likely remained relatively dim with M$_V \\approx -$14. ",
        "introduction": "The Pegasus dwarf irregular galaxy  (DIG) was first identified by A. Wilson (see Holmberg 1958) and its status as a nearby dwarf was confirmed only with the detection of HI by Fisher \\& Tully (1975).  Despite its small size, Pegasus supports ongoing star formation as evidenced by the presence of luminous blue stars (Hoessel \\& Mould 1982, Christian \\& Tully 1983, Sandage 1986) and small HII regions (Hunter, Hawley, \\& Gallagher 1993). Pegasus is fairly typical of the least luminous DIGs detected in the Local Group and a few other nearby galaxy groups (cf. Karachentseva et al. 1987; Miller 1996).  Pegasus presents an interesting structural combination of a dwarf irregular system, with a chaotic appearance due to star formation, and a relatively symmetric dwarf elliptical (dE) or dwarf spheroidal (dSph) galaxy, typically without young stars.  Pegasus has a smooth outer envelope with elliptical isophotes and a brighter core that contains at least two OB associations (Ivanov 1996). It also has a relatively low amount of HI, as indicated by its moderate (for a DIG) ratio of M$_{HI}/$L$_V =$0.4 (Hoffman et al. 1996). Pegasus may therefore be a nearby example of a transition object between a dwarf galaxy dominated by current star-formation and one dominated by past star formation.  While the evolution of very low mass galaxies is still unclear, one possibility is that brief epochs of very active star formation produce a significant fraction of the stellar mass over a few dynamical time scales, or $<$1~Gyr (e.g. Babul \\& Ferguson 1996). Such events could disrupt or eject a small galaxy's ISM, thereby inhibiting further star formation for a time (e.g. Dekel \\& Silk 1986, Sandage \\& Fomalant 1993, Marlowe et al. 1995). Under these conditions much of the metal-rich supernova ejecta will also be lost and so the remaining stellar populations are expected to be relatively metal-poor and may show little metal enrichment over time.  It is interesting to see if any indications exist for epochs of enhanced star formation activity in the recent history of the Pegasus dwarf, which might have led to its transitional morphological structure.  This paper presents the first optical study of the stellar population of the Pegasus dwarf based on observations obtained with the Wide Field Planetary Camera 2 (WFPC2) on the {\\it Hubble Space Telescope} ({\\it HST}). Supplementary ground-based data were obtained with the WIYN\\footnote{The WIYN Observatory is a joint facility of the University of Wisconsin-Madison, Indiana University, Yale University, and the National Optical Astronomy Observatories.} 3.5-m telescope. The Pegasus dwarf is highly resolved into individual stars by WFPC2.  These data sets allowed us to obtain high quality color-magnitude diagrams (CMDs) for the Pegasus dwarf, which we analyze to derive its recent star formation history (SFH).  The next section reviews the global properties of Pegasus and \\S3 describes our new observations. Our analysis of the color-magnitude diagrams is covered in subsequent sections, the derived SFH is presented in \\S7, and the results are summarized in \\S8. ",
        "conclusions": "The Pegasus dwarf irregular galaxy is a nearby, low SFR, low mass, dwarf galaxy.  Our revised distance to the Pegasus dwarf of 760 kpc places this galaxy in the M31 ``family'', but at a projected distance on the sky of 400 kpc, it is unlikely to be bound to M31.  Pegasus is probably a small, independent member of the Local Group.  The established interstellar gas reserves in Pegasus from the measured HI content suggest that the gas (including cosmic He)-to-stellar mass ratio is M$_g /$M$_* =$~0.2.  Pegasus lacks the raw material to substantially enlarge its stellar mass, but is not yet a stellar fossil. At the present time its SFR is low, SFR$_0 \\approx$3$\\times$10$^{-4} \\msun~$yr$^{-1}$, and the time scale to exhaust the gas at this rate is 13 Gyr.  The lifetime average SFR is several times higher, so it is possible that a return to normal SFRs could result in gas exhaustion within a few Gyr.  Pegasus is like many other irregular galaxies where star formation is ``down but not out'' (see Hunter 1997).  Our observed CMDs for the Pegasus dwarf obtained with WFPC2 in the equivalent of BVI colors show a sparsely populated MS and a complicated ``red plume'' consisting of a fairly broad RGB, a pronounced RC (cleanly detected here for the first time), and an extended-AGB. The extended-AGB covers much of the optical galaxy on ground-based images taken with the WIYN Telescope. Our quantitative analysis of these CMDs establishes that: (1) The reddening of E(B$-$V)$=$0.15 is higher than previously thought. This mainly accounts for our reduction in distance from $D=$960 kpc found by Aparicio (1994) to D$=$760$\\pm$100~kpc. (2) Star formation is ongoing at a low level and has produced at least one HII region, a blue MS,  and core HeB blue loop stars. (3) The RGB is prominent, densely populated, and relatively blue. (4) The presence of a pronounced RC, the width of the RGB, and the lack of a well-defined tip of the RGB are all indicators of a strong intermediate age stellar population component.  We have quantitatively modeled the SFH for Pegasus by fitting our observed CMDs following the methodologies of Tolstoy \\& Saha (1996) and Dohm-Palmer et al. (1997b). The short term trend in SFR shows a modest decline in SFR during the past 0.5~Gyr from the densities of MS and blue core HeB stars as functions of luminosity. The Pegasus dwarf therefore shows a slowly varying recent mean SFR, but nothing that would qualify as a recent star burst.  Our data are not sufficiently deep to yield a unique description of the SFH over more than $\\approx$1~Gyr. Instead we find a spectrum of possible models for the long term SFH of the Pegasus dwarf. At one extreme is a rapid evolution model that makes most of the stars between 2 and 4 Gyr before the present and assumes constant metallicity at Z$=$0.001. At the other extreme are less well-defined models where metallicity declines in older stars; these models have lower SFRs spread over longer time intervals, extending perhaps to 10 Gyr in the past.  The historical optical characteristics of Pegasus depend on its SFH. However, for models where the SFR has been roughly constant at intermediate ages of a few Gyr, it would have been most luminous (M$_V \\approx-$14) and relatively blue when its SFR began its recent decline about 2 Gyr ago.  Even at its peak Pegasus was most likely a ``dwarf'' in terms of absolute magnitude.  A galaxy like Pegasus in our most actively star-forming model would be brightest at a lookback time of $\\approx$2~Gyr, or a redshift of $z \\approx$0.2. At this distance a Pegasus-like actively star-forming dwarf would have V$\\sim$26, blue colors, and an angular size of $<$0.5 arcsec. Thus this type of object could resemble some of the low redshift, faint blue galaxies (e.g. Odewahn et al. 1996).  If the Pegasus dwarf were to lose its gas supply, it might structurally resemble a spheroidal dwarf galaxy. Lo et al. (1993) show the HI velocity field in Pegasus to be fairly chaotic. They place an upper limit of 5 km~s$^{-1}$ on gas rotation; measure a local velocity dispersion of about 5 km~s$^{-1}$; and a global dispersion of 9 km~s$^{-1}$. Interestingly, stars formed from gas with these random velocities would likely retain a velocity dispersion similar to that found in the most luminous of the Galactic dSph galaxies (e.g., Gallagher \\& Wyse 1994). Furthermore, the global rotation is sufficiently slow that the ratio of the stellar velocity dispersion to rotation velocity  also is likely to be greater than 1, and so the older stellar body of the Pegasus dwarf probably has a spheroidal rather than thin disk structure, and resembles a low luminosity dE or dwarf spheroidal galaxy. Measurements of stellar kinematics would test this possibility, and could be done through observations of the relatively bright extended-AGB stars in the Pegasus dwarf.  Stellar photometry at the level of the HB will similarly enhance our ability to constrain the long term SFH of the Pegasus dwarf. The blue HB that is seen in And I (Da Costa et al. 1996) and NGC~147 (Han et al. 1997) demonstrate the existence of old, metal-poor stellar populations, while also yielding reliable distances. Evidently these two spheroidal galaxies made most of their stars over a time span of several Gyr centered at about 8-10~Gyr before the present. In comparing the Pegasus dwarf to these galaxies we need to establish the importance of its very old ($>$8~Gyr), metal-poor stellar population component, which define the level of star forming activity during the primary epoch of giant galaxy formation. By way of contrast, in the LMC such stars are rare (Olszewski,  Suntzeff, \\& Mateo 1996), and we would like to know if this is a general distinction between the irregular and spheroidal dwarf galaxies.  This information is essential in charting possible relationships between present-epoch dwarf galaxies, and faint blue galaxies seen at moderate-to-high redshifts corresponding to lookback times exceeding several Gyr.  The improvements in the quantitative analysis of stellar populations in nearby galaxies made possible by WFPC2 on the {\\it HST} prove that despite its small mass, the Pegasus dwarf has had an extended evolutionary history; this galaxy has likely supported star formation for at least 3~Gyr. The attractive idea that very small galaxies should experience a single primary and possibly cataclysmic episode of star formation during their lifetimes does not appear to hold in the Pegasus dwarf or many other low mass Local Group galaxies. We are then left with the intriguing problems of understanding what physical mechanisms allow some small galaxies to produce their stars slowly and without violence, and how this relates to the variety of stellar age mixes and gas contents found in extreme dwarf galaxies in the nearby universe.  \\bigskip  Support for this work was provided by NASA through grant GO-5915 from the Space Telescope Science Institute, which is operated by AURA, Inc., under NASA contract NAS 5-26555. This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. ET thanks Ralph Bohlin for research support during the early phases of this project.  JSG expresses his appreciation for partial travel support from ST-ECF which made visits to Garching possible. We thank an anonymous referee for comments which improved this paper."
    },
    "9805/astro-ph9805237_arXiv.txt": {
        "abstract": "The Galactic microquasars GRS1915+105 and GRO J1655-40 were targets for dedicated monitoring programs with $RXTE$ throughout 1996 and 1997. We review recent results with particular attention to the X-ray QPOs seen in these sources. We confirm the reappearance of 67 Hz oscillations in GRS1915+105 in several different spectral states. Six intervals, each of 1 to 6 months duration, yield QPO detections with a central frequency of $66.7 \\pm 1.4$ Hz and a FWHM of $3.4 \\pm 1.0$ Hz. Comparing these observing intervals, the average X-ray luminosity varies by a factor of 4 with no significant change in the central QPO frequency. GRS1915+105 has been active in X-rays since 1992, and it shows no signs of evolving toward quiescence. In contrast, GRO J1655-40, which began its most recent X-ray outburst on 1996 April 25, is seen to fade below 20 mCrab on 1997 Aug 17. There are no reappearances of the 300 Hz QPO during the entire second half of the X-ray outburst; during this time the X-ray spectrum is dominated by the soft, thermal component. There are still several models seeking to explain the high frequency QPOs in microquasars. While it is likely that the cause invokes some effect of General Relativity, the correct model and the details of the emission mechanism remain uncertain. ",
        "introduction": "Two Galactic X-ray sources known to produce relativistic radio jets are GRS 1915+105 \\cite{Mir94} and GRO J1655-40 \\cite{Tin95,Hje95}. Optical observations of GRO J1655-40 have provided dynamical evidence for a 7 \\msun black hole \\cite{Oro97} in a 2.6 day binary orbit with a $\\sim$F4 IV companion star. GRS1915+105 is presumed to be a black hole binary, based on its X-ray high luminosity and similarities with GRO J1655-40. However, a direct measurement of the motion of its companion star has been prevented by interstellar extinction, which limits optical/IR studies of GRS1915+105 to wavelengths $> 1$ micron \\cite{Mir94}. While each source was active at radio frequencies, H I absorption measurements were combined with Galactic rotation models to derive distance measurements of 12.5 kpc and 3.2 kpc for GRS 1915+105 \\cite{Mir94} and GRO J1655-40 \\cite{Tin95}, respectively.  GRS 1915+105 is a transient X-ray source, and the BATSE light curve (20--100 keV) indicates that bright X-ray emission began during May 1992 \\cite{Har97}. Before the launch of the $Rossi ~X$-$ray ~Timing ~Explorer$ ($RXTE$), observations in soft X-rays were sporadic, and GRS1915+105 may have persisted as a bright source in soft X-rays since 1992. When the All Sky Monitor (ASM) on $RXTE$ established regular coverage on 1996 Feb 22, the source was bright and highly variable, and it has remained so throughout 1996 and 1997. The ASM light curve, which is shown in Figure~\\ref{fig:asm19}, illustrates both the extent of the intensity variations and also the repetitive character of particular variability patterns. The early ASM light curve was used to initiate $RXTE$ pointed observations (PCA and HEXTE instruments), which began on 1996 April 6.  Since then the source has been observed once or twice per week, and most of the data are available in a public archive. At the higher time resolution provided by PCA light curves, there are again dramatic and repetitive patterns of variations \\cite{Gre96}.  These results are one of the extraordinary chapters in the history of high-energy astronomy.  \\begin{figure*} \\centerline{\\psfig{figure=asm1915.ps,width=16cm,height=16cm} } \\caption{ASM light curve (2--12 keV) of GRS1915+105 for 1996 and 1997. The Crab Nebula, for reference, yields 75.5 ASM c/s. The ASM hardness ratio, $HR2$ is defined as the count rate in the 5--12 keV band relative to the rate in the 3--5 keV band. The time intervals that correspond with our groups of combined X-ray power spectra (see Table 1) are shown above the light curve.} \\label{fig:asm19} \\end{figure*}  Fourier analyses of the first 31 PCA observations \\cite{Mor97} of GRS1915+105 revealed 3 different types of oscillations: a quasi-periodic oscillation (QPO) with a constant frequency of 67 Hz; dynamic, low-frequency (0.05 to 10 Hz) QPO with a large variety of amplitudes and widths; and complex, high-amplitude dip cycles ($10^{-3}$ to $10^{-1}$ Hz) that are related to the extreme X-ray variations noted above. The combined characteristics of the power spectra, light curves, and energy spectra were interpreted as representing four different emission states \\cite{Mor97}, none of which resemble the canonical states of black hole binaries \\cite{Van95}.  The other microquasar, GRO J1655-40, was first detected with BATSE on 1994 July 27, and the correlation between hard X-ray activity and the ejections of relativistic radio jets \\cite{Har95} was an important step in establishing the relationship between accretion changes and the formation of jets. During late 1995 and early 1996, GRO J1655-40 entered a quiescent accretion state, permitting optical spectroscopy of the companion star, which led to our knowledge of the binary constituents and mass of the black hole~\\cite{Oro97}, as noted above.  The ASM recorded a renewed outburst from GRO J1655-40 \\cite{Lev96} that began on 1996 April 25. The ASM light curve is shown in Figure ~\\ref{fig:asm16}. With great fortune a concurrent optical campaign was in progress, and it was determined that optical brightening preceded the X-ray turn-on by 6 days, beginning first in the I band and then accelerating rapidly in the B and V bands. These results provide concrete evidence favoring the accretion disk instability as the cause of the X-ray nova episode.  \\begin{figure*} \\centerline{\\psfig{figure=asm1655.ps,width=16cm,height=16cm} } \\caption{(top) ASM light curve (1.5--12 keV) of GRO J1655-40 for 1996 and 1997. The tick marks above the light curve show the times of RXTE pointed observations, either from the public archive (1997) or our guest observer program (1996). (bottom) The ASM hardness ratio, $HR2$ as defined previously.} \\label{fig:asm16} \\end{figure*}  The $RXTE$ observations of GRO J1655-40 indicate a more stable form of accretion. X-ray spectral variations (see Fig.~\\ref{fig:asm16}) resemble the canonical ``soft/high'' and ``very high'' states in black hole binaries \\cite{Rem98,Van95}. There are X-ray QPOs in the range of 8--22 Hz, and there is also a transient, high-frequency QPO at 300 Hz~\\cite{Rem98}. This QPO is detected only when the X-ray power-law component reaches its maximum strength.  The efforts to explain the 67 Hz QPO in GRS1915+105 and the 300 Hz QPO in GRO J1655-40 commonly invoke effects rooted in General Relativity (GR). There are at least 4 proposed mechanisms that relate the QPO frequency to a natural time scale of the inner accretion disk in a black hole binary.  These are: the last stable orbit \\cite{Sha83,Mor97}, diskoseismic oscillations \\cite{Per97,Now97}, frame dragging \\cite{Cui98}, and an oscillation in the centrifugal barrier \\cite{Tit98}. The physics of all of these phenomena invokes GR effects in the inner accretion disk. It has also been proposed that the high frequency QPOs may be caused by an inertial-acoustic instability in the disk \\cite{Che95} (with non-GR origin), although the oscillation in GRO J1655-40 would extend this application to higher frequencies than had been argued previously.  In this paper we advertise some recent work that associates jet formation in GRS1915+105 with features in the X-ray light curve.  We then turn to the topic of X-ray QPOs. New results are presented on the reappearance of 67 Hz oscillations in GRS1915+105.  Finally we describe the various QPO tracks that appear in GRO J1655-40, and we explain how they behave in response to the strength of the power-law component in the X-ray spectrum. ",
        "conclusions": ""
    },
    "9805/astro-ph9805371_arXiv.txt": {
        "abstract": "We present the results of VLA observations in the ground-state hydroxyl (OH) transition at 1720 MHz toward 20 supernova remnants (SNRs).  We detect compact emission from four objects.  For three of these objects (G32.8$-$0.1, G337.8$-$0.1, and G346.6$-$0.2), we argue that the emission results from masers which are shock-excited due to the interaction of the SNR and an adjacent molecular cloud.  We observe a characteristic Zeeman profile in the Stokes V spectrum, which allows us to derive a magnetic field of 1.5 and 1.7 mG for G32.8$-$0.1 and G346.6$-$0.2, respectively.  The velocity of the masers also allows us to determine a kinematic distance to the SNR. Our criteria for a maser to be associated with an SNR along the line of sight are that the position and velocity of the maser and SNR must agree, and the OH(1720) emission must be unaccompanied by other OH lines. ",
        "introduction": "Supernova remnants (SNRs) are a primary source of energy and heavy elements in the ISM. While this is well accepted, it is difficult to demonstrate specific instances of interaction between the SNR and cold dense gas in the ISM. Massive stars evolve quickly and are thought to influence the molecular clouds (MCs) from which they formed (see Elmegreen 1998 for a review).  Circumstantial evidence, such as morphological signatures, has been the primary evidence of this interaction (e.g. Routledge et al.  1991, Landecker et al.  1987). Interpretation of such observations is inconclusive because of confusion due to projections along the line of sight.  Thus there is a need for an indicator of SNR-MC interaction.  In a series of recent papers, we have utilized the 1720 MHz line of the hydroxyl molecule (OH) as a diagnostic for such an interaction. The ground state of the OH molecule has four transitions: the main lines at 1665.4018 and 1667.3590 MHz and the satellite lines at 1612.231 and 1720.530 MHz.  Main line OH masers are commonly found near compact H\\thinspace{II} regions, while 1612 MHz OH masers are found in the circumstellar shells of evolved stars.  Goss \\& Robinson (1968) first detected the satellite 1720 MHz OH maser emission toward the SNRs W28 and W44.  They found strong narrow-line 1720 MHz emission and broad-line 1612, 1665, and 1667 MHz {\\it absorption}, which distinguished the 1720 MHz masers from those found in other environments.  The remnants W28 and W44 later became prototypes for SNRs interacting with molecular clouds.  Frail, Goss, \\& Slysh (1994) reobserved W28, discovering 26 maser spots, and argued that the 1720 MHz radiation is maser emission due to a strong inversion of the line from collisional excitation by H$_{2}$ behind the SNR shock.  The location of the maser spots and the physical characteristics of the SNR and MC are consistent with the shock excitation model of Elitzur (1976).  In addition to being an unambiguous sign of SNR-MC interaction, OH masers can tell us more about the host SNR and its physical characteristics.  In systems where the velocity has been measured by other means, there is strong agreement between the velocity of the masers and the systemic velocity of the remnant.  This allows a kinematic distance to be derived.  Polarimetric observations reveal Zeeman profiles in the Stokes V signal which allow us to measure the magnetic field in several of the maser spots (Claussen et al.  1997). We can already loosely constrain the temperature and density of the environment of the maser via Elitzur's (1976) theory.  Furthermore, in the future it may be possible to use maser theory to infer accurate physical parameters of the postshock gas from the OH masers.  The initial success with W28 prompted Frail et al.  (1996) to perform a survey for 1720 MHz OH maser emission toward 66 SNRs, using the Green Bank and Parkes telescopes.  A total of 19 detections were made. Follow-up observations with an interferometer were made to distinguish between compact maser emission ($\\theta < 1''$) and spatially extended thermal line emission.  This resulted in the detection of OH(1720 MHz) masers toward five new SNRs.  This survey work has continued toward the Galactic Center (e.g. Yusef-Zadeh et al.  1996) and in the Galactic plane (Green et al.  1997).  Green et al.  (1997) reported that of the $\\sim$160 SNRs searched for OH emission, 17 have yielded maser emission, or 10\\% of the sample. However, they report higher detection rates for remnants within the ``molecular ring'' of the Galaxy.  Many of these individual SNRs have several maser spots.  For instance, W28, W44 and IC 443 have 41, 25, and 6 masers, respectively (Claussen et al.  1997).  The 10\\% detection rate in a random sample of Galactic SNRs suggests that more SNR-MC interactions have yet to be discovered in the remaining systems.  In the Green et al.  survey of 75 SNRs, there were 33 detections of OH. However, 16 were not re-observed with an interferometer to determine whether the detections result from compact maser emission or extended Galactic thermal emission.  The goal of the present study was to determine which of the remaining detections visible from the VLA\\footnote{The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.} were due to shock-excited OH maser emission.  In addition, we also observed several other sources which had not been sampled by the Green et al.  survey, but were likely cases for SNR-MC interaction.  Two of these were Cas A and CTB109. Also, we have noticed a strong correlation between OH maser detections and members of J. Rho's (1996) thermal composite SNRs.  This may suggest molecular cloud interactions as an origin for this class of SNRs.  Therefore, we observed three SNRs from this category (3C396, CTB104A, and HB3). ",
        "conclusions": ""
    },
    "9805/hep-ph9805284_arXiv.txt": {
        "abstract": "We examine the propagation of electromagnetic radiation through a strong magnetic field using the method of characteristics.  Owing to nonlinear effects associated with vacuum polarization, such waves can develop discontinuities analogous to hydrodynamical shocks.  We derive shock jump conditions and discuss the physical nature of these non-linear waves ",
        "introduction": "The nonlinear properties of electromagnetic waves traveling through a magnetized vacuum is of particular interest in the study of neutron stars.  \\jcite{Heisenberg and Euler}{Heis36} and \\jcite{Weisskopf}{Weis36} first derived nonlinear corrections to the Maxwell equations of the electromagnetic field.  \\jcite{Lutzky and Toll}{Lutz59} and \\jcite{Zheleznyakov and Fabrikant}{Zhel82} applied the weak-field expansion to show that shock waves can develop in the electromagnetic field.  \\jcite{Bialynicka-Birula}{Bial81} used the full expression for the nonlinear correction to the Lagrangian to study the generation of harmonics and other nonlinear phenomena in the propagation of EM radiation.  In this paper, we use the method of characteristics to study the evolution of waves governed by an arbitrary Lagrangian and then apply these techniques to the Heisenberg-Euler-Weisskopf-Schwinger Lagrangian \\cite{Heis36,Weis36,Schw51}. We find a concise expression for the opacity to shocking for disturbances travelling through an arbitrarily strong mangetic field.  After deriving a form for the energy-momentum tensor for an arbitrary electromagnetic Lagrangian, we use relativitistic fluid mechanics to derive the shock jump conditions and the long-term evolution of electromagnetic disturbances propagating in magnetic fields. ",
        "conclusions": "We have developed a relativistic fluid dynamic description of the electromagnetic field to derive the characteristic equations for electromagnetic waves in the presence of a strong external magnetic field.  By using a analytic expressions for the effective Lagrangian of QED, we have obtained simple expressions to estimate the opacity of waves to shocking for arbitrary magnetic field strengths.  Furthermore, by deriving a concise expression for the energy-momentum tensor for an arbitrary electromagnetic Lagrangian, we have calculated the shock jump conditions and find that the discontinuities travel at the speed of light in the medium relative to the external magnetic field.  By extending this fluid dynamic description, we follow the eventual decay of a electromagnetic disturbance travelling through an intense magnetic field.  For shocks to form from electromagnetic waves, not only is a strong external field required but also a source of coherent radiation.  A prime location for electromagnetic shocks is the vicinity of a neutron star with field strengths approaching and exceeding the critical value and coherent electromagnetic Alfven waves.  The study of the nonlinear corrections to the propagation of radiation through a plasma is beyond the scope of this work; however, we expect shock formation to be a hallmark of the nonlinear corrections of quantum electrodynamics and possibly an important process in the energy transmission near neutron stars."
    },
    "9805/astro-ph9805064_arXiv.txt": {
        "abstract": "We report new results for the elemental and isotopic abundances of the normally rare elements mercury and platinum in HgMn stars. Typical overabundances can be 4 dex or more.  The isotopic patterns do not follow the fractionation model of White et al (1976). ",
        "introduction": "The HgMn stars correspond in $T_{\\rm eff}$ to the main-sequence between A0 and B6 (11\\,000--16\\,000\\,K). Their abundance anomalies include both overabundances (e.g., Mn, P, Ga, Sr, Hg, Pt) and underabundances (e.g., He, Al, Ni, Co). They have no detectable ordered magnetic fields like the classical Ap stars (SrEuCr or Si types). While normal A0 stars have a typical rms \\vsini\\,$\\sim$\\,164\\,km\\,s$^{-1}$ (Dworetsky, 1974), HgMn stars have typical \\vsini\\ of 10--20\\,km\\,s$^{-1}$; some are as low as 2--3\\,km\\,s$^{-1}$. Smith (1996) is recommended for a more detailed overview. ",
        "conclusions": "Our results contradict what was previously a largely accepted concept, namely, that the strong Hg and Pt isotopic abundance anomalies could be parameterised in a fractionation model. This is clearly not the case in at least several stars.  One (unlikely) {\\it ad hoc} explanation is that the original abundances in these stars were different from cosmic. The fact that the Pt anomalies `shadow' the Hg may support this view. However, while surely involving radiative diffusion and gravitational settling, any complete theory needs to take into account other factors such as the detailed flux profile of the stellar atmosphere. These exotic abundances are now established results in need of a theoretical explanation.  \\begin{figure} \\begin{center} \\vspace*{-2mm} \\leavevmode \\psfig{figure=T18f3.ps,width=110mm,height=65mm} \\caption{Observed data versus synthetic fractionation profile for HR 7775: (a) \\hgii $\\lambda$3984; (b) \\ptii $\\lambda$4061 } \\label{fig3} \\vspace*{-3mm} \\end{center} \\end{figure}"
    },
    "9805/astro-ph9805122_arXiv.txt": {
        "abstract": "This is the first part of a study of the  detailed X-ray properties of the cores of nearby clusters. We have used the flux-limited sample of 55 clusters of Edge et al. (1990) and  archival and proprietary data from the {\\it ROSAT} observatory. In this paper an  X-ray spatial analysis based on the surface-brightness-deprojection technique is applied to the clusters in the sample with the aim of studying their cooling flow properties. We determine the fraction of cooling flows in this sample to be 70-90 percent and estimate the contribution of the flow region to the cluster X-ray luminosity. We show that the luminosity within a strong cooling flow can account for up to 70 percent of a cluster X-ray bolometric luminosity. Our analysis indicates that about 40 percent of the clusters in the sample have flows depositing more than 100 \\Msunpyr throughout the cooling region and that these possibly have been undisturbed for many Gyr, confirming that cooling flows are the natural state of cluster cores. New cooling flows in the  sample are presented and previously ambiguous ones are clarified. We have constructed a catalogue of some intracluster medium properties for the clusters in this sample. The profiles of the mass deposited from cooling flows are analyzed and evidence presented for the existence of breaks in some of the profiles. Comparison is made to recent optical and radio data. We cross-correlate our sample with the Green Bank, NVSS and FIRST surveys and to the volume-limited sample of brightest cluster galaxies  presented by Lauer and Postman (1994). Although weak trends exist, no strong correlation between optical magnitude or radio power of the brightest cluster galaxy and the strength of the flow is found. ",
        "introduction": "\\label{intro}  The X-rays emitted from clusters of galaxies represent a loss of energy of the intracluster gas. Thermal bremsstrahlung and line emission depend on the square of the gas density, which rises towards the cluster centre. As the density rises the loss rate increases and the resultant cooling timescale, t$_{\\rm cool}$ decreases. Within the cooling radius ($r_{\\rm cool}$), the cooling time becomes smaller than the age of the cluster ($t_{\\rm cool}$ $<$ $t_{\\rm age} \\approx$ {\\it H}$_0^{-1}$) and, under gravity and thermal pressure, the  gas cools and flows inward to maintain pressure equilibrium.  This simple picture describes the physics of a homogeneous cooling flow. In a more realistic situation initial inhomogeneities in the intracluster medium (ICM) lead to the establishment of a multiphase atmosphere which cools over a wide range of radii depositing cool material throughout the central $\\sim$200 kpc of a cluster of galaxies. (For reviews on cooling flows, see Fabian 1994, and Fabian, Nulsen \\& Canizares 1984).  Although  the mass cooling out of the X-ray band  represents only a negligible amount when compared to the total hot phase of the ICM, it is comparable to the mass of a brightest cluster galaxy (BCG) for flows which are undisturbed for about a Hubble time. Therefore understanding the physical processes of a cooling flow and of the fate of the cooled  material is of great importance for understanding the evolution of the core of a cluster.  Edge et al. (1992) stress that the characterization of a cluster as a cooling flow depends heavily on the spatial resolution and sensitivity of the instruments used to image the cluster. This was the main limitation for early work.  {\\it ROSAT} offers, so far,  the best opportunity to resolve the cores of the clusters and thus establish in a reliable manner the fraction of cooling flows in a flux-limited sample. It also offers a unique opportunity to study the occurrence of substructure in cooling flows (Peres, Buote \\& Fabian, in preparation) and to resolve mass deposition and cooling time profiles. The low background contamination of one of its detectors, the Position Sensitive Proportional Counter (PSPC) allows surface brightness profiles to be extracted to a large radius. The good image capabilities of the other instrument, the High Resolution Imager (HRI), is important for the study of substructure and, in the context of the present work, fundamental for determining the properties of more distant flows.  In the absence of a volume-limited sample, statistical studies of cooling flows are best pursued with the use of a flux-limited sample, such as the one presented by Edge et al. (1990). The clusters in this sample are the brightest 55 over the sky in the 2-10 keV band (B55 sample henceforth), and all have fluxes above 1.7 $\\times$ 10$^{-11}$ \\ergpcmsqps. The clusters were selected from observations with the Einstein and EXOSAT observatories and the HEAO-1 and Ariel-V satellites.  The work of Ebeling et al. (1996) on the X-ray Brightest Abell Cluster Sample (XBACS) showed that the completeness the B55 sample was satisfactory. A remaining issue in the analysis of cooling flow properties in this sample was the ability to image the cooling region in the centres of clusters. Here we analyse {\\it ROSAT} pointed observations  of the clusters in the B55 sample and deproject their surface brightness profiles. This allows us to build a catalogue of the ICM and cooling flow properties of the B55 clusters, and to compare these to recent optical and radio data. The structure of the paper is as follows. In sections 2 and 3 we describe the selection of the observations and the method of analysis, respectively. In section 4 we present our results, and in section 5 we summarize our conclusions. We assume throughout that {\\it H}$_0$ = 50 km s$^{-1}$ Mpc$^{-1}$ and  {\\it q}$_0$ = 0.5. ",
        "conclusions": "\\label{conclusion}   Our analysis of the X-ray properties of cluster cores with {\\it ROSAT} showed that a cooling flow is the natural state of the cores of nearby clusters of galaxies. They constitute 70-90 percent of the clusters in our flux-limited sample and deposit more than  100 \\Msunpyr in about 40 percent of the clusters. Large  cooling flows can contribute more than 70 percent of the cluster bolometric luminosity.  From our catalogue of ICM properties of the B55 sample we note that cooling times at the central bin are small and cooling times at 250 kpc from the cluster centre do not exceed $\\sim$4 times the age of the Universe (13 Gyr).  We show that mass deposition in cooling flows have an approximatelly linear dependence on radius and that clear breaks in some of the integrated mass deposition (IMD) profiles exist.  The cross-correlation of optical and radio properties with the strength of the flow yielded weak trends only. We noted that overall large cooling flows tend to have luminous BCGs and powerful radio sources at their centres, whereas no such rule exists for non-cooling-flows."
    },
    "9805/astro-ph9805191_arXiv.txt": {
        "abstract": "I have carried out spectrum synthesis of $R = 120,000$ spectra of several A and B stars having $v \\sin i$ less than about 6 km s$^{-1}$. The following conclusions emerge: (1) As $T_e$ descends from 12,000 to 8,000 K, microturbulent velocity $\\xi$ deduced from abundance analysis rises steadily from 0 to about 5 km s$^{-1}$. (2) Stars with $\\xi \\geq 1$ km s$^{-1}$ show direct evidence in their line profiles of the presence of macroscopic gas motions in the form of line asymmetry (bisector curvature) which grows with increasing $\\xi$. (3) Above $T_e \\approx 9000$ K, both weak and strong spectral lines can be modelled with reasonable accuracy by conventional LTE spectrum synthesis with a single assumed model atmosphere, abundance table, $v \\sin i$, and an appropriate (constant) value of $\\xi$. (4) In contrast, at $T_e \\approx 8,000$ K the weak spectral lines are much narrower than the strong lines. If the synthesis model is constrained in $v \\sin i$ and $\\xi$ by the weak lines, {\\em no} satisfactory model can be found for the strong spectral lines. Consequently, chemical abundances for such stars based only on strong lines may be significantly in error. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/gr-qc9805090_arXiv.txt": {
        "abstract": "Brans-Dicke gravity is remarkable not only in that General Relativity and Mach's Principle find a common enlarged scenario where they are mutually consistent, but also in that it provides a very interesting quantum cosmological model within the inflationary paradigm. The interplay between the Brans-Dicke scalar $\\Phi$ and the inflaton field $\\sigma$ plays an important r\\^{o}le during the course of inflation, and although the dynamics as such is governed by the potential, the onset and the end of inflation are determined by the values of both fields jointly. The relative position of the beginning-- and end-of-inflation curves (BoI and EoI respectively) is the most relevant factor in determining the resulting quantum cosmological scenario. The classification of potentials that is given in this paper is based on the criterion of whether the BoI and EoI boundaries enclose a finite or infinite area in the ($\\sigma$,$\\Phi$) plane where inflation takes place. It is shown that this qualitative classification distinguishes two classes of potentials that yield very different cosmologies and it is argued that only those theories in which BoI and EoI enclose a finite area in the ($\\sigma$,$\\Phi$) plane are compatible with our observable universe. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805185_arXiv.txt": {
        "abstract": "Seven high-resolution and high S/N CCD spectra were used to derive elemental abundances in the atmosphere of VW Ari A (T$_{eff}$=7200, $log~g$=3.7) which is the primary component of a visual binary system. The synthetic spectrum technique applied in the analysis allowed to reveal the following feature: the atmosphere of this star is strongly deficient in some metals, while light elements have solar-like abundances. Taking into account these results, one can suggest that VW Ari A is a $\\lambda$ Boo-type star. Another argument supporting this supposition is the following: on the diagrams ``$(b-y)-c_{1}$'', ``$(b-y)-m_{1}$'' and ``$(b-y)-\\beta$'' (Paunzen et al. 1997) VW Ari A falls exactly in the region occupied by the $\\lambda$ Boo stars.  Note also, that a previous analysis (Andrievsky et al. 1995) has shown that the secondary component of VW Ari has a normal metallicity. Differences in chemical compositions of the two components appear to be due to the specific evolution of the primary VW Ari A. ",
        "introduction": "VW Ari (HD 15165, BDS 1269) is a remarkable visual binary system consisting of two components: VW Ari A (V=6.$^{m}$71, A-type) and its companion (V=8.$^{m}$33, F-type). The primary VW Ari A is a multiperiodic pulsating star (probably of $\\delta$ Sct-type) having non-radial modes. This star shows the spectrum typical of very metal-deficient stars. The rather high $v\\sin i$ value found for this star, makes it difficult to derive accurate elemental abundances. A first attempt was undertaken by Andrievsky et al. (1995), who showed that calcium and iron are strongly deficient in the atmosphere of VW Ari A, while the secondary component possesses a solar-like chemical composition. Such a strange discrepancy between the metallicities of the two components can be explained by several hypotheses. For example, these stars possibly do not constitute a physical pair or, in case they do, such an unusual stellar system could be formed as a result of stellar capture.  Nevertheless, taking into account that 1) with a rather high probability VW Ari is a binary system and 2) the probability of stellar capture in the field is too small, we propose that the difference in chemical composition of both components could appear simply due to the peculiar evolution of VW Ari A as a $\\lambda$ Boo-type star.  The atmospheres of this type of stars are known to be strongly deficient in some heavy metals, while CNO-elements exhibit solar-like abundances (see e.g. St\\\"urenburg, 1993).  To check this hypothesis, we performed a detailed spectroscopic analysis of VW Ari (primary component of the system) based on the spectral synthesis technique. ",
        "conclusions": "The abundance pattern in the atmosphere of VW Ari resembles that of $\\lambda$ Boo-type stars (see, e.g. St\\\"urenburg, 1993, Andrievsky et al., 1998): normal abundances (or slight underabundances) of carbon and oxygen and strong deficiency of other elements.  An additional confirmation that VW Ari could be a $\\lambda$ Boo star is its position in $(b-y)-m_{1}-c_{1}-\\beta$ diagrams. This star possesses photometric characteristics which place it exactly in the region occupied by $\\lambda$ Boo stars.  Supposing that VW Ari belongs to the $\\lambda$ Boo group, one can also easily explain the remarkable difference between the metallicities of this star and of its companion F-star with solar-like abundances (Andrievsky et al., 1995)."
    },
    "9805/astro-ph9805300_arXiv.txt": {
        "abstract": "Based upon the results of VLA observations, we report the detection of two unresolved radio sources that are coincident with the reported optical position of SN~1923A in M83.  For the source closest to the SN position, the flux density was determined to be $0.30 \\pm 0.05$ mJy at 20 cm and $0.093 \\pm 0.028$ mJy at 6 cm.  The flux density of the second nearby source was determined to be $0.29 \\pm 0.05$ at 20 cm and $0.13 \\pm 0.028$ at 6 cm.  Both sources are non-thermal with spectral indices of $\\alpha = -1.0 \\pm 0.30$ and $-0.69 \\pm 0.24$, respectively. SN 1923A has been designated as a Type II-P.  No Type II-P (other than SN~1987A) has been detected previously in the radio. The radio emission from both sources appears to be fading with time. At an age of approximately 68 years when we observed it, this would be the oldest radio supernova (of known age) yet detected. ",
        "introduction": "Some Type II supernovae have exhibited radio emission within a few years of explosion, including SNe 1970G in M101 (Gottesman et al. 1972; Allen et al. 1976), 1978K in NGC 1313 (Ryder et al. 1993), 1979C in M100 (Weiler et al. 1986), 1980K in NGC 6946 (Weiler et al. 1986, 1992), 1981K (van der Hulst et al. 1983; Weiler et al. 1986), 1986J in NGC 891 (Rupen et al. 1987), 1988Z in MCG +03-28-022 (Van Dyk et al. 1993b) and 1993J in NGC 3031 (Van Dyk et al. 1994).  In addition SN 1987A was detected in the radio shortly after outburst (see Turtle et al. 1987), but at a level significantly below that seen from the more distant Type II supernovae and from the Galactic SNR Cas A.  Radio emission was also detected for a brief period from two Type Ibs, 1983N and 1984L (Sramek, Panagia, \\& Weiler 1984; Weiler et al. 1986; Weiler \\& Sramek 1988), and from the Type Ic supernova 1990B (Van Dyk et al. 1993a).  No Type Ia SNe have been detected in the radio despite several searches (e.g. \\cite{eck95}).  Radio emission from supernova remnants (SNRs) is typically observed long after the supernova phase.  Such factors as the density of the local interstellar medium affect the turn-on time in models such as those of Cowsik \\& Sarkar (1984) based upon the Gull (1973) piston model.  These models typically suggest a minimum of 100 years for the formation and brightening of an SNR. SN~1987A is already indicating an increase in brightness with time (Gaensler et al. 1997), but it is entering a phase of delayed circumstellar interaction.  We (Cowan \\& Branch 1982, 1985) have defined intermediate-age supernovae, from $\\sim$~10--300 years old, as spanning the period well after the optical emission fades (typically about 2 years) and before the turn-on of radio emission from an SNR (assumed to take at least 100 years).  This time period is critical in understanding the later stages of stellar evolution. In particular, the circumstellar mass loss rate for the supernova progenitors is a critical component in the initiation and duration of radio emission in the models of Chevalier (1982, 1984).  To study this transition period, we and others have attempted to detect radio emission from intermediate-age supernovae. While there have been a number of unsuccessful searches (see \\cite{eck95,eck96,eck98}), SNe 1950B \\& 1957D in M83 (\\cite {cow85,cow94}), SN 1968D in NGC 6946 (Hyman et al. 1995) and SN 1970G in M101 (\\cite{cow91}) have been detected in the radio more than a decade after explosion. (SN 1961V was also detected [\\cite{bra85,cow88}], but there is uncertainty about whether it really was a supernova [Filippenko et al. 1995].) We also note that while the radio emission from SN~1980K has abruptly dropped after approximately ten years (\\cite{wei92,mon98}), SN~1979C (at a greater distance than SN~1980K) is still emitting at detectable levels (\\cite{wei91})  In this {\\it Letter} we report the probable detection of SN~1923A in M83, which at an age of $\\sim$ 68 years would be the oldest intermediate-age supernova of known age. ",
        "conclusions": "Although the detected sources are relatively weak, the non-thermal nature and the apparent positional coincidence with the location of SN~1923A (as shown in Table 2) make it probable that we have detected radio emission from this supernova.  The sources are separated by about $3\\fasec 5$ which, at the distance to M83 (4.1 Mpc), corresponds to almost 70 pc, thus the sources cannot both be from SN 1923A. Past optical studies of the site of SN~1923A have shown evidence for an H~II region at or near the supernova site. Rumstay \\& Kaufman (1983) list an H II region --- no. 59 in their paper --- within $1\\asec-2\\asec$ of the SN position as based upon offsets with respect to the center of its parent galaxy.  Richter \\& Rosa (1984) refer to Rumstay \\& Kaufman H II region no. 59 as an H II region associated with and lying $1\\asec$ from SN 1923A.  Pennington et al. (1982) also note that SN 1923A appears to be coincident with an H~II region.  A map of H II regions by de Vaucouleurs, Pence, \\& Davoust (1983), when overlaid with the scaled radio map, has our source for SN~1923A directly over the H II region. Since the progenitor star of SN~1923A has been estimated to have been massive, $\\simeq$ 18 M$_{\\odot}$ (Pennington et al. 1982), it should not have moved much prior to explosion, and the H~II region near the radio source may be associated with the supernova's progenitor. The presence of an H~II region at the SN site reduces the chances that the two sources are actually background sources.  There is evidence for other similar associations of radio supernovae (RSNe) and H II regions (Van Dyk 1992), including SN~1957D (also estimated to have resulted from a massive star) in M83 (CRB).  Clearly new examinations of the area surrounding the site of SN~1923A in M83 to search for an optical counterpart to the radio source are warranted.  On the basis of what is known about the shape of its light curve, SN~1923A has been tentatively designated as a sub-luminous Type II-P (Patat et al. 1994, Schaefer 1996). No Type II-P (other than SN~1987A) has been detected previously in the radio.  In Figure 2, we plot the radio luminosity at 20 cm of the new source assuming it to be SN~1923A, along with several other extra-galactic RSNe and two Galactic SNRs, as a function of time since outburst.  Data and fits (solid lines) in Figure 2 for the well-studied Type II-L SN 1979C were taken from Weiler et al. (1986, 1991), for the Type II-L SN 1980K from Weiler et al. (1986, 1992), for the Type Ib SN 1983N from Weiler et al. (1986) and Cowan \\& Branch (1985), for the Type II-L SN 1970G from Cowan et al. (1991), for the Type II-L SN 1968D from Hyman et al. (1995) and for SNe 1950B \\& 1957D from CRB.  The distance to M83 (4.1 Mpc) was taken to be the Cepheid--based distance to NGC 5253, (Saha et al. 1995), a fellow member of the Centaurus group.  As Figure 2 illustrates, the luminosity of SN~1923A is comparable to, but slightly below, the two other intermediate-age supernovae we have detected in M83, SNe 1957D and 1950B.  (These two supernovae are suspected to have had massive progenitors but their actual supernova types are unknown.)  SN 1957D may actually be somewhat less luminous than plotted, because emission from an associated H II region may have contributed to the observed flux. The radio emission from SN~1923A falls between that of Cas A and the Crab.  At an age of approximately 68 years when last observed, SN~1923A would be the oldest radio supernova yet discovered.  Detectable radio emission from supernovae decades after explosion (but before the SNR phase) may in fact be uncommon, as evidenced by the small class of such known objects.  Observations of several other supernovae over a number of years also support that conclusion.  While SN~1979C is still detectable, SN~1980K has dropped off sharply after a decade of being followed. Recently Montes et al. (1997) have reported early-time radio emission from SN~1986E, while we were unable to detect this supernova at an intermediate-age despite a deep VLA search (Eck et al. 1996). Montes et al. (1997) argue that SN~1986E is a typical Type II-L, similar to SN~1980K and the fading radio emission can be adequately explained in terms of the Chevalier model.  What is the cause of the radio emission of SN~1923A?  No Type II-P events have been observed to undergo a prompt, bright circumstellar interaction such as that of radio supernovae.  SN~1987A, having been a sub-luminous Type II-P, underwent a prompt but dim circumstellar interaction that would not have been detectable in a galaxy beyond the local group, but now it is beginning what promises to be stronger, delayed interaction with a detached circumstellar shell that originated back in its red giant days.  If SN~1923A was a sub-luminous Type II--P, it may now be fading from the kind of delayed interaction that SN~1987A is just beginning.  The key radio observations needed now (apart from more firmly establishing the presence of a non-thermal radio source at the site of SN~1923A) are to trace the radio evolution of these sources, one of which is likely to be SN 1923A, the oldest radio supernova yet detected.  While we noted above that theoretical models have suggested a minimum time of 100 years for the onset of the SNR (and a brightening) phase, radio emission from the supernovae at this age has never been previously detected or studied.  Our observations at one wavelength seem to indicate that the source closest to the SN position is still fading with time, but our uncertainties are large. Additional observations of SN 1923A will help to understand more about the nature of radio emission as supernovae evolve from intermediate-age to the SNR phase."
    },
    "9805/astro-ph9805136_arXiv.txt": {
        "abstract": "This study combines data from a set of seven Cepheid-calibrated galaxies in Virgo, Leo I and Fornax and four supernov\\ae\\ with known metallicities, including metallicity corrections to the Cepheid distances, in order to determine a consistent mean value for the Hubble constant.  We derive Virgo and Coma distances over different paths, starting from the metallicity corrected Cepheid distances of Virgo and Leo I galaxies, and we relate them to the Virgo and the Coma recession velocities, paying special attention to thereby introduced correlations.  We also determine the velocity of the Fornax cluster and relate this velocity to its distance.  In addition we use the standard constant maximum brightness SN Ia method, as well as results from multicolor light curve shape (MLCS) analysis of SNe Ia, to derive the Hubble constant.  Using the requirement of statistical consistency, our data set determines the value of the metallicity coefficient (the change in magnitude per factor of ten in metal abundance) to be $0.31^{+0.15}_{-0.14}$ in good agreement with other known determinations.  We also use consistency as a criterion to select between mutually exclusive data sets, such as the two known, but conflicting, Virgo recession velocities (in favor of the lower value), as well as between two different analyses of (partly) the same supernov\\ae\\ (in favor of MLCS analysis). When combining data we pay attention to the statistically correct handling of common errors.  Our result is a Hubble constant with the value $H_0 = 68\\pm 5$ km s$^{-1}$ Mpc$^{-1}$. ",
        "introduction": "The value of the Hubble constant $H_0$ has been notoriously controversial, not only when determined by different methods, but even values resulting from the application of a single method to different objects are often in glaring conflict. The statistical error is of course large for small data samples, but in addition most determinations depend on secondary parameters which are ill known or guessed and which lack reliable error estimates, systematic errors for short.  In this paper we address ourselves to the problem of determining $H_0$ using data on galaxies with Cepheid-calibrated distances. Then two important sources of systematic errors can be discerned: the effects of the metallicity of the Cepheid  on the period-luminosity (PL) relation, and the determination of the distance to the Large Magellanic Cloud (LMC). Our main interest in this paper is to examine the metallicity effect, and we keep the LMC distance modulus at its classical value 18.5 $\\pm$ 0.1 mag (Madore \\& Freedman, 1991). In Section \\ref{disc} we discuss recent developments of determining LMC distance, and check how these will affect our results.  The Cepheid PL relation has been calibrated in the LMC by Madore \\& Freedman (1991). For quite some time it has been speculated that the Cepheid brightness and therefore the zero point of the PL relation is affected by the metallicity of the calibrating Cepheids (for a good review see Sasselov et al., 1997). Sometimes observations have refuted and sometimes confirmed the existence of this effect. The idea is that metal richer Cepheids would be brighter and therefore lie further away than what the standard PL analysis implies. However, the brighter Cepheids are also thought to be intrinsically redder, which will complicate the analysis, and make the change of distance to depend on wavelength, or in practise, the bandpasses used in the photometric analysis. Throughout this paper we talk about the metallicity effect to Cepheid distance modulus obtained using V and I bands, because that is the choice of Hubbe Space Telescope teams.  Although  the shift in distance depends in theory (Stothers, 1988) on the abundance of heavy elements, the only actually observable \"metallicity\" is usually the oxygen abundance of surrounding HII regions. Therefore in this paper we actually address ourselves to the systematic shift due to the [O/H] dependence of the distance moduli. Assuming that the ratio of oxygen abundance of HII regions to the corresponding value of LMC, equals the corresponding ratio of heavy element abundances, we interpret the [O/H] - effect as a metallicity effect.  For the metallicity correction of the distance modulus of a Cepheid we use a form \\be \\delta\\mu = \\gamma \\ ({\\rm [O/H] - [O/H]_{LMC}}). \\ee $\\delta\\mu$ is the correction needed to the distance modulus obtained by standard VI photometry ($\\mu_{true} = \\mu_{standard \\ VI} + \\delta\\mu$). [O/H] denotes the logarithm of the abundance of oxygen (by number) relative to hydrogen, in Solar units. The logarithmic dependence in the above formula follows Sasselov et al., (1997). The constant $\\gamma$ gives the correction needed for the distance modulus in magnitude per factor of ten in metal abundance (mag/dex).  The early results (Freedman \\& Madore, 1990) indicated a small effect, which was later contested by Gould (1994) who found a much larger $\\gamma$.  At that time the color variation was not accounted for.  More recently, three new determinations have become available: Beaulieu et al.  (1997) and Sasselov et al.  (1997) in the EROS project analyzed very large samples of Cepheids in the LMC and the SMC and published a $\\gamma$ value for the Z difference. We use $\\gamma = 0.48^{+0.1}_{-0.2}$ mag/dex (Sasselov, priv.comm.) which corresponds to the [O/H] difference between LMC and SMC. In statistical agreement with this Kennicutt et al.  (1998) have reported $\\gamma = 0.24 \\pm 0.16$ mag/dex from a study targeting Cepheids in two fields in M101.  The third result, $\\gamma = 0.4 \\pm 0.2$ mag/dex, is due to Kochanek (1997) who used multicolor photometry to determine 694 Cepheid distances in 17 galaxies, including the effects of temperature, extinction and metallicity. His simultaneous fitting approach to actual photometry data is perhaps the currently best way to address the question of metallicity effects on the cepheid PL relation, but it is rather complicated (a multi-parameter fitting procedure in a Bayesian approach). The approach adopted by us is different, we do not consider the systematic errors of the actual photometry, but instead we put emphasis on the error correlations of photometry results (magnitudes) and on error correlations arising from the metallicity correction in the averaging procedure.  Beaulieu et al.  (1997) and Sasselov et al.  (1997) applied their metallicity correction to reevaluate the Hubble constant for Cepheids in HST-observed galaxies with reported metallicities, and demonstrated that some of the previous disagreement between the $H_0$ values was indeed removed.  In the present paper we investigate whether the inclusion of the metallicity correction is enough to produce a statistically consistent $H_0$ average for a larger sample of galaxies, or whether inconsistencies remain, indicating the presence of yet further systematic errors.  Our sample consists of 7 galaxies with known distances, velocities and metallicities, and 4 supernov\\ae\\ with known V and B peak absolute magnitudes and metallicities. ",
        "conclusions": "\\label{conc}  The existence of well measured distances to galaxies for which also the metallicities are known, permits one to attempt to form a consistent set of data from which a combined value of the Hubble constant can be obtained. We have used four Virgo galaxies, two Leo I galaxies, one galaxy in Fornax, and two partially overlapping analyses of four supernov\\ae\\ (all listed in Section 2. and in Table 1) to study the metallicity dependence of the Hubble constant derived from these data.  The analysis introduces several types of correlations, and some of the input errors are correlated as well.  We take care of handling all such correlations statistically correctly.  We conclude that we can select between the two supernova analyses: the MLCS set (Riess et al, 1996) is consistent with all our other data sets, whereas the standard constant maximum brightness analysis of type Ia SNe (Sandage et al, 1996) set is not. We can also select between the two conflicting Virgo velocity values in the literature (in favor of the low velocity).  Our main results are two (using $\\mu_{LMC} = 18.5 \\pm 0.1$): a new value for the Hubble constant, $H_0 = 68\\pm 5$ km s$^{-1}$ Mpc$^{-1}$, and a new value for the metallicity coefficient (the correction for distance modulus in magnitude per factor of ten in metal abundance) of $0.31^{+0.15}_{-0.14}$ mag/dex, in excellent agreement with previous results. Thus the metallicity effect is clearly significant -- combining our $\\gamma$ value with that of Sasselov et al. (1997) for [O/H], and Kennicutt et al. (1997), we obtain $\\gamma=0.33^{+0.09}_{-0.10}$ which is $3\\sigma$ away from zero. Note the the analysis by Kochanek, yielding $\\gamma = 0.4 \\pm 0.2$, is in excellent agreement with this.  Taking into account the current knowledge of the LMC distance, we see that our best value is even more robust (see Section \\ref{disc}).  The general agreement of the final data set with our criterion of statistical consistency permits us to conclude that no further systematic errors are needed beyond those accounted for here."
    },
    "9805/astro-ph9805246_arXiv.txt": {
        "abstract": " ",
        "introduction": "\\label{intr} \\vspace{-2mm} 29\\,Cyg is the first among $\\lambda$\\,Boo-type stars for which definite evidence of pulsations has been found (Gies\\,\\&\\,Percy, 1977). Since the discovery of variabilty, 29\\,Cyg has been studied photometrically by Cooper \\& Walker (1994), Kusakin\\\\ \\& Mkrtichian (1996), Paunzen\\,\\&\\,Handler (1996).  The 1996 multisite two-stage  photometric and spectral campaign on 29 Cyg was organized by the CAN group (see Mkrtichian et al., 1998) between 5--17 August and 15--25 September against the background of continuous monitoring of the star from the CAN site TSAO between 27 July and 30 September. This campaign was a natural resumption of the 1995 photometric investigation of 29 Cyg (Kusakin\\,\\&\\,Mkrtichian, 1996).  \\vspace{-2mm} ",
        "conclusions": ""
    },
    "9805/astro-ph9805070_arXiv.txt": {
        "abstract": "The hydrogen-line spectrum of eight $\\lambda$ Boo type stars is studied. The observed H$_\\delta$ and H$_\\gamma$ profiles are compared with Kurucz's theoretical profiles. The existence of weak emission-like details in the cores of five $\\lambda$ Boo stars is demonstrated. The Inglis--Teller formula is used to calculate the electron densities. It is found that electron densities in the atmospheres of stars with peculiar H-line profiles are twice lower than in stars with normal profiles. The conclusion is made that stars with peculiar profiles exhibit some of the characteristics usually observed in stars with extended atmospheres. ",
        "introduction": "According to Gray (1988), $\\lambda$ Boo type stars can be divided into two groups. Stars in the first group have normal hydrogen-line (NHL) profiles, typical of the A dwarfs. The second group contains stars with peculiar hydrogen-line profiles (PHL). They have weak cores and broad, but shallow wings. St\\\"urenburg (1993) has found a significant correlation between Gray's classification and his gas shell index. This index is related to the presence of sharp circumstellar absorption details in the cores of some metallic lines like Ca K and Na D. The PHL stars have strong signs of a gas shell, while the NHL stars show only weak indications. Thus, the observational data clearly point to a connection between the $\\lambda$ Boo phenomenon and circumstellar gas. Here we report on the effective temperatures and gravities obtained from fits to theoretical hydrogen-line profiles, as well as on the gas shell signs found directly in the hydrogen-line spectrum of some PHL $\\lambda$ Boo stars. ",
        "conclusions": "\\vspace{-2mm}  The main result form the last subsection is that $N_{\\rm e}$ at the level $\\tau \\approx 0.1$ in the atmosphere of PHL type stars is nearly twice lower than we derived for $\\lambda$~Boo stars with NHL. This could be interpreted as some evidence that $\\lambda$~Boo stars with PHL are closer to the end of their evolution {\\em on~the~MS}, because in the spectral range A0--F0 $n_{\\rm max}$ reaches 17.0 for ZAMS, 18.5 for the middle of the main sequence, and 20.0 for the luminosity classes IV--III (Kopylov 1966). An additional support to this interpretation could be seen even in the fact that the $\\log g$ values of PHL stars determined from photometric data are in general smaller than those of NHL stars. The contradiction with the suggestion of Venn \\& Lambert (1990) and gas/dust separation scenario developed by Waters et al. (1992) (both of them require $\\lambda$ Boo stars to be young and near the ZAMS) is obvious.  The presence of gas shells or envelopes which surround many $\\lambda$ Boo stars can change the situation upside-down, making PHL stars to bear a resemblance to the well-known Be or Ae/shell stars. For example, Pleione shows a B8V spectrum while in its quiescent mode, while during the shell phases $n_{\\rm max}$ has reached 40. Thus, the larger $n_{\\rm max}$ in the spectra of PHL stars can be considered as yet another shell symptom. It is important to note that the influence of gas shells on the hydrogen-line spectrum of Ae/shell stars which show weak or no emissions usually is much more visible in the highest Balmer line numbers. This fact could explain why the $\\beta$ indices and the corresponding effective temperatures of PHL $\\lambda$ Boo stars can be close to those obtained from spectroscopy, while the Balmer discontinuity, and, therefore, luminosity indices like $c_{\\rm 1}$, {\\em d\\ } or $\\delta$ can be affected by the shell. The conclusion is that $\\log g$ values of PHL stars obtained by photometry seem to need a revision, because stronger evidences for circumstellar shells imply more disturbed $\\log g$ values. Unfortunately, hydrogen-line profiles offer no alternative solution, since the effective temperatures of these stars are mostly below the critical limit of about 8\\,300\\,K."
    },
    "9805/hep-ph9805259_arXiv.txt": {
        "abstract": "Oscillations between ordinary and sterile neutrinos can generate large neutrino asymmetries in the early universe.  These asymmetries can significantly affect big bang nucleosynthesis (BBN) through modification of nuclear reaction rates.  We study this phenomenon within a model consisting of the three ordinary neutrinos plus one sterile neutrino that can be motivated by the neutrino anomalies and the dark matter problem.  We calculate how the lepton asymmetries produced evolve at temperatures where they impact on BBN.  The effect of the asymmetries on primordial helium production is determined, leading to an effective number of neutrino flavours during BBN of either about $2.7$ or $3.1$ depending on the sign of the lepton asymmetry. ",
        "introduction": "Models with sterile neutrinos are currently favoured to accommodate all the neutrino oscillation measurements.  It is difficult to simultaneously account for the solar, atmospheric and LSND anomalies with just three ordinary neutrinos and hence only two mass-squared differences.  Therefore we require at least one additional neutrino if we wish to explain all three of the observed neutrino anomalies. Constraints from the Z-width suggest that any additional neutrinos must be sterile. Oscillations between ordinary and sterile neutrinos in the early universe have been studied in \\cite{enqvist,early,kirilova,rr1,shi} and found to generate large neutrino asymmetries\\cite{rr1,shi}, which can significantly affect big bang nucleosynthesis (BBN) reaction rates, and can result in an effective neutrino number of less than three during BBN \\cite{rr}. In this paper, the techniques developed in Ref.\\cite{rr} will be applied to the four neutrino model of Ref.\\cite{model} which has almost degenerate $\\nu_\\mu$ and $\\nu_\\tau$. This model is distinct from the four neutrino model of Ref.\\cite{rr} where a mass hierarchy between $\\nu_{\\mu}$ and $\\nu_{\\tau}$ was assumed.  A motivation for studying how neutrino oscillations can affect $N_{\\nu}^{eff}$, the effective neutrino number during nucleosynthesis, arises from discrepancies between the observed deuterium abundance and standard BBN. The primordial D/H ratio can be used to give a sensitive determination of the baryon to photon ratio $\\eta$ which, given the inferred primordial $^4$He mass fraction, can be used to predict the number of light neutrinos $N_{\\nu}^{eff}$.  Although there are numerous measurements of the deuterium abundance, obtaining an upper bound to the abundance is quite difficult, and there are conflicting observations.  A high deuterium measurement of $D/H=(1.9\\pm0.4)\\times10^{-4}$ was found in Ref.\\cite{rugers}, leading to $\\eta\\sim2\\times10^{-10}$, while in Ref.\\cite{tytler} a low deuterium result of $D/H=(2.3\\pm0.3\\pm0.3)\\times10^{-5}$ was obtained, suggesting that $\\eta\\sim7\\times10^{-10}$.  From these values of $\\eta$, and the helium mass fraction $Y_P$, the effective number of neutrinos can be calculated.  By way of example, the results \\begin{eqnarray} N_{\\nu}^{eff}=2.9\\pm0.3  \\text{   high D/H},  \\nonumber \\\\ N_{\\nu}^{eff}=1.9\\pm0.3  \\text{   low D/H}, \\end{eqnarray} were obtained in Ref.\\cite{hata}.  (For other calculations of $N^{eff}_{\\nu}$ see Ref.\\cite{others}.) Hence, if the low deuterium result is correct, standard BBN would require an effective neutrino number of less than three (provided that the value of $Y_P$ used in Ref.\\cite{hata} is correct). Clearly $N_{\\nu}^{eff} < 3$ would imply new physics. Further observations are required to pin down the D/H value (and also $Y_P$) although it now seems as though the low D/H result is favoured \\cite{tytler,turner}.  One possibility for the new physics is to look at the neutrino sector and study mechanisms by which the effective number of neutrinos can be made less than three.  A possible way of doing this is to create an electron-neutrino asymmetry, as this would directly affect the reaction rates which determine the n/p ratio just before nucleosynthesis. The n/p ratio controls the helium mass fraction $Y_P$ via the equation \\begin{equation} \\frac{dY_P}{dt}=-\\lambda(n\\rightarrow p)Y_P+\\lambda(p\\rightarrow n)(2-Y_P), \\end{equation} where the reaction rates \\begin{eqnarray} \\label{eq:rates} \\lambda(n\\rightarrow p)\\simeq \\lambda(n + \\nu_e \\rightarrow p + e^-) + \\lambda(n + e^+ \\rightarrow p + \\bar{\\nu}_e), \\nonumber \\\\ \\lambda(p\\rightarrow n) \\simeq \\lambda(p + e^- \\rightarrow n + \\nu_e) + \\lambda(p + \\bar{\\nu}_e \\rightarrow n + e^+), \\end{eqnarray} depend on the momentum distributions of the species involved. The processes in Eq.(\\ref{eq:rates}) for determining $n\\leftrightarrow p$ are valid for temperatures above about 0.4 MeV, below which the weak interaction rates become frozen out and neutron decay becomes the dominant factor affecting the n/p ratio.  An excess of $\\nu_e$ over $\\bar{\\nu}_e$ would reduce the n/p ratio, due to the modification of the neutrino momentum distributions through non-zero chemical potentials, thereby changing the rates for the processes in Eq.(\\ref{eq:rates}). Neutron decay is not significantly altered by lepton asymmetries.  A change in the helium mass fraction $Y_P$ can be related to a change in the effective number of neutrinos by \\cite{walker} \\begin{equation} \\delta Y_P \\simeq 0.012\\times \\delta N_{\\nu}^{eff}. \\end{equation} A large neutrino asymmetry could simply be postulated, but it is more interesting to look at physics of the neutrino sector which will allow a large neutrino asymmetry to evolve as a dynamical variable.  A four-neutrino model was studied in Ref.\\cite{rr}, involving the three ordinary neutrinos and one sterile neutrino, where the mass hierarchy $m_{\\nu_{\\tau}}\\gg m_{\\nu_{\\mu}}, m_{\\nu_e}, m_{\\nu_s}$ was considered by way of example.  In this case oscillations between $\\bar{\\nu}_{\\tau}$ and $\\bar{\\nu}_s$ resulted in an excess of $\\nu_{\\tau}$ over $\\bar{\\nu}_{\\tau}$ thereby generating a large tau-neutrino asymmetry and hence a large tau-lepton number.  $\\bar{\\nu}_{\\tau}-\\bar{\\nu}_e$ oscillations also occurred, and it was shown that this allowed some of the tau-neutrino asymmetry to be transferred to an electron-neutrino asymmetry. The effective number of neutrinos found in Ref.\\cite{rr} was either 2.5 or 3.4. The two values arise because there is an ambiguity involving the sign of the asymmetry and hence the prediction for $N^{eff}_{\\nu}$ (see Ref.\\cite{rr1} for a discussion of this issue).  For a positive asymmetry, $\\delta N_{\\nu}^{eff} \\simeq -0.5$ was obtained over a range of mass differences $\\delta m_{\\tau s}^2\\sim 10-3000\\text{eV}^2$, while for a negative asymmetry the result was $\\delta N_{\\nu}^{eff} \\simeq +0.4$. So we have the interesting situation that $N_{\\nu}^{eff}<3$ can be achieved although the precise value of $N_{\\nu}^{eff}$ is model dependent as this paper will illustrate. If future measurements can pin down $N_{\\nu}^{eff}$ precisely enough, then this quantity can be used to help discriminate various competing models explaining the observed neutrino physics anomalies. ",
        "conclusions": "We conclude that the neutrino oscillations in the four neutrino model of Ref.\\cite{model} can modify the effective number of neutrinos during nucleosynthesis and give $N^{eff}_{\\nu} < 3$. However, it turns out that the effect is relatively small, $\\delta N^{eff}_{\\nu} \\simeq -0.3$ (if $L_e > 0$). If future measurements confirm the low D/H result as well as the $Y_P$ value used in Ref.\\cite{hata}, then this model will become cosmologically disfavoured. On the other hand, if future measurements favour $N_{\\nu}^{eff}\\simeq 3$ then this model will remain viable.  There is also the interesting possibility that the neutrino mass favoured for dark matter \\cite{primack} will have to be recalculated. In fact, from Eq.(\\ref{eq:n}) the number of heavy neutrinos is actually about 1.5 because about $50\\%$ of the $\\bar{\\nu}_{\\tau}$ and $\\bar{\\nu}_{\\mu}$ have been converted into light $\\bar{\\nu}_s$. This would suggest that the optimal mass for hot+cold dark matter senarios is somewhat heavier, about $m_{\\nu_{\\tau}}=m_{\\nu_{\\mu}}=3.1\\text{eV}$ rather than $2.4\\text{eV}$, for this particular model.  This implies that the $\\delta m^2$ relavant for LSND is about $\\text{9eV}^2$ rather than $\\text{6eV}^2$ \\cite{footnote}."
    },
    "9805/nucl-th9805035_arXiv.txt": {
        "abstract": "We construct the equation of state (EOS) of nuclear matter using the relativistic mean field (RMF) theory in the wide density, temperature range with various proton fractions for the use of supernova simulation and the neutron star calculations. We first construct the EOS of homogeneous nuclear matter. We use then the Thomas-Fermi approximation to describe inhomogeneous matter, where heavy nuclei are formed together with free nucleon gas. We discuss the results on free energy, pressure and entropy in the wide range of astrophysical interest. As an example, we apply the resulting EOS on the neutron star properties by using the Oppenheimer-Volkoff equation. ",
        "introduction": "Supernova is a spectacular stellar phenomenon happening when massive stars ($M \\geq 8M_{\\odot}$) use up the nuclear fuel. Due to the gravitational core collapse, supernova explosion may lead to the formation of a neutron star or black hole in the central part of the massive star. In order to clarify the mechanism of the explosion and the whole phenomena of supernova explosion, it is necessary to study them by performing numerical simulations. For this purpose, we need the equation of state (EOS) of nuclear matter in the wide density and temperature range with various proton fractions, which appear during the collapse and the explosion.  There have been several EOS worked out so far for supernova study \\cite{BV81,OS82,HW85,LLPR85,LS91}. Most of them are constructed based on the non-relativistic Skyrme Hartree-Fock (SHF) framework. Although some of these EOS are carefully constructed in the form for the supernova simulations, there are many places where the EOS are discontinuous or not available, and the supernova simulations stop there due to the lack of the completeness of the EOS table. Sometimes, they are not even documented and it is difficult to fix the problem. Lattimer and Swesty worked out the EOS in the compressible liquid drop model based on the non-relativistic framework \\cite{LS91}. This EOS table is almost the unique one, with which many simulations are made at present.  Recently, there is a tremendous development in the description of nuclei and nuclear matter in terms of the relativistic many body theory \\cite{BM90,LMB92,EO96}. The relativistic Br\\\"uckner-Hartree-Fock (RBHF) theory provided the saturation close to the experimental point in nuclear matter. The relativistic effect provides the strong density dependent repulsion over the non-relativistic results due to the three-body Z-graph process. The RBHF theory provides also the strong spin-orbit splittings, which are the fundamental quantities of the shell model. Based on this finding of the RBHF theory, the phenomenological model was developed using the relativistic mean field (RMF) theory \\cite{SW86,BT92,ST94,SST97}. There are many works performed on nuclear ground states using the RMF theory. The results compare very well with experimental data on not only stable nuclei but also unstable nuclei, which are being obtained using radioactive nuclear beams \\cite{T95}.  Having success in describing nuclear properties using the RMF theory, it is extremely important to work out the EOS for astrophysical purposes. We have to work out consistent calculations for high density nuclear matter, inhomogeneous nuclear matter and finite nuclei under the wide range of density, proton fraction and temperature, which appear inside neutron stars and supernovae. For this purpose, we study the properties of dense matter with both homogeneous and inhomogeneous distribution in the RMF framework adopting the parameter set TM1 for the RMF lagrangian. The RMF theory with the TM1 is known to provide excellent properties of the ground states of heavy nuclei including unstable nuclei \\cite{ST94,HS97}. The TM1 was also used  for the giant resonances within the RPA formalism and demonstrated to provide good descriptions of giant resonances \\cite{MT97}.  For the description of inhomogeneous nuclear matter, which appears below $\\rho \\sim \\rho_0 /3 $ with $\\rho_0$ being the normal nuclear matter density, we use the Thomas-Fermi approximation \\cite{O93}. We assume that the inhomogeneous matter is composed of a lattice of spherical nuclei. At finite temperatures, a proton gas, as well as a neutron gas, may be present outside the nuclei. We minimize the free energy of the system with respect to the cell volume, proton and neutron radii and surface diffusenesses of the nucleus, and densities of the outside neutron and proton gases. We compare the calculated free energy with that of homogeneous matter to determine the phase of dense matter. This has to be done for each density, temperature and proton fraction in the wide range necessary for the astrophysical use.  This paper is arranged as follows. In Sect.2, we describe the RMF theory and the TM1 parameter set. In Sect.3, we introduce the Thomas-Fermi method to be used with the Wigner-Seitz approximation. In Sect.4, we show the numerical results. In Sect.5 we show the neutron star profiles using the EOS table with the Oppenheimer-Volkoff equation. Sect.6 is devoted to the conclusions and discussions. ",
        "conclusions": "This relativistic EOS of nuclear matter is designed for the use of supernova simulations and the neutron star calculations in the wide density and temperature range with various proton fractions. We have adopted the RMF theory with the non-linear $\\sigma$ and $\\omega$ terms, which was demonstrated to be successful in describing the properties of both stable and unstable nuclei as well as reproducing the similar self-energies in the RBHF theory. The relativity plays an essential role in describing the nuclear saturation and the nuclear structure \\cite{BM90}, it also brings some distinctive properties in the EOS comparing with the case in the non-relativistic framework. Therefore, it is very interesting and important to study the astrophysical phenomena such as supernova explosion, neutron star cooling and neutron star merger using the relativistic EOS through extensive comparison with the previous studies with the non-relativistic EOS.  We have performed the Thomas-Fermi calculation to describe the non-uniform matter which appears below $\\rho_B \\sim \\rho_0/3$. The same RMF theory as the one used at high density has been adopted as the input of the Thomas-Fermi numerical code. Hence, all the thermodynamical quantities such as free energy, entropy and pressure are found smooth in the resulting EOS. The non-uniform matter phase appears in a wide $\\rho_B-Y_p$ region in the phase diagram at low temperature, while it completely disappears as temperature increases beyond $15 MeV$. The formation of heavy nuclei in non-uniform matter  provides remarkable behavior in the EOS.  It is also important to pay attention to the experimental data of unstable nuclei, which are being obtained using radioactive nuclear beams, so that we can get more information about the properties of asymmetric nuclear matter from experiments \\cite{O98}. It is definitely important and meaningful to study unstable nuclei and astrophysical phenomena simultaneously in the same framework."
    },
    "9805/astro-ph9805051_arXiv.txt": {
        "abstract": "In this article the current knowledge of the group of \\LB stars is reviewed. These metal poor objects are quite outstanding compared to other chemically peculiar stars of the upper main sequence. Up to now no theory has been developed which is able to explain the majority of observational results. This article is mainly focused on the work which needs to be done in the future in order to clarify the \\LB phenomenon. ",
        "introduction": "This group of metal poor stars was introduced with the identification of an `abnormal' spectrum for \\LB (Figure \\ref{spectrum}) in the classification survey of Morgan et al. (1943). Meanwhile, more stars with the same peculiarities were discovered, in particular when the available spectral region was extended to the red and beyond the optical region towards the IR and to the UV.  This flood of additional information is helpful on the one hand in providing more physical evidence needed to understand the nature and evolution of this group of stars, on the other hand it resulted also in obscuring the group of \\LB stars by objects which definitely have not much in common with the prototype. Confusion peaked in the eighties, when the group of \\LB stars degenerated to a sort of trash can for stars which could not be classified otherwise. This development is described in reviews on \\LB stars by Gerbaldi \\& Faraggiana (1993) and Gray (1997).  This was also the reason for a general discussion about the definition of \\LB stars at this workshop. In the following I will give my personal point of view summarizing the observational facts and proposed theories.  \\begin{figure}[hbt] \\centerline{ \\psfig{figure=T29f3.ps,height=11.5cm}} \\caption{Intermediate resolution spectrum of \\LB and of a corresponding MK standard star.} \\label{spectrum} \\end{figure} ",
        "conclusions": ""
    },
    "9805/astro-ph9805098_arXiv.txt": {
        "abstract": "We discuss mechanisms for angular momentum transport in the clumpy medium of the circumnuclear disk at the Galactic center. The viscosity due to clump-clump collisions is found to be less than some critical viscosity; this meets the conditions at which a collective mode of nonaxisymmetric shear perturbations in the disk is able to grow until going into the saturation regime where fully developed turbulence is established. We find that the angular momentum transfer due to this turbulent viscosity turns out to comparable to the transport due to magnetic torques. Taken together, the turbulent and magnetic transfer of angular momentum are able to provide the inflow of mass into the central parsec with a rate of about $10^{-2}$\\moyr, consistent with the available data. ",
        "introduction": "Transport of angular momentum in the disks of spiral galaxies is one of the central issues in galactic dynamics. It is also of prime importance for fueling the central engines in active galactic nuclei (AGN). With regard to our own Galaxy this topic has been recently explored in a series of papers (von Linden et al. 1993a,b; Biermann et al. 1993) with the conclusion that the energy input from the SN explosions can feed the turbulence of the interstellar matter so as to provide  the effective viscosity high as required to feed the star formation.  An immediate implication is that the viscous transport of angular momentum outwards is accompanied by inflow of gas inward with the rate that might be as high as $10^{-2}$\\mo/yr on the scale of hundreds pc. Earlier, a similar mass inflow rate was inferred to exist on the scale of the circumnuclear disk, i.e. at 1.5 pc $\\lax R\\lax 10$ pc (Genzel \\& Townes 1987, Jackson et al. 1993).  We also note that the star formation in the inner region of the Galaxy requires such a mass supply.  However, appropriate physical mechanism(s) which would be able to provide  the necessary rate of momentum transfer within the central 10 pc or so is (are) obscure so far. The present paper aims at consideration of such mechanisms. In Sec. 2 and Appendix, we discuss viscosity in a clumpy disk due to clump interactions, both with and without self-gravity included. Sec. 3 deals with short-wave instability of the clumpy, viscous disk; the necessary condition for developing small-scale turbulence is established here. In Sec. 4, we apply this mechanism to the circum-nuclear ring at the Galactic center. The results of the paper are discussed in Sec. 5. ",
        "conclusions": "\\bigskip  A qualitative argument presented at the end of Sec. 2 shows that the viscosity in a cloudy disk is, in general, much larger than that for a continuous disk. This assertion could be easily confirmed by numerical estimates when we address the circum-nuclear ring (CNR): in a continuous disk with similar global parameters, one would have $\\nu\\simeq c_sl\\lax c_sh \\simeq 10^{23}~{\\rm cm}^2{\\rm s}^{-1}$, where $c_s\\simeq 1$ km/s is the sound speed of the gas (G\\\"usten et al. 1987) and $h\\simeq 0.5~{\\rm pc}$ is the thickness of the disk. What provides a much larger viscosity than it would be possible in a continuous disk is (although only partly) the clump-clump collisions. As is shown in Sec. 3, the collective mode of instability would dominate the dynamics of a clumpy disk if viscosity in the latter is less than some critical viscosity [Eq. (3.11)] whose value is given by Eqs. (3.15) or (3.16). This condition is (marginally) met in the CNR, which results in the turbulent viscosity coefficient $\\nu_{\\rm turb}\\simeq 1~ 10^{24}~{\\rm cm}^2{\\rm s}^{-1}$. We note in passing that G\\\"usten et al. (1987) suspected the existence of turbulence in the clumpy CNR, although they did not evaluate its viscosity coefficient.  It is worth mentioning that, in spite of the differences in the specific physical parameters of the clumps in the CNR and active galactic nuclei, or AGN (for the latter, see Netzer 1990), the value of $\\nu_{\\rm turb}$ evaluated above turns out to be in the range of the values estimated by Fridman \\& Ozernoy (1992) for cloudy disks in AGN. This might be relevant to the issue whether the CNR could be considered as a prototype for circum-nuclear tori around some types of AGN. (It is interesting that a large value of the viscosity in the CNR implies its rather large scale height, which is a required geometry for the AGN tori). In any case, the angular momentum transport in the CNR seems to be a template while considering similar issues both for quiescent and active galactic nuclei.  The analysis performed above has not accounted for the magnetic field in the CNR. Meanwhile several observational techniques revealed the field strength in the clouds to be $\\sim 1$ mG (for a recent review, see Genzel et al. 1994). A magnetic field as strong as this cannot be ignored in the transport of angular momentum. Conservation of angular momentum transported by both viscous and magnetic stresses can be written for the CNR in which $v_{\\varphi}\\approx$ const in the form (Ozernoy \\& Genzel 1998): $$\\dot M= {2\\pi r\\Sigma\\over v_{\\varphi}}\\left (\\xi v_A^2 + \\nu_{\\rm eff} \\Omega \\right )\\left[1-\\left({R_i\\over r}\\right )^{1/2}\\right ]^{-1}, \\eqno (5.1)$$ where $\\dot M$ is the mass inflow rate, $\\Sigma$ is the surface density of the CNR, $\\xi \\sim 1$ is the $(-B_r/B_{\\varphi})$ averaged over the $z$-coordinate, $v_A$ is the Alfv\\'en velocity, $\\nu_{\\rm eff}$ is an effective viscosity, and $R_i\\simeq 1.5$ pc is the inner radius of the ring. If we consider the CNR as a magnetized disk for which $v_A\\simeq h\\Omega \\simeq 30$ km/s  and $\\nu_{\\rm eff}$ is given by Eq. (4.4) the two terms in parentheses in the r.h.s. of Eq. (5.1) turn out to be comparable. This implies that while evaluating the angular momentum transport in the CNR, both magnetic and turbulent viscosity need to be accounted for. One can see that with the parameters listed above and $\\Sigma\\simeq 2~ 10^{-2}$ g cm$^{-2}$ the total mass inflow rate given by Eq. (5.1) at a fiducial distance of $r=2$ pc amounts to $\\dot M\\simeq 10^{-2}$\\moyr.  This result is consistent with a naive, by order-of-magnitude, estimate of $\\dot M\\sim M/t_{tr}$, where $M$ is the CNR mass and $t_{tr}=R_i^2/\\nu_{\\rm turb}$ is the characteristic time for the angular momentum transport: Taking $\\nu_{\\rm turb}\\simeq 1~ 10^{24}~{\\rm cm}^2{\\rm s}^{-1}$ and $M\\sim 3~ 10^4$\\mo, i.e. somewhere in between $10^4$ and $10^5$\\mo, the current estimates for the CNR mass (Genzel et al. 1995), one gets $t_{tr}\\simeq 6~ 10^5$ yr and  $\\dot M\\sim 5~ 10^{-2}$\\moyr.  This is consistent as well with the inflow rate toward the Galactic center inferred from the observational data (e.g. Blitz et al. 1993, Genzel et al. 1994).  \\blankline  \\noindent{\\bf Appendix A:}  \\noindent{\\bf Shear Viscosity in a Clumpy Disk at Elastic/Inelastic Encounters}  The shear viscosity in a differentially rotating clumpy disk, with accounting for gravitational interactions between the clumps of mass $m$, is given by Eq. (60) in Stewart \\& Kaula (1980): \\begin{eqnarray*} -r\\Omega^ \\prime\\nu &=&A_0\\sigma_v^2, \\\\ A_0 &=&{9\\sqrt 2~ \\Omega\\omega_G^{-1}\\over (36/5)+55\\left(\\Omega\\omega_G^{-1} \\right)^2}~, \\qquad ~~~~~~(A1) \\end{eqnarray*} where $\\Omega^\\prime \\equiv {\\rm d}\\Omega/{\\rm d}r$, and $\\omega_G$ is frequency of clump-clump collisions given by $$\\omega_G={3.35~ G^2m^2n\\over \\sigma_v^3}~. \\eqno (A2)$$ \\noindent Eq. (A2) corresponds to Eq. (57) of SK when one puts $\\ln [1+(\\sigma_v^3/ Gm\\Omega)^2]^{1/2}=1$ because in a flat disk the long-range gravitational interactions could be neglected compared to the short-range ones. The value of $\\omega_G$ is related to the characteristic time of gravitational encounters, $t_G$, given by Eq. (2.3) simply by $\\omega_G^{-1}=t_G/\\sqrt 2$. By substituting Eq. (A2) into Eq. (A1), the latter could be written in the form: $$\\nu={\\sigma_v^2\\over \\Omega}{A_G\\tau\\over \\left(1.95~\\tau\\right)^2+1}~, \\eqno (A3)$$ where $A_G=0.83$ for a Keplerian disk, $A_G=1.25$ for a solid-body rotating disk, and $\\tau=\\Omega t_G$.  The viscosity coefficient for clump-clump collisions, when the clump gravity is negligible [Goldreich \\& Tremaine 1978, Eq. (46)], can be written in the form analogous to Eq. (A3): $$\\nu={\\sigma_v^2\\over \\Omega}{0.46~\\tau\\over \\left(0.97~\\tau\\right)^2+ 1}~.  \\eqno (A4)$$  Eq. (2.1) unifies the representation of viscosity both in the case when interactions are inelastic, while gravitation is negligible [Eq. (A4)] {\\it and} in the case when gravitational encounters are dominating [Eq. (A3)].  \\blankline  \\noindent{\\bf Appendix B:}  \\noindent{\\bf Character of Perturbations in a Differentially Rotating vs. a Solid-body Rotating Disk}  Let us consider an incompressible, differentially rotating, liquid cylinder. The linearized equations of motion take the form: \\begin{eqnarray*} {\\rm div}\\tilde{ \\vec v} &= & 0, \\\\ {{\\partial \\tilde v_r}\\over{\\partial t}} ~+~\\Omega {{\\partial \\tilde v_r}\\over{\\partial \\varphi}}~-~2\\Omega \\tilde v_\\varphi & = & -{{1}\\over\\rho} {{\\partial \\tilde P}\\over{\\partial r}}, \\\\ {{\\partial \\tilde v_\\varphi}\\over{\\partial t}} ~+~\\Omega {{\\partial \\tilde v_\\varphi}\\over{\\partial \\varphi}}~+~{{\\kappa^2}\\over{2\\Omega}} \\tilde v_r & = & -{{1}\\over{\\rho r}} {{\\partial \\tilde P}\\over{\\partial \\varphi}}. \\end{eqnarray*}  Substituting a perturbation in the form of the wave packet $f$ $\\sim$ $\\exp[-i\\omega+m \\varphi+\\int k_r dr]$  results in the following dispersion relation ($k_\\varphi\\equiv m/r$): $$ \\hat \\omega ~\\equiv~ \\omega - m\\Omega ~=~ -i (2\\Omega - \\kappa^2/2\\Omega) {{k_r k_\\varphi}\\over{k^2}}. $$ As always $\\kappa^2$ $\\le$ $4\\Omega^2,$ and therefore the condition of instability is $k_r k_\\varphi$ $<$ 0, which corresponds to the growth of the leading spirals. Under this condition the instability is monotonous without any oscillations with epicyclic frequency. If the rotation is solid-body, one obtains  $\\hat \\omega$ = 0, i.e. the perturbations are stationary.  The above solution has a drawback since the $r$-component of the group velocity for perturbations under consideration is zero. A more detailed analysis should be performed to find out what is the role of $r$-dependence for $\\Omega$. This difficulty can be overcome if one considers the system in a corotating frame of reference. In this case, an exact time-dependent solution can be derived (see Fridman, 1989) which demonstrates a qualitatively the same behaviour as is obtained above. Specifically, in any differentially rotating disk whose angular velocity decreases with radius, the leading spirals grow monotonously. As for a solid-body rotating disk, all perturbations are stationary.  Note that if one would consider absolutely compressible case, i.e. $P=0$ in the above equations, the continuity equation takes the form: $$ {{\\partial \\tilde \\rho}\\over{\\partial t}}~+~\\rho {\\rm div}\\tilde{ \\vec v}~+~v_r{{d\\rho}\\over{dr}}~=~0, $$ and we immediately obtain $$ \\hat \\omega^2 ~=~ \\kappa^2. $$  The contrast to other work is illuminating.  Julian \\& Toomre (1966) considered perturbations in a very thin collisionless self-gravitating stellar disk. In fact, for such perturbations a collisionless {\\it stellar} disk is absolutely compressible, in contrast to incompressible character of perturbations with wavelengths much smaller than the disk's width for a {\\it gaseous} disk considered here. Goldreich \\& Lynden-Bell (1965) studied dynamics of perturbations in a gaseous disk, but only at high frequencies. Really, perturbations considered in their paper demonstrate (asymptoticaly) oscillations with sound frequency $kc$, where $k$ is the wave number and $c$ is the sound speed. As the scale of perturbations is smaller than the disk width, i.e. $kh>1$, their frequency is higher than the epicyclic one. But as the equations of motions contain three derivatives with respect to time, they describe three modes of perturbations. Besides the two sonic high frequency modes, one mode of low frequency perturbations exists as well. What is examined in our paper is the role of {\\it low frequency} mode in dynamics of the circumnuclear disk. As was shown in Fridman (1989), this mode does not demonstrate any oscillations and is stationary in a solid-body rotating disk.  \\blankline  \\noindent{\\bf Appendix C:}  \\noindent{\\bf Velocity Perturbations in a Plane Shear Layer}  We start from linearized equations for short-wave, low-frequency perturbations in a barotropic gaseous disk in the corotating frame of reference (for derivation, see Fridman 1989): \\begin{eqnarray*} k_r(t)\\tilde v_r~+~k_\\varphi \\tilde v_\\varphi & = & 0,\\\\ {{\\partial \\tilde v_r}\\over{\\partial t}} ~-~2\\Omega_0 \\tilde v_\\varphi & = & - i k_r(t)\\tilde \\chi - \\nu k_\\perp^2(t) \\tilde v_r, \\\\ {{\\partial \\tilde v_\\varphi}\\over{\\partial t}} ~+~(2\\Omega_0-A) \\tilde v_r & = & - i k_\\varphi\\tilde \\chi - \\nu k_\\perp^2(t) \\tilde v_\\varphi. \\end{eqnarray*} Here $ \\Omega_0$ = const is the angular velocity of the corotating frame of reference; $A$ $\\equiv$ $-r_0(d\\Omega/dr)_0$  characterizes the value of the shear; $k_r(t)$ $\\equiv$ $k_r$ $+$ $tAk_\\varphi$; $k_\\perp^2(t)$ $=$ $k_r^2(t)$ $+$ $k_\\varphi^2$.  This system has an exact solution in the form (Fridman, 1989): $$ \\tilde v_r(t)~=~{{k_\\perp^2(0)}\\over{k_\\perp^2(t)}} \\tilde v_r(0) \\exp\\left[-\\nu\\int\\limits_0^t k_\\perp^2(t)dt\\right].  \\eqno (C1) $$ For small $t$ the exponential term is negligent and we come up with vorticity conservation: $$ {\\rm curl}_z(\\vec v)~=~i(k_rv_\\varphi-k_\\varphi v_r)~=~-{{ i}\\over {k_\\varphi}} k_\\perp^2 v_r~=~{\\rm const}.  \\eqno (C2) $$  For a plane shear layer, a similar set of equations can be derived. The simpliest way to do this would be to put in the above system $\\Omega$ $\\rightarrow$ $0$, $r$ $\\rightarrow$ $\\infty$, $\\Omega r$ $\\rightarrow$ $v(x)$. As a result, we obtain the following system: \\begin{eqnarray*} k_x(t)\\tilde v_x~+~k_y \\tilde v_y & = & 0,\\\\ {{\\partial \\tilde v_x}\\over{\\partial t}} & = & - i k_x(t)\\tilde \\chi - \\nu k_\\perp^2(t) \\tilde v_x, \\\\ {{\\partial \\tilde v_y}\\over{\\partial t}} ~+~B \\tilde v_x & = & - i k_y\\tilde \\chi - \\nu k_\\perp^2(t) \\tilde v_y. \\end{eqnarray*} Here $B$ $\\equiv$ $(dv/dx)_0$ characterizes the value of the shear; $k_x(t)$ $\\equiv$ $k_x$ $-$ $tBk_y$; $k_\\perp^2(t)$ $=$ $k_x^2(t)$ $+$ $k_y^2$. One can easily see that this system is almost similar to the previous one and it has a similar solution: $$ \\tilde v_x(t)~=~{{k_\\perp^2(0)}\\over{k_\\perp^2(t)}} \\tilde v_x(0) \\exp\\left[-\\nu\\int\\limits_0^t k_\\perp^2(t)dt\\right].  \\eqno (C3) $$ Obviously, in a plane layer there are no any Coriolis forces at all. Nevertheless, the solution (C3) has the same form as (C1). What is relevant to the real physical meaning of our solution is conservation of vorticity (eq. C2), which has been emphasized here.  \\def\\apj    {{ApJ{\\rm,}\\ }} \\def\\apjl   {{ApJ~(Letters){\\rm,}\\ }} \\def\\apjs   {{ApJ~Suppl.{\\rm,}\\ }} \\def\\aa     {{A\\&A{\\rm,}\\ }} \\def\\anrev  {{Ann.~Rev.~Astr.~Ap.{\\rm,}\\ }} \\def\\baas   {{Bull.~AAS{\\rm,}\\ }} \\def\\mnras  {{M.N.R.A.S.{\\rm,}\\ }} \\def\\ref#1  {\\noindent \\hangindent=24.0pt \\hangafter=1 {#1} \\par} \\def\\v#1  {{{#1}{\\rm,}\\ }}"
    },
    "9805/astro-ph9805273_arXiv.txt": {
        "abstract": "It is shown that the matter concentration observed through stellar motion at the galactic center (Eckart \\& Genzel, 1997, MNRAS, 284, 576 and Genzel et al., 1996, ApJ, 472, 153) is consistent with a supermassive object of $2.5 \\times 10^6$  solar masses composed of self-gravitating, degenerate heavy neutrinos, as an alternative to the black hole interpretation. According to the observational data,  the lower bounds on possible neutrino masses are $m_\\nu \\geq 12.0$ keV$/c^2$ for $g=2$ or $m_\\nu \\geq 14.3$ keV$/c^2$ for $g=1$, where $g$ is the spin degeneracy factor. The advantage of this scenario is that it could naturally explain the low X-ray and gamma ray activity of Sgr A$^*$, i.e. the so called \"blackness problem\" of the galactic center. ",
        "introduction": "The idea that some of the galactic nuclei are powered by matter accretion onto supermassive black holes is based on strong theoretical arguments (Salpeter 1964; Zel'dovich 1964; Lynden-Bell 1969, 1978; Lynden-Bell \\& Rees 1971; see Blandford \\& Rees 1992 for reviews) and observation of rapid time variability of the emitted radiation which implies relativistic compactness of the radiating object. However, so far, there is no compelling proof that supermassive black holes actually do exist, as the spatial resolution of current observations is larger than $10^5$ Schwarzschild radii. The standard routine in the investigation of the nature of the dark mass distribution at the centers of active galaxies is to observe stellar and gas dynamics. However, gas dynamics is usually regarded as less conclusive since it is responsive to non-gravitational forces such as e.g. magnetic fields.  As an alternative to the black hole scenario, Moffat (1997) considered general relativistic models of stellar clusters with large redshifts, and he investigated whether such objects are long-lived enough from the point of view of evaporation and collision timescales and stability criteria. He then showed that, in certain cases, stellar clusters with masses $\\geq 10^6$ $M_\\odot$ could mimic the behavior of supermassive black holes. A good unprejudiced review on the subject can be found in the paper by Kormendy \\& Richstone 1995.  The identification of a central supermassive object in the Milky Way has been a source of continuous debate in the literature. The crucial issue is whether the center of our galaxy harbors a supermassive black hole or any other compact dark matter object. Theoretical papers, which are exclusively devoted to the black hole explanation of the galactic center, include Lynden-Bell \\& Rees 1971; Rees 1987; Phinney 1989 and de Zeeuw 1993. However, even within this theoretical framework, there is no full agreement: the general consensus is that the supermassive black hole should have mass $\\sim 10^6$ $M_\\odot$, while Ozernoy 1992 and Mastichiadis \\& Ozernoy 1994 argue that the black hole mass can be as low as $\\sim 10^3$ $M_\\odot$. The main motivation for a $\\sim 10^3$ $M_\\odot$ black hole at the galactic center  is that it would emit less X-rays and gamma-rays than a $\\sim 10^6$ $M_\\odot$ black hole, behaving basically like a scaled-down active galactic nucleus. Although earlier observations (Gehrels \\& Tueller 1993; Watson et al. 1981; Skinner et al. 1987; Hertz \\& Grindlay 1984; Pavlinsky, Grebnev \\& Sunyaev 1994) have shown that the central region actually does emit X-rays and gamma-rays, the supposed true center, usually assumed to be Sgr A$^*$, does not emit strongly, at least up to energies of 30 keV (Skinner et al. 1987; Hertz \\& Grindlay 1984; Pavlinsky, Grebnev \\& Sunyaev 1994). Sgr A$^*$ radiation emission data at higher energies have been presented by Goldwurm et al. 1994. They find no source associated with Sgr A$^*$, and the inferred upper bound  implies that the hard X-ray luminosity of Sgr A$^*$ is a factor of $4 \\times 10^7$  less than that expected for a black hole of $\\sim 10^6$ $M_\\odot$ accreting matter at the maximum stable rate. An unavoidable source of accretion is the wind from IRS 16, a nearby group of hot, massive stars. Since the density and velocity of the accreting matter are known from observations, the accretion rate is basically a function of the assumed black hole mass only. This value represents a reliable lower limit to a real rate, given the other possible sources of accreting matter. Based on this and on the theories about shock acceleration in active galactic nuclei, Mastichiadis \\& Ozernoy 1994  have estimated the expected production of relativistic particles and their hard radiation. Comparing their results with available X-ray and gamma-ray observations which show that Sgr A$^*$ has a relatively low activity level, the authors conclude tentatively that an assumed black hole in the galactic center cannot have a mass greater than approximately $6 \\times 10^3$ $M_\\odot$.  Other scenarios to explain low X-ray and gamma-ray activity of Sgr A$^*$ include so-called advection dominated models (Narayan, Yi \\& Mahadevan 1995; Narayan et al., 1997; Mahadevan, Narayan \\& Krolik, 1997) which can live with  a $\\sim 10^6$ $M_\\odot$ black hole that accretes matter at a realistic accretion rate of $\\dot{M} \\sim 10^{-5}$ $M_\\odot$ yr$^{-1}$. However, most of the energy released by viscosity is carried along with the gas and lost into the black hole, while only a small fraction is actually radiated.  The purpose of this paper is to present an alternative model, based on the idea that the dark matter concentration in the galactic center could be a ball of degenerate, self-gravitating heavy neutrinos, which is consistent with present observational data. ",
        "conclusions": "We have shown that a neutrino ball of total mass $2.5 \\times 10^6 M_\\odot$ which is composed of self-gravitating, degenerate neutrinos and antineutrinos of mass  $m_\\nu \\geq 12.0$ keV$/c^2$ for $g=2$ or $m_\\nu \\geq 14.3$ keV$/c^2$ for $g=1$,  surrounded by a stellar cluster with a central density of $\\rho (R=0)=4 \\times 10^6 M_\\odot /{\\rm pc}^3$ and a core radius of $R_{\\rm core}=0.38$ pc, is consistent with the currently available observational data. A neutrino ball with the abovementioned physical parameters would be virtually indistinguishable from a black hole with the same mass,  as far as the current observational data is concerned.  Many models were  put forward to explain the low X-ray and gamma ray emission of the Sgr A$^*$. Another possible solution to this \"blackness problem\" could be the presence of a neutrino ball which is  consistent with current observational data, instead of the supermassive black hole. In fact, in the neutrino ball scenario, the accreting matter would experience a much shallower gravitational potential, and therefore less viscous torque would be exerted. The radius of a neutrino ball of total mass $2.5 \\times 10^6 M_\\odot$, which is composed of self-gravitating, degenerate neutrinos and antineutrinos of mass  $m_\\nu = 12.0$ keV$/c^2$ for $g=2$ or $m_\\nu = 14.3$ keV$/c^2$ for $g=1$,  is $1.06 \\times 10^5$ larger than the Schwarzschild radius of a black hole of the same mass. In this context, it is important to note that the accretion radius $R_{\\rm A}=2GM/v^2_{\\rm w}$ for the neutrino ball, where $v_{\\rm w}\\simeq 700$ km/sec is the velocity of the wind from the IRS 16 stars, is approximately 0.02pc (Coker \\& Melia, 1997), which is slightly less than the radius of the neutrino ball, i.e. 0.02545 pc (for $m_\\nu = 12.0$ keV$/c^2$ for $g=2$ or $m_\\nu = 14.3$ keV$/c^2$ for $g=1$).  The accretion radius is the characteristic distance from the center within which the matter is actually gravitationally captured. Therefore, in the neutrino ball scenario, the captured accreting matter will always experience a gravitational pull from a mass less than the total mass of the ball. We do not discuss  this issue  any further, since the direct comparison of the emitted X-ray spectra with the black hole or with neutrino ball instead would require to go into details of current models of X-ray emission from a compact object. The ultimate goal of this paper was to demonstrate that our model of the mass distribution at at the galactic center is consistent with the current observational data.  It is worthwhile to note that a possible way to distinguish between the supermassive black hole and neutrino ball scenarios is to track a single star, which is moving on a bound orbit inside the radius of the neutrino ball, over a significant part of the orbiting period. The star trajectory, in general, would be an open path between the classical turning points $r_{\\rm min}$ and $r_{\\rm max}$. The trajectory would be closed only in the case of $1/r$ (black hole) and $r^2$ (uniform density distribution) potentials. In the case of the black hole, the star would orbit on an ellipse, with the black hole located at the focus, whereas in the case of a uniform density distribution, the center of the ellipse would coincide with center of ball. In the neutrino ball scenario, the trajectory of a star would be somewhat intermediate  between the black hole and uniform density orbits. The period of a star on an elliptical orbit around the black hole is  $T=2 \\pi  \\sqrt{a^3/G M}$, where $a$ is the semi-major axis of the ellipse. Putting $2a\\approx 2.545 \\times 10^{-2}$ pc, with $2a$ being the radius of the neutrino ball for $m_\\nu = 12.0$ keV$/c^2$ for $g=2$ or $m_\\nu = 14.3$ keV$/c^2$ for $g=1$,  and $M=2.5 \\times 10^6 M_\\odot$ we thus obtain $T=85.1$ yr. The difference between the black hole and neutrino ball scenarios is that in the case of a neutrino ball the period will remain roughly constant for any orbit within the neutrino ball, as it is well represented by an extended object with uniform density distribution with average density about $1/6$ of the actual central density of the neutrino ball (Viollier, 1994), while in the black hole scenario $T$ would scale as $T \\propto a^{3/2}$. In summary, future stellar proper motion studies on an appreciable fraction of this time-scale may by practical in discriminating between the two scenarios.   Another possible characteristic signature of a neutrino ball at the galactic center would be the X-ray emission line at the energy $\\sim m_\\nu c^2/2$ which has a width of about the Fermi energy ($\\varepsilon_F=p^2_F/2 m_\\nu= (6 \\pi^2/g)^{2/3} (\\hbar^2/2 m_\\nu) n_\\nu^{2/3}  $). This X-ray emission, a direct consequence of the standard electroweak interaction theory, is due to the decay of the heavy neutrino into a photon and massless neutrino species, both with energies $\\sim m_\\nu c^2/2$ (Viollier, 1994). For Dirac neutrinos, this would generate a luminosity of $$ L_\\gamma=2.27 \\times 10^{31}\\left({{m_\\nu c^2}\\over{17.2 {\\rm keV}}}\\right)^5 |U_{\\tau \\nu_\\tau} U^*_{\\tau \\nu_i} |^2 {{M_\\nu}\\over{M_\\odot}} \\,\\,\\, {\\rm erg/sec}, $$ where $U_{\\tau \\nu_i}$ denotes the Cabibbo-Kobayashi-Maskawa matrix element and $M_\\nu$ is the mass of neutrino ball. Thus, putting $M_\\nu=2.5 \\times 10^6 M_\\odot$  and the experimental upper limit $|U_{\\tau \\nu_\\tau} U^*_{\\tau \\nu_i} |^2 \\leq 10^{-3}$, we obtain $L_\\gamma \\leq 1.45 \\times 10^{34}$ erg/sec.  The galactic center has been observed in the 2-10 keV range by Koyama et al. (1996). They find that the X-ray flux from inside the Sgr A shell (an oval region of $\\simeq 2^\\prime \\times 3^\\prime$) is approximately $10^{-10}$ erg cm$^{-2}$ sec$^{-1}$ in the 2-10 keV band. After correcting for the observed absorption by a column of approximately $7\\times10^{22}$ H-atom cm$^{-2}$, they obtain a luminosity of $\\simeq 10^{36}$ erg/sec for an assumed distance of 8.5 kpc to the galactic center. To detect  X-rays emitted by the neutrino ball a much higher angular resolution is needed. It would suffice to make observations of $0.6^{\\prime \\prime} \\times 0.6^{\\prime \\prime}$ (about the size of the neutrino ball) region around Sgr A$^*$. The diffuse luminosity expected from an area corresponding to the area of the neutrino ball would be $(0.6^{\\prime \\prime} \\times 0.6^{\\prime \\prime})/(2^\\prime \\times 3^\\prime) \\times 10^{36}$ erg/sec $ \\approx 1.67 \\times 10^{31}$ erg/sec. This number could be even lower since it includes contributions from all energies from 2 to 10 keV. Thus, it seems possible to detect X-ray line of $E_\\gamma \\geq 6.0$ keV ($g=2$) or $E_\\gamma \\geq 7.1$ keV ($g=1$) due to the radiative decay of the neutrino in the neutrino ball. However, it might be that the energy of the emitted X-rays is too close to the fluorescent iron lines to be detected with the current energy resolution of CCD cameras and gas-imaging spectrometers.  We would like to emphasize that the idea that Sgr A$^*$ may be an extended object rather than a supermassive black hole is not new (see e.g. Haller et al., 1996; Sanders, 1992). To our knowledge all previous such  models  assume that the extended object is of a baryonic nature, e.g.  a very compact stellar cluster. However, it is commonly accepted that these models face problems with stability and it has been questioned whether such clusters are long-lived enough, based on evaporation and collision time-scales stability criteria (for different point of view see Moffat, 1997). It is interesting to note that in the context of a different object, the center of the NGC 4258 galaxy, based on the similar criteria, Maoz (1995) has shown (the similar study has been performed later also for our Galaxy and other galaxies (Maoz, 1998)) that an object composed of elementary particles would not contradict  with the observational data, which agrees with our conclusions.   Finally, we would like to make a comment on the neutrino mass  necessary for our model to work. We are particularly interested in neutrinos with masses between 10 and 25 keV$/c^2$, as these could form supermassive, degenerate neutrino balls which may explain, without invoking the black hole hypothesis, some of the features observed around the supermassive compact dark objects of masses ranging from $10^{6.5} M_\\odot$ to $10^{9.5} M_\\odot$, which have been reported to exist at the centers of a number of galaxies (Kormendy \\& Richstone, 1995) including our own (Genzel, Hollenbach \\& Townes, 1994; Eckart \\& Genzel, 1996; Morris, 1996; Tsiklauri \\& Viollier, 1998a, 1998b) and quasi-stellar objects. A 10 to 25 keV$/c^2$  neutrino is neither in conflict with particle and nuclear physics nor with astrophysical observations (Viollier, 1994). On contrary, if the conclusion of the LSND collaboration which claims to have detected ${\\bar \\nu}_\\mu \\to {\\bar \\nu}_e$ flavor oscillations (Athanassopoulos et al., 1996)  is confirmed, and the quadratic see-saw mechanism involving the up, charm and top quarks (Gell-Mann et al., 1979; Yanagida, 1979), is the correct mechanism for neutrino mass generation, the $\\nu_\\tau$ mass may very well be within the cosmologically forbidden range between 6 and 32 keV$/c^2$ (Bili\\'c \\& Viollier 1997d). It is well known that such a quasi-stable neutrino would lead to an early matter dominated phase, which may have started as early as a few weeks after the Big-Bang. As a direct consequence of this, the Universe would have reached the current microwave background temperature  much too early to accommodate the oldest stars in globular clusters, cosmochronology and the Hubble expansion age. It is conceivable, however, that, in the presence of such heavy neutrinos, the early Universe might have evolved quite differently than described in the Standard Model of Cosmology (Kolb \\& Turner, 1990, 1991; B\\\"orner, 1988). Neutrino balls might have been formed in local condensation process during a gravitational phase transition, shortly after the neutrino matter dominated epoch began. The latent heat produced in such a first-order phase transition, apart from reheating the gaseous phase, might have reheated the radiation background as well. Annihilation of the heavy neutrinos into light neutrinos via the $Z^0$ boson will occur efficiently in the interior of the neutrino balls, as the annihilation rate is proportional to the square of the number density, which is order of $10^{25}$ particles per cm$^3$ at the center of a few $10^9 M_\\odot$ neutrino ball. Both these processes will decrease the contribution of the heavy neutrinos to the critical density today and therefore, increase the age of the Universe (Kolb \\& Turner, 1991). Thus, a quasi-stable neutrino in the mass range between 10 and 25 keV$/c^2$ is not excluded by astrophysical arguments (Viollier, 1994). In fact, it has been established in the framework of the Thomas-Fermi model at finite temperature (Bili\\'c \\& Viollier 1997a) that such neutrino balls can form via a first-order gravitational phase transition, although, the mechanism through which the latent heat is released during the phase transition and dissipated into observable or perhaps unobservable matter or radiation remains to be identified. At this stage, however, it is still not clear whether an energy dissipation mechanism can be found within the minimal extension of the Standard Model of Particle Physics or whether new physics is required in the right-handed neutrino sector, in order to generate this efficient cooling of the neutrino matter."
    },
    "9805/astro-ph9805045_arXiv.txt": {
        "abstract": "Between 1994 July and 1997 February we monitored the optical spectrum of RX J0019.8+2156. This supersoft X-ray source is one of only two accreting white dwarfs in the Galaxy that are thought to be burning hydrogen on their surface as a consequence of a high rate of mass transfer from a binary companion. Accurate orbital ephemerides are derived from radial velocity measurements for the white dwarf, which are obtained from strong He\\,II emission lines with a stable velocity semi-amplitude of $K = 71.2 \\pm 3.6$ km \\ps. We report the discovery of transient, low-velocity, bipolar jets. These jets are respresented by redshifted-blueshifted pairs of emission lines from H and He\\,II with an outflow velocity of $v$ cos($i$) $\\sim$ 815 km \\ps, where $i$ is the binary inclination angle. When present, the jet lines seen in $H\\alpha$ also exhibit an orbital modulation of 71 km \\ps\\, which strengthens the interpretation that this is the orbital velocity of the white dwarf and also indicates that the jets are oriented nearly perpendicular to the orbital plane. On most occasions, the H emission line profiles are further altered by P Cygni absorption effects, and the strength of this absorption is also dependent on binary phase.  We show that the jets and the P Cygni features have very different temporal characteristics and binary phase dependence; thus, we conclude that the outflowing material and the absorbing wind must have essentially different geometries. Finally, the measured mass function is combined with binary evolution models to suggest a limit on the inclination angle, $i < 40^\\circ$.  A particular model invoked to explain a high rate of mass transfer requires $16^\\circ < i < 25^\\circ$. However, at such small inclination it is difficult to explain the large amplitude of the orbital light curve ($\\sim 0.5$ mag). Alternatively, the results could signify substantial vertical structure in the accretions disks of supersoft X-ray sources. By contrast, a simple model fit to the jet outflow lines indicates an orbital inclination angle of $35^\\circ < i < 60^\\circ$. ",
        "introduction": "Supersoft X-ray sources exhibit extremely steep spectra in the range of 0.1 - 2 keV. They can be divided into three general subclasses: low and high luminosity galactic objects, and AGN. In this paper we concern ourselves with luminous galactic supersoft X-ray sources (hereafter SXS), which have characteristic luminosities of $\\sim$\\tento{36} - \\tento{38} erg \\ps\\ and effective temperatures of $\\sim$2-\\scinot{6}{5} K ($kT \\simeq 17-50$ eV).  The first luminous supersoft sources were discovered during a survey of the Large Magellanic Cloud with the Einstein satellite (\\cite{cal83:disc}).  The ROSAT X-ray survey (\\cite{rass:summary}) has now established that the SXS constitute a distinct astronomical class.  Greiner (1996) \\nocite{sxscat:greiner} has cataloged and summarized the properties of 34 known SXS.  Only seven of these objects are in the Galaxy, while seven are in the LMC, four are in the SMC, and sixteen have been located in M31.  A particular model for at least some of the long-lived SXS invokes steady or cyclical nuclear burning of matter on the surface of a $\\sim$1 \\msun\\ white dwarf that is accreting mass from a main-sequence or subgiant companion star of $\\sim 1 - 2$ \\msun\\ at a rate of \\scinot{3}{-8} $-$ \\tento{-6} \\mspyr\\ (\\cite{nbwd:hbnr92}, hereafter HBNR; \\cite{sxs:popsynth:rds94}, hereafter RDS; and references therein).  Such high transfer rates are a natural consequence of thermal-timescale mass transfer via Roche lobe overflow, which characteristically occurs in this type of system.  The transfer is potentially unstable because the donor star is more massive than the accreting white dwarf; however, the rate of transfer is limited by the thermal time scale of the donor star.  Models for the evolution of these systems (RDS) indicate that the orbital periods should lie in the range of 8 hr to 1.4 d.  Moreover, these studies predict that there should be some $10^3$ such systems in the Galaxy and in M31.  In spite of the greater distance, it is apparently easier to detect these supersoft sources (at least in the soft X-ray band) in nearby external galaxies at relatively high $b^{II}$, such as the LMC ($-33^\\circ$) and M31 ($-21^\\circ$).  In the plane of our own Galaxy, a hydrogen column density of only ${\\sim}10^{21}$ cm$^{-2}$ will absorb most of the very soft X-radiation.  In the Galactic plane this corresponds to distances of less than a kpc.  Five of the seven SXS in the Milky Way are {\\it un}likely to be the type of nuclear burning white dwarf (NBWD) discussed above.  Two are recently exploded novae, two are symbiotic novae, and one is unclassified. Thus, there are only two good NBWD candidates in the Galaxy: RX~J0925.7-4758 (\\cite{0925:disc}) and RX~J0019.8+2156 (hereafter RXJ0019; \\cite{0019:disc}).  Since RX~J0925.7-4758 is obscured by the Vela Sheet molecular cloud, and its intrinsic optical spectrum is heavily reddened (\\cite{0925:disc}; \\cite{0925:absorber}), RXJ0019 remains as the one prime galactic NBWD candidate for detailed optical study.  Beuermann \\etal\\ (1995) \\nocite{0019:disc} discovered RXJ0019 as part of the ROSAT all-sky survey in 1990 December. They subsequently initiated a battery of short ROSAT PSPC, ROSAT HRI, IUE, and optical observations in 1992-93 (\\cite{0019:disc}). The following is a brief summary of their results. The optical counterpart to RXJ0019 is a moderately bright object with $V=12.2$. The light curves of the X-ray, optical, and UV data show modulations with a period of 15.85 hr, interpreted as the binary period, while smaller, aperiodic fluctuations are seen on time scales of hours. They adopt a distance to RXJ0019 of 2 kpc on the basis of the calcium absorption features and the X-ray column density ($N_H \\sim$ \\scinot{4.4}{20} cm$^{-2}$). The X-ray spectrum is extremely soft and luminous. For an assumed LTE model atmosphere (\\cite{nbwd:lte-model}), the ROSAT spectrum yields $kT = 34$ eV and an estimated bolometric luminosity, $L_{bol} \\simeq$ \\scinot{4}{36} erg \\ps\\, with significant uncertainty caused by the limited spectral constraints.  The optical spectrum shows strong hydrogen lines, exceptionally strong He\\,II emission, and P Cygni absorption associated with the Balmer lines. Beuermann \\etal\\ also report that the radial velocity of the He\\,II $\\lambda$4686{\\AA} line is modulated at the orbital period with a semi-amplitude of $67 \\pm 3$ km \\ps.  Greiner \\& Wenzel (1995) \\nocite{0019:hps} examined archival photographic plates dating between 1890 and 1990 from the Harvard plate stacks and the Sonnenberg Observatory. They detected the orbital period, low amplitude variations on time scales as short as 2 hr, and $\\sim$1 mag variations over intervals of $\\sim$20 years.  Will \\& Barwig (1996) \\nocite{0019:photometry} studied RXJ0019 with UBVRI photometry from 1992 to 1995 using a multichannel photometer on the Wendelstein Observatory 80-cm telescope.  RXJ0019 shows both primary and secondary minima; the primary minimum is a $\\sim 0.5^{\\rm m}$ eclipse which covers nearly 40\\% of the orbital phase (FWHM). Will \\& Barwig's analysis of mid-eclipse times provides an accurate orbital ephemeris with a period of $15.85133 \\pm 0.00017$ hr.  G\\\"{a}nsicke, Beuermann, \\& de\\,Martino (1996) \\nocite{0019:uvspec} observed RXJ0019 with the IUE telescope using both the short-wavelength prime (SWP; 1250-1900 \\AA) and long wavelength prime (LWP; 2000-3200 \\AA) spectrographs.  The 15.85 hr periodicity, and both primary and secondary dips are present in the UV light curves; primary minima were deeper in the LWP spectra (32\\% modulation) than in the SWP spectra (22\\%).  The He\\,II $1640$\\AA\\ line is present and shows a stronger modulation than the continuum (43\\%).  Fits to the UV spectrum using a simple model for an irradiated donor and accretion disk (\\cite{0513:irrad}) yield parameters consistent with masses of $\\sim$1.0 and $\\sim$1.5 \\msun\\ for the white dwarf and donor star, respectively (\\cite{0019:uvspec}).  In this work we present our spectroscopic observations of RXJ0019. The radial velocity work of Beuermann \\etal\\ (1995), \\nocite{0019:disc} was completed in four days. In contrast, our observations span $\\sim$ 2.5 yr and allow us to probe the spectral features and the stability of radial velocity measurements over a much longer baseline.  In \\S 2, we present our observations and the data reduction.  In \\S 3 we present the results from the spectral analysis and the radial velocity study, including the P Cygni profiles and transient high-velocity emission lines, which we interpret as emission from bipolar jets. Similar high-velocity lines have been seen in recent infrared spectra reported by Quaintrell \\& Fender (1998).  In \\S 4 we discuss our attempts to model the complete binary system using the constraints of these and previously published data on RXJ0019. ",
        "conclusions": "\\begin{enumerate}  \\item  We have made spectroscopic observations of RX J0019.8+2156 over a 2.5 year baseline, collecting over 200 ks of data on this source.  \\item Radial velocity studies of the He\\,II $\\lambda$4686\\AA\\ emission line have yielded a Doppler curve with a velocity half-amplitude of $71.2 \\pm\\ 3.6$ km \\ps (see Table~\\ref{tab:ephtable}). The binary mass function is $0.0247 \\pm 0.0040$ \\msun.  \\item The period as determined from our radial velocity study is in agreement with the photometrically determined value obtained from the literature.  However, the phases determined by the two methods disagree by 28$^\\circ$ ($6 \\sigma$ significance), assuming that the binary orbit is circular and the accretion disk is circularly symmetric.  \\item The summed rest-frame spectrum shows extraordinary He\\,II series emission.  Most notably, the $n{\\rightarrow}5$ transition series is visible for values of $n$ up to 22.  \\item A model-dependent population synthesis study indicates that if the binary parameters are to be consistent with the measured mass function, then $16^\\circ \\la i \\la 40^\\circ$.  Furthermore, if the system also has a mass ratio which is capable of driving high mass transfer rates (in the context of the van den Heuvel \\etal\\ model), then the orbital inclination angle is $16^\\circ \\la i \\la 30^\\circ$. However, such a low inclination is difficult to reconcile with published photometry, which shows a pronounced primary and secondary dip.  \\item We discovered the existence of low-velocity bipolar jets that are revealed as red-shifted and blue-shifted emission lines that flank H$\\alpha$, H$\\beta$, and He\\,II $\\lambda$4686.  These lines are transient and were detected only during the 1994 Dec, 1996 May, and 1996 Dec. After correcting for the gamma velocity of the system, these lines are nearly symmetrical around the central line with a velocity of $\\pm$ 815 km \\ps.  \\item The outflow lines also exhibit orbital Doppler motion with the same amplitude and phase as that determined by the He\\,II lines.  This implies that the jets are oriented nearly perpendicular to the orbital plane.  \\item The profiles of the outflow lines potentially possess a great deal of information. While the limited spectral resolution limits our analysis, we are nevertheless able to conclude that the range of inclination angles is likely to be $ 35^\\circ \\la i \\la 60^\\circ$, and the corresponding range of opening half-angles $40^\\circ \\ga \\alpha \\ga 15^\\circ$, if the intrinsic width in the line profiles is dominated by the opening geometry of the jet.  \\item In order for both red and blue outflow lines to be observable at all orbital phases and also exhibit the Doppler curve of the white dwarf, the emission must originate from a distance $H$ from the disk in the range $1 \\la H / a_x \\la 6$, where $a_x$ is the semimajor axis of the orbit of the compact object.  \\item We discovered that the strength of the P Cygni profiles is variable with orbital phase. The $\\sim100^\\circ$ opening half-angle of the absorbing gas and its asymmetry with respect to the line joining the stars is distinctly different from the geometry of the gas emanating from the jets.  For this reason, and because of the transient nature of the outflow lines, it is unlikely that the absorption and jet components have the same origin.  \\end{enumerate}  Acknowledgements: We thank John Thorstensen, Jules Halpern, Mike Eracleous, and Bob Barr for helping with the observations at MDM Observatory. Jim Peters and Perry Berlind made observations for us at FLWO, and Susan Tokarz helped with the reduction of FLWO spectra. Travel support to RR and CB for these optical observations was provided by NSF grant AST-9315074. This work was also supported, in part, by NASA grant NAG5-3011.  \\clearpage"
    },
    "9805/astro-ph9805335_arXiv.txt": {
        "abstract": "We investigate the effects on extrasolar giant planets [EGPs] of intense irradiation by their parent stars, describing the issues involved in treating the model atmosphere problem correctly. We treat the radiative transfer in detail, allowing the flux from the parent star to interact with all relevant depths of the planetary atmosphere, with no need for a pre-assumed albedo. We present a low-resolution optical and near-IR spectrum of a close-in EGP, focusing on the differences from an isolated planet.  In our dust-free planetary atmospheres we find that Rayleigh scattering increases the EGP's flux by orders of magnitude shortward of the \\ion{Ca}{2}~H\\&K doublet (3930 ${\\rm \\AA}$), and the spectral features of the parent star are exactly reflected. In the optical and near-IR the thermal absorption of the planet takes over, but the absorption features are changed by the irradiation. The inclusion of dust increases the reflected flux in the blue; the stellar spectral lines can be seen blueward of H$\\beta$ (4860 ${\\rm \\AA}$). ",
        "introduction": "Within the past 2 years, detection of planets orbiting sun-like stars has exploded. To date 9 have been detected, 4 of which orbit amazingly close to their parent stars, less than 0.1 AU, which is 4 times closer than Mercury is to the sun. These include 51 Peg b (Mayor \\& Queloz 1995), 55 Cnc b, $\\tau$ Boo b, and $\\upsilon$ And b (Butler et al. 1997). A more distant, yet still close-in EGP, $\\rho$ CrB b at 0.23 AU was detected by Noyes et al. (1997). Because these planets are being bombarded by intense radiation from their parent stars, their spectra and temperature structures are radically altered compared to planets at Jupiter-like distances from their parent stars.  There is a strong need to theoretically model the close-in EGPs' atmospheres because ironically they may be the first of the extrasolar planets to be detected directly. The flux emitted from EGPs represents only a small fraction of the flux of the parent star at any wavelength. However, due to dust formed high in their atmospheres, the close-in EGPs' reflected light in the optical may add significantly to the observed flux, which may be enough to make the EGPs directly detectable. For example, Noyes (1998) has developed a spectral separation technique that uses the Doppler shift of the spectrum of the primary and takes advantage of the large orbital velocity of the close-in EGPs. With reflected light, close-in EGPs may be bright enough for spectral separation. In addition, models of close-in EGPs are needed for evolutionary and interior calculations, which are critically sensitive to the detailed treatment of the planetary atmosphere.  Recent work on spectral modeling, colors, interior modeling, and evolution of EGPs and brown dwarfs (e.g. Burrows et al. 1997b; Allard et al. 1997) has been extremely successful. So far there have been only limited attempts at stellar irradiation modeling. Due to the intense radiation field from the nearby star, the atmospheres of close-in EGPs must be treated by detailed radiative transfer solutions, which is a new problem for planetary physics. In particular, planets have outer convection zones, while close-in EGPs develop outer radiative zones due to the strong external heating, which inhibits convection (Guillot et al. 1996). The first models for irradiated EGPs were calculated by Saumon et al. (1996) and for close-in EGPs by Guillot et al. (1996). They assumed for simplicity that the EGPs reflect the light of the parent star like a grey body, that the thermal emission of the EGP is that of a blackbody, and that a fraction (with a Bond albedo = 0.35) of the flux from the parent star is absorbed by the planet. Their calculations also reveal the large sensitivity of the EGPs to the amount of irradiation.  In this paper we present model atmosphere calculations of the effects of strong irradiation on an EGP with particular emphasis on the issues involved in treating irradiation accurately. As an illustration we compute preliminary models for the EGP orbiting $\\tau$ Boo. ",
        "conclusions": "This first paper on spectrum formation in close-in EGPs shows that strong irradiation results in a radically altered temperature structure and change in absorption features as compared to an isolated EGP with the same $T_{\\rm eff}$. Because spectral features, including reflection from dust, are highly sensitive to irradiation, model atmospheres of close-in EGPs must treat the incident radiation accurately with no need for a pre-assumed albedo. This is especially relevant for detection techniques (e.g. Noyes 1998) which use a narrow range of frequency and detailed information about high resolution line shapes."
    },
    "9805/astro-ph9805103_arXiv.txt": {
        "abstract": "Eighteen candidates for emission line galaxies were discovered in a narrow-band infrared survey that targeted the redshifts of damped \\lya or metal lines in the spectra of quasars. The presence of emission lines is inferred from the photometric magnitudes in narrow and broad band interference filters, corresponding to \\ha at redshifts of 0.89 (6 objects) and 2.4 (10 objects), and \\oii at a redshift of 2.3 (2 objects). Most of the candidates are small resolved objects, compatible with galaxies at the redshifts of the absorbers. Because a similar survey targeted at the redshifts of quasars themselves uncovered only one emission-line galaxy in a larger volume, the results imply substantial clustering of young galaxies or formation within filaments or sheets whose locations are indicated by the redshifts of absorption along the lines of sight to more distant quasars. ",
        "introduction": "Exploration of the high redshift universe and discovery of the most distant objects is still in its infancy. Only recently have the tools been available to detect normal galaxies at redshifts larger than one, when the first galaxies were created (\\cite{pescarelle96,hu96,cowie98,steidel96}). It seems likely that young galaxies will have a variety of different signatures so that it will be necessary to use several diverse techniques to uncover all of them: searches at optical, infrared, x-ray, and radio wavelengths, for example. In particular, basing the statistical studies of the high redshift galaxies only on objects detected in the rest-frame UV could miss many young galaxies (\\cite{franceschini98,guiderdoni97}), and sampling longer wavelength ranges is necessary.  We carried out a survey for infrared emission-line galaxies by imaging through narrow (${\\Delta\\lambda \\over \\lambda} \\sim 0.01$) and broad band filters between 1 and 2.5\\,$\\mu$m, identifying objects that appeared brighter in the narrow filters (\\cite[hereafter TMB96]{tmb96}). Our first survey was designed to uncover emission lines at the redshifts of quasars within each survey field, in case there is substantial clustering marked by quasars. In an area of 276\\,\\sq\\arcmin , only one emission-line galaxy was discovered (\\cite{beckwith98}).  The surface density of such objects implied by these results is similar to that inferred from other surveys (\\cite{cowie94,graham96,malkan96,bechtold97}) and suggests that the infrared emission-line galaxies constitute at most a modest population of young galaxies at high redshift.  Using the same instruments, we undertook a second infrared survey for emission-line galaxies targeted at the redshifts of damped \\lya or metal absorption lines in the spectra of quasars.  Damped \\lya absorbers are thought to contain as much baryonic matter as seen in all spiral galaxies today (\\cite{wolfe86}) and may, therefore, mark sites of ongoing star formation. Several other groups (\\cite{lowenthal91,macchetto93,wolfe92,moller93,djorgovski96,francis97}) carried out similar surveys at optical wavelengths looking for \\lya emission-line galaxies in these regions. They discovered only a few such emission-line (non-AGN) galaxies, but \\cite{wolfe93} showed that the implied volume density was significantly higher than in the general field. Metal absorption systems also indicate that star formation has taken place, and are identifiable from the ground at lower redshifts than \\lya alone.  We selected damped \\lya systems or metal absorbers whose redshifts put the main optical lines H$_\\alpha$, H$_\\beta$, \\Oiii, and \\Oii\\ into standard narrow-band filters in the J, H and K bands.  The resulting redshift ranges are: $0.5<z<1.9$, $2.1<z<2.5$ and $3.1<z<3.8$.  Special emphasis was given to the \\ha\\ line expected to be the brightest in young star systems and the least affected by dust. This letter describes the results of the new survey. ",
        "conclusions": "The most striking result of this survey is the large number of candidates discovered in a small area implying that by choosing the right redshift intervals, it is possible to detect emission-line objects rather easily. The total sky coverage in this survey was 228\\,\\sq\\arcmin\\ compared with 276\\,\\sq\\arcmin\\ in TMB96, yet 18 candidates were discovered in the second survey compared to a single emission-line galaxy in the first.  Taken at face value, the new results suggest that damped \\lya and metal absorbers pinpoint the regions where galaxy formation is taking place.  Confirmation of the exact nature of these objects will require spectroscopic follow up and perhaps additional imaging with HST. Since not all the objects are resolved at the coarse pixel scale used, some could be stars or other nearby objects.  A few objects are near the 3$\\sigma$ detection limit. Statistical comparison must, therefore, be regarded with some caution until further data are available.  Nevertheless, we believe that the main result of this survey is robust: there are more emission-line galaxies associated with damped \\lya and metal absorber redshifts than with the redshifts of quasars or at arbitrary redshifts.  First, the observational method employed was identical to that of TMB96, but the resulting number of candidates is almost 20 times larger. Second, the one candidate, cK39, discovered in the first survey by TMB96 was also coarsely sampled but is, indeed, an emission line galaxy (\\cite{beckwith98}) spectroscopically confirmed and easily resolved with the Keck telescope. Third, the HST images of two of our {\\em least} statistically significant candidates show them to be galaxies of exactly the size expected for objects at redshifts greater than a few tenths and with the kind of morphology expected of objects in late stages of assembly.  These results may be compared with similar studies of \\lya emitting galaxies in the neighborhood of damped \\lya absorption line systems. By the number of detected objects (\\cite{lowenthal91,wolfe92,macchetto93,moller93,djorgovski96}) and observed fields (\\cite{smith89,deharveng90,lowenthal95}) an average comoving density of these systems of about 7$\\times10^{-4}$ Mpc$^{-3}$ is derived down to a limiting SFR of about 5\\,\\sfr, assuming case B recombination and no dust. The infrared technique gives similar results, i.e., detections in about 1/4 of the fields and a density of $9\\times 10^{-4}$\\,Mpc$^{-3}$. The range of SFR implied by our observations 6 - 200\\,\\sfr , is somewhat higher than that usually obtained by the optical searches, 5 - 20\\,\\sfr , but not dramatically so. If there is modest local extinction to the star formation regions in high redshift galaxies, optical derived SFRs must be increased by a factors of about 3 to get the true SFR (\\cite{pettini97}), making their range almost coincident with ours. Our results are perhaps more consistent with the limiting SFR of 30 - 80\\,\\sfr\\ along sightlines toward eight $z>2$ damped \\lya absorbers as derived from redshifted H$\\alpha$ (\\cite{bunker98}).  On the other hand, the lines could be produced by active galactic nuclei. There are several good reasons to believe this could be the case for the one emission-line object discovered in the survey of TMB96 (\\cite{beckwith98}), and that it could be a widespread phenomenon (\\cite{francis97}). If so, the density of such objects in these regions is considerably higher than the cosmic average.  The density of known quasars at these redshifts (e.g., \\cite{andreani92}) implies about $4-8\\times10^{-5}$ quasars per field, meaning that we might be seeing far more AGNs than expected. Although the damped \\lya redshifts mark regions with many galaxies, the nature of each single galaxy will remain a mystery until proper follow up observations will be able to distingish between emission due to star formation and AGNs.  Cold dark matter simulations of early galaxy formation produce filaments of galaxies (\\cite{white94}) with groups of high redshift metal and damped \\lya absorbers spanning up to several hundred kpc (\\cite{rauch97}).  If the damped \\lya absorbers actually trace the positions of these filaments, then the high detection rate of emission line galaxies in this survey may support the overall structure predicted by CDM.  This conclusion can be made quantitative by measuring the average overdensity of objects near damped \\lya and metal systems with respect to TMB96.\\\\  \\vspace*{3mm}"
    },
    "9805/astro-ph9805179_arXiv.txt": {
        "abstract": "We have studied finite temperature corrections to the baryon transport cross sections and diffusion coefficients. These corrections are based upon our recently computed renormalized electron mass and modified state density due to the background thermal bath in the early universe. It is found that the optimum nucleosynthesis yields computed using our diffusion coefficients shift to longer distance scales by a factor of about 3. We also find that the primordial $^4 He$ abundance decreases by $\\Delta Y_p \\simeq 0.01$ while $D$ and $^7 Li$ increase. Effects of these results on constraints from primordial nucleosynthesis are discussed. In particular, we find that a large baryonic contribution to the closure density ($\\omegabh2 \\lsim 0.4$) may be allowed in inhomogeneous models corrected for finite temperature.  \\vskip 0.5cm ",
        "introduction": "It is generally believed that a cosmological phase transition from quark-gluon plasma to hadronic matter would have occurred when the temperature of the universe was about $T \\simeq 100 \\sim 200$ MeV \\cite{witten,qhpt}. This quark-hadron phase transition involves a rich variety of physical phenomena, particularly if it is first order. One possible consequence of this phase transition (or other mechanism \\cite{ibbn} is the formation of isothermal baryon density fluctuations \\cite{qhpt,AH,ibbn,AHS}. In particular, Applegate and Hogan \\cite{AH} suggested that, once formed, such fluctuations could lead to the segregation of neutrons and protons. Baryon density inhomogeneities which persist to  $T \\sim 0.1$ MeV could affect primordial big-bang nucleosynthesis (BBN) yields. The primordial abundances of light elements in such inhomogeneous big-bang nucleosynthesis (IBBN) models would be quite different from those of the standard homogeneous big-bang nucleosynthesis (HBBN) model.  In the IBBN model, neutrons can diffuse relatively easily through the primordial plasma. Applegate, Hogan and Scherrer (AHS) \\cite{AHS} have shown that the different diffusion lengths for neutrons and protons could lead to the formation of high-baryon-density proton-rich regions and low-baryon-density neutron-rich regions. Before the freeze-out from weak reaction equilibrium (at $T \\approx 1$ MeV), there is no segregation between neutrons and protons because the baryons only diffuse efficiently during the fraction of the time they spend as neutrons. After weak equilibrium freeze-out, however, protons and neutrons diffuse independently.  Since protons and neutrons have different diffusion coefficients, the result can be a spatial segregation of these species.  In view of the importance of using the light-element yields from BBN to constrain both the baryon-to-photon ratio and various cosmological and particle-physics theories \\cite{kolb}, such IBBN models must be examined seriously. It is therefore important to evaluate the diffusion coefficients of neutrons and protons as accurately as possible in order to quantify the degree to which the primordial abundances of the light elements can be modified in the IBBN model. In this regard, AHS have derived the baryon diffusion coefficients using a mobility formula and the Einstein relation between mobility and the diffusion coefficient. After that, Banerjee and Chitre \\cite{banerjee} used relativistic kinetic theory to calculate the baryon diffusion coefficients in the lowest order Chapman-Enskog approximation.  In calculations of the baryon diffusion coefficients, it is necessary to know the baryon transport cross sections for scattering with the background primordial plasma \\cite{AHS}. However, in all previous works, the transport scattering cross sections were calculated in vacuum and the electron mass was neglected. Therefore, in the present work we take into account finite temperature effects in the calculation of baryon diffusion coefficients and explore their effects on primordial nucleosynthesis.  Finite temperature effects on elementary processes are significant from the point of view of cosmology and astrophysics. The early universe is usually described as a hot gas of particles in nearly thermodynamic equilibrium. Finite temperature effects enter through the statistical distribution functions which renormalize the masses and wave functions. These renormalized masses and wave functions affect scattering processes and decay rates.  Several authors \\cite{donoghue} have generalized the electron mass and wave-function renormalization to all temperatures and densities. As an application of finite temperature effects, Dicus et al. \\cite{dicus}, and independently Cambier et al. \\cite{cps}, included the finite temperature effects on weak reaction rates in the calculation of standard big-bang nucleosynthesis. They obtained the corrected light-element abundances and found that the corrections are only of order of a few percent. In subsequent work, Saleem \\cite{saleem} included the effects of the electron mass shift at finite temperature on BBN, and Baier et al. \\cite{baier} examined the finite temperature radiative corrections to the weak neutron-proton decay rates. More recently, Fornengo et al. \\cite{cwkim} have considered the finite temperature effects on the neutrino decoupling temperature which is important in the evolution of the early universe.  In this paper we calculate the baryon diffusion coefficients at finite temperature and apply these new diffusion coefficients to an IBBN model. We investigate finite temperature effects on the yields of primordial $D,\\; ^3 He, \\; ^4 He,$ and $^7 Li$ in the IBBN model. In the calculation of light element yields, we use a multi-zone computer code \\cite{mathews90} which couples the diffusion equation with a nuclear reaction network and the cosmic expansion.  The plan of the paper is as follows. In Section \\ref{sec:finite}, we discuss how to include finite temperature effects in the calculation.  We briefly discuss the effective mass of an electron with special attention to the $T \\lsim~1$ MeV temperature range which is most relevant for the neutron, proton diffusion before and during BBN. We also evaluate finite temperature effects on the baryon transport cross section in the scattering of baryons from the background electron-positron plasma. In Sec.~\\ref{sec:diffusion} we calculate the baryon diffusion length with new diffusion coefficients for neutrons and protons and compare these to those estimated in AHS. In Sec.~\\ref{sec:ibbn} we apply our results to IBBN. Finally, we summarize our results and discuss the effects of finite temperature on light-element constraints from BBN. We shall employ units in which $\\hbar = k_B = c = 1$, except when specific units must be attached to a result. ",
        "conclusions": "\\label{sec:conclusion}  We have studied the leading finite-temperature effects on baryon diffusion and nucleosynthesis in the early universe. The major motivation has been to quantify the degree to which finite-temperature effects could affect the nucleosynthesis constraints. We have calculated the baryon diffusion lengths based upon our previously calculated scattering cross sections \\cite{isuh} which included: (1) finite temperature Dirac spinors which are recast into the form of an effective electron mass; (2) finite temperature modifications to the phase space distribution of the electrons; and (3) the modified neutrino temperature.  The baryon diffusion coefficients which affect baryon inhomogeneities before or during big-bang nucleosynthesis are changed significantly by our temperature dependent electron-hadron transport cross sections. We have reevaluated the upper limits to $\\omegabh2$ in inhomogeneous primordial nucleosynthesis models which take into account finite temperature effects. The optimum conditions are shifted to larger fluctuation distance scales by a factor of about 3 in condensed spherical fluctuation geometry. It also is found that the value of $^4 He$ decreases by $\\Delta Y_p \\simeq 0.01$. The limits on the baryon to photon ratio are constrained, however, by increased $^{7}Li$ which limits the allowed range of $\\omegabh2$.  With the larger value of the $D$ constraints, allowed values of $\\omegabh2$ are nearly unchanged when compared to previous works. However, for the preferred smaller value of the $D$ constraints, it is possible to have an allowed region of $\\omegabh2$ as large as 0.12 to 0.4 in inhomogeneous models.  \\vspace{1cm}  {\\bf Acknowledgments.} \\noindent The author would like to thank Prof. M. Orito for his careful reading of an early version of this manuscript. I.S.S. also acknowledges the Korea Research Foundation (KRF) for financial support. This work supported in part by DOE Nuclear Theory Grant DE-FG02-95ER40934."
    },
    "9805/astro-ph9805209_arXiv.txt": {
        "abstract": "We report  the  results  of abundance  determination for the components of the SB2 star 66 Eri ($M_{\\rm A}/M_{\\rm B}$=0.97) from high resolution CCD echelle spectra  with S/N{$\\ge$}100  taken with the 1-m telescope of Special Astrophysical Observatory (Zelenchuck, Russia). The atmospheric parameters of the components were determined  using all available photometric, spectrophotometric and spectral data. The abundances of~ 27 elements were found. The abundances of components  are  different. The B component, previously  classified as an Hg-Mn star, does not show anomalies typical of this group such as deficit of He,  Al and excess of P, Ga  but shows  overabundances  of heavy elements which amount  up to 4-5 dex. The A component also shows moderate Mn and Ba excess. Lines of other heavy elements were not detected. Estimates of upper limits to their abundances do not permit to exclude completely the presence of fainter anomalies in the A component either. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805319_arXiv.txt": {
        "abstract": "A model of the gravitationally evolved dark matter distribution, in the Eulerian space, is developed.  It is a simple extension of the excursion set model that is commonly used to estimate the mass function of collapsed dark matter haloes.  In addition to describing the evolution of the dark matter itself, the model allows one to describe the evolution of the Eulerian space distribution of the haloes.  It can also be used to describe density profiles, on scales larger than the virial radius, of these haloes, and to quantify the way in which matter flows in and out of Eulerian cells.  When the initial Lagrangian space distribution is white noise Gaussian, the model suggests that the Inverse Gaussian distribution should provide a reasonably good approximation to the evolved Eulerian density field, in agreement with numerical simulations.  Application of this model to clustering from more general Gaussian initial conditions is discussed at the end. ",
        "introduction": "The hypothesis that, in comoving coordinates, initially denser regions contract more rapidly than less dense regions, and that sufficiently underdense regions expand, is simple, reasonable, and powerful.  As a consequence of this expansion and contraction, the density distribution in the initial Lagrangian space will be different from that in the evolved, Eulerian space. Suppose that, as the universe evolves, the number of expanding and contracting regions is conserved---only their comoving size changes---and the mass within each such region is also conserved.  If we have a model for the way in which the evolution of the size of a region depends on its initial size and density, and we also have a model for the initial number of regions as a function of initial size and density, then we can compute the distribution of sizes and densities at some later time.  For example, suppose that the initial Lagrangian density distribution is a Gaussian random field, and that the evolution of regions is given by the spherical collapse model.  Then it should be possible to construct a model for $p(M_0|R,z)$, where $p(M_0|R,z)$ is the fraction of regions of size $R$ that, at $z$, contain mass $M_0$.   The quantity $p(M_0|R,z)$ is often called the Eulerian probability distribution function.  In the Press-Schechter (1974) approach, at any given time, all matter is in the form of collapsed objects, usually called haloes, and the distribution of halo masses evolves with time.  At any time, the matter within a randomly placed cell $R$ is divided among many collapsed haloes.  Thus, $p(M_0|R,z)$ depends both on the halo mass function, and on the spatial distribution of the haloes. Bond et al. (1991) showed how to estimate the evolution of the halo mass function if the initial Lagrangian distribution is Gaussian (also see Lacey \\& Cole 1993). They did not show how to estimate the spatial distribution of these haloes, but Mo \\& White (1996) showed how this might be accomplished.  This paper develops a model that combines, self-consistently, the Bond et al. (1991) excursion set approach with the Mo \\& White (1996) model for the Eulerian space halo distribution.  The model developed here allows one to simultaneously describe both the distribution of the dark matter, i.e. $p(M_0|R,z)$, and that of the haloes. See, e.g., Mo \\& White (1996) for why such a model is useful.  Section~\\ref{model} describes the model. It shows why the distribution of first crossings of a barrier whose height is not constant, by Brownian motion random walks, is useful. The shape of the barrier associated with the spherical collapse model is given in Section~\\ref{smodel}. The relation between the first crossing distribution and the Eulerian distribution $p(M_0|R,z)$ is discussed in Section~\\ref{prelim}. Section~\\ref{halod} shows that, in the context of the model developed here, the halo mass function is related to the small cell size limit (i.e. $R\\to 0$) of $p(M_0|R,z)$. It discusses the Bond et al. (1991) excursion set results in this context, and then shows how the model can be used to describe the spatial distribution of the haloes as well.  The results of Mo \\& White (1996) are discussed in Section~\\ref{mowhite}. Section~\\ref{twobare} shows that the associated two barrier problem can be used to provide information about the evolved density profile, and also about the way in which matter flows in and out of Eulerian cells.  The first crossing distribution associated with the spherical collapse barrier must be obtained numerically. Therefore, to illustrate the usefulness of our approach, Section~\\ref{linear} shows the results of assuming that the initial distribution is white-noise Gaussian, and that the barrier shape is simpler than that associated with the spherical collapse model.  In Section~\\ref{linear}, the barrier is assumed to be linear for a number of reasons. Firstly, this linear barrier can be understood as arising from a simple variant of the spherical collapse model (Section~\\ref{scoll}). Secondly, the barrier crossing distribution can be computed analytically (e.g. Schr\\\"odinger 1915).  The details of the derivation are presented in Appendix~\\ref{deriv}. Thirdly, the Eulerian probability distribution associated with the first crossings of this linear barrier is Inverse Gaussian (Section~\\ref{pdfe}), and the Inverse Gaussian provides a good fit to the Eulerian distribution measured in numerical simulations of clustering from white noise initial conditions (Section~\\ref{sims}). Finally, studies of clustering from Poisson initial conditions also suggest that this barrier shape is a useful approximation (Sheth 1998).  The Bond et al. (1991) results are derived, within the context of this linear barrier model, in Section~\\ref{const}. Analytic expressions for the Eulerian distribution function as well as the halo--mass and halo--halo correlations are derived for all scales, and for all times, in Sections~\\ref{pdfe} and~\\ref{halobar}. Section~\\ref{twobar} provides various formulae associated with the two-linear-barrier problem. A final section summarizes the results.  It also discusses how the model can be extended to describe clustering from more general initial conditions. ",
        "conclusions": "\\label{discuss} \\subsection{Summary of the model} This paper develops a model that allows one to provide an approximate description of the spatial distribution of dark matter, as well as dark matter haloes, in a self-consistent way. It can also be used to estimate the density profiles of haloes, and to quantify how matter flows in and out of Eulerian cells. The model is described in Section~\\ref{model}.  The model assumes that, at the epoch $z$, the comoving Eulerian size $R$ of an initial Lagrangian region $R_0$, is related to its initial overdensity $\\delta_0$.  This assumption is motivated by the spherical collapse model.  The additional assumption that the number of such regions is conserved, only their comoving size changes, allows one to compute statistics in the Eulerian space by taking appropriate averages over the relevant Lagrangian space quantities.  Since the initial Lagrangian distribution can be computed (e.g. Mo \\& White 1996; Sheth \\& Lemson 1998), so can the Eulerian distribution.  The algorithm for taking these appropriate averages is related to the solution of an excursion set, barrier crossing problem, in which the barrier shape is given by the spherical collapse model.  A constant barrier is a special case of the barrier studied here. The first crossing distribution associated with a constant barrier can be used to estimate the distribution of virialized halo masses (Bond et al. 1991).  In the context of this paper, this special case corresponds to the limit of vanishing Eulerian cell size; the halo mass function is simply the Eulerian probability distribution function in the limit of vanishing Eulerian cell size.  Thus, the model here shows how the Bond et al. (1991) construction that yields the halo mass function can be extended to yield the Eulerian probability distribution function. The results of Mo \\& White (1996) are also easily understood within the context of this model.  In Section~\\ref{linear}, the linear barrier was used to show explicitly how the model works.  This was done for a number  of reasons. Firstly, the spherical collapse barrier problem must be solved numerically, whereas, for the linear barrier, most interesting quantities can be computed analytically. Secondly, within the context of this model, the linear barrier can be understood as arising from a simple variant of the spherical collapse model (Section~\\ref{scoll}). For this barrier shape, the associated Eulerian distribution is Inverse Gaussian (Section~\\ref{pdfe}), and this is good agreement with numerical simulations of clustering from white noise initial conditions (Fig.~\\ref{pdf0}). This provides yet another reason for considering the linear barrier.  In the model, correlations between haloes and mass arise from the simple fact that a cell which is known to contain a halo certainly contains that halo's mass.  Correlations between haloes arise primarily from volume exclusion effects.  That is, in the initial Lagrangian space, haloes occupy a volume that is proportional to their mass.  Since no other halo can occupy the region taken up by a given halo, this exclusion effect gives rise to (anti)-correlations, at least on small scales.  This is probably a generic consequence of the white-noise initial conditions, and not specific to the linear barrier.  For the linear barrier model, the evolution of halo--halo correlations can be written analytically (Section~\\ref{halobar}). Section~\\ref{twobar} showed that the two-linear-barrier problem can also be solved analytically; in the model, this can be used to provide information about the density run around random positions in Eulerian space.  Although the necessary formulae were all derived there, we did not pursue this further.  \\subsection{Scaling properties of the model} The Inverse Gaussian distribution was derived here by considering the distribution of crossings of a linear barrier by Brownian motion.  The relation between this barrier shape (equation~\\ref{linear}) and that required by the spherical collapse model (equation~\\ref{mow}) was discussed in Section~\\ref{scoll}.  The thick solid lines in Fig.~\\ref{barrier} show this spherical collapse barrier (equation~\\ref{mow}) for representative values of $V\\propto R^3$.  Curves for larger $R$ look just like those for smaller $R$ except that are shifted to the left (since the $x$-axis of the plot shows the negative of the logarithm of $V_0$). Curves for the same comoving $R$ but at higher redshift are simply multiplied by $(1+z)$, so the zero crossing remains the same. Dashed curves show the corresponding quantities for the linear barrier used in Section~\\ref{linear}. This figure shows that changing the comoving size $R$ at fixed $z$ is equivalent to rescaling the $S_0$ axis, and changing $z$ at fixed comoving radius $R$ is equivalent to rescaling the $\\delta_0$ axis.  This provides a strong constraint on the form of the distribution of first crossings.  \\begin{figure} \\centering \\mbox{\\psfig{figure=igdexf4.ps,height=6cm,bbllx=89pt,bblly=65pt,bburx=608pt,bbury=456pt}} \\caption{Dependence of the barrier shape on comoving Eulerian size $R$ and redshift $z$.  Solid curves show $B=\\delta_0(R_0|R,z)$ of equation~(\\ref{mow}), dashed curves show equation~(\\ref{linbar}). For white noise, $S_0\\propto 1/V_0\\propto 1/R_0^3$.} \\label{barrier} \\end{figure}  Namely, it implies that $f(S_0,\\delta_0|R)$ should have a self-similar form, in that it should be a function of two parameters, one which is related to the scaling of the $S_0$ axis, and another which relates the scaling of the $S_0$ axis to that of the $\\delta_0$ axis.  Since the first crossing distribution is a function of only two parameters, the associated Eulerian distribution is also.  It is convenient, then, to think of these parameters as being the mean and variance of the Eulerian distribution. This is an attractive feature of the model, since Colombi et al. (1997) find that, to a good approximation, the Eulerian distributions measured in $N$-body simulations of clustering from white noise initial conditions are functions of the mean and the variance only.  \\subsection{Scale--free initial conditions} Initial Gaussian random fields with scale-free power spectra, $P(k)\\propto k^n$ and $-3<n<1$, are a simple generalization of white-noise (for which $n=0$). Equation~(\\ref{slin}) shows that, for these spectra, \\begin{equation} S_0\\propto M_0^{-\\alpha},\\qquad {\\rm where}\\qquad \\alpha = (n+3)/3, \\label{s0n} \\end{equation} which means that $\\Delta = M_0/V = (S_{\\rm V}/S_0)^{1/\\alpha}$, where $S_{\\rm V}\\propto V^{-\\alpha}$. This means that, when plotted as a function of $(\\delta_0,S_0)$, the shape of the spherical collapse barrier depends on the initial power spectrum.  Although the barrier shape when $n\\ne 0$ is different than for white noise, it still has the same scaling as the white noise barrier, since changing the Eulerian scale $R$ is still equivalent to a simple shift of the barrier along the $\\log(S_0)$ axis.  So, the barrier crossing and the associated Eulerian distributions will still be functions of two parameters. This is in qualitative agreement with the results of Colombi et al. (1997).  For initial conditions that are not scale-free, changing the Eulerian scale $R$ is no longer equivalent to a simple shift of the barrier along the $\\log(S_0)$ axis.  Thus, the barrier crossing distribution is no longer self--similar, and the associated Eulerian distribution is no longer a function of just two parameters.  Of course, two parameter fits may remain a good approximation; at issue is the rate of change of the slope of the power spectrum over the scales at which most of the trajectories cross the barrier.  While these are encouraging features of the model, extending it to describe clustering from initial conditions that are different from white noise is not straightforward.  The problem is one of ensuring correct normalization. For scale--free initial conditions, equations~(\\ref{s0n}) and~(\\ref{fspd}), and the requirement that the associated Eulerian distribution have unit mean (equation~\\ref{norm}), means that the first crossing distributions must satisfy \\begin{equation} \\int_0^\\infty \\!\\!\\! f(S_0,\\delta_0|R)\\,{\\rm d}S_0 = \\int_0^\\infty \\!\\!\\! \\left({S_0\\over S_{\\rm V}}\\right)^{1\\over\\alpha} f(S_0,\\delta_0|R)\\,{\\rm d}S_0 = 1. \\label{fsalpha} \\end{equation} The previous section showed that, for a white-noise Gaussian field, the distribution of first crossings of the linear barrier has an associated Eulerian distribution which has unit mean, as required.  However, in general, not all barrier shapes are compatible with this requirement.  For example, recall that equation~(\\ref{bnrdo}) with $\\delta_{\\rm c0}=1.5$ is a good approximation to the spherical collapse model.  When $n=-1$, $\\alpha = 2/3$, so the associated barrier, when written as a function of $S_0$, is the same (linear) as it was in Section~\\ref{linear}. Now, for a linear barrier, the first passage distribution is Inverse Gaussian.  However, when $\\alpha=2/3$ and $f(S_0,\\delta_0|R)$ is Inverse Gaussian, then the second equality of equation~(\\ref{fsalpha}) is not satisfied.  Thus, it appears likely that, for arbitrary power spectra, the excursion set approach developed here will not work if at least some parametrizations of the spherical collapse barrier are used.  This means that it may be more fruitful to consider the inverse problem to that considered in this paper.  \\subsection{The inverse problem and the Generalized Inverse Gaussian distributions} In this paper, the barrier shape was specified, then the first crossing distribution was obtained, and finally, the associated Eulerian distribution was derived.  It may be more useful to first specify an Eulerian distribution.  Then, transform it to a first crossing distribution, from which the shape of the barrier can be inferred.  This ensures that all normalization requirements have been satisfied.  Once the barrier shape is known, the halo--mass and halo--halo correlations can be worked out just as they were in this paper.  Finally, this barrier shape can be compared with that required by the spherical collapse model.  In this context, it is worth mentioning that there is a class of distributions which provides convenient generalizations of the Inverse Gaussian, and may provide useful approximations to the Eulerian distributions measured in simulations of gravitational clustering from scale--free initial conditions.  The Generalized Inverse Gaussian (GIG) distribution with index $\\lambda$, scale parameter $\\eta$, and concentration parameter $\\omega$ is \\begin{equation} f_\\lambda(x|\\eta,\\omega)\\,{\\rm d}x = {(x/\\eta)^\\lambda\\over 2\\,K_\\lambda(\\omega)}\\, {\\rm e}^{-{\\omega\\over 2}({x\\over\\eta}+{\\eta\\over x})}\\ {{\\rm d}x\\over x}, \\qquad x\\ge 0, \\end{equation} where $K_\\lambda(\\omega)$ is a modified Bessel function of the third kind with index $\\lambda$.  For a given $\\lambda$, these are functions of just two parameters, $\\eta$ and $\\omega$, so they have the self--similar property discussed above. For these modified Bessel functions, \\begin{eqnarray} K_\\lambda(\\omega) &=& K_{-\\lambda}(\\omega) \\qquad{\\rm and} \\nonumber \\\\ K_{\\lambda+1}(\\omega) &=& 2\\,(\\lambda/\\omega)\\,K_\\lambda(\\omega) + K_{\\lambda-1}(\\omega). \\end{eqnarray} Therefore, when $\\lambda = -(2\\alpha)^{-1}$, then the GIG distributions, with $x=S_0$, $\\eta=S_{\\rm V}$ and $\\omega=\\delta_{\\rm c0}^2/S_{\\rm V}$ satisfy equation~(\\ref{fsalpha}), so the associated Eulerian distributions are correctly normalized and have unit mean.  When $n=0$, then $\\alpha=1$, so $\\lambda = -1/2$, and this distribution is the same as the Inverse Gaussian distribution considered in section~\\ref{linear}.  The symbols in Fig.~\\ref{skew} show the values of $S_3$ and $S_4$ (defined similarly to equation~\\ref{esen}) plotted versus variance for these GIG distributions when $n=0$ (filled circles), $-1$ (open circles), and $-2$ (stars). The curves were computed using the extended perturbation theory fitting functions of Colombi et al. (1997). The solid curves show their fits to the values of $S_3$ and $S_4$ measured in numerical simulations of clustering from the corresponding scale-free initial conditions, and dashed curves give an estimate of the allowed range of values. While the GIG values do not fit the numerical results, they do show similar trends.  \\begin{figure} \\centering \\mbox{\\psfig{figure=igdexf5.ps,height=6cm,bbllx=84pt,bblly=63pt,bburx=608pt,bbury=459pt}} \\caption{$S_3$ and $S_4$ versus variance for the Generalized Inverse Gaussian distributions associated with initial power spectra with slopes $n=0$ (filled circles), $-1$ (open circles), and $-2$ (stars). The solid curves show the values measured in numerical simulations of clustering from these initial conditions, as parametrized by Colombi et al. (1997).  Dashed curves show their estimates of the allowed range of values.} \\label{skew} \\end{figure}  The problem, then, is to determine the shape of the barrier for which the GIG distribution is the first crossing distribution, and to then compute the associated halo distribution. The solution of this inverse problem is the subject of ongoing work, where the barriers associated with the distributions given by the extended perturbation theory of Colombi et al. (1997) are also used as the Eulerian distributions."
    },
    "9805/astro-ph9805296_arXiv.txt": {
        "abstract": "We present optical broad band B--I colour maps of a further sample of 10 Seyfert 2 galaxies. In these bands, the contribution from emission lines to the total flux is small, and hence the images predominantly trace the continuum distribution. As in our earlier colour maps of a sample of Seyferts type 1 and 2, we detect extended blue continuum components in the circumnuclear region of several galaxies. These components are either elongated (in Mkn 533, Mkn 607, Mkn 1066, NGC 5347, NGC 5953 and NGC 7319) or form a double structure across the nucleus (in NGC 5929). They are closely aligned with the radio and emission line axes of these galaxies and probably arise from scattering of nuclear continuum light by extranuclear mirrors. Similar blue elongations (in Mkn 1 and NGC 7212) and double structures (in NGC 788) are less well aligned with the radio and line emission, and their relationship with scattering regions must be considered uncertain. The colours of the blue maxima are consistent with those expected from scattering off dust or electrons, a conclusion strengthened by the combined sample of the two papers in this series. Our findings thus strongly support the current unified models of AGN. ",
        "introduction": "In the unified models of active galactic nuclei (AGN), the observed differences between various types of AGN are explained in terms of collimated radiation, obscuration and viewing angle effects on intrinsically similar objects (\\eg Antonucci 1993). Specifically for Seyfert galaxies, there is increasing evidence (for references see Section 4 and Kotilainen \\& Ward 1997; hereafter KW97), that in Seyfert 2s optically thick material, in the form of a thick disk or a dusty torus, blocks our direct view of the compact nucleus and the broad line region (BLR), and collimates the ionising photons along the minor axis of the torus into two oppositely directed cones.  The emphasis in studies of the nuclear regions of Seyferts has been on the emission line properties (\\eg integral field spectroscopy and emission line imaging). However, although the nuclear torus is expected to be too small to be resolved, its outer parts, \\eg in the form of a flattened red dust distribution, may be detectable in continuum colour maps. Evidence for these obscuring regions, perpendicular to the bipolar extended narrow line region (ENLR) has been detected in \\eg NGC 5252 (Tadhunter \\& Tsvetanov 1989; Kotilainen \\& Prieto 1995) and Mkn 348 (Simpson \\etal 1996). Similarly, continuum colour maps have revealed blue morphological features coincident with the high-excitation gas in the ENLR (\\eg Pogge \\& De Robertis 1993, 1995; Mulchaey, Wilson \\& Tsvetanov 1996a; KW97). In KW97, we presented high resolution colour maps of three Seyfert 1 and six Seyfert 2 galaxies. These maps revealed the existence of blue elongations and double structures in NGC 1068, NGC 3227, NGC 4151, Mkn 3 and Mkn 573, that were not discernible in single filter images, and that we interpreted as scattered nuclear light. In this paper, we present similar colour maps of the circumnuclear regions of 10 more Seyfert 2 galaxies. We briefly describe the observations, data reduction and the methods used in the construction of the colour maps in section 2. In section 3 we present and discuss the morphological features detected for each galaxy, including comparison with available multiwavelength data. In section 4 we discuss the relevance of our findings in the enlarged sample of Seyfert galaxies, in view of current unified AGN models. Conclusions are presented in section 5. Throughout this paper we use Hubble constant H$_{0}$ = 50 km s$^{-1}$ Mpc$^{-1}$, and deceleration parameter q$_{0}$ = 0. ",
        "conclusions": "Continuum colour maps provide a powerful method to reveal scattered nuclear light in the circumnuclear regions of Seyfert 2 galaxies. As in our previous paper (KW97), we have detected elongated or double blue morphological structures in the B--I maps of several Seyfert 2 galaxies and found a good agreement with previous emission line and radio studies of these galaxies. In Mkn 533, Mkn 607, Mkn 1066, NGC 5347, NGC 5929, NGC 5953 and NGC 7319, we interpret the colour structure as arising from scattering of the nuclear continuum light from electrons or dust in extranuclear mirrors. In Mkn 1, NGC 788 and NGC 7212, the blue structures are less well in agreement with the radio and line emission, and their relationship with scattering regions remains uncertain. The colours of the blue maxima are consistent with those expected from scattering off dust or electrons, this conclusion strengthened by the combined sample of the two papers in this series. Our findings thus strongly support the current unified models of AGN."
    },
    "9805/astro-ph9805069_arXiv.txt": {
        "abstract": "We present preliminary results from a programme aimed at acquiring linear spectropolarimetry of magnetic A and B stars. Linear polarization in the spectral lines of these objects is due to the Zeeman effect, and should provide detailed new information regarding the structure of their  strong magnetic fields.  To illustrate the impact of these new data, we compare observed circular and linear polarization line profiles of 53~Cam with the profiles predicted by the magnetic model by Landstreet (1988).  Linear polarization in the spectral lines of {\\em all} stars studied is extremely weak; in most cases, below the threshold of detectability even for very high SNRs. In order to overcome this problem, we employ the Least-Squares Deconvolution (LSD) multi-line analysis technique in order to extract low-noise mean line profiles and polarization signatures from our \\'echelle spectra. Tests show that these mean signatures can be modelled as real spectral lines, and have the potential to lead to high-resolution maps of the magnetic and chemical abundance surface distributions. ",
        "introduction": "Recent observations of the broadband linear polarization variability of magnetic Ap stars have been shown to be very sensitive to the detailed topology of their magnetic fields. Using these data, Leroy et al. (1996) have presented models of the magnetic field structure of several stars which resolve clear departures from the simple multipolar geometries assumed in the past. Following these studies, we have taken an important next step by acquiring systematic linear {\\em spectropolarization} measurements of a number of strong-field Ap stars. Due principally to the additional information contained in line profile shapes, these data promise important new constraints on the structure of the magnetic fields of these objects, and also offer the possibility of constructing mutually consistent, detailed magnetic field and chemical abundance distribution maps for individual stars. This will in turn provide new information about the interaction of the magnetic field with the atmospheric processes responsible for abundance inhomogeneities, as well as motivate new theoretical investigation into the mechanisms responsible for magnetic fields in early-type stars. ",
        "conclusions": "The results presented herein show that spectral line linear polarization measurements provide important new constraints on the magnetic field structure of A and B stars. Least-Squares Deconvolution is shown to be an effective technique for the extraction of mean Stokes signatures which can be accurately modelled as real line profiles. Such modelling has the potential to lead to the construction of mutually consistent, high-resolution maps of the magnetic and chemical surface structure of these stars."
    },
    "9805/astro-ph9805225_arXiv.txt": {
        "abstract": "A newly discovered instability in rotating neutron stars, driven by gravitational radiation reaction acting on the stars' $r$-modes, is shown here to set an upper limit on the spin rate of young neutron stars. We calculate the timescales for growth of linear perturbations due to gravitational radiation reaction, and for dissipation by shear and bulk viscosity, working to second order in a slow-rotation expansion within a Newtonian polytropic stellar model.  The results are very temperature-sensitive: in hot neutron stars ($T>10^9$~K), the lowest-order $r$-modes are unstable, while in colder stars they are damped by viscosity.  (Similar results have also been obtained independently by Lindblom {\\em et al} (1998), working however to first order in rotation, which omits key terms needed for the viscosity calculation.) These calculations have a number of interesting astrophysical implications. First, the $r$-mode instability will spin down a newly born neutron star to a period close to the initial period inferred for the Crab pulsar, probably between 10 and 20~ms. Second, as an initially rapidly rotating star star spins down, an energy equivalent to roughly 1\\% of a solar mass is radiated as gravitational waves, which makes the process an interesting source for detectable gravitational waves. Third, the $r$-mode instability rules out the scenario whereby millisecond pulsars are formed by accretion-induced collapse of a white dwarf: the new star would be hot enough to spin down to much slower rates.  Stars with periods less than perhaps 10~ms must have been formed by spin-up through accretion in binary systems, where they remain colder than the Eddington temperature of about $10^8$~K. More accurate calculations will be required to define the limiting spin period more reliably, and we discuss the importance of the major uncertainties in the stellar models, in the initial conditions after collapse, and in the physics of cooling, superfluidity and the equation of state. \\endabstract ",
        "introduction": "One of the puzzles of pulsar astrophysics is that neutron stars seem to be formed with rather small initial spin rates. The best studied young pulsar is the Crab (PSR0531+21), whose initial period is estimated to have been about 19~ms (Glendenning 1996).  Even the recently discovered young 16 ms X-ray pulsar in the supernova remnant N157B (Marshall et al 1998) probably had an initial period no shorter than 6-9 ms (assuming a braking index typical of young pulsars). This contrasts with the shortest known period of a recycled pulsar of 1.6 ms, and with the theoretical lower limit on the period of about 0.5 to 2 ms, depending on the equation of state (Friedman et al 1989).  If angular momentum is conserved in the collapse event that forms the neutron star it would not be surprising to find young pulsars spinning at or close to the theoretical maximum.  In this paper we offer an explanation for the observed slow initial rotation rates by showing that, if a new neutron star has a period shorter than some critical value $P_f$, then within a year its period will increase to $P_f$.  During this spindown period it will lose angular momentum to gravitational radiation, through the growth of a radiation-driven nonaxisymmetric instability associated with the $r$-modes. The final rotation period $P_f$ depends on details of cooling rates, viscosity, and perhaps on initial conditions following collapse. For the simple models of neutron stars that we compute in this paper, this period is about 20 ms, consistent with the estimated initial spin of the Crab pulsar, given the various uncertainties in our calculation (see below). These uncertainties are, in fact, probably sufficient to include the presumed initial period of the recently discovered pulsar in N157B.  We will discuss this further later in the paper. The instability operates only in hot neutron stars ($T>10^9$~K); neutron stars that are spun up in accreting systems do not reach this temperature, and we do not expect the instability to put a strong constraint on their speeds.  An alternative explanation for the slow initial rotation of pulsars was recently suggested by Spruit and Phinney (1998).  They argue that pulsars are likely to form slowly rotating because magnetic locking between the core and the envelope of the progenitor star prevents the core from spinning rapidly. If this is the case, it has implications for the evolution of the angular momentum of the degenerate cores of pre-supernova giants, and it also reduces expectations that gravitational collapse may be followed by a strong burst of gravitational radiation. However, even if the mechanism suggested by Spruit and Phinney is operating and explains the slow rotation of most pulsars, we believe that  the 16~ms pulsar in N157B is clear evidence that there is a class of pulsars that form rapidly spinning (with periods shorter than say 10-20 ms). For these pulsars the r-mode instability should be relevant, and it is the rotation rate of such young short-period pulsars that we discuss below.  Besides explaining the slow rotation of young pulsars, the $r$-mode instability has another important consequence.  Millisecond pulsars (with periods shorter than $P_f$) can only reach their rotation speeds by being spun up again after they cool down.  The alternative model, that these stars are formed by accretion-induced collapse of a white dwarf, is inconsistent with this instability: the collapse would form a star hot enough to spin down within a year.  So the $r$-mode instability gives further weight in support of the standard picture that millisecond pulsars are a distinct class of recycled pulsars that acquired their angular momentum through accretion in binary systems.  The $r$-mode instability is a member of the class of gravitational-radiation-driven instabilities called CFS instabilities (Thorne 1987), after Chandrasekhar (1970), who discovered them, and Friedman \\& Schutz (1978), who showed that they are generic in rotating neutron stars.  The criterion for and mechanism of this instability are simple: if the pattern speed of the mode is forward-going as seen from far away, but backward-going with respect to the rotation of the star, then when the mode radiates away angular momentum the star can find a rotation state of lower angular momentum {\\em and} lower energy.  This allows the mode amplitude to grow.  For a mode with angular pattern speed $\\sigma$ in an ideal inviscid uniformly rotating star of angular velocity $\\Omega$, this means there will be an instability if \\begin{equation}\\label{eqn:criterion} 0<\\sigma<\\Omega. \\end{equation} The mechanism is much the same as that which drives the Kelvin-Helmholz instability, raising waves in the sea (Schutz 1983). Viscosity, however, acts on the mode in a different way and, if large enough, will damp out the instability.   Many families of oscillation modes of neutron stars contain members that, for sufficiently high stellar rotation rate, satisfy the instability criterion (\\ref{eqn:criterion}).   Until recently it was believed that viscosity in neutron stars stabilized all of them except in very limited regimes (Lindblom 1995). But the picture changed when Andersson (1998) pointed out that a hitherto unstudied family of modes, the $r$-modes, which are generalizations of the Rossby waves familiar to oceanographers, should also exhibit the instability.  He noticed  that,  uniquely among the families of neutron star normal modes, the $r$-modes satisfy the instability criterion (\\ref{eqn:criterion}) for any stellar rotation rate, no matter how small. He attempted a preliminary estimate of the growth rate of these modes, which showed that it might be interesting and was worth further investigation.  Friedman and Morsink (1998) then showed that the instability is generic: every $r$-mode is in principle unstable in every rotating star, in the absence of viscosity.  To show that the instability is interesting for neutron star evolution, however, it is necessary both to calculate accurately the growth rate of the mode due to gravitational radiation reaction and to estimate the damping caused by viscosity.  In this paper, and in independent work by Lindblom et al (1998), it is shown that the the $r$-mode instability is surprisingly effective.  In a rapidly rotating newly formed star, the mode grows rapidly and remains unstable during the time it takes the star to cool to about $10^9$~K, about 1 year using conventional cooling rates.  During this time the gravitational radiation emitted by the mode will remove most of the star's initial angular momentum, until it spins down to roughly the period $P_f$, where viscosity dominates gravitational radiation for any temperature.  The $r$-modes are unusual in several respects.  They are primarily perturbations of the velocity field in the star, with little disturbance in the star's density.  In a nonrotating star, they are neutral rotational motions.  In a rotating star, Coriolis effects provide a weak restoring force that gives them genuine dynamics.  The gravitational radiation that $r$-modes emit comes primarily from their time-dependent {\\em mass currents}.  This is the gravitational analogue of magnetic multipole radiation.  We believe this instability is unique among expected astrophysical sources of gravitational radiation in radiating primarily by gravitomagnetic effects, the same sector of relativistic gravitation theory that is responsible for the Lense-Thirring effect near rotating bodies and the Penrose process near spinning black holes (see, for example, Misner et al 1973). The present paper and Lindblom et al (1998) give the first calculations of the strength of gravitomagnetic radiation from stellar perturbations.  Like magnetic multipole radiation, gravitomagnetic radiation of any multipole is $(v/c)^2$ weaker than radiation from the corresponding mass (electric) multipole of the same amplitude, where $v$ is the typical velocity associated with the perturbation.  Despite this, the quadrupole ($l=2$) $r$-mode is more strongly unstable to gravitational radiation than any other mode so far examined in neutron stars, such as the density-dominated $f$-mode (Lindblom 1995).  The reason is that the $r$-modes are unstable in {\\em any} inviscid star, while the quadrupole $f$-mode is never more than marginally unstable in the most rapidly rotating stars, and the higher-order $f$-modes do not radiate any more strongly than the $r$-mode.  Our initial calculation of the strength and the importance of the $r$-mode instability necessarily involves approximations. The neutron stars are treated as Newtonian polytropes, the modes are treated in a slow-rotation approximation, damping effects due to viscosity are assumed uniform throughout the star, and the stars are assumed to be uniformly rotating. An important difference between our calculations and the independent ones of Lindblom et al (1998), which were completed somewhat earlier, is that we carry the calculation to a sufficiently high order in the star's angular velocity $\\Omega$ to fully calculate the small density perturbation associated with the mass motions of the $r$-modes.  This is important for estimating correctly the effects of bulk viscosity, which dominates the evolution of the spin of the star at high temperatures.  Reassuringly, despite this and other smaller differences, the conclusions of the two studies are broadly the same.  It seems unlikely to us that when more accurate calculations are performed the results will be markedly different qualitatively, but there could be quantitative differences in the limiting period and in details of the spindown behavior.  The evolution of the spin of the neutron star and the precise value of the limiting period $P_f$ depend also on the physics of neutron-star viscosity, cooling, and stratification.  We have adopted the best available models of cooling and viscosity (see the references in the body of this paper), but there are important uncertainties, particularly concerning the onset of superfluidity.  We have ignored stratification (differentiation of the crust from the interior, for example, or differential rotation of the interior).  These uncertainties can also have a significant effect on the value of $P_f$.  We hope that the fact that $P_f$ is in principle an observable, and that gravitational radiation from the spindown may eventually also be directly observed, will encourage further work on the critical physics issues here.  In a separate paper (Owen et al 1998), the implications of this instability for gravitational radiation are examined.  That paper contains a first attempt at developing a time-dependent model of the evolution of the instability, including the nonlinear phase where it is radiating away most of the angular momentum of the star in a steady way.  The conclusions are that (i) the radiation emitted by a young neutron star in the Virgo Cluster can be directly detected by the advanced versions of the LIGO (Abramovici et al 1992) and VIRGO (Bradaschia et al 1990) gravitational wave detectors now under construction, and (ii) the radiation emitted by all the neutron stars created since the beginning of star formation should form a random background of gravitational radiation that may be detectable by the same instruments.  Such detections could come as early as 2005.  We will then have direct means of investigating the physics that affects the evolution of this instability in neutron stars.  The plan of this paper is as follows.  In Section 2 we describe the perturbation formalism and the nature of the $r$-modes, and we calculate the effects of gravitational radiation, bulk viscosity, and shear viscosity on them.  The fundamental limiting period $P_f$ is derived in this section.  Section 3 draws our main astrophysical conclusions and discusses the chief physics uncertainties that affect the computation of the limiting period $P_f$. We examine various possible spindown scenarios, and identify the key uncertainties that can be addressed in future studies. ",
        "conclusions": ""
    },
    "9805/astro-ph9805013_arXiv.txt": {
        "abstract": "s{% Four Hubble diagrams are combined to test for the linearity of the cosmic expansion field. The expansion rate, $H_0$, is found to decrease by $\\sim5\\%$ out to $18\\,000\\kms$. Beyond this distance the mean value of $H_0$ is close to the value at $10\\,000\\kms$. The absolute value of $H_0$ is derived in two different ways. The one leads through Cepheids from {\\sl HST\\/} and blue SNe\\,Ia to $H_0=57\\pm7$ (external error) at $30\\,000\\kms$. The other uses various distance indicators to the Virgo cluster whose distance can be extended to $10\\,000\\kms$ by means of only relative cluster distances; the result of $H_0$ is closely the same. Independent distances from purely physical methods (SZ effect, gravitational lenses, and MBW fluctuations) cluster about $H_0\\;$(cosmic)$\\approx58$. } ",
        "introduction": "\\label{sec:1} At small scales the cosmic expansion field is highly disturbed. The negative recession velocity of the Andromeda nebula (M\\,31) is telling proof, as well as the local peculiar motion of $630\\kms$ relative to the MWB comprising a volume of radius $2000\\kms$ at least. The question therefore arises out to what distances one has to go that the observed expansion rate, i.\\,e.\\ the Hubble constant $H_0$, has truly cosmic significance. This problem is addressed in Section~2, where four Hubble diagrams are combined to test for the {\\em linearity\\/} of the expansion field in function of distance. In Section~3 distance determinations via Cepheid distances from Hubble Space Telescope ({\\sl HST\\/}) and blue SNe\\,Ia are pushed out to $30\\,000\\kms$. The distance of the Virgo cluster is derived in Section~4 using different methods, and this distance is transported to $10\\,000\\kms$ by means of relative cluster distances. Purely physical determinations of $H_0$ are compiled in Section~5. The conclusions are given in Section~6. ",
        "conclusions": "\\label{sec:6} Two separate routes to the Hubble constant require consistently $55<H_{0}<60$. The one is based on Cepheid-calibrated blue SNe\\,Ia and their Hubble diagram out to $30\\,000\\kms$. The other relies on the Virgo cluster distance from various methods and relative cluster distances out to $10\\,000\\kms$. The conclusion that $H_{0}\\,({\\rm cosmic})=57\\pm7$ (external error) is in perfect agreement with present physical distance determinations like the Sunyaev-Zeldovich effect, gravitational lenses, and the CMB fluctuation spectrum.  It is not understood in the face of this evidence why values of $H_0$ as high as $\\sim\\!70$ are still occasionally proposed (cf. \\citeauthor{Freedman:etal:97} 1997; \\citeauthor{Freedman:etal:98} 1998). Principal reasons are the neglect of the Malmquist correction for selection effects and the adopted too small Virgo cluster distance. The latter is based only on the Cepheid distance of M\\,100 \\cite{Ferrarese:etal:96}, which actually lies on the {\\em near side\\/} of the Virgo cluster (cf. Table~\\ref{tab:2}). The small Virgo distance is only seemingly supported by planetary nebulae \\cite{Ciardullo:etal:98} and the surface brightness fluctuation method \\cite{Tonry:97}, i.\\,e.\\ two distance indicators which have been included in Table~\\ref{tab:3} for the determination of {\\em relative\\/} distances, but have utterly failed for {\\rm absolute\\/} distances in several cases (cf. \\citeauthor{Tammann:98} 1998).  The determination of $H_0$ through blue SNe\\,Ia has presently the highest weight. Forthcoming additional calibrators with Cepheid distances from {\\sl HST\\/} will further strengthen the calibration. If {\\em more\\/} SNe\\,Ia will become available with redshifts up to $\\sim30\\,000\\kms$ and good photometry near maximum, they will lead to an even tighter Hubble diagram. In this way it will be possible not only to better map the variations of $H_0$ with distance, but also to push the error of $H_0$ well below $10\\%$.  \\mpspacing{0.9} \\bigskip \\noindent {\\small {\\bf Acknowledgement.} Support of the Swiss National Science Foundation is gratefully acknowledged.\\ The author thanks the hidden team in Baltimore who make $HST$ so successfully operational.\\ He also thanks Mr.~Bernd Reindl for computational help and the outlay of the manuscript.}  \\mpspacing{1.0} \\clearpage"
    },
    "9805/astro-ph9805363_arXiv.txt": {
        "abstract": "We present VLA radio maps in total intensity and polarization at 1.4, 4.9 and 8.4 GHz of fifteen 3CR radio galaxies for which good maps showing the large-scale radio structure have not previously been available. Previously unknown cores are detected in several sources and a bright one-sided jet in 3C\\,287.1 is mapped for the first time; several other jet-like features are also imaged. Total and core fluxes are tabulated and radio core positions are listed and compared to optical positions.  The galaxy at the optical position listed for 3C\\,169.1 is found to lie farther from the radio core position than another dimmer, bluer galaxy.  We discuss individual sources in some detail. ",
        "introduction": "This paper is the first in a series dealing with a comparison of the optical and radio properties of a sample of 3CR radio galaxies with $0.15 < z < 0.65$.  The sample was taken from the Revised 3C Catalog of Radio Sources of Smith, Spinrad \\& Smith (1976) as updated by \\cite{Spin85} and \\cite{Spinrad}.  In total, there are 66 objects in the sample and good quality radio maps showing the large-scale radio structure were found in the literature for 51 of these objects.  We observed the remaining 15 objects with the VLA.  In future papers, results of the radio-optical comparison will be presented.  Here we discuss the new VLA images that were obtained.  Listed in Table \\ref{sample} are the observed sources along with their optical positions, redshifts, 178 MHz flux densities on the scale of \\cite{Baars}, and spectral indices between 178 and 750 MHz.  (The spectral index $\\alpha$ is defined throughout in the sense $S \\propto \\nu^{-\\alpha}$.) The optical positions and redshifts are taken from \\cite{Spinrad}. The 178 MHz flux densities are from Kellermann, Pauliny-Toth \\& Williams (1969) except for 3C\\,268.2 and 3C\\,306.1 which are taken from \\cite{Pilk} and Gower, Scott \\& Wills (1967) (both taken from \\cite{NED}), respectively, because their values in \\cite{Keller} are likely contaminated with confusing flux.  All fluxes were adjusted to the scale of \\cite{Baars} using the corrections given in \\cite{LaiPea}.  The spectral indices are also consistent with the scale of \\cite{Baars}.  They were taken from Laing, Riley \\& Longair (1983) if available (3C\\,16, 3C\\,220.1, 3C\\,401), or computed using the 178 MHz flux densities given here and 750 MHz flux densities from \\cite{Keller} adjusted to the scale of \\cite{Baars} using the corrections given in \\cite{LaiPea}.  Also included in Table \\ref{sample} are the rest frame luminosity at 178 MHz, the scale, the largest angular size (measured from the best available maps), and the linear size of the source. A Friedmann cosmology with H$_0$ = 50 km s$^{-1}$ Mpc$^{-1}$ and q$_0$ = 0 was used in the calculation of these quantities and is assumed throughout this paper.  Table \\ref{radio-opt} contains the position of the radio core for each source, the frequency of the map used to determine the core position, the reference for the core position, other radio map references, the optical spectral type of the object (i.e. the ``optical class'') and the reference for this type.  The optical type was taken directly from \\cite{Jack} who use slightly different notation than that used here.  Their ``quasar/weak quasar'' (Q/WQ) classification is equivalent to our broad-line (B) type. Similarly, their ``high-excitation galaxy'' (HEG) corresponds to our narrow line (N) type and their ``low-excitation galaxy'' (LEG) is our low-excitation (E) type.  The references for the optical spectra on which these classifications are based can be found in \\cite{Jack}.  References giving additional spectral information that are not found in \\cite{Jack} are included in the Optical Reference column. ",
        "conclusions": ""
    },
    "9805/astro-ph9805155_arXiv.txt": {
        "abstract": "We present an extensive photometric and spectroscopic study of two high-redshift clusters of galaxies based on data obtained from the Keck 10m telescopes and the Hubble Space Telescope.  The clusters CL0023+0423 ($z=0.84$) and CL1604+4304 ($z=0.90$) are part of a multi-wavelength program to study nine candidate clusters at $z \\simgreat 0.6$ (Oke, Postman \\& Lubin 1998).  Based on these observations, we study in detail both the field and cluster populations. From the confirmed cluster members, we find that CL0023+0423 actually consists of two components separated by $\\sim$2900 km s$^{-1}$. A kinematic analysis indicates that the two components are a poor cluster with $\\sim$3 $\\times 10^{14}$ M$_\\odot$ and a less massive group with $\\sim$10$^{13}$ M$_\\odot$.  CL1604+4304 is a centrally concentrated, rich cluster at $z = 0.8967$ with a velocity dispersion of 1226 km s$^{-1}$ and a mass of $\\sim$3 $\\times 10^{15}$ M$_\\odot$.  A large percentage of the cluster members show high levels of star formation activity.  Approximately 57\\% and 50\\% of the galaxies are active in CL0023+0423 and CL1604+4304, respectively.  These numbers are significantly larger than those found in intermediate-redshift clusters (Balogh \\etal 1997). We also observe many old, red galaxies. Found mainly in CL1604+4304, they have spectra consistent with passive stellar evolution, typical of the populations of early-type galaxies in low and intermediate-redshift clusters. We have calculated their ages by comparing their spectral energy distributions to standard Bruzual \\& Charlot (1995) evolutionary models. We find that their colors are consistent with models having an exponentially decreasing star formation rate with a time constant of 0.6 Gyr.  We also observe a significant luminosity brightening in our brightest cluster galaxies. Compared to brightest cluster galaxies at $z \\sim 0.1$, we find a luminosity increase of $\\sim 1$ mag in the rest $M_B$ and $\\sim 0.8$ mag in the rest $M_V$.  In the field, we find that $\\sim 76\\%$ of the galaxies with $z > 0.4$ show emission line activity. These numbers are consistent with previous studies (e.g.\\ Hammer \\etal 1997).  We find that an exponentially decaying star formation rate is required to produce the observed amount of star formation for the majority of the galaxies in our sample. A time constant of $\\tau = 0.6$ Gyr appears to be optimal. We also detect several interesting galaxies at $z > 1$.  Two of these galaxies are extremely luminous with strong MgII$\\lambda$2800 absorption and FeII resonance line absorption. These lines are so strong that we conclude that they must be generated within the atmospheres of a large population of young, hot stars. ",
        "introduction": "The study of cluster galaxies at moderate look-back times of $50-70$\\% of the cosmic age ($0.7 \\simless z \\simless 1.5$) can provide important constraints on the nature and duration of the processes which have yielded the current epoch distribution of galaxy properties. This intermediate redshift range is likely to be a period where clusters are undergoing (or have recently completed) virialization and where many galaxies are only 1 or 2 Gyr past the peak in the cosmic star formation history (Madau, Pozzetti \\& Dickinson 1998). From a practical point of view, it is a regime which has only recently become readily accessible, both photometrically and spectroscopically, from space and ground-based optical and near-IR telescopes.  In the redshift range from $z = 0.3$ to $0.6$ there have been numerous photometric and/or spectroscopic studies of both optically and X-ray selected clusters (e.g.\\ \\cite{koo81}; \\cite{cou83}; \\cite{ell85}; \\cite{cou85}; \\cite{cou87}; \\cite{cou91}; \\cite{fab91}; \\cite{hen92}; \\cite{dres92}; \\cite{oke96}). Galaxies in these systems can be studied easily down to 5\\% of the characteristic luminosity and contamination by interlopers is not a significant problem.  A growing number of clusters with $z \\leq 0.5$ have been extremely well-studied.  With spectra of hundreds of cluster members, detailed modeling of the infall patterns and intracluster chemical composition gradients is now available (\\cite{abr96}; \\cite{elling97}).  The above studies indicate that the cores of rich clusters are typically dominated by a population of luminous early-type, red galaxies which produce a remarkably narrow ridge (or ``red locus'') in the color-magnitude (CM) relation. At $z \\simless 0.4$ both the mean color and the CM relation are consistent with those of present-day ellipticals (e.g.\\ Arag\\'on-Salamanca \\etal 1991; Dressler \\& Gunn 1992; Stanford, Eisenhardt \\& Dickinson 1994,1997).  The bulk of these galaxies have spectra which show no obvious signs of current or recent star formation. However, there is a non-negligible fraction which show post-starburst spectra with strong Balmer lines in absorption (e.g.\\ Gunn \\& Dressler 1988; Dressler \\& Gunn 1992; Poggianti 1997). In conjunction with this population is a fraction of blue cluster members which is increasing with redshift, a phenomenon known as the Butcher-Oemler effect (Butcher \\& Oemler 1984). Most of these galaxies appear as normal spirals or have peculiar morphologies (Dressler \\etal 1994; Couch \\etal 1994; Oemler, Dressler \\& Butcher 1997). A large fraction of this blue, spiral population exhibits exceptionally strong Balmer lines and/or [OII] emission which indicates that a significant fraction of the cluster members have recently undergone or are currently undergoing a high level of star formation activity (e.g.\\ Lavery \\& Henry 1988; Gunn \\& Dressler 1998; Lavery, Pierce \\& McClure 1992; Dressler \\etal 1994; Poggianti 1997).  Complimenting the ground-based work are several Hubble Space Telescope (HST) programs to quantify the morphology of galaxies in these clusters (e.g.\\ \\cite{cou94}; \\cite{dres94}; \\cite{sma97}; \\cite{oem97}; \\cite{ell97}). HST enables morphological classifications which can be made on scales of $\\sim 1~{\\rm kpc}$, thus providing a direct comparison to ground-based classifications of nearby galaxies. In addition to relating the photometric and spectral characteristics of a galaxy to its morphology, the HST studies can be used to examine the overall morphological distribution in these clusters. This work may indicate that the morphological composition of clusters is evolving with redshift (Dressler \\etal 1997; Oemler, Dressler \\& Butcher 1997), though these results are not yet certain (e.g.\\ Stanford, Eisenhardt \\& Dickinson 1997; Lubin \\etal 1998). These evolutionary changes are also apparently reflected in the evolution of the morphology--density relation (\\cite{dres80}; \\cite{pg84}). This relation in intermediate-redshift clusters which are centrally-concentrated and compact is qualitatively similar to that in the local universe; however, unlike present-day clusters, the relation is non-existent in the loose, open clusters (Dressler \\etal 1997).  Studies of clusters of galaxies with redshifts greater than $z = 0.6$ are substantially more difficult because (1) the galaxies are approaching the sensitivity limits of optical spectrographs on 4m class telescopes, (2) the interloper contamination becomes substantial, and (3) the well understood rest wavelength region redward of 4000\\AA\\ moves into the near infrared.  Despite these difficulties, several studies of high-redshift clusters have been made. These studies indicate that the Butcher-Oemler effect continues to strengthen up to $z = 0.9$ (Arag\\'on-Salamanca \\etal 1993; Rakos \\& Schombert 1995; Lubin 1996).  In addition, the red envelope of the early-type cluster population moves bluewards with redshift. At $z \\sim 0.9$ there are few galaxies with colors as red as present-day ellipticals (Arag\\'on-Salamanca \\etal 1993; Rakos \\& Schombert 1995; Oke, Gunn \\& Hoessel 1996; Lubin 1996; Ellis \\etal 1997; Stanford, Eisenhardt \\& Dickinson 1995,1997). This color evolution is consistent with passive evolution of an old stellar population formed at an early cosmic age.  The amount of color evolution is similar from cluster to cluster at a given redshift and is independent of the cluster richness or X-ray luminosity. These results indicate that the history of early-type galaxies may be insensitive to environment; that is, these galaxies appear to be coeval with a common star formation history (Bower \\etal 1992a,b; Arag\\'on-Salamanca \\etal 1993; Dickinson 1995; Stanford, Eisenhardt \\& Dickinson 1995,1997; Ellis \\etal 1997).  There have also been many studies of high-redshift field galaxies. Hamilton (1985) obtained spectra of 33 very red, field galaxies in the redshift range of $z = 0.2$ to $z = 0.8$.  He found that the 4000 \\AA\\ break changed by less than 7\\% over the observed $z$ range.  Songaila \\etal (1994) obtained $BVK$ photometry and spectra of a nearly complete sample of 298 galaxies.  The redshifts are nearly all at $z < 1.0$. They find no $K$--band luminosity evolution. From measurements of emission-line strengths and the 4000 \\AA\\ break, they infer that galaxies are undergoing significantly more star formation at $z=1$ than at the present epoch.  The Canada-France-Redshift-Survey (CFRS) group has used CFHT to carry out a very extensive spectroscopic survey out to redshifts of 1.3 (\\cite{CFRS}, \\cite{ham97}).  They find that the fraction of galaxies with significant emission lines (EW of [OII] $> 15$ \\AA) increases from about 13\\% locally to over 50\\% at $z >0.5$.  The fraction of luminous, quiescent galaxies (no significent [OII] emission) decreases with redshift from 53\\% at $z=0.3$ to 23\\% for $z>0.5$.  They also find evidence that the metal abundance is lower in emission-line galaxies at high redshifts than locally.  In addition, Cohen \\etal (1996a,b) have obtained spectra of high redshift galaxies in the fields of both the Hubble Medium Deep Survey and the Hubble Deep Field survey. They find that the redshifts are highly clumped; the velocity dispersion in these clumps are similar to those found in local groups of galaxies.  Further supporting this observational evidence of strong velocity structure at high redshift, Koo \\etal (1996) carried out photometry and spectroscopy of 35 galaxies with redshifts of 0.3 to 1.6.  They found that half of the redshifts in their sample are actually in two structures at $z=0.81$ and $z=1.0$.  Because studies of both field and cluster galaxies indicate that the high-redshift universe is a place of substantial evolution, we have undertaken an extensive program to study nine candidate clusters of galaxies at $z \\simgreat 0.6$ (Oke, Postman \\& Lubin 1998; hereafter Paper I). With the commissioning of the Keck 10 meter telescopes, this detailed survey is now possible. The first paper in the series describes the sample selection, data acquisition, and data reduction procedures of the survey (Paper I). In this paper, we present our analysis and interpretation of the spectra and photometry for the first two clusters to be completed, CL0023+0423 ($z = 0.84$) and CL1604+4304 ($z = 0.90$). In the third paper of this series (Lubin \\etal 1998; hereafter Paper III), the HST observations and the resulting morphological composition of these two clusters are described. ",
        "conclusions": "In this paper, we have presented the results from a deep photometric and spectroscopic survey of two fields which are centered on the candidate, high-redshift clusters of CL0023+0423 at $z = 0.84$ and CL1604+4304 at $z = 0.90$.  Based on these observations, we conclude the following :  \\newcounter{discnt}  \\begin{list} {\\arabic{discnt}.}  {\\usecounter{discnt}}  \\item We have spectroscopically observed magnitude-limited (R $\\le 23.3$) samples of two fields with candidate clusters of galaxies at $z > 0.8$.  Both candidates are confirmed as real cluster-like density enhancements, but only $\\sim20$\\% of the observed galaxies are actually members of the clusters. The large foreground/background contamination is a common hazard when working in this redshift range and implies that one requires a minimum of $\\sim 100-150$ redshifts per 15 square arcminutes if accurate kinematic parameters are desired.  \\item We find that the main cluster in the CL0023+0423 field consists of two components at nearly identical redshifts of $z = 0.8274$ and $0.8453$ (corresponding to a radial velocity difference of $\\sim 2900~{\\rm km~s^{-1}}$).  Their calculated velocity dispersions and corresponding masses indicate that these two components are more similar to groups or poor clusters of galaxies. In addition, the morphological analysis of Paper III indicates that these systems have a morphological composition which more closely resembles that of the field.  The fraction of spiral galaxies is 66\\% or more.  Though it is possible that these two groups are simply a chance projection on the sky, they may also be in the process of merging. The dynamics of the merging scenario is presented in another paper of this series (Lubin, Postman \\& Oke 1998).  \\item We have confirmed that there is a centrally-concentrated, rich cluster at $z=0.8967$ in the CL1604+4304 field. The velocity dispersion and implied mass are consistent with an Abell richness class 2 cluster.  In addition, the morphological composition of CL1604+4304 is characteristic of a normal, present-day rich cluster. Early-type galaxies comprise 76\\% of the galaxies in the central $\\sim$0.5 $h^{-1}~{\\rm Mpc}$.  In this population, the ratio of S0 to ellipticals is $\\sim$1.7, consistent with galaxy populations found in local clusters (Paper III).  \\item The mass-to-light ratios in both clusters are similar to those seen locally but the uncertainties are high due to the sensitivity of the derived luminosity to the amount of color and luminosity evolution.  Our k-corrected $R$-band M/L values are, none the less, significantly less (by $3 - 4\\sigma$) than the zero redshift $R$-band M/L of $\\sim1300$ expected for an $\\Omega = 1$ universe.  \\item Defining active star formation as a rest equivalent width of [OII] $\\lambda3727$ which is greater than 15 \\AA, 79\\% of the field (non-cluster) galaxies with $0.4 \\le z < 0.9$ are active, while 62\\% with $z \\ge 0.9$ are active.  The fraction of field galaxies which have active star formation appears to be relatively independent of redshift.  Of cluster members, 57\\% and 50\\% are active in the CL0023+0423 and CL1604+4304, respectively. These numbers are significantly higher than those of intermediate-redshift clusters (Balogh \\etal 1997). Even though the cluster masses are very different in our two cases, the fractions of active galaxies are approximately the same; this suggests that star formation activity may not be a strong function of the cluster mass.  \\item The traditional estimator D(4000) of the 4000 \\AA\\ break has been supplemented by a measurement $J$ which is color independent, measures either the Balmer jump or the 4000 \\AA\\ break, and is able to distinguish between the two.  For blue objects which have a Balmer jump, the measured jump is smaller than that predicted by standard Bruzual \\& Charlot models.  For the old, red galaxies our measurements of D(4000) are consistent with the measured colors of the galaxies. Because changes in D(4000) are accompanied by compensating changes in color when the metallicity is varied, it is not possible to derive the metallicity from our data.  \\item We have compared our photometric and spectroscopic observations with various families of Bruzual \\& Charlot evolutionary models.  The equivalent widths of the [OII]$\\lambda3727$ line show that models with a single burst of star formation followed by passive evolution (ssp models) are in most cases not satisfactory.  However, models with an exponentially decreasing star formation rate after time zero (tau models) do predict the correct equivalent widths of [OII]; a time constant of 0.6 Gyr appears to be optimum.  Our parameter $J$ also indicates that an exponentially decaying star formation model is better than the assumption of a single, monolithic burst.  \\item Fitting the observed spectra to the tau0.6 synthetic models indicate that the red galaxies, found mainly in the CL1604+4304 cluster at $z=0.8967$, have ages of $\\sim 5$ Gyr for an assumed solar metallicity.  For a cosmology of $H_0=65$ and $q_0=0.1$, this derived age allows nearly 1 Gyr between the time of creation and the first onset of star formation.  Of the other Bruzual \\& Charlot models that we have examined, the tau1.0 models, which are otherwise acceptable, predict ages for these early-type galaxies which appear to be too old for this cosmology. The ages derived using ssp models, which we have shown are not generally appropriate for other observational reasons, give ages which are only $\\sim 3$ Gyr. The morphological analysis presented in Paper III reveals that these old galaxies belong to the Hubble class of spheroid or early spiral galaxies. The derived ages are consistent with previous studies of high-redshift, early-type cluster members (Dickinson 1995; Stanford, Eisenhardt \\& Dickinson 1997).  \\item Comparing the brightest cluster galaxies in our clusters with those in Abell clusters of similar richness at redshifts of $z \\sim 0.1$, we find a luminosity brightening of 1.02 mag for rest $M_B$ and 0.85 mag for rest $M_V$.  For the same cosmological parameters above, the tau0.6 models predict a brightening of 0.8 and 0.7 mag for $M_B$ and $M_V$, respectively.  Increasing $q_0$ from 0.1 to 0.5 increases the predicted brightening to 1.1 and 0.9 mag, respectively.  At a given $q_0$, increasing $H_0$ also slightly increases the predicted brightening.  Unlike the vast majority of their low-redshift counterparts, these two BCGs are disk systems (early-type spirals).  \\item Most of the field galaxies at $z > 0.9$ have absolute luminosities which are not brighter than the predicted luminosity of a brightest cluster galaxy at the same redshift; however, there are two interesting exceptions. One galaxy at $z=1.5029$ in the CL0023+0423 field and one galaxy at $z=1.3757$ in the CL1604+4304 field are overluminous by nearly 1 mag compared with what is expected for brightest cluster galaxies at these redshifts. These galaxies have strong MgII$\\lambda2800$ absorption and FeII resonance line absorption.  Although one argument suggests that in other galaxies these lines are mainly interstellar, they are too strong in these two galaxies for this to be the case.  Therefore, they presumably come from late B-type stars in the galaxies.  \\end{list}  \\vskip 0.5cm"
    },
    "9805/hep-th9805081_arXiv.txt": {
        "abstract": "I show that de~Sitter space disintegrates into an infinite number of copies of itself. This occurs iteratively through a quantum process involving two types of topology change. First a handle is created semiclassically, on which multiple black hole horizons form.  Then the black holes evaporate and disappear, splitting the spatial hypersurfaces into large parts.  Applied to cosmology, this process leads to the production of a large or infinite number of universes in most models of inflation and yields a new picture of global structure. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805141_arXiv.txt": {
        "abstract": "We report results of a near--infrared imaging survey of L977, a dark cloud in Cygnus seen in projection against the plane of the Milky Way. We use measurements of the near--infrared color excess and positions of the 1628 brightest stars in our survey to measure directly dust extinction through the cloud following the method described by Lada et al.  (1994).  We spatially convolve the individual extinction measurements with a square filter 90$^{\\prime\\prime}$ in size to construct a large--scale map of extinction in the cloud.  We integrate over this map to derive a total mass of $M_{L977} = (660 \\pm 30) \\ (D/500 pc)^2$ M$_\\odot$ and, via a comparison of source counts with predictions of a galactic model, estimate a distance to L977 of 500 $\\pm$ 100 pc.  We find a correlation between the measured dispersion in our extinction determinations and the extinction which is very similar to that found for the dark cloud IC 5146 in a previous study. We interpret this as evidence for the presence of structure on scales smaller than the 90$^{\\prime\\prime}$ resolution of our extinction map.  To further investigate the structure of the cloud we construct the frequency distribution of the 1628 individual extinction measurements in the L977 cloud.  The shape of the distribution is similar to that of the IC 5146 cloud.  Monte Carlo modeling of this distribution suggests that between 2 $<$ A$_V$ $<$ 40 mag (or roughly 1 $<$ $r$ $<$ 0.1 pc) the material inside L977 is characterized by a density profile $\\rho \\,(r) \\propto r^{-2}$.  Direct measurement of the radial profile of a portion of the cloud confirms this result.    At the lower galactic latitude of L977, we find both the mean and dispersion of the infrared colors of field stars to be larger than observed toward IC 5146.  This produces an increase of about a factor of 2 in the minimum or threshold value of extinction that can be reliably measured toward L977 with this technique.  Nevertheless the accuracy in an extinction map pixel is not significantly different toward L977 due to the increased number of field stars at this latitude.  We also find an increase in the number of detected giant stars at the lower galactic latitude of the survey by almost a factor of two.  Most of these excess stars suffer extraneous extinction and are probably red giants seen along the disc of the Milky Way up to distances $\\sim$ 15 kpc and reddened by unrelated background molecular clouds along this direction of the Galaxy.  We discuss a possible application of this observable to galactic structure studies on the plane of the Galaxy. ",
        "introduction": "Knowledge of the distribution of mass within molecular clouds is critical for the understanding of how molecular clouds form, evolve, and ultimately produce stars and planets.  Dust is the most reliable tracer of the undetectable molecular hydrogen gas, which is the primary component of the mass of molecular clouds.  The most straightforward way to measure the dust content of a molecular cloud is through direct measurements of dust extinction of background starlight.  This can be optimally done at near infrared wavelengths where the dust opacity is modest and the reddening law appears universal (Jones \\& Hyland 1980; Cardelli et al.  1989; Martin \\& Whittet 1990; Whittet et al.  1993).  The determination of gas mass from dust extinction is relatively straightforward and relies on basically one observationally well established assumption:  a constant gas--to--dust ratio (Lilley 1955; Jenkins \\& Savage 1974; Bohlin et al.  1978).   Recently, Lada et al.  (1994) demonstrated that large scale near-infrared imaging surveys of dark clouds can directly trace the distribution of dust extinction and column density through a molecular cloud at considerably higher angular resolution and greater optical depth than previously possible from pure star counts (Bok 1937).  In fact, Lada et al.  (1994) developed a method for measuring and mapping extinction which is more direct and fundamentally more powerful than star counts made at any wavelength.  Their method uses:  1) measurements of infrared (H[1.6$\\mu$m]$-$K[2.2$\\mu$m]) color excess, 2) knowledge of the near--infrared extinction law, and 3) techniques of star counting to derive directly and map extinction through a cloud.  Also, since the method is based on pencil beam measurements of the extinction, it can give information about both the large--scale and small--scale structure of a molecular cloud.   We have two main motivations for the present paper.  First, we want to apply the Near--Infrared Color Excess method used by Lada et al. (1994) (hereafter, NICE method) to a different molecular cloud to compare its physical conditions with those derived for IC 5146. Second, because the information on the spatial distribution of dust derived from this method is a function of both the surface density and colors of stars detected through the target cloud, we want to test the applicability of the method against rich background star fields at low galactic latitudes.   L977 \\footnote {Two dark clouds from the updated Lynds catalogue (Lynds 1962; Nagy 1979) fall inside the Near--Infrared (NIR) survey presented in this paper:  L977 and L981.  We choose L977 as the representative cloud since it has the largest  area (0.200 vs. 0.019 deg$^2$).  In the original version of the catalogue this cloud has the number 978.}  is part of dark globular filament GF7 in the Cygnus complex (Schneider \\& Elmegreen 1978) and lies against a rich background of field stars towards the warped plane of the Galaxy ($l = 90$\\deg, $b = 2$\\deg).  There is no obvious sign of ongoing star formation in this cloud.  Although three IRAS point sources lie in the direction of L977, their colors are indicative of infrared cirrus emission rather than pre--main sequence objects.  This dark cloud is in the Dobashi et al.  (1994) $^{13}$CO large scale survey of the Cygnus region as an isolated cloud and free from confusion with other molecular clouds.  Apart from this survey L977 is virtually unstudied.  We describe the acquisition and reduction of the observations in Section 2,  the results in Sec. 3,  the analysis and discussion in Sec.  4, and we summarize our conclusions in Sec. 5 of the paper. ",
        "conclusions": "\\subsection{Application of the NICE Extinction Determination}  The line--of--sight extinction to an individual star can be directly determined from knowledge of its color excess and the extinction law. The infrared color excess is proportional to dust extinction and thus column density and can be directly derived from the observations through:  \\[ E(H-K)\\  = \\ (H-K)_{observed} \\ - \\  (H-K)_{intrinsic} \\]  \\noindent The intrinsic $(H-K)$ color of main--sequence and giant stars have a narrow enough range (Koornneef 1983, Bessel \\& Brett 1992) that even without further information a reasonable determination (1$\\sigma \\sim 2.5$ mag of visual extinction) of the line--of--sight extinction (e.g., A$_K$) towards any significantly extincted star can be made.  This determination can be considerably improved through observations of control fields around the target cloud since we expect the colors of field stars to be dominated primarily by K--giants, assuming the differential reddening is small.  The infrared color excesses derived in this way can then be converted to an extinction A$_\\lambda$ at a given wavelength $\\lambda$, via the extinction law.   We start with the assumption that stars observed in our off cloud control fields can serve as a surrogate for the stellar population directly behind the cloud.  To test the validity of this assumption we can search for possible color variations across the background by comparing our control fields on both sides of L977.  The positions of our control fields were carefully selected to be close to the molecular cloud but distant enough to be free from contamination by it (see Figure 1).  Although the two fields are 1\\deg\\ apart in the sky (and almost 1\\deg\\ apart in Galactic latitude) our photometry (Figure 3) shows no sign of a color variation.  Both fields have virtually the same mean $(J-H)$ and $(H-K)$ colors, dispersions, and similar spatial distributions on the color--color plane.  The angular size scale for color variations on the near-infrared colors of field stars must then be greater than 1\\deg\\ in this direction of the Galaxy and we can assume with a high level of confidence that the $(H-K)$ colors of stars behind the cloud are well represented by the $(H-K)$ colors of stars in the control fields.  The mean value of the latter can then be taken more securely to characterize the typical intrinsic $(H-K)$ color of stars behind the cloud:  \\[(H-K)_{intrinsic}\\  \\equiv \\ <(H-K)>_{control}\\  = \\ 0.20 \\pm 0.13 \\ \\rm mag \\]  \\noindent where the quoted uncertainty is the dispersion in $(H-K)$ control field colors.  We used the above equations to derive individual color excesses to the 1628 stars observed through L977.  We converted the individual measurements to equivalent visual extinctions, A$_V$, using a standard reddening law (Rieke \\& Lebofsky 1985):  \\[ A_V = 15.87 \\times E(H-K) \\]  Although the individual A$_V$'s derived in this manner are proportional to the true dust column density along the line--of--sight to each star they only accurately reflect the true visual extinctions as long as the assumed reddening law is appropriate for this cloud. Grain growth in cold clouds alters the extinction law at wavelengths $\\ll 1\\mu$m.  Conversion to a near--infrared extinction (e.g., A$_K = 1.76$ E$(H-K)$) would avoid this problem since the NIR reddening law does not vary significantly with grain growth in dark clouds (e.g., Mathis 1990).  However, to facilitate comparison with previous work we follow convention and express our dust opacities in terms of equivalent visual extinctions.  Because of dust extinction, stars that will be detected through the increasingly denser regions of L977 can only be the brighter background stars.  It is important then to know how the mean $(H-K)$ color and dispersion will vary with $K$ brightness as this variation might affect the correct determination of extinction from color excesses.  To search for such possible variation we present in Figure 5 the color--magnitude diagram for all the stars in the control fields.  The open circles represent sources with $(J-K) < 0.5$ mag (the blue group, described in Sec.  3.1) and the filled circles represent sources with $(J-K) > 0.5$ mag (the red group).  To evaluate how the mean $(H-K)$ color and dispersion vary as a function of $K$ brightness we binned the data in $K$ (2 magnitudes bins) and determined the mean $(H-K)$ color and dispersion for each bin.  This mean $(H-K)$ color and dispersion is represented in Figure 5 as the open squares and the error bars, respectively.  We conclude that while the dispersion is essentially the same for all bins there is a slight increase in the mean $(H-K)$ color from fainter to brighter magnitudes.  This is caused by the larger number of red sources (most likely red background giants; see Appendix) that populate the bright end of the color--magnitude diagram.  Nevertheless, this increase ($\\sim$ 0.05) is less than half the average dispersion ($\\sim$ 0.13) and has a negligible effect on the determinations of extinction.  In Figure 6 we plot the frequency distribution of A$_V$'s derived toward L977.  For comparison we also plot the distribution of A$_V$'s for IC 5146 using the data of Lada et al.  (1994).  The systematic offset between both distributions is caused by the richer background of L977 at lower galactic latitudes.  The shape of these distributions is a function of both the detailed distribution of matter within the cloud and the detection limits of the infrared survey at high extinctions.  The distributions in both clouds fall off steeply with increasing extinction in a fashion roughly consistent with a power--law.  This indicates that the amount of material at high extinctions is relatively small in both these filamentary dark clouds. At high extinctions the steepness of the distribution is enhanced by the inability to detect heavily reddened faint stars imposed by the sensitivity of the NIR survey.  Since the sensitivity of both surveys is the same, the similarity of the shapes of the extinction frequency distributions suggests that both L977 and IC 5146 are characterized by a similar intrinsic structure, i.e., a comparable proportion of high column density to low column density gas.  In Section 4.5 of this paper we will model this frequency distribution for various clouds configurations.   To characterize the large--scale global structure of the cloud we spatially convolve the individual pencil--beam measurements (which represent a random spatial sampling of the cloud extinction) with a square filter 90$^{\\prime\\prime}$ in size and produced an ordered map of the extinction uniformly sampled at the Nyquist frequency.  The map is shown in Figure 7 superposed on a digitized POSS red plate of the region.  The 1$\\sigma$ confidence level on the derived extinction measurements toward individual stars is set by the rms dispersion in the $(H-K)$ color of background stars measured in the control fields (Figure 3) and is equal to $\\sim$ 2 magnitudes of visual extinction.  The 1$\\sigma$ confidence level on a map pixel is nevertheless $\\sim$ 1 magnitude (see discussion in Section 4.6). Stars suspected to be foreground to the cloud (see next section) were removed from the database.  Contours start at 4 visual magnitudes of extinction (4$\\sigma$) and increase in steps of 2 magnitudes up to 24 magnitudes of visual extinction.  The discontinuity near the bright star SAO 050355 is due to lack of one SQIID field in our data.  We decided not to take this field as a precautionary measure since this SAO star, a M3 star (Neckel 1974) with an estimated $K=$ 3.8 mag, would potentially affect the SQIID detector.  The overall shape of the map correlates well with the shape of the more opaque regions of the molecular cloud (see Figure 7).     \\subsection{Distance Determination}  The only reference in the literature regarding the distance towards L 977 appears in Dobashi et al.  (1994) $^{13}$CO survey.  These authors estimate a distance of 800 pc from association to nearby astronomical objects with known distances.  We used the Wainscoat et al.  (1992) infrared model of the Galaxy to determine the expected number of foreground stars seen toward the opaque regions of L977 and independently determined a distance to this cloud.  The molecular cloud can be used as a wall to discriminate between foreground/background stars as foreground stars are easily detected against the regions of high extinction.  We searched for these stars in regions that presented an average extinction above 8 magnitudes (the central region of the L977 filament, $\\sim$ 36 arcmin$^2$) and compared the number of stars with derived extinctions A$_V < 2$ mag (5 stars) with the expected number of foreground stars from the galactic model.  The $K-$band surface density of stars of the control fields was used to calibrate the galactic model.  In Figure 8 we present the expected number of foreground stars toward the denser (A$_V >$ 8 mag) regions of L977 as a function of distance.  We estimate a distance to L977 molecular cloud of 500 $\\pm$ 100 pc.  Extra information regarding a lower limit to the distance to L977 can be obtained from studied bright stars clearly foreground to the cloud. There are two such bright stars seen in projection against the molecular cloud complex:  \\begin{enumerate} \\item Less than 1\\deg\\ North of L977 (in front of L988, a molecular cloud associated in projection and velocity with L977) lies SAO 033091 (HR 8072), a K0III star with a visual magnitude of m$_V = $ 6.37 (Breger 1968, Strassmeier et al.  1994).  This star was observed with the astrometric satellite HIPPARCOS (Perryman et al.  1997) and an accurate distance of 119$\\pm$8 pc was derived from trigonometric parallax (a spectrophotometric parallax would have given a distance $\\sim$ 150 pc).  The distance determined by HIPPARCOS can be taken as a conservative lower limit of the distance to L977.  \\item The second bright star, not as well studied as SAO 033091, is seen in projection against L977 (SAO 050355, see Figure 1) .  It was classified as a M3 star by Neckel (1974) but there is no information available on its luminosity class.  If we assume a main-sequence, giant, or supergiant luminosity we can determine a distance of d$_V = \\,\\,\\sim 8$ pc, d$_{III} = \\,\\,\\sim 580 $ pc, or d$_I = \\,\\,\\sim 4800$ pc respectively to this star.  In any case the derived extinction is negligible which puts this star in the foreground of L977.  We can readily dismiss the derived distance of 4800 pc:  1) at this distance we would be seeing this star at least through two of the Galaxy's spiral arms ($l = 90$\\deg, $b = 2$\\deg) which is not compatible with the negligible extinction measured towards this star and 2) at this distance there would be no contrast to distinguish the shape of L977 against the background field due to foreground stars which is clearly not the case.  The other two derived distances cannot be dismissed but while d$_V = \\,\\,\\sim 8$ pc would bring no further constraints to this discussion, the one determined for the giant star case (580 pc) is still consistent, within the uncertainties, with a distance of 500 $\\pm$ 100 pc derived from the galactic model.  \\end{enumerate}  The uncertainty associated with the distance determination will be the largest source of systematic error in the derived properties of L977. We will assume 500pc as the distance to L977 throughout this paper.   \\subsection{Derived Mass}   A precise mass determination (and accurate if not for the uncertainty associated with the distance determination) can then be obtained for L977 through spatial integration of the extinction map in Figure 7. Using the standard gas--to--dust ratio (i.e., N(H + H$_2$) = 2 $\\times$ 10$^{21}$ cm$^{-2}$ mag$^{-1}$; Lilley 1955, Bohlin et al.  1978), and typical mass fractions ($\\sim$ 63\\% H$_2$, $\\sim$ 36\\% He, $\\sim$ 1\\% dust; e.g., Rohlfs \\& Wilson 1996) we derive a mass of the surveyed area of the cloud:  $M_{L977} = (660 \\pm  30) \\ (D/0.5 kpc)^2\\ $ \\msun\\ where $D$ is the true distance to this molecular cloud.       \\subsection{The $\\sigma_{disp}$ versus A$_V$ Relation}  Our ordered map of the distribution of extinction in L977 provides useful information about the overall large--scale structure and distribution of mass in the cloud.  However this is done at the expense of angular resolution.  Since each extinction measurement is made along an individual pencil--beam through the cloud the observations represent a random, but highly undersampled map of extinction made at infinitely high angular resolution.  One of the most interesting results of the first application of the NICE method to a dark cloud was the demonstration that useful information concerning cloud structure on small angular scales could be derived from the randomly sampled extinction measurements.  Lada et al.  (1994) showed that the relation between $\\sigma_{disp}$, the dispersion of extinction measurements within a square map pixel, and A$_V$, the mean extinction derived for the map pixel, can be used to characterize cloud structure on scales smaller than the resolution of the map (i.e., the size of the map pixels).  In IC 5146 Lada et al.  (1994) found that $\\sigma_{disp}$ increased in a systematic fashion with increasing A$_V$.  At the same time the dispersion in this relation was also found to increased with A$_V$.  Through modeling, this relation was found to be related to the form of the distribution of the small--scale cloud structure.  In Figure 9 we present this same relation for L977.  The same trend observed in IC 5146 is found for L977.  A least--squares fit over the entire data set returns:  \\[\\sigma_{disp} \\:\\, = 1.93 \\pm 0.11 +(0.40 \\pm 0.02)\\,A_V \\,\\,\\,\\, (mag) \\]  \\noindent The slope of the relation is virtually the same  as that found in the IC 5146 study ($0.40 \\pm 0.01$).  The increase in the value of the intercept (from 0.73 in IC 5146 to 1.93 in L977) is solely related to the higher dispersion in the colors of background stars in the present study (Lada et al.  1994).  We note, however, that this new data set (L977) is not as robust as the IC 5146 data set since there are not as many measurements at high extinction.  This is related to the fact that the total area surveyed in the L977 study is smaller than that in IC 5146.  Apparently, the same general behavior of the $\\sigma_{disp}$ versus A$_V$ relation is observed for L977.  As was shown by Lada et al.  (1994), this indicates that significant structure must be present down to scales smaller than the extinction map resolution, and moreover the dust cannot be distributed uniformly or in discrete high--extinction clumps on scales smaller than our resolution (0.22pc).  In a recent paper, Padoan, Jones, \\& Nordlund (1997) found that the form of the observed $\\sigma_{disp}$ versus A$_V$ relation is consistent with cloud structure models characterized by supersonic random motions.  However, it is not clear whether the shape of the $\\sigma_{disp}$ versus A$_V$ relation is due to random spatial fluctuations in the cloud structure or to systematic, unresolved gradients in the distribution of extinctions (e.g., Thoraval, Boiss\\'e, \\& Duvert 1997).  Higher sensitivity and angular resolution observations are needed to resolve this issue.    \\subsection{Modeling and Density Structure}  It is possible to predict the frequency distribution of detected background stars through the target cloud as presented in Figure 6. The decrease in the observed number of background stars for increasing cloud depth results from the combined effect of two causes:  1) extinction of the faintest stars in the background luminosity function to below the limits of detection and 2) the large--scale distribution of mass, as traced by dust extinction, inside the molecular cloud. The existence of a rising luminosity function determines that for the same detection limit only the intrinsically brighter (and fewer) stars can be detected as dust extinction (e.g., A$_V$) increases.  The way dust extinction is distributed inside the cloud will also affect the final shape of the observed frequency distribution.  We constructed Monte Carlo models with different cloud configurations to mimic the infrared data and predict the shape of the frequency distribution of stars detected through the cloud.  Because of the cloud's filamentary nature we adopted a cylindrically symmetric synthetic cloud with a finite radius $R$.  For every model the distribution of mass inside the model cloud was characterized by a density profile of the form $\\rho \\,(r) = \\rho_0 ({r\\over r_0})^{-\\alpha}$ with $\\alpha =$ 0, 1, 2, 3, and 4 where $r$ is the orthogonal distance from the major axis of the cylindrical cloud.  The density profile was integrated to produce a column density (or visual extinction) profile for the synthetic cloud, i.e.:   \\[ A_V(p) = 10^{-23} \\times \\rho_0r_0^\\alpha \\times \\int\\limits_{0}^{S} (s^2 + p^2)^{-{\\alpha\\over 2}}ds\\  \\,\\,  (mag)\\]  \\noindent where  \\[S = (R^2 - p^2)^{1\\over2}\\]  \\noindent and where $p$ is the projected distance of the line--of--sight from the major axis of the cloud, $s$ is the distance through the cloud along the line--of--sight and $\\rho_0$ and $r_0$ the appropriate scaling constants for the density distribution. This column density profile was then scaled to fit the data: the total radius of each cloud model was taken as the one that best reproduced the observed contrast from the highest measurements of extinction (A$_V \\sim 40$ mag, at $\\sim$ 40$^{\\prime\\prime}$ from the center of the filament) to the lowest (1$\\sigma$) measurements (A$_V \\sim 2$ mag, at $\\sim$ 275$^{\\prime\\prime}$ from the center of the filament).  We then constructed synthetic luminosity functions for the background population from randomly sampling the $K$ luminosity function of the Control Fields.  Stars from the synthetic background luminosity function were randomly assigned a projected distance from the center of the cloud.  This assumes a uniform spatial distribution of background stars which is supported by the similarity between the East and West Control Fields (see Section 3).  Next, the corresponding extinction A$_V$, for a given density structure model, was assigned to each star.  Noise, simulating extinction measurement errors (see next Section) was also added to each star's extinction value.  The extincted brightness of each star was then calculated.  Stars whose extincted brightness was less than the limiting sensitivity of our SQIID survey were rejected from further consideration.  One realization was completed when the total number of extincted detectable stars (with A$_V \\le 2$) reached the observed number:  889. To mimic the effects of the foreground stellar population we randomly assigned a zero extinction (plus noise) to five stars (see Section 4.2) in each realization.  A total of 1000 realizations for each cloud model ($\\alpha =$ 0, 1, 2, 3, and 4) were performed.  To compare the different model clouds and the observed frequency distribution the resulting extinction (unbinned) distribution from each individual realization was compared to the observed (unbinned) distribution via the KS test.  In Figure 10 we present the comparison between the observed frequency distribution of extinction measurements and the predictions from clouds models with density structures having $\\alpha =$ 1 (dashed line), 2 (solid line), 3 (dotted line), and 4 (dash--dotted line). The represented predictions correspond to the average of the 1000 realizations done for each cloud model.  Both the data and the results from the models are binned in 2 mag wide bins.  The average KS probability for all realizations in each model was {\\it p}$_{KS} \\,(\\alpha=1) \\approx 10^{-3}$, {\\it p}$_{KS}\\, (\\alpha=2) \\approx 0.43$, {\\it p}$_{KS} \\,(\\alpha=3) \\approx 0.03$, and {\\it p}$_{KS} \\,(\\alpha=4) \\approx 10^{-3}$ with rms dispersions of 10$^{-2}$, 0.24, 0.05, and 10$^{-2}$ respectively.  The constant density case, $\\alpha=0$, predicts a slowly increasing frequency distribution with a zero probability of being, together with the observed distribution, randomly drawn from the same parent distribution.  This model cloud is not represented on Figure 10 for the sake of clarity.  Our Monte Carlo simulation suggests that the large--scale distribution of matter inside L977 is most consistent with a density structure $\\rho \\,(r) \\propto r^{-2}$.  We note that an isothermal cylinder has a density structure $\\rho \\,(r) \\propto r^{-4}$ (Ostriker 1964).  In Figure 11 we present the averaged extinction profile of the southern portion of L977 where the center to edge contrast in extinction is the highest.  Each point in this figure represents the average extinction along the North--South direction of 7 adjacent extinction map pixels covering a declination of $\\sim 5$ arcmin on the southern part of the NIR survey, where the filament is roughly aligned to the North-South direction.  The error bars represent the rms dispersions in the average of the pixels.  The ``bump'' around 300 arcsec is caused by a small secondary condensation within L977.  The dotted lines are column density profiles corresponding to radial density laws $\\rho \\,(r) \\propto r^{-\\alpha}$, for $\\alpha=$0, 1, 2, and 3.  The best fit between the observed radial profile and the various model profiles occurs for a $\\rho \\,(r) \\propto r^{-2}$ density structure.  The 1$\\sigma$ uncertainty in the averaged extinction measurements is now A$_V\\sim 0.3$ mag.  We conclude from our Monte Carlo simulation and from the observed radial profile of the cloud that the distribution of matter inside L977 between 2 $<$ A$_V$ $<$ 40 mag (or 1 $<$ $r$ $<$ 0.1pc) is consistent with a density profile of $\\rho \\,(r) \\propto r^{-2}$.   We do not expect the density in the inner part of L977 to reach infinity as $r \\rightarrow 0$.  Clearly a break at some small $r_0$ ($r_0$\\ltsima 0.1pc; from Figure 11) must occur.  With the present data we cannot probe the very inner part of L977 because we do not detect stars through it.  In fact, even with new and more sensitive near--infrared arrays this will not be an easy task since the probability of having a fairly bright background star within a smaller solid angle (the denser regions of the cloud) is also increasingly small.     \\subsection{The Effect of Galactic Latitude on the NICE Method}  It was a goal of this article to analyze the effects of a richer but more complicated background of field stars --- at lower galactic latitude --- on the application of the NICE method.  An extensive discussion on the limitations and uncertainties in this method is given by Lada et al.  (1994).  Here we will only be concerned about the limitations that result from applying the NICE method to low Galactic latitude molecular clouds by comparing the previous study (IC 5146, $b = -5$\\deg) with the present one ($b = 2$\\deg).  The angular resolution of the dust extinction map is limited by the surface density of stars detected through the target cloud at both the $H$ and $K-$bands.  For the same detection limit it is clear that better resolution, hence more complete information on the small--scale spatial distribution of dust, can be obtained for clouds that lie in the foreground of rich star fields close to the Galactic plane.  On the other hand it was unclear before the present study what was the contribution of distant luminous giant stars, suffering extraneous extinction, to the increase in the dispersion of the mean ($H-K$) color of control field stars at these low Galactic latitude fields. The NICE method is sensitive to this mean color since it assumes that the intrinsic $(H-K)$ color of stars behind the cloud is accurately represented by the mean color of the control fields.  A successful application of the method requires, to first order, the dispersion on the mean ($H-K$) color for the control fields to be significantly smaller than the observed range in $(H-K)$ colors for stars seen through the target cloud.  A direct consequence of observing at low Galactic latitudes is an appreciable increase in the mean and dispersion of the observed $(H-K)$ (and ($J-H)$) color of stars:  \\[<(H-K)>_{IC5146} \\:\\, = 0.13 \\pm 0.08 \\:\\, mag \\,\\,\\,\\,\\,\\,\\,\\, (b = -5^\\circ ) \\]  \\[<(H-K)>_{L977} \\:\\, = 0.20 \\pm 0.13 \\:\\, mag \\,\\,\\,\\,\\,\\,\\,\\,\\, (b = 2^\\circ ) \\]  \\noindent While the value of the mean observed $(H-K)$ color will function as a zero--point for the derived extinction measurements with the NICE method, the dispersion in this value --- $\\sigma(H-K)_{intrinsic}$ --- sets the threshold (1$\\sigma$ confidence level) above which extinctions toward background stars can be reliably measured.  The observed increase in dispersion, caused by distant reddened giants detected at low Galactic latitude, is going to raise this threshold.  For the IC 5146 study the threshold value was $\\sim$ 1 magnitude of visual extinction and it was similar to the measurement uncertainty on a pixel of the derived extinction map:  \\[ \\sigma_{pixel}\\ =\\ {15.9/N} \\sqrt{\\Sigma_{i=1}^N (\\sigma(H-K)^2_i + \\sigma(H-K)^2_{intrinsic})} \\,\\,\\,\\,\\,\\,\\,\\, (mag)\\]  \\noindent where $\\sigma(H-K)_i$ is the photometric error in the derived color for a star in a given map pixel and N is the number of stars in that pixel.  For the L977 study the larger value of $\\sigma(H-K)_{intrinsic}$ raises the threshold to $\\sim$ 2 magnitudes of visual extinction while the uncertainty in the extinction map is still $\\sim$ 1 magnitude of visual  extinction.  The uncertainties on a map pixel are similar in both studies because the average number of stars inside each pixel map is now 8 (instead of 5 for the IC 5146 study) which compensates for the larger $\\sigma(H-K)_{intrinsic}$ observed in the present study.  We conclude that the consequence of applying the NICE method at low Galactic latitudes is an increase of as much as a factor of 2 in the 1$\\sigma$ confidence level in the derived extinction measurements toward background stars.  This is directly related to the detection at low galactic latitudes of a large number of giant stars along the plane of the Galaxy affected by random, unrelated, amounts of extinction and affects somewhat the method's aptitude to address small--scale cloud structure (e.g., the increase in random dispersion around the $\\sigma_{disp}$ versus A$_V$ relation).  Nonetheless, and because of the comparatively higher number of stars detected at these very low galactic latitudes $(b = 2^\\circ)$ \\footnote {The warping of the galactic disk makes this latitude essentially $\\sim 0^\\circ$.}, the capacity of the method to study the large--scale distribution of dust in a molecular cloud is essentially the same as an application of the method at the same galactic longitude but at a galactic latitude $(b = -5^\\circ)$ free from ``contamination'' of distant giant stars. The NICE method remains a powerful method for mapping the distribution of dust through a molecular cloud."
    },
    "9805/astro-ph9805188_arXiv.txt": {
        "abstract": "There is a widespread assertion in the literature that the optical Ga lines give much higher abundances than the UV lines.  We have determined Ga abundances in HgMn stars taking the observed hyperfine structure of the optical Ga {\\sc{ii}} lines into account.  This reduces these abundances to within 0.2 dex of the values from the resonance lines. ",
        "introduction": " ",
        "conclusions": ""
    },
    "9805/astro-ph9805007_arXiv.txt": {
        "abstract": "Quasi-periodic X-ray brightness oscillations (QPOs) with frequencies $\\gta1$~kHz have now been discovered in more than a dozen neutron stars in low-mass X-ray binary systems using the {\\em Rossi X-ray Timing Explorer}. There is strong evidence that the frequencies of the kilohertz oscillations are the orbital frequencies of accreting gas in nearly circular orbits around these stars. Some stars that produce kilohertz QPOs may have spin frequencies $\\gta400$~Hz. For spin rates this high, first-order analytic treatments of the effects of the star's rotation on its structure and the spacetime are inaccurate. Here we use the results of a large number of fully relativistic, self-consistent numerical calculations of the stellar structure of rapidly rotating neutron stars and the interior and exterior spacetime to investigate the constraints on the properties of such stars that can be derived if stable circular orbits of various frequencies are observed. We have computed the equatorial radius of the star, the radius of the innermost stable circular orbit, and the frequency of the highest-frequency stable circular orbit as functions of the stellar spin rate, for spin rates up to the maximum possible and for several illustrative equations of state. Our calculations show that the upper bounds on the stiffness of neutron star matter implied by a given orbital frequency are typically significantly stricter for stars with spin frequencies $\\gta 400$~Hz than for slowly rotating stars. ",
        "introduction": "The successful launch of the {\\em Rossi X-ray Timing Explorer\\/} (\\rxte) has made it possible to investigate, for the first time, the X-ray variability of neutron stars and black holes at frequencies $\\gta300$~Hz. One of the most important discoveries made with \\rxte is that many neutron stars in low-mass X-ray binaries produce high-frequency brightness oscillations with frequencies in the range $\\sim$300--1200~Hz (see van der Klis 1998). High-frequency oscillations are observed both during type~I (thermonuclear) X-ray bursts and in the persistent X-ray emission. The discovery of these oscillations has made possible derivation of interesting upper bounds on the masses and radii of these neutron stars and significant new constraints on the equation of state of neutron star matter (Miller, Lamb, \\& Psaltis 1998a, 1998b; Lamb, Miller, \\& Psaltis 1998; Miller \\& Lamb 1998; Strohmayer et al.\\ 1998).  Only a single oscillation has been observed from each source during a type~I X-ray burst, and the oscillations in the tails of bursts appear to be highly coherent (see, e.g., Smith, Morgan, \\& Bradt 1997), with frequencies that are always the same in a given source (see, e.g., Strohmayer 1997). The burst oscillations are thought to be caused by the existence of one or two brighter regions on the stellar surface that produce oscillations at the stellar spin frequency or its first overtone as the star spins (see Strohmayer et al.\\ 1997b for compelling evidence in favor of this interpretation). The frequency of the burst oscillations ranges from $\\sim$330~Hz in \\fu{1702$-$42} (Swank 1997) to 589~Hz in an unidentified source in the direction of the Galactic center (Strohmayer et al.\\ 1997a).  The kilohertz quasi-periodic oscillations (QPOs) observed in the persistent emission have high amplitudes and relatively high coherences. The frequencies of the two QPOs often observed simultaneously in a given source have a frequency separation that is almost constant in all sources (see Wijnands \\& van der Klis 1997; Psaltis et al.\\ 1998a, 1998b), although the frequencies of the QPOs themselves vary by hundreds of Hertz (see Wijnands et al.\\ 1998 and van der Klis 1995). The separation frequencies of the two kilohertz QPOs seen in \\fu{1728$-$34} (Strohmayer et al.\\ 1996, 1997b) and \\fu{1702$-$42} (Swank 1997) are consistent with the frequencies of their burst oscillations. The separation frequencies in \\fu{1636$-$536} (Wijnands et al.\\ 1997; Zhang et al.\\ 1997) and \\ks{1731$-$260} (Smith et al.\\ 1997; Wijnands \\& van der Klis 1997) are consistent with half the frequencies of their burst oscillations.  The presence of only two simultaneous kilohertz QPOs in a given source, the approximately constant frequency separation $\\Delta\\nu$ between them, and the consistency of $\\Delta\\nu$ with the stellar spin frequency inferred from burst oscillations is strong evidence that the stellar spin is generating the frequency difference, that only one sideband of the primary QPO frequency is being generated, and that one of the two QPOs is therefore caused by the beat of the spin frequency with the other frequency (see Lamb et al.\\ 1998; Miller et al.\\ 1998a). This implies that in addition to the spin frequency there is only one other primary frequency and that this frequency is a rotational frequency such as an orbital frequency. This excludes neutron star surface and photon bubble oscillations as explanations for the primary kilohertz QPO frequency and makes disk oscillations an improbable explanation (see van der Klis 1998; Lamb et al.\\ 1998).  In all sources the frequencies of the kilohertz QPOs fall within the expected range of orbital frequencies near a neutron star and can vary by several hundred Hertz in less than a few hundred seconds while remaining highly coherent ($\\nu/\\Delta\\nu \\sim 100$). This is strong further evidence against disk oscillations and in favor of orbital motion of inhomogeneities in the accretion disk as the cause of the primary kilohertz QPO (see Lamb et al.\\ 1998). The accreting gas exerts a strong torque on the neutron star and hence the star is expected to be spinning in the same sense as the orbital motion of the accreting gas. The beat frequency must therefore be the lower of the two kilohertz QPO frequencies whereas the orbital frequency is the higher.  There are two candidates for the orbital frequency: the frequency at the radius where the accreting gas first couples strongly to magnetic field of the neutron star and the frequency at the sonic radius where radiation forces or general relativistic effects cause the radial motion of the gas to increase sharply and become supersonic. The general properties of the kilohertz QPO sources and the specific properties of the kilohertz QPOs themselves strongly indicate that the relevant frequency is the orbital frequency at the sonic point (Miller et al.\\ 1998a; see also Lamb et al.\\ 1998). In either case, the frequency $\\nu_{\\rm QPO2}$ of the higher-frequency kilohertz QPO is the orbital frequency of gas in a nearly circular orbit around the neutron star, whereas the frequency $\\nu_{\\rm QPO1}$ of the lower-frequency QPO is the beat of the neutron star spin frequency with an orbital frequency near $\\nu_{\\rm QPO2}$.  In the following discussion we shall describe a circular orbit as stable or unstable according to its properties as determined by solving the geodesic equation for a test particle moving in that orbit in the spacetime of interest. However, it is important to bear in mind that {\\em there are no closed, circular orbits in the vicinity of an accreting neutron star}, because the motion of gas near such a star is affected not only by the curvature of spacetime but also by radiation, magnetic, and viscous forces, which cause the gas to spiral inward even at radii where, in their absence, closed, stable orbits would be possible (Miller \\& Lamb 1993, 1996). However, the distinction between stable circular orbits (SCOs) and unstable circular orbits is still relevant. In particular, the innermost stable circular orbit (ISCO) is still physically significant when the effects of radial gas pressure forces on the ISCO can be neglected (which should be valid when the luminosity of the source is much less than the Eddington critical luminosity), because under these conditions gas {\\em inside\\/} the ISCO spirals inward so quickly that it cannot produce a wavetrain with the coherence observed for the kilohertz QPOs, regardless of whether it is acted on by radiation, magnetic, and viscous forces (Miller et al.\\ 1998a).  Identification of the higher-frequency kilohertz QPO with the frequency of an SCO has made it possible to derive upper bounds on the masses and radii of the neutron stars in the kilohertz QPO systems (Miller et al.\\ 1998a, 1998b). These bounds follow from the requirement that the radius $R_{\\rm orb}$ of the orbit be larger than the radius $R_{\\rm ms}$ of the ISCO as well as larger than the equatorial radius $R_{\\rm eq}$ of the star (if $R_{\\rm eq} < R_{\\rm orb} < R_{\\rm ms}$, orbits with the required frequency exist but are unstable).  For nonrotating stars, observation of a given orbital frequency can be used to derive upper bounds on the mass and radius that are independent of the equation of state assumed (Miller et al.\\ 1998a). For rotating stars, the situation is more complicated. In general, both the structure of the star and the spacetime are affected by the star's rotation, and there are no general analytical expressions for the relevant quantities. However, the exterior spacetime of a slowly and uniformly rotating fluid star is unique to first order in the dimensionless angular momentum $j\\equiv cJ/GM^2$, where $J$ and $M$ are the star's angular momentum and gravitational mass, and can be expressed analytically to this order (Hartle \\& Thorne 1968). The leading corrections to the expressions for the orbital frequency and the radius of the ISCO are first-order in $j$. Using these expressions, one can demonstrate that observation of a given orbital frequency also implies upper bounds on the mass and radius of a slowly rotating star (Miller et al.\\ 1998a). For a given stellar spin frequency, these upper bounds depend on the moment of inertia and hence on the equation of state assumed.  Many of the kilohertz QPO sources appear to have spin frequencies $\\sim$250--350~Hz (see Miller et al.\\ 1998a). Examples of such sources include \\fu{0614$+$091}, \\fu{1608$-$52}, \\fu{1820$-$30}, \\cyg2, \\sco1, \\gx{5$-$1}, and \\gx{17$+$2}, all of which have kilohertz QPO separation frequencies in this range, as well as \\fu{1728$-$34} and \\fu{1702$-$42}, which not only have kilohertz QPO separation frequencies in this range but also have burst oscillation frequencies that are consistent with these separation frequencies. The $\\sim$520 and $\\sim$580~Hz frequencies of the burst oscillations seen in \\ks{1731$-$260} and \\fu{1636$-$536} are thought to be twice their spin frequencies, although this is not certain. For spin frequencies in this range, $j$ is $\\sim$0.1--0.3, depending on the assumed equation of state and the mass of the star, and hence an analysis that is first-order in $j$ is quite accurate for such stars. Such an analysis shows that spin rates $\\sim$300~Hz can increase the upper bound on the stellar mass by as much as $\\sim$10--20\\% but typically increase the upper bound on the radius by only $\\sim$2--5\\% (Miller et al.\\ 1998a, 1998b).  On the other hand, some neutron stars that show kilohertz QPOs may turn out to have spin frequencies $\\gta400$~Hz. For example, oscillations with frequencies $\\sim$550 and $\\sim$590~Hz have been seen during X-ray bursts from, respectively, \\aql1 and the unknown source in the direction of the galactic center, possibly indicating that these neutron stars have spin frequencies this high (Miller et al.\\ 1998a). The recent discovery using \\rxte that the source SAX J1808.4--3658 has a coherent 401~Hz oscillation indicates that this accreting neutron star is spinning rapidly (Wijnands \\& van der Klis 1998a, 1998b). For stars spinning this fast, the effect of the star's spin on its equilibrium structure (which is second-order) and on the spacetime can be substantial. In order to obtain accurate results for such high spin rates, the equilibrium stellar structure and the interior and exterior spacetime must be computed self-consistently, which can only be done numerically.  Here we use the results of a large number of fully relativistic, self-consistent numerical calculations of the structure of rapidly rotating neutron stars and the interior and exterior spacetime to investigate the constraints on the properties of such stars that can be derived if SCOs of various frequencies are observed. We have computed the equatorial radius of the star, the radius of the ISCO, and the frequency of the highest-frequency SCO as functions of the stellar spin rate and gravitational mass, for spin rates up to the maximum possible and for several illustrative equations of state. Comparison of these results with the highest observed kilohertz QPO frequency in a given source can be used to derive bounds on the mass and radius of the neutron star in that source, for a given equation of state. We also report the frequency of the highest-frequency SCO as a function of the stellar spin rate, for stars of any mass constructed using a given equation of state. These curves can be used to check whether a particular equation of state is consistent with the frequency of a given kilohertz QPO.  Our calculations show that the upper bounds on the stiffness of neutron star matter implied by a given orbital frequency are typically significantly stricter for stars with spin frequencies $\\gta400$~Hz than for slowly rotating stars.  In \\S~2 we describe our assumptions and calculational methods. In \\S~3 we present our results and discuss the implications for constraining the properties of neutron star matter. Our conclusions are summarized in \\S~4. ",
        "conclusions": "Our results show that deviations from a first-order treatment of the effects of spin on the structure of neutron stars and on circular orbits around them are typically significant for spin frequencies $\\gta400$~Hz. The Kerr spacetime is generally a poor approximation to the exterior spacetime of neutron stars spinning this fast or faster.  Our results demonstrate that the upper bounds on the stiffness of neutron star matter implied by the high frequencies and coherences of the kilohertz QPOs are tightened significantly if the star is rotating rapidly. The constraints on the equation of state become much tighter if observations conclusively identify a QPO frequency as the orbital frequency at the ISCO. For example, if the orbital frequency at the ISCO is 1100~Hz, the required mass is $2.0\\msun$ for nonrotating stars and substantially more for rotating stars. Such a high mass would rule out many of the softer equations of state and would imply that the three-nucleon interaction is strongly repulsive at high densities (see Pandharipande et al.\\ 1998).  As shown in \\S~3.3, even if the spin rate of the star is unknown, the area of the mass-radius plane allowed by observation of a high-frequency SCO can be quite small. If the spin rate is known, the range of radii allowed for a given equation of state is usually very small.  Results of the kind presented in this work will be still more constraining if measurements of SCO frequencies can be combined with constraints on other quantities, such as the stellar compactness $M/R$ (see Strohmayer 1997; Miller \\& Lamb 1998; Lamb et al.\\ 1998; Strohmayer et al.\\ 1998) or the radius of the star (see Strohmayer et al.\\ 1997b; Strohmayer et al.\\ 1998).   Additional observations of kilohertz QPO sources are extremely important, because these observations could provide very tight and robust constraints on the fundamental properties of neutron stars and on the equation of state of neutron star matter."
    },
    "9805/astro-ph9805231_arXiv.txt": {
        "abstract": "We present new radio continuum data at 4 frequencies on the supermassive, peculiar galaxy NGC~1961. These observations allow us to separate the thermal and the nonthermal radio emission and to determine the nonthermal spectral index distribution. This spectral index distribution in the galactic disk is unusual: at the maxima of the radio emission the synchrotron spectrum is very steep, indicating aged cosmic ray electrons. Away from the maxima the spectrum is much flatter. The steep spectrum of the synchrotron emission at the maxima indicates that a strong decline of the star formation rate has taken place at these sites. The extended radio emission is a sign of recent cosmic ray acceleration, probably  by   recent star formation. We suggest that a violent event in the past, most likely a merger or a collision with an  intergalactic gas cloud, has caused the various unusual features of the galaxy. ",
        "introduction": "NGC~1961 (Arp 184) is a very massive (dynamical mass $10^{12} - 10^{13}\\ms$, Gottesman et al. 1983) Sb spiral galaxy with a peculiar and asymmetric optical appearance. It is among the  most massive and largest known spiral galaxies (Romanishin 1983); its  optical diameter of $D_{25}=4.6$ arcmin (de Vaucouleurs et al. 1991, RC3) translates to 110 kpc (adopting a distance of 82.9 Mpc for $H_0 = 50$ km s$^{-1}$ Mpc$^{-1}$). It is a member of a group of 6 galaxies where it is by far the largest and most massive member.  Despite its size and peculiar appearance, NGC~1961 has received relatively little attention. A detailed study of its \\hi structure and kinematics by Shostak et al. (1982, SHSVV) found an unusual asymmetric morphology with an extensive ``wing'' of gas extending 30 kpc to the north-west and a sharp edge to the south-east. SHSVV suggested that the disturbed appearance is due to stripping of the gas in NGC~1961 by a hot intergalactic medium (IGM). They discarded the possibility of tidal interaction with another galaxy since there is no galaxy with a large enough mass sufficiently close to NGC~1961. They also dismissed the possibility of a merger with another galaxy or an intergalactic gas cloud since they found no trace of a merger remnant such as a second nucleus, or disturbed velocity distribution in the \\hi gas. In favour of the stripping hypothesis SHSVV noted the following points. i) The displacement of the \\hi with respect to the optical disk in the south of the galaxy. ii) The presence of increased radio continuum emission in the south-east which they interpreted as being caused by an enhancement of the magnetic field or by an increase in the cosmic ray (CR) density due to a high star formation rate (SFR). iii) The blue colour of the optical emission  of the NW wing and the south-eastern ridge indicating recent star formation. iv) A curved  X-ray emission feature in the south-east of the galaxy detected by the {\\it Einstein} satellite (Shostak et al. 1982) which they interpreted as emission from hot IGM. However, recent ROSAT observations have shown that the X-ray emission of NGC~1961 is fairly typical for a spiral galaxy and comes mainly from the central region (Pence \\& Rots 1997). The cause of the asymmetric \\hi distribution therefore remains unclear.  In this paper we present multi-frequency radio continuum data which we use to identify thermal and nonthermal (synchrotron) emission. The thermal radio emission reveals the sites and intensity of present day star formation, whereas the synchroton emission, and in particular its spectral shape, contains information about the star formation in the past (on time-scales of a few $10^7$ yr which is related to the life-time of the CR electrons and of supernova (SN) progenitors), and also about the acceleration and propagation mechanisms of CRs. ",
        "conclusions": "We have presented radio data at 4 frequencies of the supermassive spiral galaxy NGC 1961. Using these data we fitted a combined aged-synchrotron plus thermal emission spectrum to each point in the galaxy. The most recent star formation, that is traced by the thermal radio emission, is mainly taking place in the centre and the south-eastern arm of the galaxy, whereas the underlying extended radio emission is predominantly nonthermal.  The  spectral index distribution between 1.46 and 4.92~GHz over the disk of the galaxy shows a peculiar behaviour. At the maxima of the radio emission the nonthermal spectral index is very steep whereas the extended radio emission has a much flatter spectrum. We discuss various possibilities to explain this peculiar behaviour. The most likely explanation is that the steep spectrum is due to the end of the CR injection about $10^7$ years ago. Assuming that CRs are accelerated mainly in SNRs this means that a phase  of intense SF has ceased  $\\la 10^8$ yr ago. The flat spectrum of the extended emission is a sign of  recent CR acceleration in these regions. This could be due to CR acceleration in the shocks of SNRs of a  recently formed stellar population. The radio-to-FIR ratio exceeds the average value of spiral galaxies by a factor of 3-4. An overall increase of the magnetic field with respect to the radiation field could account for this high FIR/radio ratio. Spatial variations of the magnetic field, which are likely to be present if the magnetic field is frozen into the gas, are however unable to give the observed distribution of the spectral index.  The discovery of a second radio nucleus leads us to suggest that NGC~1961 is most likely the result of a merger rather than collision with a massive intergalactic gas cloud."
    },
    "9805/astro-ph9805356_arXiv.txt": {
        "abstract": "We present the results of a detailed analysis of the newly discovered 15$'$ large radio galaxy WNB\\,0313+683. It has been discovered in the WENSS and NVSS radio surveys, and has been identified with an optical galaxy at a redshift of $0.0901\\pm0.0002$. The linear size of the radio structure is 2.0~Mpc, which places it in the class of Giant Radio Galaxies (GRGs).\\\\ Radio observations have been carried out with the WSRT at 1.4~GHz and with the 92-cm broadband system, with the 100-m Effelsberg telescope at 10.45~GHz and with the VLA at 1.4~GHz and 5~GHz. At 10.45~GHz the core is the dominant source structure, emitting $\\sim 25\\%$ of the total flux. It has an inverted spectrum with a spectral index of $\\alpha=+0.42 \\pm 0.03$ (S$_{\\nu}\\propto \\nu^{\\alpha}$) between 1.4~GHz and 10.45~GHz, suggesting a very compact structure, although we can not rule out variability of the core luminosity.\\\\ The Rotation Measure distribution has been mapped using a new method and has been found to be very uniformly distributed over the source. It is therefore probably galactic in origin, with only a small contribution from the (surroundings of the) source itself.\\\\ From the spectral index distribution we have determined an upper limit on the spectral age of $1.4 \\pm 0.1 \\times 10^8$~yrs. The particle density of the ambient medium, using ram-pressure equilibrium at the hotspots, is $\\ga 1.6 \\times 10^{-6}$~cm$^{-3}$ for the southern lobe and $\\ga 5.8 \\times 10^{-7}$~cm$^{-3}$ for the northern lobe. An independent measure of the external density has been determined using the amount of depolarization towards the southern lobe. This gives a density, averaged along the line of sight, which is a factor of 10 higher ($2.0\\times10^{-5}$~cm$^{-3}$) than the density near the head of the lobes found using the ram-pressure arguments. This discrepancy might at least partly be the result of a contribution from internal depolarization, which we can not exclude on basis of our radio data.\\\\ From spectroscopical observations of the host galaxy we find that the H$\\alpha$ emission line has a broad component, and that the extinction must be large with colour index $E(B-V) = 0.98 \\pm 0.10$~mag. Since the galactic latitude is $+9\\fdg8$, the extinction is probably mostly galactic in origin. We further find that WNB\\,0313+683 has a very large optical emission line flux with respect to its estimated jet power, when compared with the correlation between these two properties found by Rawlings \\& Saunders (1991). We argue that this, together with the relatively high radio power and the inverted radio spectrum of the radio core, is suggestive of a new phase of radio activity in WNB\\,0313+683.\\\\ ",
        "introduction": "Giant Radio Galaxies (GRGs, e.g. Saripalli et al. 1986, Subrahmanyan et al. 1996) are the largest members of the radio galaxy population, with  a (projected) linear size $\\ga 1.0$~Mpc\\footnote{We use ${\\rm H}_0 = 50$~km~s$^{-1}$~Mpc$^{-1}$ and ${\\rm q}_0 = 0.5$ throughout this paper.}. At low redshift ($z \\la 0.3$) some thirty of these large sources are known. The most extreme case is the radio source 3C\\,236 at a redshift of $\\sim 0.09$, which has a projected linear size of 5.7 Mpc (Willis et al. 1974, Strom \\& Willis 1980, Barthel et al. 1984).\\\\ GRGs are interesting objects to study. First, at low redshift their angular size is several arcminutes or larger, which allows detailed studies of the different components of their radio structure, such as their jets (e.g. NGC\\,6251, Perley et al. 1984) and their lobe emission (e.g. Mack et al. 1997a) using a variety of radio instruments. Secondly, because of the large physical size of GRGs, they have expanded well out of the denser central regions of the clusters that they reside in, into a much less dense Intergalactic Medium (IGM). This makes GRGs a powerful and unique tool to probe the low-density medium at a large distance from their host galaxies, which is inaccessible to current X-ray instruments. Thirdly, their active galactic nucleus (AGN) must be approaching the endstage of its radio active phase. GRGs can therefore provide important information on the properties of old AGN. Lastly, several of these large sources have been identified with quasars (e.g. 4C\\,34.47, J{\\\"a}gers et al. 1982) or Broad Line Radio Galaxies (e.g. 0319+412, de Bruyn 1989). This makes them important test cases for the orientation dependent unification schemes for radio loud AGN (e.g. Barthel 1989). These state that quasars and radio galaxies are mostly alike, but that quasars have their radio jets oriented closer to the line of sight than radio galaxies. A natural consequence of this is that quasars are not expected to have large projected linear sizes.\\\\ There are now many indications that the ambient medium of the radio lobes of GRGs has a very low density ($\\sim\\!10^{-6} - 10^{-5}$~cm$^{-3}$). Probably the strongest evidence is provided by observations of the Rotation Measure (RM) and the depolarization towards these sources. The measured RMs are small, with values which are usually $\\la 20$~rad~m$^{-2}$ (e.g. Klein et al. 1994, Strom \\& Willis 1980). Depolarization occurs only at wavelengths above 10-20~cm (Willis \\& Strom 1978, Strom \\& Willis 1980, J{\\\"a}gers 1986, Klein et al. 1996). This combination can only be explained satisfactorily by a very low density of thermal electrons along the line of sight.\\\\ Further indications for a low density environment result from spectral age analyses of GRGs. Typical spectral ages for GRGs have been found to be $10^7 - 10^8$~yrs (e.g. Mack et al. 1998), which translates into expansion velocities of $0.01c - 0.1c$. Using the assumption of ram pressure equilibrium at the head of the jet (e.g. Miley 1980), external densities of $\\sim10^{-6} - 10^{-5}$~cm$^{-3}$ are commonly found for GRGs. This method suffers from many assumptions that have to be made, such as equipartition between the energy density of the relativistic particles and the magnetic field, the filling factor of the radio lobes, the fraction of energy in heavy particles, the area of the bow shock, and so on. It is therefore often criticized (e.g. Eilek et al. 1997) and its result should be interpreted with care.\\\\ Many clusters have large and bright X-ray haloes, often extending to distances of more than 1~Mpc from their centers and containing cooling flows (e.g. Fabian 1994). X-ray studies of some GRGs (e.g. NGC~6251; Mack et al. 1997b), however, have found only weak thermal X-ray emission around the host galaxies. GRGs are therefore probably not inside rich clusters with dense cores, and the lobes are thus not likely to be found in dense environments. Subrahmanyan et al. (1996) studied a small sample of GRGs on the southern hemisphere. By studying the surface density of optical galaxies in the neighbourhood of the GRG host galaxies using the UKST plates, they conclude that they do not reside in rich clusters.\\\\ GRGs have relatively low radio powers (Saripalli et al. 1986; Subrahmanyan et al. 1996), usually around or below the luminosity which divides FRI and FRII type radio galaxies ($\\sim 2 \\times 10^{26}$~W~Hz$^{-1}$ at 178~MHz; Fanaroff \\& Riley 1974). Because of their large sizes and relatively low radio power, the surface brightness of GRGs is low. This is why they are so difficult to find in most radio surveys. In recent years, the Westerbork Northern Sky Survey (WENSS, Rengelink et al. 1997) has mapped the sky above +28\\degr~declination at a frequency of 327~MHz. WENSS has a sensitivity of 18~mJy (5$\\sigma$) and a beamsize of $54\\arcsec \\times 54\\arcsec~{\\rm cosec}\\,\\delta$, with $\\delta$ the declination. At low frequency, the bulk of emission of radio galaxies originates in the extended radio lobes. This makes the WENSS ideally suited to find large, low surface brightness radio galaxies such as GRGs. The first discovery of such an object (WNB\\,1626+5153) has been reported by R\\\"{o}ttgering et al. (1996). This initiated a project aimed at finding and studying a large, uniformly selected, sample of giant radio sources from the WENSS.\\\\ Here we report on the discovery and subsequent analysis of the radio source WNB\\,0313+683, which we have identified as an FRII-type radio galaxy with a linear size of 2.0~Mpc. It has several remarkable properties, among which there is a large flux asymmetry of the radio lobes and a prominent, inverted spectrum, radio core. In Sect. 2, we will present the radio and optical data that we have collected on this object. Section 3 describes a first analysis of these data, including a new way to measure Rotation Measures. It also presents some of the derived physical properties. In Sect. 4 we derive the advance velocities of the hotspots and the age of the source from a spectral index analysis. Section 5 then discusses the observed depolarization towards WNB\\,0313+683. A discussion on the properties of the radio core is given in Sect. 6. We argue that WNB\\,0313+683 may currently be in a new phase of radio activity. Finally, a summary and our conclusions are presented in Sect. 7. ",
        "conclusions": "We have investigated the newly discovered large radio source WNB\\,0313+683. This 15\\arcmin~source was discovered in the WENSS. The radio core has been identified with a weak optical galaxy. It has a redshift of $0.0901 \\pm 0.0002$, and thus a projected linear size of 2.0~Mpc. The optical spectrum is typical of a narrow-line radio galaxy. The H$\\alpha$-line has a broad ($\\sim 6000$~km~s$^{-1}$) component. The color index, determined from the (narrow) H$\\alpha$/H$\\beta$ line-ratio, is large: $E(B-V) = 0.98 \\pm 0.10$~mag, where we have assumed that the extinction is mostly galactic (the galactic latitude is +9.8$^{\\circ}$). The HI column density towards the host galaxy, measured in the Leiden-Dwingeloo galactic HI survey, translates into a color index of $0.51$~mag. But the large variations observed on scales of 30\\arcmin~in the HI survey maps suggest that denser regions may well exist within a single beam.\\\\ In addition to the WENSS and the NVSS 1.4-GHz maps, we have observed WNB\\,0313+683 with the 100-m Effelsberg telescope at 10.4~GHz, with the WSRT using the 92-cm 8-channel broadband system, and with the VLA at 1.4 and 5 GHz. The VLA and Effelsberg observations show that WNB\\,0313+683 is a core-dominated radio galaxy. The radio spectral index of the core is $+0.42 \\pm 0.03$, at least up to 10.4 GHz, at which frequency it contributes $\\sim 25\\%$ to the total flux of the source. The prominence of the core, if due to relativistic beaming, suggests that the radio axis of the (sub)arcsec core structure is oriented at an angle $\\la 50\\degr$ to the plane of the sky. If the radio axis of the source as a whole is similar to the radio axis of the core, the deprojected linear size would be $\\ga$ 2.6 Mpc.\\\\ Spectral index profiles have been made along the radio axis. They have been used to estimate the age and advance velocities of the two lobes. We find that the source has a maximum age of $1.4 \\pm 0.1 \\times 10^8$ yrs. The found advance velocities of the lobes are $(0.027 \\pm 0.002)c$ for the northern lobe, and $\\sim (0.03 \\pm 0.01)c$ for the southern lobe. The ambient densities have been estimated using ram-pressure equilibrium at the heads of the jet, and by assuming that the impact area is limited to the observed size of the hotspots. We find lower limits to the particle density of $1.6 \\times 10^{-6}$~cm$^{-3}$ for the material surrounding the southern hotspot and $4.8 \\times 10^{-7}$~cm$^{-3}$ around the northern hotspot. The contrast in the external density of the two lobes is reflected in the lobe luminosities. The southern lobe is much brighter (by a factor of $\\sim\\,5$) and appears more confined than the northern one.\\\\ We have studied the Rotation Measure and depolarization towards WNB\\,0313+683. The Rotation Measures have been determined by a new technique, called Rotation Measure Mapping. We find a very smooth RM distribution over the southern lobe, with a mean value of $-10.6 \\pm 0.2$~rad~m$^{-2}$. We believe that this is of galactic origin. After subtracting this mean value from the observed values, we only find significant residual RM values towards the central bulge ($-1.9 \\pm 0.5$~rad~m$^{-2}$).\\\\ Depolarization towards WNB\\,0313+683 is only important at wavelengths above 21cm, since we measure almost no depolarization between the Effelsberg and the NVSS data but find significant depolarization between the NVSS and the 92-cm WSRT data. We have discussed the origin of the depolarization and conclude that it could either be caused inside the radio source itself, or by a large magneto-ionic cluster-like halo surrounding the source. The current data do not allow us to distinguish between these two scenarios. We have investigated the distribution of the depolarization, and the derived Faraday depth $\\Delta$, towards the southern lobe. We find no strong increase of $\\Delta$ near the core, as would be expected if the host galaxy lies at the center of a cluster halo with a dense core, but the values are significantly higher than those near the southern hotspot. Overall, $\\Delta$ is close to 1.0~cm$^{-3}~\\mu$G~pc.\\\\ In the case that a large halo surrounding the radio source is responsible for the depolarization, we estimate a lower limit for the electron density, averaged along the line of sight towards the southern lobe, of $2 \\times 10^{-5}$~cm$^{-3}$. This is an order of magnitude above the values we find in the spectral index analysis. Part of this might be caused by a contribution from internal depolarization. Our data do not exclude this possibility.\\\\ In many respects, WNB\\,0313+683 is like any other GRG. Its size is close to the median value of the `classical' GRGs ($\\sim 2$~Mpc), the density of the ambient medium is low, of the order of $10^{-6}$~cm$^{-3}$, the spectral age is large ($\\sim 10^8$~yrs), there is only little depolarization, and the Rotation Measures, corrected for the galactic contribution, are low as well ($\\la 2$~rad~m$^{-2}$).\\\\ The inverted spectrum radio core and the very strong optical emission lines as compared to the estimated jet power, suggest an interrupted radio activity of the AGN. VLBI observations are necessary to investigate the structure of the radio core in more detail.\\\\"
    },
    "9805/astro-ph9805160_arXiv.txt": {
        "abstract": "I review some results of estimation of the power spectrum of density fluctuations from galaxy redshift surveys and discuss advances that may be possible with the Sloan Digital Sky Survey. I then examine the realities of power spectrum estimation in the presence of Galactic extinction, photometric errors, galaxy evolution, clustering evolution, and uncertainty about the background cosmology. ",
        "introduction": "\\vskip-7.30truecm \\hskip6.1truecm{$^1$} \\vskip7.0truecm  \\vskip-12.00truecm {$\\>$}\\\\ {\\it To appear in {``Ringberg Workshop on Large-Scale Structure''},\\\\ ed. D. Hamilton (Kluwer, Amsterdam) } \\vskip11.00truecm  The advent of deep redshift surveys of $10^4-10^6$ galaxies, such as the Las Campanas Redshift Survey (LCRS; Shectman \\etal~1996) and the AAT 2df survey (Colless 1998), and multiband photometric and spectroscopic surveys such as the Sloan Digital Sky Survey (SDSS; Gunn \\& Weinberg 1995; Szalay 1998), marks the beginning of a new era of investigations in large-scale structure.  Rather than treat galaxies as indistinguishable tracers of mass in a static distribution, we will study the detailed dependence of clustering on galaxy type and cosmic epoch.  In fact, as I will discuss, we must study the species and redshift dependence of clustering to precisely differentiate among cosmological models. In somewhat shallower wide-angle surveys (Huchra \\etal~1983; Giovanelli \\& Haynes 1984; Geller \\& Huchra 1989; da Costa \\etal~1994; Fisher \\etal~1995) it has been possible to begin study of these effects; at redshift $z=0.1$ and larger it becomes critical to do so.  In addition to precisely characterizing galaxy clustering at the present epoch, we will use the anisotropy of clustering, as well as the evolution of this clustering, to constrain cosmological parameters.  A key issue that I examine is the limiting precision with which we can hope to measure galaxy clustering at the present epoch.  Although much has been made of the power of large galaxy surveys to measure cosmological parameters, most of this discussion has focussed on finding an analysis method that gives ``optimal'' uncertainties on model parameters, where the uncertainties are due to cosmic variance and shot noise (Vogeley \\& Szalay 1996; Tegmark \\etal~1998). Much less has been said about the uncertainties caused by random and systematic errors in photometry or Galactic extinction corrections, and systematic uncertainties that arise from issues of cosmology, and evolution of galaxies and their clustering.  In this paper I discuss the impact of these effects on analysis of galaxy redshift samples from the SDSS. ",
        "conclusions": "The large sample volume and dense sampling of the galaxy distribution in the SDSS redshift survey will set lower bounds to the uncertainties that could, in principle, allow detection of small deviations from smooth spectra and easily differentiate between similar cosmological models.  Clever use of the anisotropy of clustering would allow simultaneous estimation of both $\\Omega_M$ and the real-space power spectrum.  Further, we could use the growth rate of clustering to constrain $\\Omega_M$ and $\\Omega_{\\Lambda}$ in a different way. However, all of these dreams hinge on our ability to cleanly separate true galaxy clustering from contamination by extinction and photometric error, and to differentiate between clustering evolution and evolution of the galaxies themselves.  Thus, the future of large-scale structure lies in studying evolution, both of galaxies and of the clustering pattern itself. This evolution is strongly species-dependent and will require estimation of multivariate luminosity functions to interpret.  In a few years' time, the SDSS, together with the 2MASS infrared survey and FIRST radio surveys, which will also cover (really, have already begun to cover) the same region of sky, will provide a database of optical, near-IR, and radio information about millions of objects and will be ideal for just this sort of investigation.  The author acknowledges the collaborative effort of many of the participants in the SDSS, in particular Andy Connolly and Alex Szalay, that contributed to the results presented in this volume.  Support for this work was provided by NASA through grant HF-01078.01-94A from the Space Telescope Science Institute, which is operated by AURA, Inc. under NASA contract NAS5-26555,"
    },
    "9805/astro-ph9805210_arXiv.txt": {
        "abstract": "We describe the ultraviolet spectra of the Ap star $\\alpha^2$ CVn that have been acquired with the Hubble Space Telescope and the nature of the optical region analyses for its rotational velocity and elemental abundances that have been undertaken in support of the ultraviolet data. ",
        "introduction": "\\label{intr} The spectrum of the prototypical magnetic star $\\alpha$$^2$ CVn (= HR 4915, A0p Si, Hg, Eu, Cr) has been recognized as peculiar for one hundred years (Maury 1897).  As a result of its brightness it has been an important target for several lines of inquiry into warm chemically peculiar stars. Yet even though $\\alpha^2$ CVn has been a centerpiece for Ap star study, there is still much to learn from its spectrum, particularly at ultraviolet (UV) wavelengths. We have undertaken to study its UV spectrum in order to more fully identify those ions that are present and better quantify the rich line opacity that leads to flux redistribution. The UV atlas consists of 35 wavelength settings of the Goddard High Resolution Spectrograph (GHRS) onboard the Hubble Space Telescope (HST). Three first-order gratings, characterized by spectral resolving powers between R = $\\lambda$/$\\Delta\\lambda$ = 20000 and 35000, were used to collect high quality (S/N $>$ 100) data over the wavelength interval 1510 to 2640 \\AA. The data were collected on three separate dates, each corresponding to magnetic phase 0.0 +/- 0.1, the phase of maximum UV opacity and Eu\\,{\\sc ii} line strength at optical wavelengths.  However, the nature of the UV spectrum can not be studied independently of the optical region, and it was deemed necessary to acquire optical region spectra to address issues such as rotational velocity and elemental abundances. Optical region spectra were obtained during several observing campaigns with the 2.6-m Nordic Optical Telescope (NOT). The SOFIN echelle spectrograph was used to obtain complete optical coverage at a resolving power of R = 25000 at a single phase, along with limited wavelength coverage observations for many phases at a resolving power of R = 80000. ",
        "conclusions": ""
    },
    "9805/astro-ph9805026_arXiv.txt": {
        "abstract": "We discuss HD and MHD compressible turbulence as a cloud-forming and cloud-structuring mechanism in the ISM. Results from a numerical model of the turbulent ISM at large scales suggest that the phase-like appearance of the medium, the typical values of the densities and magnetic field strengths in the intercloud medium, as well as the velocity dispersion-size scaling relation in clouds may be understood as consequences of the interstellar turbulence. However, the density-size relation appears to only hold for the densest clouds, suggesting that low-column density clouds, which are hardest to observe, are turbulent transients. We then explore some properties of highly compressible polytropic turbulence, in one and several dimensions, applicable to molecular cloud scales. At low values of the polytropic index $\\gamma$, turbulence may induce the gravitational collapse of otherwise linearly stable clouds, except if they are magnetically subcritical. The nature of the density fluctuations in the high Mach-number limit depends on $\\gamma$, and in no case resembles that resulting from Burgers turbulence. In the isothermal ($\\gamma=1$) case, the dispersion of $\\ln (\\rho)$ scales like the turbulent Mach number. The latter case is singular with a lognormal density pdf, while power-law tails develop at high (resp. low) densities for $\\gamma <1$ (resp. $\\gamma >1$). ",
        "introduction": "One of the main features of turbulence is its multi-scale nature (e.g., \\cite{scalo87}; \\cite{lesieur90}). In particular, in the interstellar medium (ISM), relevant scale sizes span nearly 5 orders of magnitude, from the size of the largest complexes or ``superclouds'' ($\\sim 1$ kpc) to that of dense cores in molecular clouds (a few $\\times 0.01$ pc), with densities respectively ranging from $\\sim 0.1$ cm$^{-3}$ to $ \\simgt 10^6$ cm$^{-3}$. Moreover, in the diffuse gas itself, even smaller scales, down to sizes several $\\times 10^2$ km are active (see the chapters by Spangler and Cordes), although at small densities. Therefore, in a unified turbulent picture of the ISM, it is natural to expect that turbulence can intervene in the process of cloud formation (\\cite{hunter79}; \\cite{hunt_fleck82}; \\cite{elm93}; \\VS, Passot \\& Pouquet 1995, 1996) through modes larger than the clouds themselves, as well as in providing cloud support and determining the cloud properties, through modes smaller than the clouds (\\cite{chandra51}; \\cite{bonazzola87}; \\cite{Leorat90}; \\cite{VS_gaz95}). Moreover, another essential feature of turbulence is that all these scales interact nonlinearly, so that coupling is expected to exist between the large-scale cloud-forming modes and the small-scale cloud properties.  In this chapter we adopt the above viewpoint as a framework for presenting some of the most relevant results we have learned from two-dimensional (2D) numerical simulations of the turbulent ISM in a unified and coherent fashion, as it relates to the problems of cloud formation, the phase-like structure of the ISM and the topology of the magnetic and density fields, as well as internal cloud properties, such as their virialization and scaling relations (\\S\\ \\ref{2D}). Next we discuss recent results from multi-dimensional simulations and a simple heuristic model of one-dimensional polytropic turbulence, as a first attempt to gain more physical insight into the mechanisms responsible for the generation of the statistics of the density fluctuations in compressible turbulence (\\S\\ \\ref{polytropes}). Finally, we present a summary and conclusions in \\S\\ \\ref{conclusions}. ",
        "conclusions": "In this Chapter we have discussed a scenario in which turbulence plays a fundamental role in the production and determination of interstellar cloud properties. Large-scale turbulent modes intervene in the former, while small-scale modes seem to  participate in determining cloud scaling relations. ISM features that appear naturally in our simulations are the phase-like appearance (a consequence of turbulent density fluctuation production together with an effective polytropic exponent $\\gamef<1$), density and magnetic field topologies and field strength ranges, and the velocity dispersion-size relation and cloud mass spectrum. However, the suggestions are made that the Larson (1981) density-size scaling relation may be an artifact of surveys which do not integrate for long enough times, and that the logatropic equation of state (\\cite{liz_shu89}) is not verified in highly dynamic situations.  Given the possible nature of clouds as turbulent density fluctuations, their production in polytropic flows was also discussed. The density jump across shocks and a criterion for the collapse of these fluctuations were advanced. Finally, a model for the probability density function of the fluctuations was described, which satisfactorily explains the pdf shapes observed in isothermal, polytropic and Burgers cases. Future work in this area will address the fully thermodynamic and magnetic cases, aiming at explaining other forms of the pdf, which have been observed numerically, but not produced by the model."
    },
    "9805/hep-ph9805333_arXiv.txt": {
        "abstract": "We present a formalism to calculate the probability distribution function of a scalar field coarse-grained over some spatial scales with a Gaussian filter at finite temperature. As an application, we investigate the role of subcritical fluctuations in the electroweak phase transition in the minimal standard model. It is  concluded that the universe was in a mixed state of true and false vacua already at the critical temperature. ",
        "introduction": "\\label{sec:introduction}  \\indent  Quantum field theories with spontaneous symmetry breaking usually exhibit symmetry restoration at high temperature \\cite{KL}, and hence our universe is supposed to have experienced several phase transitions in the early stage of its evolution as the cosmic temperature decreased with expansion.  Much work has been done on these phase transitions and their generic properties such as the transition order and the critical temperature, $T_c$, have been extracted from the effective potential.  The dynamics of phase transitions, on the other hand, is less understood, partly because it is difficult to clarify the role of thermal fluctuations.  Thermal fluctuations often play a conspicuous role on the dynamics of phase transitions. For example, at formation of topological defects associated with some symmetry breaking \\cite{Kib}, magnitude of thermal fluctuations fixes the temperature when the phase distributed randomly on each coherent volume becomes stabilized. Also, their statistical property is important to set the initial distribution of topological defects. Furthermore,  thermal fluctuations called subcritical bubbles have been considered  important in the context of the electroweak phase transition, where the one-loop improved finite-temperature effective potential shows that it is of first order \\cite{DLHLL}-\\cite{She}. Gleiser, Kolb, and Watkins \\cite{GKW} \\cite{GK}, however, suggested that the barrier between the two minima of the effective potential at $T_c$ was so shallow that thermal fluctuations called subcritical bubbles might be dominant and that the standard bubble nucleation picture of the first order phase transition might not work.  Thus, it is indispensable to understand the property of thermal fluctuations in order to elucidate the real dynamics of the phase transition. Among the property of thermal fluctuations, their typical scale and amplitude are particularly important. The former gives us the scale we should focus on, and the latter determines whether thermal fluctuations dominate the dynamics. First, we should notice that in evaluating the magnitude of thermal fluctuations, we should pay attention to fluctuations coarse-grained over some characteristic coherent scale of thermal fluctuations rather than raw fluctuations. If we estimate the typical amplitude out of raw fluctuations, random infinitesimal ones make a dominant contribution to them and we cannot obtain a meaningful value to understand the role of thermal fluctuations.  In this paper we present a formalism to calculate the probability distribution function (PDF) of a coarse-grained scalar field at finite temperature. The PDF of coarse-grained fluctuation has also been studied by Hindmarsh and Rivers \\cite{HR} and by Bettencourt \\cite{Bet}, but they gave only the upper limit of probability.  As an application, we apply our formalism to the electroweak phase transition which may be cosmologically important for the possibility of baryogenesis during it \\cite{KRS}. For successful baryogenesis, it should be of first order to realize an out-of-equilibrium condition \\cite{Sak}. However, if subcritical fluctuations dominate the dynamics of the phase transition it proceeds in a similar manner to a second-order transition and baryogenesis becomes impossible.  Here we evaluate PDF of the Higgs field on the relevant scales in the minimal standard model and investigate whether the universe is homogeneously in the symmetric phase at the critical temperature.  We have also studied this issue in the previous paper \\cite{YY}, solving a phenomenological Langevin equation for the {\\it classical} expectation value of the Higgs field numerically.  But the PDF we formulate here is based on finite-temperature field theory and quantum effects are taken into account more appropriately.  In the next section, we present a formalism of evaluating the PDF for the coarse-grained fluctuations. In Sec.\\ \\ref{sec:application}, the method established in Sec.\\ \\ref{sec:formalism} is applied to the electroweak phase transition. ",
        "conclusions": "\\label{sec:remark}  \\indent  In this paper, we formulate the probability distribution function for the coarse-grained fluctuations, which can be evaluated by calculating perturbatively the connected generating functional $W[J]$. In application, we apply our formalism to the electroweak phase transition. We evaluate the probability distribution function of the field coarse-grained over $2\\pi / \\beta_{c}$ so that we find that the thermal fluctuations called subcritical bubbles play a significant role and that the universe is not homogeneously in the symmetric phase at the onset of the electroweak phase transition. We also calculate the probability $P[\\varphi_{R} \\ge \\varphi_{\\rm inf}]$ and the root-mean-square at several coarse-graining scale. The result is depicted in Table.\\ \\ref{tab:prob}. We can observe that as the coarse-graining scale is larger, the probability $P[\\varphi_{R} \\ge \\varphi_{\\rm inf}]$ becomes smaller and also the RMS smaller. Thus, the coarse-graining scale is very important in investigating the property of thermal fluctuations. Unless the coarse-graining scale is specified properly, the discussion becomes meaningless and brings out confusion.    \\begin{table}[tcb] \\caption{Property of thermal fluctuations at several scales} \\label{tab:prob} \\begin{center} \\begin{tabular}{rcc} \\hline $R~~~~~~~~~~~$ & $P[\\varphi_{R} \\ge \\varphi_{\\rm inf}]$ & $\\sqrt{\\la \\varphi_R^2\\ra}$ \\\\ \\hline $~~\\beta/(2\\pi) \\simeq 0.159\\beta$  & $44\\%$ & $0.76T_c$ \\\\ $10\\beta/(2\\pi) \\simeq 1.59\\beta$ & $22\\%$   & $0.15T_c$ \\\\ $20\\beta/(2\\pi) \\simeq 3.18\\beta$ & $12\\%$  & $9.8\\times10^{-2}T_c$ \\\\ $30\\beta/(2\\pi) \\simeq 4.77\\beta$ & $5.9\\%$ & $7.3\\times10^{-2}T_c$ \\\\ $40\\beta/(2\\pi) \\simeq 6.37\\beta$ & $2.6\\%$ & $5.8\\times10^{-2}T_c$ \\\\ $50\\beta/(2\\pi) \\simeq 7.96\\beta$ & $0.99\\%$& $4.8\\times10^{-2}T_c$ \\\\ $60\\beta/(2\\pi) \\simeq 9.55\\beta$ & $0.32\\%$& $4.1\\times10^{-2}T_c$ \\\\ $70\\beta/(2\\pi) \\simeq 11.1\\beta$ & $8.4\\times10^{-2}\\%$ & $3.5\\times10 ^{-2}T_c$ \\\\ $80\\beta/(2\\pi) \\simeq 12.7\\beta$ & $1.8\\times10^{-2}\\%$ & $3.0\\times10 ^{-2}T_c$ \\\\ $1/M \\simeq 13.7 \\beta$           & $1.6\\times10^{-3}\\%$ & $2.8\\times10 ^{-2}T_c$ \\\\ \\hline \\end{tabular} \\end{center} \\end{table}    \\subsection*"
    },
    "9805/astro-ph9805095_arXiv.txt": {
        "abstract": "Examination of the process of silicon burning, the burning stage that leads to the production of the iron peak nuclei, reveals that the nuclear evolution is dominated by large groups of nuclei in mutual equilibrium. These quasi-equilibrium (QSE) groups form well in advance of the global Nuclear Statistical Equilibrium (NSE).  We present an improved ``minimal'' nuclear network, which takes advantage of quasi-equilibrium in order to calculate the nuclear evolution and energy generation while further reducing the computational cost compared to a conventional \\alp-chain network.  During silicon burning, the resultant \\emph{QSE-reduced} \\alp\\ network is twice as fast as the full \\alp\\ network it replaces and requires the tracking of only half as many abundance variables, without significant loss of accuracy. When the QSE-reduced \\alp\\ network is used in combination with a conventional \\alp\\ network stretching from He to Si, the combined \\alp 7 network provides an accurate approximation for all of the burning stages from He burning to NSE, while tracking only 7 abundances.  These reductions in computational cost and the number of species evolved make the \\alp 7 network well suited for inclusion within hydrodynamic simulations, particularly those in multi-dimension. ",
        "introduction": "Nucleosynthesis calculations perform two functions in astrophysical models. The most important function, in the context of the evolution of the encompassing hydrodynamic model, is the calculation of the rate of thermonuclear energy generation. Where nucleosynthesis occurs, it is often the dominant source of energy, making careful determination of this rate of energy generation vital to accurate hydrodynamic modeling.  The second function of a nucleosynthesis calculation is to determine the evolution of the nuclear species.  For nuclear astrophysics and our understanding of the origin of the elements, this is of paramount importance.  Often the best comparisons of model calculations with observations also depend on detailed knowledge of the abundances of nuclear species.  Unfortunately, accurate calculation of the nuclear evolution is computationally expensive.  For this reason, it is common to include within a hydrodynamic model only a limited approximation for nuclear burning which estimates the rate of energy generation.  The detailed nuclear evolution is then calculated at a later time using the thermodynamic trajectories from the hydrodynamic model.  Such a scheme is referred to as post-processing nucleosynthesis.  The success of such a post-processing scheme depends largely on the accuracy with which the limited approximation which is included within the dynamical calculation can estimate the rate of energy generation. Frequently this inline nuclear burning calculation is a small network, trading a limited ability to follow the nuclear evolution for speed.  Tracking the nuclear evolution from helium burning through to NSE requires a network that includes nuclei from \\alp-particles to Zn.  Silicon burning presents a particular problem as material proceeds from silicon to the iron peak not via heavy ion captures but through a chain of photodisintegrations and light particle captures.  The minimal nuclear set which can follow this evolution is the set of \\alp-particle nuclei; \\alp, \\nuc{C}{12}, \\nuc{O}{16}, \\nuc{Ne}{20}, \\nuc{Mg}{24}, \\nuc{Si}{28}, \\nuc{S}{32}, \\nuc{Ar}{36}, \\nuc{Ca}{40}, \\nuc{Ti}{44}, \\nuc{Cr}{48}, \\nuc{Fe}{52}, \\nuc{Ni}{56}, \\nuc{Zn}{60}. For convenience we will label this full set \\calF\\ and refer to its abundance as \\yf.  Silicon burning in fact presents a larger problem, as the nuclear flow from silicon to the iron peak nuclei does not generally procede through nuclei with N=Z, especially when significant neutronization has occurred (\\cite{HiTh96} : henceforth HT96).  In some models, however, such compromise is made necessary by the computational limitations.  The nature of nuclear network calculations has been extensively reviewed elsewhere (see, \\eg, \\cite{Woos86}, \\cite{ThNH94} or \\cite{Arne96}), so we will discuss it only briefly here.  From a set of nuclear abundances, and the reaction rates for the reactions which link them, one can calculate the time derivatives of the abundances, $\\dot {\\vec Y}$.  With these derivatives, the abundances of the included nuclei, at time $t$, and over timestep $\\Delta t$, are evolved according to the finite difference prescription \\begin{equation} {{\\vec Y(t+\\Delta t)- \\vec Y(t)} \\over {\\Delta t}} = \\begin{array}{ll} \\dot {\\vec Y}(t) & (explicit) \\\\ \\dot {\\vec Y}(t+\\Delta t) & (implicit)\\ . \\\\ \\end{array} \\label{eq:deriv} \\end{equation} For the stiff set of non-linear differential equations which form most nuclear networks, a fully implicit treatment is most successful.  Solving the implicit version of Eq.~\\ref{eq:deriv} is equivalent to finding the zeros of the set of equations \\begin{equation} \\vec \\calZ(t+\\Delta t)\\equiv {{\\vec Y(t+\\Delta t)- \\vec Y(t)} \\over {\\Delta t}} - \\dot {\\vec Y}(t+\\Delta t) =0 \\ . \\end{equation} This is done using the Newton-Raphson method (see, \\eg, \\cite{NumRec}), which is based on the Taylor series expansion of $\\vec \\calZ(t+\\Delta t)$, with the trial change in abundances given by \\begin{equation} \\Delta \\vec Y = \\left( \\partial \\vec \\calZ (t+\\Delta t) \\over \\partial \\vec Y (t+\\Delta t) \\right)^{-1} \\vec \\calZ \\ , \\label{eq:dely} \\end{equation} where $\\partial \\vec \\calZ / \\partial \\vec Y $ is the Jacobian of $\\vec \\calZ$.  Iteration continues until $\\vec Y(t+\\Delta t)$ converges.  The conclusion of silicon burning with nuclear statistical equilibrium (NSE) offers a hint to a more efficient means of evolving nuclear abundances, since the equilibrium distribution reduces the number of independant variables which must be tracked.  NSE grows from several groups of nuclei which form local equilibria.  The existence of these quasi-equilibrium groups frees one from accounting for the changes in the abundances of each of the members of these groups, as the changes within the groups can be accounted for by the changes of a few crucial abundances.  Previous authors, most notably \\cite{BoCF68}: henceforth BCF), have attempted to use quasi-equilibrium to simplify the calculation of silicon burning.  Having noted the computational difficulties encountered by \\cite{TrCG66} in their pioneering nuclear network study of silicon burning, BCF sought to model silicon burning with a single quasi-equilibrium group, stretching from \\nuc{Mg}{24} and \\nuc{Si}{28} through the iron peak nuclei. BCF postulated that the downward flow from this group occurred via $\\rm ^{24}Mg (\\gamma,\\alpha)^{20}Ne$ and that the flow of $\\alpha$-particles downward through \\nuc{Ne}{20}, \\nuc{O}{16}, and \\nuc{C}{12} conserved the abundances of these nuclei.  From these assumptions, they derived expressions for the abundances of these light $\\alpha$-particle nuclei and for the downward flux of $\\alpha$-particles which depended only on the abundances of $\\alpha$-particles and \\nuc{Si}{28}.  Though this model was moderately successful in replicating the abundance pattern found in the solar system from A=30 to A=60, it had clear limitations when compared to the network calculations of TCG, particularly at early times when the assumption of conservative flow was most suspect.  \\cite{WoAC73}: henceforth WAC) concluded that such efforts were at best marginally successful, in part because as matter cools the photodisintegration reactions freeze out well before the particle captures cease, thereby destroying quasi-equilibrium.  WAC found a further complication in that, at early times, their models exhibited two separate QSE groups, one around silicon and the other containing the iron peak nuclei.  WAC did demonstrate that, with the assumption of QSE, reasonable estimates of the time required to merge the two QSE groups and the time to reach silicon exhaustion could be calculated.  HT96 demonstrated additional complications as both the path of the nuclear flow from the silicon group to the iron peak group and the membership of these groups depended strongly on the amount of neutronization.  HT96 further demonstrated that for material which has undergone significant neutronization, the separation between the silicon and iron peak groups persists through a significant part of the nuclear evolution.  In spite of these complications, \\cite{HiTh98}: henceforth HT98) showed that QSE could provide a reasonable estimate of the abundances of many species during silicon burning, even under explosive conditions, where the photodisintegration reactions freeze out before their corresponding particle captures.  \\cite{Arne96} provides a more complete history of the use of QSE. ",
        "conclusions": "In this paper, we have demonstrated the ability of QSE to reduce the number of independant nuclear abundances which need to be evolved to track \\alp-chain nucleosynthesis for silicon burning.  An \\alp\\ network reduced in this fashion is capable of providing reasonable estimates of the nuclear energy generation and elemental production, even in cases where thermodynamic variations result in incomplete silicon burning or a freezeout rich in free \\alp-particles.  Computational, this reduction in the number of independent nuclear abundance variables offers two advantages.  First, within the nuclear network calculation, the smaller number of variables results in a smaller matrix to build and solve. This speeds the nucleosynthesis calculation by a factor of 2 in the case of the \\alp\\ network, with the potential for much greater speed increases for larger networks.  Second, the reduction in the number of nuclear variables also reduces the number of equations which must be solved within the encompassing hydrodynamic model, a matter of particular concern for models with large numbers of zones or grid points, especially multi-dimensional studies.  The relative importance of these 2 advantages depends on the size of the network.  As a rule of thumb, a PPM hydrodynamic code evolving 45 nuclear abundances spends roughly half its time on the network solution (\\cite{Arne98}).  Thus for small networks, like the \\alp\\ network, the reduction in necessary number of hydrodynamic equations is the paramount advantage.  In an effort to streamline the computation necessary for advanced nuclear burning stages, we have combined a smaller \\alp\\ network (for calculating the nuclear evolution from helium burning through oxygen burning) with this QSE based method.  The resulting combined network, which we have dubbed \\alp 7, can be used in place of the full 14 element \\alp\\ network from helium burning through to NSE without significant errors in energy generation or nucleosynthesis. Such a combination offers the same reductions in the computational cost discussed above for silicon burning for all burning stages beyond H burning.    In a future paper \\cite{HiFT99}, we will apply the method of QSE-reduction to larger nuclear networks, where the potential for improvement in speed and size is even greater."
    },
    "9805/astro-ph9805338_arXiv.txt": {
        "abstract": "Using a realistic model for line emission from the broad emission line regions of quasars, we are able to reproduce the previously observed correlations of emission-line ratios with the shape of the spectral energy distribution (SED). In agreement with previous studies, we find that the primary driving force behind the Baldwin Effect ($W_\\lambda\\/ \\propto L^{\\beta}$, $\\beta < 0$) is a global change in the SED with quasar luminosity, in that more luminous quasars must have characteristically softer ionizing continua. This is completely consistent with observations that show (1) a correlation between $L_{uv}$ and $\\alpha_{ox}$ and $\\alpha_{uvx}$, (2) correlations of SED shape sensitive line ratios with $\\alpha_{ox}$, $\\alpha_{uvx}$, and $L_{uv}$, and (3) correlations between line equivalent widths and $\\alpha_{ox}$, $\\alpha_{uvx}$, and $L_{uv}$.  However, to explain the complete lack of a correlation in the $W_\\lambda$(\\ion{N}{5}) -- $L_{uv}$ diagram we propose that the more luminous quasars have characteristically larger gas metallicities ($Z$). As a secondary element, nitrogen's rapidly increasing abundance with increasing $Z$ compensates for the losses in $W_\\lambda$(\\ion{N}{5}) emitted by gas illuminated by softer continua in higher luminosity quasars. A characteristic relationship between $Z$ and $L$ has an impact on the $W_\\lambda$ -- $L_{uv}$ relations for other lines as well.  For a {\\em fixed} SED, an increasing gas metallicity reduces the $W_\\lambda$ of the stronger metal lines (the gas cools) and that of Ly$\\alpha$ and especially \\ion{He}{2} (because of the increasing metal opacity), while the weaker lines (e.g., \\ion{C}{3}] 1909) generally respond positively.  The interplay between the effects of a changing SED and $Z$ with $L$ results in the observed luminosity dependent spectral variations.  All of the resulting dependences on $L_{uv}$ are within the range of the observed slopes. ",
        "introduction": "Since their discovery as the most distant and luminous discrete objects in the Universe, it has been hoped that somehow quasars could be used as ``cosmological candles'' to measure the expansion parameters ($q_o$ and $H_o$) of the universe. With the discovery of the Baldwin Effect (Baldwin 1977), whereby the emission line equivalent widths scale inversely with the quasar luminosity, that hope is half-realized. Unfortunately, the very large scatter in these correlations (Baldwin, Wampler \\& Gaskell 1989; Kinney, Rivolo \\& Koratkar 1990; Osmer, Porter \\& Green 1994) have curtailed their usefulness. What remains is to understand the mechanism of the formation of quasar emission lines well enough so that this scatter can somehow be calibrated in terms of observable parameters, and then removed.  A second major goal of studies of distant quasars is to use them as probes of the first generations of nucleosynthesis in the young Universe. There is evidence of wide variations in the atomic abundances in the BLRs of different quasars, and that the spectroscopic effects of this can be measured out to very large redshifts (Hamann \\& Ferland 1992; Hamann \\& Ferland 1993: HF93; Ferland et al 1996). Again, full interpretation requires a proper understanding of the formation of quasar emission lines.  There is mounting evidence that the line strengths strongly depend on the spectral energy distribution (SED) in the UV -- X-Ray range. (Netzer, Laor, and Gondhalekar 1992, Zheng et al. 1995, Green 1996, Wang et al.  1998). In addition, many studies have found a correlation between $\\alpha_{ox}$ and luminosity (cf.  Zamorani et al.  1981; Wilkes et al. 1994; but see also LaFranca et al 1995 and Avni et al 1995 for discussion of conflicting results).  This link through $\\alpha_{ox}$ offers a promising explanation of the observed $W_\\lambda$--$L_{uv}$ relations (Wang et al 1998; Green 1998).  Here we investigate to what extent we can reproduce these observed effects using realistic photoionization models of the BLR, and study the degree to which the additional parameter of metallicity must also be taken into account. ",
        "conclusions": "We have shown here that including a $Z$--$L$ effect in addition to a SED--$L$ relation allows us to simultaneously reproduce the observed correlations with $L$ of the equivalent widths and intensity ratios of all of the strong UV lines, particularly \\ion{N}{5}.  As a secondary element, nitrogen's rapidly increasing abundance with increasing $Z$ compensates for the losses in $W_\\lambda$(\\ion{N}{5}) emitted by gas illuminated by softer continua in higher luminosity quasars. For a {\\em fixed} SED, an increasing gas metallicity reduces the $W_\\lambda$ of the stronger lines of other heavy elements (the gas cools) and that of Ly$\\alpha$ and especially \\ion{He}{2} (because of the increasing metal opacity), while the weaker metal lines (e.g., intercombination lines such as \\ion{C}{3}] 1909) generally respond positively. Weak lines such as \\ion{O}{4}] $\\lambda$1402, formed in the He$^{++}$ zone, suffer from the significant drop in temperature and their intensities remain roughly constant with increasing $Z$.  The interplay between the effects of a changing SED and $Z$ with $L$ results in the observed luminosity dependent spectral variations.  However, the photoionization models only predict relations with $\\alpha$ and $Z$. To get to relations with $L$ we have had to adopt both an $\\alpha$--$L$ relation, and a $Z$--$L$ relation. These two relations or at least their combined effect must still be calibrated empirically, from comparing emission line parameters to continuum fluxes, before the Baldwin Effect can be used as a real luminosity indicator for cosmological tests.  Why should we expect a global relationship between $Z$ and quasar luminosity? HF93 and Ferland et al.\\ (1996) showed that the average \\ion{N}{5}/\\ion{C}{4} and \\ion{N}{5}/\\ion{He}{2} ratios increase systematically with redshift and/or luminosity. They attributed this change in line ratios to an increase in metallicity, as we do here. HF93 proposed that the luminosity dependence of metallicity is really a mass-metallicity relation.  Because of their deeper potential wells, higher-mass galaxies are better able to retain gas deposited during stellar evolution than are lower-mass galaxies, so they have a more rapid buildup of heavy elements in their nuclear regions. Coupling this scenario with observations showing that the most luminous quasars tend to reside in the most massive galaxies (McLeod \\& Rieke 1995a, 1995b; Bahcall et al.\\ 1997) leads to a natural explanation for a global relationship between quasar luminosity and gas metallicity. To the extent that luminosity evolution affects individual quasars, the redshift-metallicity dependence would also require a steady decrease in metallicity with time in individual objects. HF93 suggested that this was due to dilution by lower-$Z$ gas falling into the nuclear regions throughout the lives of these galaxies.  In addition, the simplest accretion disk models predict that more luminous quasars should in general emit softer continua (Netzer et al.\\ 1992), though the continuum emission mechanism is far from understood. Together, these form the basis of the proposed global $L$--SED--$Z$ relationship.  Can the noise in the Baldwin Effect be eliminated, much as an observer corrects for reddening?  Metallicity should scale with the mass of the host galaxy and not necessarily the quasar luminosity, thus introducing noise.  The importance of this effect can be estimated from Figure~3, which compares the observations of Osmer et al. (1994) to the range covered by our models on the $W_\\lambda$(\\ion{C}{4})--$L$ diagram as metallicity is varied from half solar to ten times solar. Just making a correction on the basis of the \\ion{N}{5}/\\ion{C}{4} ratio may remove much of the scatter at high redshifts; this is something that can be tested empirically. There are likely to be additional sources of scatter as well: a spread in $\\alpha$ at given $L$; the strength of any additional EUV bump above a simple power-law continuum; variations in the cloud distributions; line and continuum emission anisotropies; variability (the effects of line--continuum lags and the intrinsic Baldwin Effect; Kinney, Rivolo, \\& Koratkar 1990; Pogge \\& Peterson 1992). Our aim is to use the more accurate BLR model described here together with the measurements of critical line and continuum strengths to distinguish between as many of these effects as possible and guide us in their empirical calibration.  Even a factor of two reduction in the scatter would be an important gain in our ability to use these diagrams for cosmology."
    },
    "9805/astro-ph9805081_arXiv.txt": {
        "abstract": "By  correcting the redshift--dependent  distances for peculiar motions through  some peculiar  velocity field   models,  we recover the  true distances of an  extensive,  all-sky sample of nearby  galaxies ($\\sim 6400$ galaxies with  recession  velocities  $cz<5500\\;$\\ks),  which is complete    up   to   the     apparent     limiting blue     magnitude $B=14\\;$mag. Relying on available catalogs  of galaxy groups, we treat $\\sim$2700 objects  as  members  of galaxy  groups and  the  remaining objects as field galaxies.  We  invert  the  derived    redshift-distance relations   to  estimate distances for field galaxies and  groups and we overcome the ambiguity inherent  to the triple-valued   zones by using Tully-Fisher relations calibrated on  suitably defined samples  of galaxies  having distances predicted by peculiar velocity models.  We use two  independent approaches to  modeling the  peculiar velocity field: i)  a cluster dipole  reconstruction scheme that we modify with the    inclusion of a local      model of Virgocentric-infall; ii)   a multi--attractor model fitted to the Mark II  and Mark III catalogs of galaxy peculiar  velocities.  In the multi--attractor model  we assume that the velocity field is generated  by a few prominent gravitational sources      (Virgo     cluster, \"Great   Attractor\",   Perseus-Pisces Supercluster, and Shapley supercluster).  We discuss differences in   the  results  coming out from    different velocity  models  and  from  different   Mark  II and  Mark III   data subsets. In particular, according to Mark III data the Great Attractor appears to have a smaller  influence on local dynamics than previously believed, whereas the Perseus-Pisces and Shapley superclusters acquire a specific dynamical role. Remarkably, the Shapley structure, which is found  to  account for  nearly half  the peculiar  motion of the Local Group, is placed by Mark III data closer to the zone of avoidance with respect  to  its optical  position.  On  the other  hand, the modified cluster   dipole model is   characterized  by  relatively small  flows towards the  Great  Attractor and the   Shapley supercluster, together with a large Virgocentric infall.  Our multi--attractor model  based    on   Mark  III data   favors    a cosmological  density  parameter $\\Omega_0\\sim$0.5  (irrespective of a biasing factor of order unity).  The use of different peculiar velocity field models allows us to check to what extent differences in current views on cosmic flows affect the recovering  of galaxy   distances.   We find  that  differences  among distance estimates are less  pronounced in the  $\\sim 2000 - 4000$ \\ks distance  range than at  larger  or  smaller  distances.  In the  last regions these differences have a serious impact  on the 3D maps of the galaxy distribution  and on  the local  galaxy  density  ---  on small scales ($<$  1 Mpc) ---, which  is  a crucial parameter  being used in statistical   studies of environmental  effects   on the properties of nearby galaxies.  \\vspace*{6pt} \\noindent {\\em Subject headings: } galaxies: distances and redshifts --- galaxies: clusters: general --- cosmology: large--scale structure of universe ",
        "introduction": "The determination of  galaxy distances is one  of the most fundamental problem in  extragalactic astronomy.  In particular, galaxy  distances allow  us to map the  three--dimensional (3D) distribution of galaxies and hence to evaluate the local galaxy  density, which is an important characteristics of the galaxy environment.  In many previous studies which rely only on the two-dimensional (2D) projection of galaxies on the sky (see, e.g., Lahav (1987) for all-sky 2D optical maps of bright galaxies), a local galaxy density parameter is defined in terms of galaxy counts over a sky area which are statistically corrected for projection effects. With the advent of large surveys of galaxy redshifts and well-selected galaxy catalogs, it has become possible to map the spatial distribution of galaxies and hence to construct a 3D definition of the local galaxy density.  Optical galaxy samples are more suitable for mapping the galaxy density field on relatively small scales than IRAS-selected galaxy samples, which have been frequently used as tracers of the density field on large scales up to a recession velocity $cz\\sim20000\\;$ \\ks (e.g., Rowan-Robinson et al., 1990; Kaiser et al., 1991; Strauss et al., 1992a,b; Fisher et al., 1995). The latter samples have the advantage that they are homogeneously selected and penetrate to lower galactic latitudes because IRAS fluxes are much less impeded by Galactic extinction. But the IRAS galaxy samples have the serious drawback that they do not include the early-type galaxies (because of their little dust content and star formation), which usually lie in the densest regions of clusters and superclusters.  Furthermore, the density field traced by IRAS galaxies is noisy nearby, because of the sparseness of the samples and because IRAS fluxes are much less linked with galaxy mass than optical fluxes. Therefore, in the following we shall focus on all-sky optical galaxy samples.  The magnitude-limited Revised Shapley-Ames (RSA) Catalog of Bright Galaxies (Sandage \\& Tamman, 1981), which contains the first large compilation of redshift data for bright galaxies across the entire sky, was used by Yahil, Sandage \\& Tamman (1980) to delineate the 3D-density field of galaxies in the Local Supercluster (LS).  The structures of the LS region have been comprehensively described and named by Tully \\& Fisher (1987). The former author intended to include all nearby galaxies with systemic recession velocities $cz<3000\\;$\\ks (2367 objects) in his Nearby Galaxies Catalog (Tully, 1988a, hereafter NBG), which is a combination of the RSA catalog and a diameter-limited sample of late-type and fainter galaxies found in an all-sky HI survey made principally by Fisher \\& Tully (1981) and Reif et al. (1982). In the NBG catalog, which is complete up to the corrected blue total magnitude $B_T\\sim12$~mag (although it extends to fainter magnitudes), the distances of all non-cluster galaxies have been estimated on the basis of their redshifts, with an assumed Hubble constant $H_0=75\\;km s^{-1} Mpc^ {-1}$, whereas the galaxy members of systems with relatively high velocity dispersion have been given a distance consistent with the mean redshift of the system. The important fact that redshifts are not equivalent to distances because galaxies have peculiar motions (i.e. deviations from the pure Hubble flow) is simply taken into account in the NBG catalog through the Virgocentric retardation model described by Tully \\& Shaya (1984), in which the authors assume that the Milky Way is retarded by 300 \\ks from the pure Hubble flow by the mass of the Virgo cluster. Defining the local galaxy density on the basis of his NBG catalog, Tully (1988b) incorporated corrections for the catalog incompletion at large distances.  Later, similar quantifications of the local galaxy density, based on the NBG catalog data, have been exploited in statistical analyses of environmental effects on some properties of LS galaxies such as the frequency of bars (Giuricin et al., 1993), the classification of spiral arms (Giuricin et al., 1994), the frequency of active galactic nuclei (LINER and Seyfert objects) selected from optical spectroscopic surveys (Monaco et al., 1994), and the bulge-to-disk light ratio (Giuricin et al., 1995).  In an effort of going beyond the LS, Hudson (1993a,b; 1994a,b) assembled an extensive galaxy sample from a merging of the diameter-limited northern UGC catalog (Nilson, 1973) and the diameter- limited southern ESO catalog (Lauberts, 1982; Lauberts \\& Valentijn, 1989). He took into account the fairly large incompleteness in redshift of his sample as a function of angular diameter and position on the sky, by applying statistical corrections, which allowed him to reconstruct the density field of optical galaxies to a depth of $cz=8000$ \\ks.  The Lyon-Meudon Extragalactic Database (LEDA), which collects and homogenizes several data for all the galaxies of the main optical catalogues, such as the catalogs UGC (Nilson, 1973), ESO (Lauberts, 1982, Lauberts \\& Valentijn, 1989), CGCG (Zwicky et al., 1961-1968), and ESGC (Corwin \\& Skiff, 1998, in preparation), allows the extraction of an all-sky galaxy sample which has properties of completeness and is deep enough to probe the galaxy field significantly beyond the LS. In particular, Garcia et al. (1993) extracted a magnitude-limited sample which covers all the sky and contains 6392 galaxies having recession velocities $cz<5500$ \\ks. Although different optical catalogs are characterized by different limits of completeness in apparent magnitude or angular diameter, the above-mentioned galaxy sample was found to be substantially complete up to its limiting corrected total blue magnitude $B_T=14$~mag (for galactic latitudes $|b|>20^{\\circ}$) (Garcia et al., 1993).  The authors tabulated several parameters for each galaxy, such as the morphological type, the maximum velocity rotation deduced from the deprojected 21-cm hydrogen line width, the de Vaucouleurs' luminosity index, the distance modulus, the corrected angular sizes and total blue magnitudes. The authors derived the last two quantities by transforming the original raw data to the standard system of the RC3 catalog (de Vaucouleurs et al., 1991) and by applying corrections for Galactic extinction, internal extinction, and K-dimming. Only for the few galaxies which lack magnitude measures, the values of $B_T$ are rough estimates derived from their diameters through a mean magnitude-diameter relation. The tabulated distance modulus was a weighted mean derived from the combination of redshift-distances (with $H_0=75\\;km\\;s^{-1} Mpc^{-1}$) and redshift-independent distances obtained from distance indicators (DIs), i.e. from a Tully-Fisher (TF) relation (which relates the absolute magnitude $M_B$ to the logarithmic maximum rotation velocity $\\log V_m$) and a luminosity-luminosity index relation.  In particular, the authors used the former kind of distances alone for the farthest galaxies (with $cz>1500\\;$ \\ks) and the latter kind alone for the few nearest objects (with $cz<500\\;$ \\ks).  Group assignments for the galaxies of this sample have been provided by Garcia (1983). She employed two 3D methods of group identification, the percolation or {\\it friends of friends} method proposed by Huchra \\& Geller (1982) and the hierarchical clustering method (Materne, 1978; Tully, 1980, 1987). The latter method gave, on average, smaller groups than the former. The adopted final catalog of groups was defined as that one which includes only groups (as well as galaxy members) common to the two catalogs.  Accordingly, 3381 objects are field galaxies. Of the remaining grouped objects, 2703 galaxies are members of 485 systems with at least three members; most of them are groups which contain less than 10 members, and 13 systems have more than 20 members (among which the clusters Virgo, Centaurus, Perseus).  A somewhat different approach to the construction of a large and deep sample of optical galaxies with good completeness in redshift was followed by Santiago et al. (1995). Their \"Optical Redshift Survey\" is a redshift survey of an optical galaxy sample which contains 8266 galaxies with known redshift. This sample consists of two largely overlapping subsamples (which are limited in magnitude and in diameter, respectively) drawn from the catalogs UGC, ESO, ESGC. The authors selected their own sample according to the raw (observed) magnitudes and diameters (rather than the corresponding corrected values) and then quantified the effects of Galactic extinction (as well as random and systematic errors) on the optical density field. Redshift-distances were used in general, but the members of rich clusters were placed at each cluster's center. Probing the density field at large depth (up to $cz=8000$~ \\ks) requires much care in the derivation of appropriate selection functions, which quantify the loss of galaxies due to magnitude or diameter limits or other observational selection effects. Taking into account possible non-uniformities in galaxy sampling, the authors attempted to match the original three catalogs from which their sample was drawn, by using separate selection functions (Santiago et al., 1996).  The reliability of the distances given in all above-mentioned galaxy samples (and hence of the resulting galaxy density field) is weakened by the fact that simple redshift--distances are in general adopted with no corrections (or no adequate corrections) for the peculiar motions which are due to inhomogeneous mass distribution and which prevent redshift to scale linearly with distance.  In the present paper we propose to recover the true distances of a sample of nearby galaxies by treating adequately the effects of peculiar velocities of galaxies. As shown in a forthcoming paper (Marinoni et al., 1998b), in which the resulting 3D galaxy density field is presented and local galaxy density parameters (corresponding to various smoothing scales) are discussed, corrections of galaxy distances for peculiar motions appear to have a large impact on the values of the aforementioned parameters on small scales (i.e. on $<1$ Mpc scale) and, hence, on the characterization of the environment (see \\S 6).  For our purpose, we take the above-mentioned complete sample extracted from LEDA (Garcia et al., 1993; Garcia, 1993), for which all galaxy data as well as group assignments are available in the literature. Thus, we limit ourselves to the nearby universe (within $cz=5500$~\\ks), where the peculiar velocity field is best known and no large corrections for galaxy sample incompletion are required. However, the galaxy sample considered is deep enough to cover interesting regions of prominent overdensities, such as the \"Great Attractor\" (GA) region and the Perseus-Pisces supercluster.  The study of the peculiar velocity field resulting from a proper analysis of large redshift--distance samples (i.e., samples of galaxies having both redshift and redshift-independent distance estimates) has acquired a well-established role in the cosmological context (e.g., for testing the gravitational instability paradigm for the origin of large--scale structure, for deducing the relative distributions of luminous and dark matter, and for constraining the value of the cosmological density parameter $\\Omega_0$; see, e.g., the reviews by Dekel, 1994, and Strauss \\& Willick, 1995).  In general, peculiar velocity analyses proceed along two main lines of research. One can turn a distribution of galaxies or galaxy systems with known redshifts into a smoothed mass-density field (under the linear approximation of the gravitational instability theory) and then predict from it the peculiar velocity field , e.g., by means of the \"dipole analysis\". Despite the complexity of these studies, the velocity vector of the Local Group (LG) with respect to the reference frame defined by the cosmic microwave background (CMB) ($v_{LG}\\sim 630$ \\ks towards $l \\sim 277^{\\circ}$ and $b \\sim 30^{\\circ}$ (Lubin \\& Villela, 1986; Kogut et al., 1993)) has been approximately reconstructed in terms of the gravitational acceleration exerted by various populations of extragalactic objects (e.g., the references cited in Kolokotronis et al., 1996).  Alternatively, one can face the problem of deriving the true matter distribution from the observed peculiar velocity field, e.g using few-parameter toy models or the non-parametric POTENT method developed by Bertschinger \\& Dekel (1989).  In this paper, following the second kind of approach, we attempt to correct the Hubble flow for peculiar motions by means of two basic independent approaches to modeling the velocity field in the nearby universe:  i) a modified cluster dipole model, which is Branchini \\& Plionis' (1996, hereafter BP96) optical cluster dipole reconstruction scheme that we modify by including a local model of the Virgocentric infall in the LS region. We regard the cluster dipole model as a typical approach in which the peculiar velocity field is self-consistently derived through iteration techniques from the redshifts and positions of a sample of objects. ii) a multi-attractor model, which we take as an example of a parametric toy model which is fitted to catalogs of peculiar velocities. In the multi-attractor model we assume that the velocity field is generated by a few prominent gravitational sources (Virgo cluster , Great Attractor, Perseus-Pisces Supercluster, Shapley concentration).  We construct these models relying on the various data subsets contained in the Mark II and Mark III catalogs of peculiar velocities of galaxies. In particular, the Mark III catalog (Willick et al., 1997a), which is the most comprehensive homogenized catalog of peculiar velocities available today, is a merging of several data sets of spiral and elliptical galaxies.  Through these peculiar velocity field models we provide homogeneous estimates of distances for our galaxy sample. Moreover, we can evaluate to what extent differences in current views on the peculiar motions affect the recovering of galaxy distances and, hence, the 3D maps of galaxy distribution.   The outline of our paper is as follows. In \\S 2 we describe the properties of the Mark II and Mark III catalogs which are relevant to our velocity field models. In \\S 3 we present our two basic approaches to modeling the peculiar velocity field and we discuss the results relative to different data sets. In \\S 4 we address the inversion of the redshift-distance relation and, in particular, the ambiguity inherent to the triple-valued zones of this relation. We attempt to solve this problem using blue TF relations calibrated on suitably defined samples of galaxies having distances predicted by velocity field models.  In \\S 5, outlining some basic differences among the velocity field models, we discuss the resulting sets of galaxy distances. In our conclusions (see \\S 6) we mention some developments of peculiar velocity studies and focus on the effects of our galaxy distance corrections on the determination of the 3D galaxy density field.  Throughout, the Hubble constant is $75\\;h\\;km\\;s^{-1}\\;Mpc^{-1}$. In general, distances are expressed in velocity units (\\ks). ",
        "conclusions": ""
    },
    "9805/astro-ph9805048_arXiv.txt": {
        "abstract": "The modern Doppler Imaging (DI) technique allows the reconstruction of different stellar surface structures based on accurate calculation of spectra of specific intensity. New applications like the mapping of the magnetic field vector put very stringent requirements on the radiative transfer (RT) solver which should be accurate, fast, and robust against numerical errors. We describe the evaluation of three different algorithms for our new magnetic DI code INVERS10. We also show the first results of numerical experiments made with the new code. ",
        "introduction": "{\\em Doppler Imaging} is a method that allows to reconstruct the stellar surface structures from the rotational modulation of spectral line profiles (e.g. Piskunov \\&\\ Rice 1993). DI is an {\\em iterative process}. On each iteration the observational data (spectra taken at different rotational phases) are computed for the current surface distribution at each observed wavelength $\\lambda$ and rotational phase $\\phi$. This is done by integrating the specific intensity $I_\\lambda(M)$ over the visible hemisphere. {\\bf $I_\\lambda(M)$ must be computed at each iteration for each $\\lambda$, $\\phi$, and surface element $M$ in a spectral range large enough to accommodate rotational Doppler shifts}. Therefore, spectral synthesis of specific intensities is the most time consuming part of any DI algorithm!  With the improvement of computer performance and the expansion of the DI applications the need for a new RT became obvious as other solutions (e.g. tables of pre-computed local profiles) have been exhausted. ",
        "conclusions": "Although many more experiments will be required to investigate all the properties of the new code, even now it is clear that we can {\\em reliably reconstruct the vector magnetic field} and that {\\em the observations of all four Stokes parameters are required}. It is also clear that the MDI problem must be solved in a consistent way rather then by separately imaging magnetic field and abundance (or temperature), since (at least with incomplete observations) one of the variables can successfully mimic the other."
    },
    "9805/astro-ph9805032_arXiv.txt": {
        "abstract": "Observational data on rotation of Ap stars suggest that the bulk of their rotation rates form a separate Maxwellian distribution with an average value 3-4 times lower than the normal star distribution. No evidences for a significant angular momentum (AM) loss on the main sequence (MS) have been found. It is thus concluded that Ap stars must lose a large fraction of their initial angular momentum (AM) in the pre-MS phase of evolution, most probably as a result of the interaction of their primordial magnetic fields with accretion disks and stellar winds. The observationally most acceptable values of accretion rate from the disk, $10^{-8}  M_{\\sun}$ /year, of mass loss rate via a magnetized wind, $10^{-8} M_{\\sun}$ /year, and of the surface magnetic field, 1 kG on the ZAMS, result in the AM loss in full agreement with observations.  There exists a separate group of extremely slowly rotating Ap stars, with periods of the order of 10-100 years. They are too numerous to come from the distribution describing the bulk of Ap stars. It is conjectured that their extremely low rotation rates are the result of additional AM loss on the MS. ",
        "introduction": "\\label{intr}  The A and B type stars with peculiar spectra have been recognized as having unusually sharp spectral lines as soon as their spectra were analyzed with large enough dispersion. In fact, low \\v values were the reason why Babcock (1947) included them into his search for stellar magnetic fields. He believed at that time that strong magnetic fields are coupled with rapid rotation but because of instrumental reasons he selected early type stars with sharp lines, assuming that the low values of \\v are the result of low $i$, not $v$ values. Later, when the variation periods of Ap stars (typically of several days) were identified as their rotation periods,  according to the oblique rotator theory (Stibbs, 1950), it became apparent that the low \\v values are the result of the low rotation rates of these stars. The presently known values of the rotation periods of peculiar stars are in agreement with the conclusion that the stars rotate on average much slower that normal stars of the same spectral types (Catalano \\& Renson, 1997 and references therein).  The observed chemical peculiarities of Ap stars are explained as resulting from the diffusion of different ions under the influence of the radiative force (Michaud, 1970; Michaud \\& Proffitt, 1993). A slow rotation rate was assumed to be {\\it condicio sine qua non} for a diffusive segregation of elements to occur in atmospheres of these stars although the role of the magnetic field must also be very important for the development and distribution of over- and under-abundances over the stellar surface (Pyper, 1969; Babel 1993). Recently Abt \\& Morrel (1995) went one step further suggesting that the slow rotation is also a sufficient condition for the Ap star phenomenon to occur. They propose the existence of a threshold for A stars, such that all stars rotating slower than that should be chemically peculiar, even if not yet recognized as such.  \\begin{table}[t] \\small \\begin{center} \\caption{Ap stars with extremely long periods.} \\label{t1} \\begin{tabular}{llll} \\hline\\hline HD & Sp & $P$ (years)\\\\ \\hline 9996 & A0 SiCrSr & 22\\\\ 94660& A0 EuCrSi & $\\ga$ 7.5\\\\ 110066& A1 SrCrEu& 13.5 or 27\\\\ 137949& F0 SrCrEu& $\\ga$ 75\\\\ 187474& A0 EuCrSi& 6.5\\\\ 201601& A9 SrEu &  75\\\\ \\hline\\hline \\end{tabular} \\end{center} \\end{table}  The rotation distribution of Ap stars cannot be interpreted as a slow rotation tail of the normal star distribution. A statistical investigation of rotation velocities of peculiar stars indicates that their distribution can be approximated by the Maxwellian one with an average value 3-4 times lower that for normal stars (Preston, 1970; Wolff, 1981; Abt \\& Morrel, 1995). In addition, several Ap stars with extremely long rotation periods of more than 5 years are known (Table 1). Note that all belong to cooler Ap stars of the CrSrEu type. The periods of these stars, if interpreted as rotation periods, cannot be part of the Maxwellian distribution describing the bulk of the rotation periods of peculiar stars, because the probability of finding even one rotator with such a long period in the known sample of Ap stars is exceedingly small (Preston, 1970). We have to assume that they form a separate population of Ap stars in which a special mechanism of spin down is (or has been) operating.  The slow rotation of A and B peculiar stars can be a result of one or more of the following circumstances: (i) they are formed from protostellar clouds with particularly low angular momentum (AM), (ii) they lose extra AM in the pre main sequence (PMS) phase of evolution, (iii) they lose AM on the MS.  The first possibility seems unlikely. Open clusters contain a substantial number of peculiar stars which have been formed simultaneously with the other cluster members. Had they been formed from a low rotation tail of the protostellar cloud distribution, they would have become part of the same stellar distribution, which contradicts the observations.  The second and third possibility were a subject of investigation of several authors in the past. Abt (1979) measured \\v values of Ap stars in a number of stellar clusters of known ages and found a significant decrease of the rotation rate of at least hot peculiar stars with age. That was confirmed by Wolff (1981). Unfortunately, their conclusion relied heavily on the data points from the Orion association for which only three \\v values had been measured. Wolff (1981) discussed also possible mechanisms for AM loss of Ap stars, based on the interaction of their magnetic fields with the stellar environment. More accurate and numerous data on the rotation periods of peculiar stars in young clusters obtained later, showed that the period distribution of peculiar cluster members is indistinguishable from the distribution of field Ap stars, assumed to be much older (Borra et al. 1985; North, 1984a, 1987). North (1984b) discussed the dependence of the observed period of field Ap stars on gravity, treated as a measure of age for a given mass. He concluded that the rotation period increases when $\\log g$ decreases just as expected from the conservation of AM during the MS life, without ``...the least suggestion of any braking mechanism''. But recently Pyper et al. (1998) showed that the short-period Ap star CU Vir abruptly increased its period by about $5 \\times 10^{-5}$ of its value. Because the reason for this change seems at present completely obscure, and it is not clear what is its relation to possible evolutionary period changes during the MS life, that case will be ignored and it will be assumed in the following that the observational data do not show significant AM loss of Ap stars during their MS life.  This leaves us with the hypothesis that progenitors of peculiar stars are born with normal rotation rates but they lose a large fraction of their initial AM in the PMS phase, hence they rotate 3-4 times slower than normal stars when they land on the ZAMS. This hypothesis will be the subject of the rest of the paper. ",
        "conclusions": "The calculations have been carried out for two values of the stellar mass, 2 and 3 $M_{\\sun}$. The time scales of the PMS evolution, and the dependence of the moment of inertia and stellar radius on time, i.e. $I(t)$ and $R(t)$, were taken from models computed by Palla \\& Stahler (1993) (see also B\\\" ohm \\& Catala, 1995).  Based on the observational results about the accretion rate and mass loss via winds of Herbig Ae/Be stars given in Section 1, the following values have been adopted as typical: ${\\dot M}_{acc} = 10^{-8} M_{\\sun}/{\\rm year}$ and ${\\dot M}_{wind} = 10^{-8} M_{\\sun}/{\\rm year}$. The value of the magnetic moment was adopted as $\\mu = 2.7\\times 10^{36}$ in cgs units, which corresponds to the 1 kG dipole field on a 2 $R_{\\sun}$ star. To allow for a possible variation of these values as well as other factors modifying the efficiency of all the considered mechanisms, arbitrary weights were added to the the first three terms in equation (5).  Figure 1 (upper) compares the variation of the rotation period of a 2 $M_{\\sun}$ star when its AM is preserved during the PMS evolution (solid line) and when the consecutive AM change mechanisms, described by the first three terms in brackets on the right hand side of equation (5) are added. It is assumed that the star emerges from the protostellar phase after $10^6$ years with a rotation period of 5 days, and the PMS phase ends after $8\\times 10^6$ years (Palla \\& Stahler, 1993). When AM is conserved in a rigidly rotating star its rotation period decreases down to a value of 0.55 of a day on the ZAMS (solid line). When only the accretion is added, the ZAMS period is even shorter, and it is equal to 0.15 of a day (dotted line), because the accretion of a high AM matter from the disk increases the stellar AM, hence spins up the star. When only a wind is added, the resulting ZAMS period reaches a value of 1.35 day (dotted-broken line). The most powerful mechanism influencing the rotation period of a PMS star is the interaction of its magnetic field with the disk. If the disk is massive enough, it will force the stellar rotation in a relatively short time to an approximate value of the corotation period at the edge of the magnetosphere (Armitage \\& Clarke, 1996). In our case this is close to 5 days, hence the rotation period of the considered star stays close to this value through the whole PMS evolution (broken line). For the initial rotation periods shorter than 5 days the field-disk linkage slows down the rotation but for values longer that that it spins up the star, so that the final value of about 5 days is always reached. \\begin{figure}[t] \\psfig{figure=T34f1.eps,height=12cm} \\caption{The results of model calculations for a 2 $M_{\\sun}$ and different mechanisms for stellar AM change (see text)} \\label{fp} \\end{figure}  Figure 1 (lower) demonstrates the result of a simultaneous action of all the considered mechanisms. The constant AM case is repeated from the upper part of the figure for comparison as a dotted-broken line. The solid line describes the evolution of the rotation period of the considered star when all terms in equation (5) are taken into account with weights equal to 1. The resulting ZAMS rotation period is equal to 2.1 days, about four times longer than in case of constant AM. This agrees well with the observations. In addition, the evolution of the rotation period is shown in the case when the importance of the accretion and wind is decreased by a factor of 10, i. e. the corresponding terms in equation (5) were taken with weights equal to 0.1. Physically, this corresponds to e. g. lower accretion and mass loss rates, or shorter time scales of both phenomena. The resulting ZAMS rotation period is equal now to 3.9 days. This can be compared with the value of 1.25 days expected when the considered star preserves AM in shells during its approach to the ZAMS.  Similar calculations have been obtained for a 3 $M_{\\sun}$ star. The main difference between the two cases is the time scale of the PMS evolution which is 7 times shorter for the 3 $M_{\\sun}$ star than for the 2 $M_{\\sun}$ star. To get a 3-4-fold increase of the rotation period, a relatively more efficient AM loss is required. This can be achieved e. g. by an increase of the weight of the term describing the field-disk interaction up to a value of 2 -- equivalent to the increase of $\\mu$ by a factor of 1.4, see equation (5). Note that this corresponds to a surface magnetic field of about 1.5 kG which would show up in observations as a longitudinal magnetic field of only about 500 G (Preston, 1971).  One concludes that the presence of a moderate primordial magnetic field can explain the observed difference between the average rotation rate of Ap and normal stars. The required values of the parameters describing the interaction between a star and its environment are within the observed ranges. Because the time scales of the existence of disks and winds, mass loss and accretion rates, and the intensity of the magnetic field are expected to vary randomly from one PMS star to another, the resulting ZAMS rotation period is not expected to be strongly correlated with any single parameter.  The discussed AM loss mechanism cannot, however, explain the extremely long rotation periods observed in some stars (Table 1). The required values of the considered parameters are unreasonable. Therefore, it is suggested that those stars lose AM also on the MS because of some exceptional circumstances. Because the accretion disks are not observed around young MS stars we reject this mechanism. A continuous mass loss via a magnetized wind is a more realistic possibility. With the same intensity as adopted here for the PMS phase, the wind can spin down the star to a rotation period of the order of 100 years just in $2\\times 10^7$ years, which is a tiny fraction of the total MS lifetime of a star with a mass below 2 $M_{\\sun}$. A less intense wind would need, of course, a correspondingly longer time."
    },
    "9805/astro-ph9805204_arXiv.txt": {
        "abstract": "In order to provide atomic data needed for astrophysical investigations, a set of electron-impact broadening parameters for  ionized rare-earth element lines should be calculated.  We are going to calculate the electron-impact broadening parameters for more than 50 transitions of ionized rare-earth elements. Taking into account that the spectra of these elements are very complex, for calculation we can use the modified semiempirical approach -- MSE or simplified MSE. Also, we can estimate these parameters on the basis of regularities and systematic trends. ",
        "introduction": "The spectral lines of rare-earth elements are present in Solar as well as in stellar spectra (see e.g.  Grevesse \\& Blanquet 1969, Molnar 1972, Adelman 1987, Mathys \\& Cowley 1992, Sadakane 1993, Bidelman et al. 1995, Cowley et al. 1996, etc.). Principally, these lines  originate in layers of stellar atmospheres with higher electron density (photosphere or subphotosphere). Consequently, electron-impact broadening mechanism can be important, especially for hot (A and  B) stars as well as for white dwarfs. So, it is important to have a set of electron-impact broadening data for the lines of ionized rare-earth elements. For some transitions of La\\,{\\sc ii} and La\\,{\\sc iii} we have calculated Stark widths (Popovi\\'c \\& Dimitrijevi\\'c 1997) by using the modified  semiempirical approach (Dimitrijevi\\'c \\& Konjevi\\'c 1980, Popovi\\'c \\& Dimitrijevi\\'c 1996a,b). Here we  present our plans and specify the number of lines for which we may calculate electron-impact broadening parameters with a satisfying accuracy and discuss the difficulties which may appear in the calculation. ",
        "conclusions": ""
    },
    "9805/astro-ph9805174_arXiv.txt": {
        "abstract": " ",
        "introduction": "The dwarf galaxy population is not merely an extension of giants to fainter luminosities. Fundamental differences between the two populations raise the issue of whether the same formation mechanism applies to both. Understanding the origin and evolution of dwarf galaxies is essential in developing the correct model of galaxy and cluster formation. Because compact groups are unique environments and active sites of galaxy interaction, they provide an opportunity to study the formation of dwarf galaxies both as a function of environment and in tidal debris.  Hickson (1982) cataloged 100 compact groups of galaxies, selecting them on the basis of population, isolation, and compactness. Radial velocity measurements (Hickson et al. 1992) and morphological studies (Mendes de Oliveira \\& Hickson 1994) suggest that most of the Hickson compact groups (HCGs) are physical associations, although Mamon (1990) has argued that the data are also consistent with many HCGs being superpositions of binary-rich loose groups and Hernquist et al. (1995) contend that many compact groups are chance projections of large filamentary structures. Diaferio et al. (1994) propose that during the collapse of a rich loose group of galaxies, compact configurations are continuously forming and that the members of such a compact group eventually merge, even as new galaxies are joining the group from the surrounding region.  Regardless of the exact fraction of groups that are physical associations, compact groups are sites of galaxy interactions and among themselves, they provide a variety of environments to examine (e.g., x-ray luminous vs. non-luminous, elliptical vs. spiral dominated, tidally interacting vs. quiescent). Compact groups are particularly intriguing because of the combination of high galaxy projected density  (similar to the centers of rich clusters) and low velocity dispersion (comparable to loose groups). This combination suggests, at least for the fraction of physical groups, that interactions and mergers occur frequently and that during such encounters, tidal forces have sufficient time to extract significant amounts of material, resulting in tidal tails and bridges. As proposed by Zwicky (1956), self-gravitating objects might develop in tidal tails and evolve into dwarf galaxies. This idea is supported by the observation of regions of active star formation at the ends of tidal tails in the Superantennae and Antennae systems (Mirabel et al. 1991, Mirabel et al. 1992) and of two dwarf galaxies in the tidal tails of Arp 105 (Duc \\& Mirabel 1994). Furthermore, numerical simulations (Barnes \\& Hernquist 1992, Elmegreen et al. 1993) confirm that gravitationally bound clumps can form in tidal tails. Therefore, compact groups represent a unique environment in which to study the formation of dwarf galaxies in tidal debris. In a previous paper (Hunsberger et al. 1996), we examined a sample of 42 HCGs and identified 47 tidal dwarf candidates in seven groups with tidal tails and tidal arms. We also estimated the fraction of compact group dwarf galaxies produced by the tidal dwarf formation mechanism over the lifetime of a group and found it to be significant ($> 30\\%$).  Previously determined luminosity functions of compact groups (Heiligman \\& Turner 1980, Mendes de Oliveira \\& Hickson 1991, Sulentic \\& Raba\\c{c}a 1994, Ribeiro et al. 1994, Zepf et al. 1997) have yielded faint end slopes of $\\alpha=0.0$, $\\alpha=-0.2 \\pm 0.9$, $\\alpha=-1.13 \\pm 0.13$, $\\alpha=-0.82 \\pm 0.09$, and $\\alpha=-1.0$, respectively. Some of these values are flatter than that observed in either clusters ($-1.4 < \\alpha < -1.0$; Ferguson \\& Sandage 1991) or the field ($\\alpha \\sim -1.0$; Loveday et al. 1992, Marzke et al. 1994, Ratcliffe et al. 1998) implying fewer dwarfs per giant in compact groups than in the other environments. The Heiligman \\& Turner (1980) luminosity function used ten compact groups (not necessarily HCGs) for which galaxy redshifts were available at that time. Knowing the completeness limits of the Palomar Sky Survey from which the groups were identified, they quantified the deficit of fainter members in their sample by a comparison to the field luminosity function. The number of ``missing'' galaxies was significant and could not readily be attributed to small sample size, inaccurate photometry, galaxy misclassification, or selection effects. Using a sample of 68 HCGs, Mendes de Oliveira \\& Hickson (1991) determined the ``best-fit'' parameters of their luminosity function with Monte Carlo simulations of compact groups selected from various Schechter distributions. They similarly concluded that there is a lack of low luminosity galaxies due to environmental factors rather than selection effects and suggested that the excess of very luminous ellipticals (the mean magnitude of HCG ellipticals is brighter than that of Virgo cluster ellipticals) coupled with a deficit of fainter galaxies indicates merger activity. Sulentic \\& Raba\\c{c}a (1994) extended the compact group luminosity function to fainter absolute magnitudes using the Hickson catalog but applying a correction factor for incompleteness. Although their adopted value for the faint end slope did not reflect the decreasing number of low luminosity galaxies of previous luminosity functions, they noted that a single Schechter function did not provide a good fit to both bright and faint ends simultaneously. Ribeiro et al. (1994) obtained CCD images of 22 HCGs and used a statistical method to create a luminosity function which probed the galaxy population fainter than the original Hickson catalog, however, the resultant compact group luminosity function did not reveal the depletion of faint galaxies suggested by earlier analyses. Zepf et al. (1997) confirmed the validity of this photometric approach by obtaining spectroscopic redshifts and producing a similar luminosity function.  Most luminosity functions are presented as averaged functions and do not address questions arising from environmental differences and tidal dwarf formation. In this paper, we present luminosity functions determined by a statistical technique, similar to that used by Ribeiro et al. (1994), to address issues regarding dwarf galaxy formation. The technique involves using galaxy counts in the outer regions of images to statistically subtract the background/foreground contribution. Such an approach bypasses the need to obtain redshifts of all faint galaxies within the compact group field; so only moderately deep imaging is required. Using galaxy counts within 17 HCGs, the Zepf et al. (1997) luminosity function extends down to $M_B=-14.5 + 5\\log h$. Our luminosity function uses a sample of 39 HCGs and extends to $M_R=-13.4 + 5\\log h$ ($M_R=-14.0 + 5\\log h_{75}$). Therefore, our limit is two magnitudes fainter assuming a typical dwarf galaxy color of $B-R=1.0$, which is the median value for Local Group dwarf galaxies \\footnote{From NED, the NASA/IPAC Extragalactic Database is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.}. We compare luminosity functions of various sub-samples of HCGs such as  spiral-dominant vs. elliptical-dominant, x-ray-rich vs. x-ray-poor, and groups with tidal dwarfs vs. groups without tidal dwarfs to discover how processes such as galaxy destruction, galaxy creation, and tidal stripping modify the luminosity function of a compact group during its evolution.  In the next section (\\S2), we describe the observational procedures and data analysis. The method for determining a luminosity function is detailed in \\S3 and we present the results in \\S4. In \\S5 the major results are summarized and discussed and we present some speculations as to the origin of the luminosity function differences. ",
        "conclusions": "Summarizing, we find that  \\begin{itemize} \\item Dwarf galaxies are detected in Hickson compact groups using a statistical method for background subtraction. The resultant luminosity function compiled from all groups is not described by a single Schechter function. When we fit the distribution with a composite of two Schechter functions (for bright and faint galaxies), we find that the faint end slope is $\\alpha = -1.17$, down to a limiting magnitude of $M_R=-14.0 + \\log h_{75}$. \\item The general shape of the luminosity function exhibits a dip at $M_R=-17.5 + \\log h_{75}$ and $\\alpha = -0.52$ for the Schechter function characterizing the bright galaxy population. The decreasing number of galaxies per magnitude with decreasing luminosity suggests a ``deficit'' of intermediate luminosity galaxies in comparison to most field luminosity functions ($\\alpha \\sim 1.0$). \\item Dwarf galaxies presently observed in tidal debris (tidal dwarf candidates) make a small contribution to the total luminosity function of all groups, but groups with tidal dwarfs have a significantly larger dwarf population than those without tidal dwarfs. \\item Groups with first-ranked E/S0 galaxies have an excess faint galaxy population over those with first-ranked spirals. \\item Groups with diffuse x-ray emission have a large dwarf population extending to two group radii. Of the 14 groups with dominant E/S0 galaxies, five of them also have detected x-ray halos. \\item The dwarf-to-giant ratio in compact groups with x-ray halos is comparable to that of the cores of rich clusters. \\end{itemize}  To understand these results, we hypothesize the following (realizing that these are not unique interpretations of the data): \\parindent=0em  1) {\\it X-ray halos are an indication of previous or current giant galaxy interactions.} \\newline A hot intra-group medium is characterized by extended, diffuse x-ray emission rather than discrete sources which are associated with individual galaxies. Both Ebeling at al. (1994) and Pildis et al. (1995) detected extended x-ray emission in HCGs and discovered a relationship between x-ray emission and the morphological type of the dominant galaxy, i.e., an x-ray  group usually has a first-ranked E or S0 galaxy. They also reported an anti-correlation between the spiral fraction of the group and the presence of an x-ray halo. A more complete survey of HCGs by Ponman et al. (1996) finds x-ray emission in some spiral-rich groups as well. Recall that five of the seven x-ray groups in our sample have first-ranked E/S0 galaxies and a low spiral fraction ($< 50\\%$) and the other two are spiral-rich and currently interacting as evidenced by tidal tails. If one assumes that elliptical and lenticular galaxies are merger remnants, then this naturally suggests a connection between interactions/mergers and x-ray halos. During the merger process, gas and stars from the outer spiral disks are being tidally stripped. Supernovae from tidally-triggered star formation and shock-heating in tidal tails may be responsible for heating the gas to x-ray temperatures. Models of merging galaxies (Mihos et al. 1998) indicate that low-mass or extended galaxy halos allow tidal debris to be expelled to large distances.  2) {\\it Many dwarf galaxies form in tidal debris during interactions.} \\newline There is observational and theoretical evidence to support the idea of dwarf galaxy formation in the tidal debris of giant galaxy interactions (cf \\S1). In a previous paper (Hunsberger et al. 1996), we compiled a list of tidal dwarf candidates for our sample of HCGs using FOCAS detections coincident with tidal features. We identifed 47 possible tidal dwarfs in seven of the groups. These tidal dwarfs are responsible for the excess of faint galaxies in Figure 11b, but the real question is what is the ultimate fate of these objects? Elmegreen et al. (1993) predict that if the perturbing galaxy has a greater mass, the tidal dwarfs will be ejected into the group at large or become satellites of the new galaxy instead of falling back into the merger remnant. Therefore, we expect a large fraction of these systems to survive.  Next, consider the similarity of the luminosity functions of groups with tidal dwarfs and groups with diffuse x-ray emission. Neither luminosity function exhibits the deficit of intermediate luminosity galaxies, and both have significant dwarf galaxy populations. If x-ray halos are a sign of recent interaction and if tidal dwarfs are expelled from giants and can survive for as long as the cooling timescale of the gas then tidal dwarfs provide a reasonable explanation for the abundant dwarf population observed in x-ray groups.  3) {\\it The size and distribution of dark matter halos profoundly affects the evolution of groups of galaxies, especially the dwarf population.} \\newline If individual galaxies have massive halos then mergers occur quite quickly because of dynamical friction (Barnes 1985). If the galaxies have low-mass halos then they present smaller cross-sections for interaction so that the merger rate within the group is slower. In general, as the mass fraction of a common group halo increases, the merger rate decreases (Bode et al. 1993). The mass distribution of a group halo changes as the group evolves. As mergers occur in the central region, the galaxies' orbital energy is transferred to the halo causing it to become less centrally concentrated. Is the dark matter in compact groups in individual galaxy halos or a common group halo? The high spatial density of galaxies in compact groups makes it unlikely that a member can retain an extended halo. The $M/L$ measurements for each HCG listed in Table 2 are based on velocity dispersions of a few giant galaxy members in each group which introduces considerable inaccuracy. Such estimates are meaningful only within the group radius and high M/L values can mistakenly be inferred from large velocity differences of galaxies viewed in projection. Furthermore, it is quite possible that compact groups are not virialized systems. Although the reported mass-to-light ratios may be unreliable, we proceed with a hypothetical discussion. If a group has a low $M/L$, i.e., below the median, then galaxies have low-mass halos and either a) there is a low-mass group halo or b) the common halo is more massive but extended and not centrally concentrated. In this situation, giant galaxy interactions can produce tidal tails (Mihos et al. 1998) and tidal debris (and tidal dwarfs) can be ejected into an intra-group medium. In the case of b), the expelled gas should trace the gravitational potential which is now shallow but more extended and be detectable as a diffuse x-ray halo. When a group has a high $M/L$ it means that a) the galaxies have massive halos, b) there is a massive, extended group halo, or c) the common halo is less massive but has a strong central concentration. For a) or c), the massive halo(s) prevent formation of tidal tails during an interaction (Mihos et al. 1998). So as dwarfs are either cannibalized by massive galaxy halos or destroyed during a merger of giant members, there is no way to replenish the population.  4) {\\it There exists some mechanism which preferentially eliminates intermediate luminosity galaxies.} \\newline First we examine what is known about compact group formation. Previous studies hint at a connection between compact groups and loose groups. Despite the original isolation criterion for HCG selection, Ramella et al. (1994) report that 29 of 38 HCGs are embedded in rich, looser systems. N-body simulations (Diaferio et al. 1994) suggest that compact configurations resembling HCGs are continually forming during the collapse of rich loose groups. Furthermore, a plot of group magnitude vs. group diameter (Sulentic \\& Raba\\c{c}a 1994) for both loose groups from the CfA survey (Geller \\& Huchra 1983) and HCGs shows a very smooth transition in parameter space from one sample to the other. If a compact group evolves from a looser system, then its initial population should be similar to giant field galaxies and their companions.  Statistical analyses of the distribution of satellites around giant galaxies (Lorrimer et al. 1994, Loveday 1997) indicate that faint companions are more strongly clustered about the primary galaxy than their brighter counterparts. It is not clear whether a variation of clustering strength or of galaxy counts is responsible for the lack of intermediate luminosity objects in the luminosity function of compact group galaxies, but in either case, ``dynamical friction'' offers a plausible explanation.  In the field and loose groups, giant galaxies have companions orbiting within large dark matter halos. Dynamical friction provides a means to decrease the orbital angular momentum of satellites and the timescale for energy loss is inversely proportional to satellite mass. Because the orbits of more massive companions decay more rapidly, they merge sooner with the primary galaxy. If luminosity is proportional to mass, then such a scenario is consistent with fewer bright companions than faint companions, especially near a giant. Recall, however, that the dip in the luminosity function is less severe for groups with tidal dwarfs and essentially disappears for groups with x-ray halos. (As an aside here, we note that 18\\% of the groups in our total sample are groups with diffuse x-rays while 36\\% of the Ribeiro et al. (1994) sample are x-ray groups. This may explain why the trough is not a strong feature in their luminosity function.) Is it possible that tidal dwarfs replenish the population of intermediate luminosity galaxies ($-19 < M_R < -17$) and/or that formation of an x-ray halo inhibits cannibalism? The tidal dwarf formation models of Barnes \\& Hernquist (1992) and Elmegreen et al. (1993) predict masses in the range $10^7$ to $10^9M_{\\odot}$. For $M/L = 1$, this corresponds to $-18.2 \\leq M_R \\leq -13.2$ which agrees with the magnitude range of tidal dwarf candidates in our sample (see Fig. 10a). The tidal dwarf galaxies cataloged by Duc \\& Mirabel (1998) exhibit a magnitude range of $-18.8 \\leq M_B \\leq -12.1$ with a median value of $M_B = -14.2$ implying that most tidal dwarfs are low luminosity objects although a few could be considered intermediate luminosity galaxies. Now assume that diffuse x-rays trace a common dark matter halo which is less concentrated than individual galaxy halos (as discussed previously). Because the dynamical friction timescale is inversely proportional to medium density, the loss of orbital energy proceeds more slowly. Once a significant group potential forms, merging times increase as well as the expected lifetimes of dwarf members.  5) {\\it Groups with first-ranked E/S0 galaxies are more evolved than other compact groups, i.e., they have undergone previous mergers.} \\newline In groups with a first-ranked E/S0 galaxy, there is a substantial population of dwarf galaxies which extends out to two group radii (Figure 12b). A possible explanation is that the dwarfs exist ``at large'' in the group and an example of such a free-floating dwarf population is reported in the Virgo cluster (Ferguson 1992). A merger scenario can explain how the dwarf population gets redistributed and enhanced. Assume that a dominant-spiral compact group is evolving into a dominant-E/S0 group. Initially dwarfs are bound to individual galaxies and their distribution traces the giant galaxy halos. As two giant members merge, some of the dwarf companions are destroyed, but more importantly the dark matter is restructured. It becomes less centrally concentrated and more extended, either creating a common group halo or making the existing one more massive. If tidal tails form during the interaction and if they are sites of dwarf galaxy formation (tidal dwarfs), then there is a mechanism by which the dwarf population can be replenished and enhanced. The conditions which allow tidally stripped gas to form a hot (x-ray emitting) intra-group medium may also distribute tidal dwarfs throughout the group. When such a merger is complete, the distribution of both the hot x-ray gas and the dwarf population will trace the extended group halo. This scenario is consistent with results from a survey of twelve poor groups, including three HCGs, conducted by Zabludoff \\& Mulchaey (Zabludoff \\& Mulchaey 1997, Mulchaey \\& Zabludoff 1997). They discovered that nine x-ray detected groups had $20-50$ dwarf members ($-16 < M_B - 5\\log h < -14$) and established that groups with a central elliptical have an x-ray halo extending $100-300 h^{-1}{\\rm kpc}$. They concluded that the longevity of the dwarf members is increased because of the distribution of dark matter in an extended common halo.  \\parindent=2em \\bigskip  Finally, combining these hypotheses, we speculate that the evolution of a typical HCG may proceed as follows: The group forms because dynamical friction of massive dark halos around field galaxies brings them together. These newly formed HCGs contain dwarf companions to the giant galaxies, but these are gradually cannibalized as interactions occur. Eventually a major merger occurs and this has several consequences: the formation of a giant E/S0 galaxy, the production of an x-ray halo, and the formation of tidal dwarfs. The ongoing interactions strip gas, dark matter, and dwarf companions from the giants and spread them into a common group halo. This redistribution extends the lifetime of the dwarf population (and the group as a whole) and creates an environment more conducive to the tidal dwarf replenishment process."
    },
    "9805/astro-ph9805342_arXiv.txt": {
        "abstract": "\\Zw\\ is the most metal poor star-forming galaxy known and is an ideal laboratory to probe stellar evolution theory at low metallicities. Using archival HST WFPC2 imaging and FOS spectroscopy we were able to improve previous studies. We constructed a continuum free \\heii\\ map, which was used to identify Wolf-Rayet (WR) stars recently found by ground-based spectroscopy and to locate diffuse nebular emission. Most of the \\heii\\ emission is associated with the NW stellar cluster, clearly displaced from the surrounding shell-like [\\ion{O}{3}] and \\Ha\\ emission. We found evidence for \\HeII\\ sources, compatible with 5--9 WNL stars and/or compact nebular \\heii\\ emission, as well as residual diffuse emission. Only one of them is outside the NW cluster. We have done an extensive comparison between our results and the recent ground-based data used by Izotov et al.\\ (1997) and Legrand et al.\\ (1997) to identify WN and WC stars in \\Zw. The differences between the various data may be understood in terms of varying slit locations, continuum fits, and contamination by nebular lines. We have calculated evolutionary tracks for massive stars and synthesis models at the appropriate metallicity ($Z$ $\\sim 0.02$ Z$_{\\odot}$). These single star models predict a mass limit $M_{\\rm WR} \\approx$ 90 M$_{\\odot}$ for WR stars to become WN and WC/WO. For an instantaneous burst model with a Salpeter IMF extending up to $M_{\\rm up} \\sim$ 120-150 \\msun\\ our model predictions are in reasonable agreement with the observed equivalent widths. Our model is also able to fully reproduce the observed equivalent widths of nebular \\heii\\ emission due to the presence of WC/WO stars. This quantitative agreement and the spatial correlation of nebular \\heii\\ with the stellar cluster and the position of WR stars shown from the ground-based spectra further supports the hypothesis that WR stars are responsible for nebular \\HeII\\ emission in extra-galactic \\hii\\ regions. ",
        "introduction": "\\Zw\\ (= Mrk~116 = UGCA~166) is described in Zwicky's catalog of compact galaxies (Zwicky 1971) as a pair of interconnected blue compact, presumably young galaxies. This otherwise unremarkable galaxy attracted immediate attention when Searle \\& Sargent (1972) established its extremely low oxygen abundance of only 2\\% the solar value. \\Zw\\ became one of the proto-types of the ``flashing galaxies'' (Searle, Sargent, \\& Bagnuolo 1973), whose blue colors suggest a brief period of intense star formation. Subsequently this galaxy class was generally referred to as ``starburst galaxies'' (Weedman 1987), with \\Zw\\ belonging to the sub-category of blue compact dwarfs (Thuan 1991). A review of its most pertinent parameters was given by Dufour et al. (1996): It has a radial velocity of 740~\\kms, is at a distance of about 10~Mpc, and has an extinction of $E(B-V) = 0.07$.  \\Zw's extreme metal deficiency was confirmed in several follow-up studies (e.g., Kunth \\& Sargent 1986; Campbell 1990; Pagel et al. 1992; Kunth et al. 1994; Stasi\\'nska \\& Leitherer 1996). Although strong efforts to detect other galaxies with very low metal content have been undertaken (e.g. Terlevich, Skillman, \\& Terlevich 1995), \\Zw\\ is still the record holder. Its low metallicity is of particular importance for studies of the primordial helium abundance (Izotov, Thuan, \\& Lipovetsky 1996; Olive, Skillman, \\& Steigman 1997; Skillman, Terlevich, \\& Terlevich 1998), and for properties of chemically unevolved galaxies in general. Despite the low oxygen abundance, however, there is clear evidence for stellar nucleosynthetic processing from its carbon abundance so that \\Zw\\ is {\\em not} a primordial galaxy (Garnett et al. 1997). The same suggestion was made earlier by Thuan (1983) on the basis of surface photometry of its underlying stellar population.  \\Zw\\ is an ideal laboratory to probe stellar evolution theory in a low-metallicity environment. The spatial resolution of WFPC2 on the Hubble Space Telescope (HST) allows the detection of individual stars down to the $\\sim$10~\\Ms\\ range (Hunter \\& Thronson 1995; Dufour et al. 1996). Analysis of deep WFPC2 images clearly demonstrates ongoing massive-star formation. Additional, indirect evidence for a population of massive stars comes from bubbles and outflows of the interstellar medium, indicative of powerful hot-star winds and supernovae (Martin 1996).  Wolf-Rayet (WR) stars are the evolved, less massive descendants of previously massive O stars (Maeder \\& Conti 1994). Stellar evolution theory predicts few WR stars to form in a low-metallicity environment such as in \\Zw\\ (e.g., Meynet 1995). Therefore recent reports (Izotov et al.\\ 1997, hereafter \\izotov; and Legrand et al.\\ 1997, hereafter \\legrand) of WR star detections in \\Zw\\  have far-reaching implications for stellar evolution theory. Furthermore \\Zw\\ is probably the best studied object among approximately 50 to 60 extra-galactic \\hii\\ regions showing nebular \\heii\\ emission (Campbell et al.\\ 1986; Izotov, Thuan \\& Lipovetsky 1994, 1997) whose origin is still poorly known and subject to recent debates. Various mechanisms reaching from photoionization by hot massive stars, photoionization by X-ray binaries to shocks have been put forward to explain the high excitation (Garnett et al.\\ 1991). Understanding the origin of nebular \\heii\\ in low-metallicity galaxies has implications for our knowledge of the far-UV flux of galaxies and may prompt a re-examination of the importance of young galaxies to the ionization of QSO absorption line systems and the ionization of the intergalactic medium (cf. Garnett et al. 1991; Pettini et al. 1997).  WR stars can be recognized, e.g., via their broad ($\\sim$100~\\AA) emission bump around $\\lambda$4686, which is generally a blend of H, \\HeII\\ and several metal lines. \\Zw\\ has been known to show {\\em narrow}, nebular \\heii\\ emission but the {\\em broad} WR emission bump was not known before the findings of \\izotov\\ and \\legrand. The blending of several stellar and nebular emission lines around 4686~\\AA\\ and the complex spatial morphology in ground-based data make it challenging to disentangle stars and gas and to derive the WR content and the nebular properties. This is evident in the apparently discrepant results of \\izotov\\ and \\legrand. The superior spatial resolution of HST may prove useful in resolving the discrepancy.  In this paper we report on an HST archival project to study the spatial morphology of the highly ionized gas in \\Zw, as observed primarily in \\heii. Our goals are to reconcile existing ground-based spectroscopic observations, to test stellar evolution models for WR stars in an extreme chemical environment, and to elucidate the nature of nebular \\HeII\\ emission in extragalactic \\hii\\ regions. The paper is organized as follows. In \\S 2 and \\S 3 we describe the data used in this work and their analysis. In \\S 4 we discuss the recent spectroscopic detection of Wolf-Rayet stars in connection with the HST imagery. The interpretation and a comparison with Wolf-Rayet star models are presented in \\S 5. Finally, \\S 6 summarizes our conclusions. ",
        "conclusions": "We have used an extensive collection of archival HST data to study the stellar content and gaseous distribution in \\Zw. Continuum subtraction and cosmic-ray removal were found to be critical for the search for faint sources. Using WFPC2 imaging and FOS spectra we were able to improve previous studies and construct a continuum free \\heii\\ map, which was used to search for stellar sources (WR stars) recently found by ground-based spectroscopy and to locate diffuse nebular emission. The \\heii\\ emission is associated with the NW stellar cluster, clearly displaced from the surrounding shell-like [\\ion{O}{3}] and \\Ha\\ emission. We found evidence for \\HeII\\ sources, compatible with 5--9 WNL stars and/or compact nebular \\heii\\ emission, as well as residual diffuse emission. Only one of these sources is not located in NW cluster but in the region in between the NW and SE condensation.  We have done an extensive comparison between our results and the recent ground-based data of \\izotov\\ and \\legrand\\ which allowed them to identify WN and WC stars in \\Zw. The differences between the various data may be understood in terms of varying slit locations, continuum fits and contamination by nebular lines. Ground-based spectra taken with better resolution and larger telescopes will allow one to probe the massive star content more accurately, to provide more stringent constraints on evolution models, and to study the interaction between the stars and their surrounding ISM in more depth. In order to avoid the difficulties in correlating spectra and image location, Fabry-Perot interferometry or other imaging spectrocopy would be the technique of choice.  We have calculated evolutionary tracks for massive stars and synthesis models at the appropriate metallicity ($Z$ $\\sim 0.02$ Z$_{\\odot}$). These single star models predict a mass limit $M_{\\rm WR} \\approx$ 90M$_{\\odot}$ for WR stars which may become WN and WC/WO. For an instantaneous burst model with a Salpeter IMF extending up to $M_{\\rm up} \\sim$ 120-150 \\msun\\ our model predictions are in reasonable agreement with the observed equivalent widths. The WR population in \\Zw\\ has properties consistent with those predicted by single-star evolution models for this metallicity. The WR/O ratio forms an extension of the trend observed in WR galaxies (Maeder \\& Conti 1994) towards very low $Z$. Our model is also able to fully reproduce the observed EW of nebular \\heii\\ emission due to the presence of WC/WO stars. This quantitative agreement and the spatial correlation of nebular \\heii\\ with the stellar cluster and the position of WR stars further supports the hypothesis by Schaerer (1996) that WR stars are responsible for nebular \\HeII\\ emission in extra-galactic \\hii\\ regions."
    }
}