Table of contents

The mass-density field as extracted from peculiar velocities in our cosmological neighborhood is mapped back in time to the cosmic microwave background (CMB) in two ways. First, the density power spectrum (P_k) is translated into a temperature angular power spectrum of subdegree resolution (C_l) and compared to observations. Second, the local density field is translated into a temperature map in a patch on the last-scattering surface of a distant observer.

A likelihood analysis of the Mark III catalog of peculiar velocities have constrained the range of parameters for P_k within the family of COBE-normalized cold dark matter (CDM) models, favoring a slight tilt in the initial spectrum, n < 1. The corresponding range of C_l is plotted against current observations, indicating that the CMB data can tighten the constraints further: only models with small tilt (n ~ 0.9) and high baryonic content (Ω_b ~ 0.1) could survive the two data sets simultaneously.

The local mass-density field that has been recovered from the velocities via a Wiener method is convovled with a Boltzmann calculation to recover 10' resolution temperature maps as viewed from different directions. The extent of the CMB patch and the amplitude of fluctuations depend on the choice of cosmological parameters, e.g., the local 100 h^-1 Mpc sphere corresponds to 90'-30' at the CMB for Ω between 1 and 0, respectively. The phases of the temperature map are correlated with those of the density field, contrary to the contribution of the Sachs-Wolfe effect alone. This correlation suggests the possibility of an inverse reconstruction of the underlying density field from CMB data with interesting theoretical implications.

We compare the independent FIRAS and DIRBE observations from COBE in the wavelength range of 100-300 μm. This cross calibration provides checks of both data sets. The results show that the data sets are consistent within the estimated gain and offset uncertainties of the two instruments. They show the possibility of improving the gain and offset determination of DIRBE at 140 and 240 μm.

The possible existence of a primordial magnetic field in the universe has been investigated in many articles. Studies involving the influence of a magnetic field in the nucleosynthetic era, as well as studies of the effects on the formation of structures during the radiation era and the matter era, have been considered. Here we assume the existence of a primordial magnetic field and study its effect, in particular, on the formation of voids. The aim of the study is twofold: to place constraints on the strength of the magnetic field during the recombination era and to preview its effects on the formation of voids.

We use high-resolution N-body simulations to study the equilibrium density profiles of dark matter halos in hierarchically clustering universes. We find that all such profiles have the same shape, independent of the halo mass, the initial density fluctuation spectrum, and the values of the cosmological parameters. Spherically averaged equilibrium profiles are well fitted over two decades in radius by a simple formula originally proposed to describe the structure of galaxy clusters in a cold dark matter universe. In any particular cosmology, the two scale parameters of the fit, the halo mass and its characteristic density, are strongly correlated. Low-mass halos are significantly denser than more massive systems, a correlation that reflects the higher collapse redshift of small halos. The characteristic density of an equilibrium halo is proportional to the density of the universe at the time it was assembled. A suitable definition of this assembly time allows the same proportionality constant to be used for all the cosmologies that we have tested. We compare our results with previous work on halo density profiles and show that there is good agreement. We also provide a step-by-step analytic procedure, based on the Press-Schechter formalism, that allows accurate equilibrium profiles to be calculated as a function of mass in any hierarchical model.

The cosmological origin of cosmic gamma-ray bursts is tested by using the method of peak alignment for the averaging of time profiles. The test is applied to the basic cosmological model with standard sources, which postulates that the difference between bright and dim bursts results from the different cosmological redshifts of their sources. The average emissivity curve (ACE_bright) of a group of bright BATSE bursts is approximated by a simple analytic function that takes into account the effect of the squeezing of the time pulses with increasing energy of photons. This function is used to build the model light curve for ACE_dim of dim BATSE bursts, which takes into account both the cosmological time-stretching of the light curves of bursts and the redshifting of photon energies. Direct comparison between the model light curve and the ACE_dim of dim bursts is performed, based on the estimated probabilities of differences between ACEs of randomly selected groups of bursts. The comparison shows no evidence for the predicted cosmological effects. The 3 σ upper limit of the average redshift z_dim of emitters of dim bursts is estimated to be as small as ~0.1-0.5, which is not consistent with values of ~1 predicted by current cosmological models of gamma-ray bursts.

We report on the discovery of Cepheids in the field spiral galaxy NGC 3621, based on observations made with the Wide Field Planetary Camera 2 on board the HubbleSpaceTelescope (HST). NGC 3621 is one of 18 galaxies observed as a part of the HST Key Project on the Extragalactic Distance Scale, which aims to measure the Hubble constant to 10% accuracy. Sixty-nine Cepheids with periods in the range 9-60 days were observed over 12 epochs using the F555W filter, and over four epochs using the F814W filter. The HST F555W and F814W data were transformed to the Johnson V and Kron-Cousins I magnitude systems, respectively. Photometry was performed using two independent packages, DAOPHOT II/ALLFRAME and DoPHOT. Period-luminosity relations in the V and I bands were constructed using 36 fairly isolated Cepheids present in our set of 69 variables. Extinction-corrected distance moduli relative to the LMC of 10.63 ± 0.09 and 10.56 ± 0.10 mag were obtained using the ALLFRAME and DoPHOT data, respectively. True distance moduli of 29.13 ± 0.18 and 29.06 ± 0.18 mag, corresponding to distances of 6.3 ± 0.7 and 6.1 ± 0.7 Mpc, were obtained by assuming values of μ₀ = 18.50 ± 0.10 mag and E(V - I) = 0.13 mag for the distance modulus and reddening of the LMC, respectively.

We find that the observed log N-log S relation of X-ray clusters can be reproduced remarkably well with a certain range of values for the fluctuation amplitude σ₈ and the cosmological density parameter Ω₀ in cold dark matter (CDM) universes. The 1 σ confidence limits on σ₈ in the CDM models with n = 1 and h = 0.7 are expressed as (0.54±0.02)Ω (λ₀ = 1 - Ω₀) and (0.54 ± 0.02)Ω (λ₀ = 0), where n is the primordial spectral index, and h and λ₀ are the dimensionless Hubble and cosmological constants. The errors quoted above indicate statistical errors from the observed log N-log S only; the systematic uncertainty from our theoretical modeling of X-ray flux in the best-fit value of σ₈ is about 15%. In the case where n = 1, we find that the CDM models with (Ω₀, λ₀, h, σ₈) ≃ (0.3, 0.7, 0.7, 1) and (0.45, 0, 0.7, 0.8) simultaneously account for the cluster log N-log S, X-ray temperature functions, and the normalization from the COBE 4 year data. The derived values assume that the observations are without residual systematic errors, and we discuss in detail other theoretical uncertainties that may change the limits on Ω₀ and σ₈ from the log N-log S relation. We show the power of this new approach, which will become a strong tool as observations attain more precision.

We derive lower bounds on the cosmic baryon density from the requirement that the high-redshift intergalactic medium (IGM) contain enough neutral hydrogen to produce the observed Lyα absorption in quasar spectra. These analytic bounds follow from a key theoretical assumption—that absorbing structures are on average no more extended in redshift space than in real space—which is likely to hold in the gravitational instability picture of the Lyα forest, independently of the details of the cosmological model. The other ingredients that enter these bounds are an estimate of (or lower limit to) the intensity of the photoionizing UV background from quasars, a temperature T ~ 10⁴ K for the "warm" photoionized IGM that produces most of the Lyα absorption, a value of the Hubble constant, and observational estimates of the mean Lyα flux decrement or, for a more restrictive bound, the distribution function P(τ) of Lyα optical depths. With plausible estimates of the quasar UV background and , the mean decrement bound implies a baryon density parameter Ω_b ≳ 0.0125 h^-2, where h ≡ H₀/(100 km s^-1 Mpc^-1). A recent observational determination of P(τ) implies that Ω_b ≳ 0.0125 h^-2 even for a conservative estimate of the quasar UV background, and Ω_b ≳ 0.018 h^-2 for a more reasonable estimate. These bounds are consistent with recent low estimates of the primordial deuterium-to-hydrogen ratio (D/H)_P, which imply that Ω_b ≈ 0.025 h^-2 when combined with standard big bang nucleosynthesis. Since the bounds account only for baryons in the warm IGM, their combination with the nucleosynthesis constraint implies that most of the baryons in the universe at z ~ 2-4 were distributed in diffuse intergalactic gas rather than in stars or compact dark objects. The P(τ) bound on Ω_b is incompatible with some recent high estimates of (D/H)_P, unless one drops the assumptions of standard big bang nucleosynthesis or abandons the idea that Lyα forest lines originate in the smooth, large-scale structures of photoionized gas that arise in gravitational instability theories.

We quantify the consequences of intergalactic dust produced by the first Type II supernovae in the universe. The fraction of gas converted into stars is calibrated based on the observed C/H ratio in the intergalactic medium at z = 3, assuming a Scalo mass function for the stars. The associated dust absorbs starlight energy and emits it at longer wavelengths. For a uniform mix of metals and dust with the intergalactic gas, we find that the dust distorts the microwave background spectrum by a y-parameter in the range (0.06-6) × 10^-5 (M_SN/0.3 M_☉), where M_SN is the average mass of dust produced per supernova. The opacity of intergalactic dust to infrared sources at redshifts of z ≳ 10 is significant, τ_dust = (0.1-1) × (M_SN/0.3 M_☉), and could be detected with the NextGenerationSpaceTelescope. Although dust suppresses the Lyα emission from early sources, the redshifts of star clusters at z = 10-35 can be easily inferred from the Lyman limit break in their infrared spectrum between 1 and 3.5 μm.

Using traditional morphological classifications of galaxies in 10 intermediate-redshift (z ~ 0.5) clusters observed with WFPC2 on the HubbleSpaceTelescope, we derive relations between morphology and local galaxy density similar to that found by Dressler for low-redshift clusters. Taken collectively, the "morphology-density" relationship, T - Σ, for these more distant, presumably younger clusters is qualitatively similar to that found for the local sample, but a detailed comparison shows two substantial differences: (1) For the clusters in our sample, the T - Σ relation is strong in centrally concentrated "regular" clusters, those with a strong correlation of radius and surface density, but nearly absent for clusters that are less concentrated and irregular, in contrast to the situation for low-redshift clusters, where a strong relation has been found for both. (2) In every cluster the fraction of elliptical galaxies is as large or larger than in low-redshift clusters, but the S0 fraction is 2-3 times smaller, with a proportional increase of the spiral fraction.

Straightforward, though probably not unique, interpretations of these observations are (1) morphological segregation proceeds hierarchically, affecting richer, denser groups of galaxies earlier, and (2) the formation of elliptical galaxies predates the formation of rich clusters and occurs instead in the loose-group phase or even earlier, but S0's are generated in large numbers only after cluster virialization.

We describe the evolution of interstellar gas in a family of low-luminosity elliptical galaxies all having M_B = -20 but with different degrees of flattening (E0, E2, and E6) and two current supernova rates, SNu = 0.01 and 0.04. The galaxies are composed of 90% dark matter, are rotationally flattened, and have isotropic stellar velocity dispersions.

The soft X-ray luminosity of the hot interstellar gas after evolving for 15 Gyr decreases dramatically with increasing galactic rotation. As the rotating hot interstellar gas loses energy in the galactic potential, it cools onto a large disk. The outer radius of the disk can be much reduced by increasing the supernova rate, which drives a gentle galactic wind transporting high angular momentum gas out of the galaxy. The total mass of cooled disk gas is less sensitive to the supernova rate.

Although the hot interstellar gas may be difficult to observe in rotating low-luminosity ellipticals, the cooled disk gas can be observed (1) in optical line emission, since part of the cooled disk gas is photoionized by stellar UV, and (2) in the optical continuum, assuming the colder disk gas forms into luminous stars. The mass of H II gas (~10⁸M_☉) may be much greater than previously realized, since rotationally supported, low-density H II contributes little to the global optical line emission. We interpret the stellar disks that are common (or ubiquitous) in low-luminosity ellipticals as stars that have formed in the cold disk gas. The total mass of cold disk gas available for star formation is similar to the masses of stellar disks observed. The high stellar Hβ photometric index observed in disky ellipticals can be understood by combining the light of young disk stellar populations with that of the old bulge population.

We investigate the form of the momentum distribution function for protons and electrons in an advection-dominated accretion flow (ADAF). We show that for all accretion rates, Coulomb collisions are too inefficient to thermalize the protons. The proton distribution function is therefore determined by the viscous heating mechanism, which is unknown. The electrons, however, can exchange energy quite efficiently through Coulomb collisions and the emission and absorption of synchrotron photons. We find that for accretion rates greater than ~10^-3 of the Eddington accretion rate, the electrons have a thermal distribution throughout the accretion flow.

For lower accretion rates, the electron distribution function is determined by the electron's source of heating, which is primarily adiabatic compression. Using the principle of adiabatic invariance, we show that an adiabatically compressed, collisionless gas maintains a thermal distribution until the particle energies become relativistic. We derive a new, nonthermal distribution function for relativistic energies and provide analytic formulae for the synchrotron radiation from this distribution. Finally, we discuss its implications for the emission spectra from ADAFs.

Assuming that supernova shocks accelerate nonthermal particles, we model the temporally evolving nonthermal particle and photon spectra at different stages in the lifetime of a standard shell-type supernova remnant (SNR). A characteristic νF_ν spectrum of an SNR consists of a peak at radio through optical energies from nonthermal electron synchrotron emission and another high-energy gamma-ray peak due primarily to secondary pion production, nonthermal electron bremsstrahlung, and Compton scattering. We find that supernova remnants are capable of producing maximum gamma-ray luminosities ≳10³⁵ ergs s^-1 if the density of the local interstellar medium is ≳10 cm^-3. This emission will persist for ≳10⁵ yr after the supernova explosion because of the long energy loss timescales for electrons with kinetic energy ~1 GeV. This long gamma-ray lifetime implies that SNRs with a wide range of ages could be gamma-ray sources and could constitute some of the unidentified EGRET sources.

Sari & Piran have demonstrated that the time structure of gamma-ray bursts (GRBs) must reflect the time structure of their energy release. A model that satisfies this condition uses the electrodynamic emission of energy by the magnetized rotating ring of dense matter left by neutron star coalescence; GRBs are essentially fast, high-field, differentially rotating pulsars. The energy densities are large enough for the power to appear as an outflowing equilibrium pair plasma, which produces the burst by baryon entrainment and subsequent internal shocks. In this paper the magnetic field and the characteristic timescale for its rearrangement—which determines the observed time structure of the burst—are estimated. There may be quasi-periodic oscillations at the rotational frequencies, which are predicted to range up to 5770 Hz (in a local frame). This model is one of a general class of electrodynamic accretion models that includes the Blandford and Lovelace model of active galactic nuclei and that can also be applied to black hole X-ray sources of stellar mass. The apparent efficiency of nonthermal particle acceleration is predicted to be 10%-50%, but higher values are possible if the underlying accretion flow is super-Eddington.

We study the cluster mass distributions that are derived from observations of the X-ray gas, Faraday rotation, and gravitational arc images. The Faraday rotation measure is inferred to be that resulting from the magnetic field, derived as the difference between the gas pressure gradient (determined from the X-ray images) and the gravitational potential (determined from the arcs). It is found that the magnetic field cannot affect the gas distribution if the gas density and the rotation measure of the cluster are not very different from those of the Coma Cluster. We estimate the rotation measure in A2218, requiring that the mass distribution be consistent with observation of the giant arc image and the X-ray surface brightness. We compare the result with observed values in nearby clusters. In conclusion we find that the magnetic pressure can contribute substantially to supporting the gas against the gravity produced by the mass distribution, which can form the giant arc.

We have observed M31 using the ROSAT HRI. We have searched for X-ray emission from supernova remnants (SNRs) previously identified from surveys using narrowband optical observations. We find that a surprisingly small number of the identified SNRs are detected in X-rays at a threshold of ≈ 10³⁵ ergs s^-1. The absence of detected X-ray flux from many of the optically identified SNRs suggests that the local ISM density in the vicinity of these SNRs is typically quite low, less than 0.1 cm^-3. The few that have been detected likely represent the upper end of the density distribution, having implied ambient densities in the range 0.1 to >10 cm^-3. Since H I observations show that all SNRs are in regions with large column densities, which implies densities of 1-10 cm^-3, a multicomponent ISM must be invoked. Our measured densities are upper limits for the dominant low-density component of the ISM.

We propose that the intensity changes and spectral evolution along the M87 jet can be explained by adiabatic changes to the particle momentum distribution function and the magnetic field. This is supported by the lack of any significant variation in the radio-to-optical spectral index along the jet and by the moderate changes in radio brightness. Assuming a simple scaling law between magnetic field and density, we use the deprojection of a 2 cm VLA intensity map by Sparks, Biretta, & Macchetto to predict the spectral evolution along the jet.

We derive limits for the magnetic field and the total pressure by comparing our results with the spatially resolved fit to the spectral data by Neumann, Meisenheimer, & Röser of a model spectrum that cuts off at ≈ 10¹⁵ Hz. To explain the weakness of synchrotron cooling along the jet, the magnetic field strength must lie below the equipartition value. Although the inferred pressure in the limit of nonrelativistic bulk flow lies far above the estimated pressure of the interstellar matter in the center of M87, bulk Lorentz factors Γ_jet in the range of 3-5 and inclination angles θ_LOS ≲ 25° lead to pressure estimates close to the interstellar medium pressure. The average best-fit magnetic fields we derive fall in the range of 20-40 μG, departing from equipartition by a factor ≈ 1.5-5.

This model is consistent with the proposal by Bicknell & Begelman that the knots in the M87 jet are weak, oblique shocks. First-order Fermi acceleration will then have a minimal effect on the slope of the radio-to-optical spectrum while possibly accounting for the X-ray spectrum.

In a new survey of nearby galaxies from stacked photographic images, seemingly regular galaxies of several types show amorphous, often asymmetrical features at very faint levels (28 mag arcsec^-2). In M87, a diffuse fan of stellar material extends along the projected southeast (major) axis out to about 100 kpc.

We suggest that accretion of a small spheroidal galaxy into a larger potential is the most likely explanation for the diffuse structure. The orbit is required to pass close to the center of the potential in order to produce a fan that is nearly aligned with the major axis and has a large opening angle, as seen in M87.

Our simulations include a rigid primary potential with characteristics similar to those derived for M87 and a populated secondary potential. We investigate the structure of the dark matter at large galactic radii by representing M87 with different potentials. The morphologies of the debris of intruder spheres and disks of different masses and orbital parameters limit the possible accretion scenarios. The total luminosity of the fan and the kinematics of debris in the center of the primary potential are analyzed and compared with substructure in M87.

The short lifetimes (t_fan ≲ 5 × 10⁸ yr) of the simulated diffuse fans and lack of observed shells indicates either that we are seeing M87 at a "special time" during its evolution or that infall from small intruder galaxies is common. Our simulations indicate that several accretion events could be hidden in galaxies. For many orbits, intruder material is quickly spread out to very low light levels. Observations of the high specific frequency of globular clusters in M87 provide evidence that the galaxy may experience frequent accretions of this type.

We present new Hα, CO, and H I images of the high gas surface density flocculent spiral NGC 4414 and compare them to recently published near-infrared observations that reveal kiloparsec-scale structures in the inner disk. The subtraction of an estimated supergiant contribution confirms that the near-infrared enhancements represent primarily variations in the surface density of the old stellar population. As material structures of kiloparsec extent cannot be maintained in a differentially rotating disk for sufficiently long times to be discernible in the old stellar population, the near-infrared observations suggest that global dynamical processes contribute to the formation of structure in NGC 4414. However, variations in the distribution of CO, H I, and Hα peaks with respect to near-infrared "arm" structures suggest that stochastic processes are also important.

We present deep spectropolarimetric observations obtained with the W. M. Keck Telescope of the very high redshift (z = 3.79786 ± 0.0024) radio galaxy 4C 41.17. We find that the bright, spatially extended rest-frame UV continuum emission from this galaxy, which is aligned with the radio axis, is unpolarized (P_2σ < 2.4%). This implies that scattered AGN light, which is generally the dominant contributor to the rest-frame UV emission in z ~ 1 radio galaxies, is unlikely to be a major component of the UV flux from 4C 41.17. The resulting total light spectrum shows absorption lines and P Cygni-like features that are similar to those detected in the spectra of the recently discovered population of star forming galaxies at slightly lower (z ~ 2-3) redshifts. It may be possible for a galactic outflow to contribute partially to the absorption line profiles of the low ionization species; however, it is unlikely that the high-velocity wings of the high ionization lines are dominated by a galactic wind since the outflow mass implied by the absorption line strengths is very large. The detection of the S V λ1502 stellar photospheric absorption line, the shape of the blue wing of the Si IV profile, the unpolarized continuum emission, the inability of any AGN-related processes to account for the UV continuum flux, and the overall similarity of the UV continuum spectra of 4C 41.17 and the nearby star-forming region NGC 1741B1 strongly suggest that the UV light from 4C 41.17 is dominated by young, hot stars. If all of the UV emission is due to starlight from a young population, the implied star formation rate is roughly 140-1100 h⁻²₅₀M_☉ yr^-1. The deep spectroscopy presented here combined with the morphology of the system at radio and optical wavelengths and the possibly comparable ages for the radio source structure and the UV stellar population suggest that star formation in 4C 41.17 was triggered by the expansion of the radio source into the ambient medium. Our current observations are consistent with the hypothesis that 4C 41.17 is undergoing its major epoch of star formation at z ~ 4, and that by z ~ 1 it will have evolved to have spectral and morphological properties similar to those observed in known z ~ 1 powerful radio galaxies.

We present new VLA observations of the H I medium of the Local Group dwarf galaxies Sag DIG, LGS 3, and Phoenix. Sag DIG is a gas-rich, blue dwarf irregular with some known recent star formation, whereas LGS 3 and Phoenix are gas-poor, red galaxies of intermediate irregular/spheroidal type with little recent star formation. These galaxies complete a small sample of Local Group and near-Local Group irregular and elliptical galaxies that have been mapped in H I and, where possible, in CO. We compare the properties and kinematics of the ISM in these different galaxy types in order to gain some insights into the relationship between galaxy properties, star formation, and the ISM.

Both Sag DIG and LGS 3 have larger H I extents and higher H I fluxes than previously known, and in both cases the H I extends significantly farther than the stellar component. Neither one shows convincing signs of rotation; both seem to derive a significant amount of their support against gravity from random motions in the gas. The dwarf galaxies of the sample support the idea that there are large variations in the dark/luminous mass ratio at a given luminosity.

The high sensitivity and high spectral and spatial resolution of these observations also make it possible to study the physical properties of the H I medium. The H I in Sag DIG is decomposed into broad (σ = 10 km s^-1) and narrow (σ = 5 km s^-1) components, with the broad component distributed throughout the galaxy and the narrow component concentrated into a small number of prominent clumps of about 8 × 10⁵M_☉. It is argued that these H I components are in fact cold and warm phases of the H I medium, as in Galactic H I and in the dwarf irregular Leo A. LGS 3, on the other hand, shows little sign of such a two-phase H I structure. This new information on the phase structure of the ISM in dwarf galaxies is consistent with theoretical models of the H I medium if the H I line width is greater than purely thermal widths. The lack of a cold H I phase may be a reason for the lack of recent star formation in LGS 3; we suggest that the presence of a cold H I phase serves as a better indicator of conditions appropriate for star formation than measures of total H I content.

The Phoenix dwarf and LGS 3 have been interpreted as two dwarf spheroidal galaxies which are unusual in that they contain H I. The presence of H I in LGS 3 is interesting, then, in the context of models that remove the gas from dwarf spheroidals by a burst of star formation. Either LGS 3 has not had a burst of star formation sufficient to remove its gas, or gas removal was not complete. It is not clear whether the Phoenix dwarf has H I. The current observations show emission in the vicinity of the galaxy at +55 km s^-1 and -23 km s^-1 (heliocentric). However, none of this emission is coincident with the optical galaxy. Until the stellar velocities in Phoenix are known, we cannot distinguish whether any of the detected H I is actually associated with the galaxy or is perhaps associated with the Magellanic Stream.

Observations made with the short-wavelength spectrometer of the InfraredSpaceObservatory are used to investigate the composition of interstellar dust in the line of sight to Cygnus OB2 No. 12, commonly taken as representative of the diffuse (low-density) interstellar medium. Results are compared with data for the Galactic center source Sgr A*. Nondetections of the 3.0 and 4.27 μm features of H₂O and CO₂ ices in Cyg OB2 No. 12 confirm the absence of dense molecular material in this line of sight, whereas the presence of these features in Sgr A* indicates that molecular clouds may contribute as much as 10 mag of visual extinction toward the Galactic center. The spectrum of Cyg OB2 No. 12 is dominated by the well-known 9.7 μm silicate feature; detection of a shallow feature near 2.75 μm indicates that the silicates are at least partially hydrated, with composition possibly similar to that of terrestrial phyllosilicates such as serpentine or chlorite. However, the 2.75 μm feature is not seen in the Galactic center spectrum, suggesting that silicates in this line of sight are less hydrated or of different composition. The primary spectral signatures of C-rich dust in the diffuse ISM are weak absorptions at 3.4 μm (the aliphatic C=H stretch) and 6.2 μm (the aromatic C=C stretch). We conclude, based on infrared spectroscopy, that the most probable composition of the dust toward Cyg OB2 No. 12 is a mixture of silicates and carbonaceous solids in a volume ratio of approximately 3:2, with the carbonaceous component primarily in an aromatic form such as amorphous carbon.

An 8.4 GHz VLA survey of 91 recently discovered lithium-rich late-type stars from the ROSAT All-Sky Survey and pointed observations is presented. These objects lie in the vicinity of the Taurus-Auriga star-forming region (d ≃ 140 pc); however, some are dispersed nearly 30° from known active star-forming cloud cores. This sample represents a spatially complete, flux-limited population of X-ray-bright young stars both within and away from the primary Tau-Aur stellar nurseries. Of the 91 sources, 29 are detected in this radio survey with a sensitivity limit of ~0.15 mJy. If they are at the distance of the star-forming clouds, we find that 32% of widely distributed young stars with L_X ≥ 5 × 10²⁸ ergs s^-1 have radio luminosity densities in excess of 3.5 × 10¹⁵ ergs s^-1 Hz^-1. This detection rate, the ranges of radio and X-ray luminosities, and the L_R/L_X ratios are consistent with known young weak-lined T Tauri stars (ages ~10⁶ yr) that reside within the Taurus molecular clouds, but they are considerably higher than a zero-age main-sequence population such as the Pleiades (age ≃7 × 10⁷ yr). The radio properties thus support the pre-main-sequence classification of the stars. They fitted well among other active young stars on the empirical L_R versus L_X diagram, implying that solar-type gyrosynchrotron activity is the radio emission mechanism.

We report the results of a rocket-borne observation of the [C II] 158 μm line and far-infrared continuum emission at 152.5 μm toward the high-latitude molecular clouds in Ursa Major. We also present the results of a follow-up observation of the millimeter ¹²CO J = 1 → 0 line over a selected region observed by the rocket-borne experiment. We have discovered three small CO cloudlets from the follow-up ¹²CO observations. We show that these molecular cloudlets, as well as the MBM clouds (MBM 27, 28, 29, and 30), are not gravitationally bound. Magnetic pressure and turbulent pressure dominate the dynamic balance of the clouds.

After removing the H I-correlated and background contributions, we find that the [C II] emission peak is displaced from the 152.5 μm and CO peaks, while the 152.5 μm continuum emission is spatially correlated with the CO emission. We interpret this behavior by attributing the origin of the [C II] emission to the photodissociation regions around the molecular clouds illuminated by the local UV radiation field. We also find that the ratio of the molecular hydrogen column density to the velocity-integrated CO intensity is 1.19 ± 0.29 × 10²⁰ cm^-2 (K km s^-1)^-1, from the FIR continuum and the CO data. The average [C II]/FIR intensity ratio over the MBM clouds is 0.0071, which is close to the all-sky average of 0.0082 reported by FIRAS on the COBE satellite. The average [C II]/CO ratio over the same regions is 420, significantly lower than in molecular clouds in the Galactic plane.

We present new narrowband near-infrared images together with K-band spectra of highly collimated bipolar jets close to the IRAS 05487+0255 source. The jets are located ~50'' west of the Herbig-Haro 110 outflow. The jets are not visible at optical wavelengths and therefore do not fall into the "standard" Herbig-Haro object classification scheme. Nevertheless, they belong to an ever growing group of molecular hydrogen jets associated with young stellar objects that are optically undetected. The jets are very well collimated, with length-to-width ratios of ~10-20. The spectra of the jet and its counterjet in the K band show a limited number of H₂ emission lines that make it difficult to obtain an accurate excitation temperature. We estimate T_ex = 1104±67 K and T_ex = 920±156 K for the red and blue jet components, respectively. The radial velocities of the jet and counterjet, based on the shift of the (1,0) S(1) 2.121 μm line, are ~-275 ± 50 km s^-1 and ~180 ± 50 km s^-1, respectively, suggesting an angle of ~30°-45° between the jet and the line of sight. The H₂ emission of the entire jet extends for at least 40'', ~0.1 pc at the distance of Orion. If the flow velocity is comparable to that of the radial velocities, then the dynamical age of the system is quite short (~500 yr), consistent with a young jet arising from an embedded source. Entrainment in a turbulent mixing layer may explain this morphology and spectral character.

The 6₂-6₁E transition of CH₃OH has been observed toward the Orion KL region with an angular resolution of 007 and frequency resolution of 12 kHz (0.15 km s^-1) at two epochs, 1 year apart. The observations have a sensitivity of 1 Jy beam^-1 (corresponding to a main beam brightness temperature of 4 × 10⁵ K). Only the brightest masers previously measured with 3'' resolution are detected; these are resolved into several bright features. The brightness temperatures of the detected masers range from 2 × 10⁶ to 3 × 10⁷ K, with apparent sizes in the range 22-56 AU.

From the trends of radial velocity versus position, the observed line widths of the masers may be explained as blends of individual narrower features, each with line widths less than the frequency resolution of these measurements. The intensities of the masers increase with decreasing apparent size. This is interpreted as evidence that the excitation process plays a larger role than path length in determining maser intensity; our measurement of narrow line widths supports this interpretation.

The CH₃OH masers are probably collisionally pumped and formed in very turbulent regions. The kinetic temperatures are ≈ 100 K, corresponding to sound speeds of order 2 km s^-1. From this, intensity variations are expected on timescales of years, or longer. At both epochs the majority of the maser features are located in the same relative positions with approximately the same flux densities. However, some of the masers appear to vary on timescales of a year. Variations in intensity of the strongest maser feature may have been detected over a year's time. Further, two weak maser features that appeared in 1990 were not detected in 1991.

The supernova shock breaks out (i.e., increases its speed beyond 10⁹ cm s^-1) at about 300 km simply because the velocity of the infalling material decreases greatly with increasing shock radius. The decrease is not due to a decreasing rate of accretion. Recombination of nucleons into α-particles plays only a minor part. The energy of the shock is directly related to the neutrino luminosity at early times, and agrees well with the computation by Wilson et al. and with the observation on SN 1987A. For the calculation of nucleosynthesis it is essential to consider the shock as a sharp discontinuity. The calculated amount of ⁵⁶Ni in SN 1987A is then 3 times the observed amount, indicating substantial fallback. Convection between entropy maximum and shock stops at a shock radius of about 1000 km; this keeps the proton fraction Y_e closer to that of the infalling material.

The recent discovery of persistent gamma-ray burst (GRB) counterparts at lower frequencies permits several important conclusions to be drawn. The spectrum of GRB 970508 is not consistent with an external shock origin for both the prompt GRB and the persistent emission, suggesting that at least the prompt radiation is produced by internal shocks. Comparisons among three GRBs with counterparts (or upper limits on them) establishes that GRBs are not all scaled versions of similar events. The angular size inferred from the apparent observation of self-absorption in the radio spectrum of GRB 970508 a week later implies that its expansion had slowed to semirelativistic speeds. This permits a remarkably low upper bound to be placed on its residual energy, suggesting either that radiation has been more than 99.7% efficient or that the initial outflow was strongly collimated. Observations of self-absorbed radio emission from future GRBs may permit direct measurement of their expansion and determination of their parameters and energetics. We estimate initial Lorentz factors of γ₀ ~ 100 for GRB 970228 and GRB 970508, and present a solution for the evolution of a blast wave with instantaneous cooling.

The relativistic Cowling approximation in which all metric perturbations are omitted is applied to nonaxisymmetric infinitesimal oscillations of uniformly rotating general relativistic polytropes.

Frequencies of lower order f-modes, which are important in analysis of secular instability driven by gravitational radiation, are investigated, and neutral points of the mode along equilibrium sequences of rotating polytropes are determined. Since this approximation becomes more accurate as stars are more relativistic and/or as they rotate more rapidly, we will be able to analyze how a rotation period of a neutron star may be limited by this instability.

Possible errors in determining neutral points caused by omitting metric perturbations are also estimated.

Observations of young clusters indicate that a significant fraction of solar-type stars are rotating very slowly, with equatorial velocities less than 10 km s^-1. So far, models have failed to reproduce a sufficiently large proportion of these stars on the zero-age main sequence. On the basis of the idea that the mixing length in convection theories could depend on the size of the convective zone (Canuto & Mazzitelli), we examine the influence of a varying mixing-length parameter α on the rotational evolution of solar-type stars. A decreasing α (owing to evolution) in the mixing-length theory (MLT) leads to a slower contraction rate and to a larger stellar moment of inertia. The stellar spin-up is consequently reduced, and this helps to increase the number of very slow rotators present in young clusters. We also investigate the possibility that α could depend on the rotation rate, and show the consequences of this parameterization for the lithium surface abundance.

As part of a long-term program of observations to search for and characterize disks of gas and dust around intermediate-mass counterparts to solar-mass T Tauri stars, we have probed the environments of seven pre-main-sequence stars of spectral type Ae using millimeter-wave continuum and molecular line aperture synthesis imaging. In each case we identify a compact region of thermal continuum emission centered on the star. Upper limits to radii are in the range 200-300 AU for five members of our sample, and 680 AU for the distant source HD 245185. We identify an elongated continuum source around HD 163296, with a semimajor axis of 110 AU. Adopting relatively high values for dust grain opacities, we obtain minimum masses of circumstellar dust and gas in the range 0.005-0.034 M_☉ for the seven sources, assuming that the observed continuum emission is optically thin. We detect molecular line emission from gas regions centered on four of the stars. Two of these regions are spatially resolved and are found to be elongated, with semimajor axes of 310 and 450 AU for HD 163296 and AB Aur, respectively. Ordered velocity gradients along the major axes of both of these structures point strongly to the presence of orbiting material in disklike configurations, and we argue that the nebular environments of our entire sample include substantial disk components.

We present new, detailed non-LTE (NLTE) calculations for model atmospheres of novae during outburst. This fully self-consistent NLTE treatment for a number of model atoms includes 3922 NLTE levels and 47,061 NLTE primary transitions. We discuss the implication of departures from LTE for the strengths of the lines in nova spectra. The new results show that our large set of NLTE lines constitutes the majority of the total line-blanketing opacity in nova atmospheres. Although we include LTE background lines, their effects are small on the model structures and on the synthetic spectra. We demonstrate that the assumption of LTE leads to incorrect synthetic spectra and that NLTE calculations are required for reliably modeling nova spectra. In addition, we show that detailed NLTE treatment for a number of ionization stages of iron changes the results of previous calculations and improves the fit to observed nova spectra.

Five rotational transitions of vibrationally excited SiC₂ in the ν₃ = 1 antisymmetric mode were sought toward IRC+10216: 4₁₄-3₁₃, 5₁₅-4₁₄, 6₁₅-5₁₄, 7₁₇-6₁₆, and 7₃₅-6₃₄. All five spectra show emission features. However, the 5₁₅-4₁₄ and 7₃₅-6₃₄ transitions are noticeably lower in intensity compared to the other three. In addition, the 4₁₄-3₁₃, 6₁₅-5₁₄, and 7₁₇-6₁₆ transitions have wider line widths than expected. For these reasons, we conclude that our detection of vibrationally excited SiC₂ is tentative. The ratio of column densities of vibrationally excited SiC₂ to ground state SiC₂ is N(ν₃ = 1)/N(ν₃ = 0) ≲ 0.024.

The positions of the two error boxes for the soft gamma repeater (SGR) 1900+14 were determined by the "network synthesis" method, which employs observations by the Ulysses gamma-ray burst and CGRO BATSE instruments. The location of the first error box has been observed at optical, infrared, and X-ray wavelengths, resulting in the discovery of a ROSAT X-ray point source and a curious double infrared source. We have recently used the ROSAT HRI to observe the second error box to complete the counterpart search. A total of six X-ray sources were identified within the field of view. None of them falls within the network synthesis error box, and a 3 σ upper limit to any X-ray counterpart was estimated to be 6.35 × 10^-14 ergs cm^-2 s^-1. The closest source is ~3' away, and has an estimated unabsorbed flux of 1.5 × 10^-12 ergs cm^-2 s^-1. Unlike the first error box, there is no supernova remnant near the second error box. The closest one, G43.9+1.6, lies ~26 away. For these reasons, we believe that the first error box is more likely to be the correct one.

The UV spectrum of the Red Rectangle and its central source, HD 44179, was examined using the Goddard High Resolution Spectrograph (GHRS) of the HubbleSpaceTelescope (HST). Spectra have been obtained of the 0-0, 0-1, and 0-2 vibronic bands of the CO fourth positive system, the 1657 Å C I multiplet, and the 1931 Å C I line. All of the lines and bands display both absorption and emission components. The components do not display P Cygni profiles, but are nearly symmetrically reversed and show different spreads of radial velocity. The widths and Doppler shifts of each component provide information about the environment of its origin, as do the rotational and vibrational distributions in CO. The 1657 Å multiplet is too complex for analysis, but the 1931 Å line can be deconvolved into two components: emission with narrow velocity dispersion, 25 km s^-1, and absorption with a broad velocity dispersion, 230 km s^-1. The radial velocity of these atomic features is consistent with the velocity of the center of mass of this known spectroscopic binary. For the CO bands, we obtained a best fit from a four-component model, consistent for all three bands. There are three emission components and one absorption component. Two emission components display narrow velocity spreads (29 km s^-1), and one shows a rotational temperature of 50 K, the other, of about 3000 K. The third emission component has a very wide velocity spread (600 km s^-1) and a rotational temperature of about 100 K. The absorption component shows a wide velocity spread (230 km s^-1) and a rotational temperature of about 100 K. The high-temperature emission displayed partially-resolved rotational line structure corresponding to levels of J = 50. The Doppler shifts of these features place the narrow, cold emission with one of the binary components and the narrow, hot emission with the center of mass. We discuss these observations with respect to the accretion disk model of this object. We also compare our UV CO observations with the work of others on the IR CO overtone spectra in young stellar objects.

We report submilliarcsecond-precise astrometric measurements for the late-type star AB Doradus via a combination of VLBI (very long baseline interferometry) and HIPPARCOS data. Our astrometric analysis results in the precise determination of the kinematics of this star, which reveals an orbital motion readily explained as caused by gravitational interaction with a low-mass companion. From the portion of the reflex orbit covered by our data and using a revised mass of the primary star (0.76 M_☉) derived from our new value of the parallax (66.3 mas < π < 67.2 mas), we find the dynamical mass of the newly discovered companion to be between 0.08 and 0.11 M_☉. If accurate photometric information can be obtained for the low-mass companion, our precise mass estimate could serve as an accurate calibration point for different theoretical evolutionary models of low-mass objects. This represents the first detection of a low-mass stellar companion using VLBI, a technique that will become an important tool in future searches for planets and brown dwarfs orbiting other stars.

We report time-resolved photometric observations of the first superoutburst confirmed since 1975 in a most enigmatic cataclysmic variable, AL Comae Berenices, during 1995 April-June. We detected double-peaked modulations in the early phase of the outburst, the period of which is slightly shorter than the superhump period. Three possibilities for the mechanism causing these oscillations are discussed. We also refer to similar modulations observed in ER UMa stars.

As seen during the 1978 outburst of WZ Sge, a dip of over 2 mag in the light curve occurred in the 1995 outburst. After recovery from this dip, AL Com again declined 0.5 mag and rebrightened, showing modulations with the same period as that of the superhump, which indicates that a "second" superoutburst occurred. These modulations disappeared before the final decline. After decline, AL Com stayed about or over 1 mag brighter than its quiescence level for at least 2 weeks. It is likely that the accretion disk remained in its hot state near the inner edge during the dip and after the end of the outburst.

We study the effect of rotation on the excitation of internal oscillation modes of a star by the external gravitational potential of its companion. Unlike the nonrotating case, there are difficulties with the usual mode decomposition for rotating stars because of the asymmetry between modes propagating in the direction of rotation and those propagating opposite to it. For an eccentric binary system, we derive general expressions for the energy transfer, ΔE_s, and the corresponding angular momentum transfer, ΔJ_s, in a periastron passage when there is no initial oscillation present in the star. Except when a nearly precise orbital resonance occurs (i.e., the mode frequency equals multiple of the orbital frequency), ΔE_s is very close to the steady state mode energy in the tide in the presence of dissipation. It is shown that stellar rotation can change the strength of dynamical tide significantly. In particular, retrograde rotation with respect to the orbit increases the energy transfer by bringing lower order g-modes (or f-modes for convective stars), which couple more strongly to the tidal potential, into closer resonances with the orbital motion of the companion.

We apply our general formalism to the problems of tidal capture binary formation and the orbital evolution of the PSR J0045-7319/B star binary. Stellar rotation changes the critical impact parameter for binary capture. Although the enhancement (by retrograde rotation) in the capture cross section is at most ~20%, the probability that the captured system survives disruption/merging and therefore becomes a binary can be significantly larger. It is found that in order to explain the observed rapid orbital decay of the PSR J0045-7319 binary system, retrograde rotation in the B star is required.

Dust from the interstellar medium (ISM) can collide with and destroy particles in the circumstellar dust disks around main-sequence stars (Vega/β Pic stars). Two current theories tying the occurrence of the Vega/β Pic phenomenon to the erosive influence of the ISM are critically reconsidered here. Using the local standard of rest frame, we find little evidence for a correlated motion (streaming) of prominent disk systems, which one theory suggests would result from a passage about 10⁷ yr ago of these stars, but not the control A-type stars, through the nearby Lupus-Centaurus interstellar cloud complex. Moreover, the prototype system of β Pic could not have retained dust produced in such a passage for much longer than 10⁴ yr. We show theoretically that the ISM sandblasting of disks has minor importance for the structure and evolution of circumstellar disks, except perhaps in their outskirts (usually >400 AU from the stars), where under favorable conditions it may cause asymmetries in observed brightness and color. The ISM neither produces the disks (as in one theory) nor depletes and eliminates them with time (as in another theory), because typical ISM grains are subject to strong radiative repulsion from A- and F-type dwarfs (a few to 100 times stronger than gravity). Atypically large ISM grains are not repelled strongly, but are unimportant on account of their small number density.

Dust production and destruction in β Pic-type disks results mainly from their collisional nature enhanced by the radiatively produced eccentricities of particle orbits, rather than from nurture in a hostile ISM. The residence times of the few-micron dust grains predominant in the densest part of the β Pic disk is only 10⁴ yr, or a few dozen orbital periods. Submicronic debris is blown out as β meteoroids, carrying away from this system an equivalent of the solar system's total mass in solids (~120 Earth masses) in only ~65 Myr. This rate of collisional erosion exceeds almost 10⁸ times that of the zodiacal light disk of our own system. A massive and relatively young (≲10² Myr) planetesimal disk appears to surround β Pic, destined to decline in dust density over time comparable to its age. Other dust disks, like those around Fomalhaut and Vega, contain much less dust and may be much older than the β Pic disk, but like the β Pic disk they are also derived from and replenished many times during their lifetimes by unseen parent bodies.

One of the oustanding questions about the architecture of the outer solar system is how the trans-Neptunian disk of comets and small planet-scale objects known as the solar system's Edgeworth-Kuiper Belt (EKB) originated and evolved to its present mass and architecture. Applying a time-dependent model of collisonal evolution of the EKB, we find that under a wide range of assumptions, collisional evolution should have depleted the mass of the 30-50 AU zone by >90% early in the history of the solar system, thereby creating a deep scar or gap in the surface mass density across a wide region beyond Neptune, much like what is observed today. Dynamical erosion may have further accelerated the depletion process. Given the fact that Neptune has had far less dynamical influence beyond 50 AU, our results also suggest that unless the solar nebula was truncated near 50 AU, then surface mass density of solids somewhere beyond ~50 AU may increase again, most likely dramatically.

The Oriented Scintillation Spectrometer Experiment (OSSE) on board the ComptonGammaRayObservatory observed the 1991 June 4 X12+ solar flare, one of the most intense nuclear gamma-ray line flares observed to date. Using these OSSE observations, we have derived time profiles of the various components of gamma-ray emission and obtained information about the accelerated particle spectra and composition and about the ambient plasma at the flare site. The main results are (1) the nuclear reactions associated with the impulsive phase of the flare continued for at least 2 hours and resulted from ions that were probably continuously accelerated rather than impulsively accelerated and trapped; (2) the total energy in these accelerated ions exceeded the energy in >0.1 MeV electrons; (3) the accelerated α/proton ratio was closer to 0.5 than to 0.1; (4) there is evidence for a decrease of the accelerated heavy ion-to-proton ratio as the flare progressed; (5) there is evidence for a temporal change in the composition of the flare plasma; (6) the ratio of electron bremsstrahlung to the flux in narrow γ-ray lines decreased as the flare progressed; (7) the high-energy (>16 MeV) component of the electron spectrum was much more impulsive than the lower energy ~MeV component; (8) a model-dependent upper limit of 2.3 × 10^-5 was obtained for the photospheric ³He/H abundance ratio; and (9) energetic ions may have been present for several hours prior to and following the impulsive phase of the flare.

We present a method for determining the amplitude of mass fluctuations on 8 h^-1 Mpc scale, σ₈. The method utilizes the rate of evolution of the abundance of rich clusters of galaxies. Using the Press-Schechter approximation, we show that the cluster abundance evolution is a strong function of σ₈: dlogn/dz∝-1/σ²₈; low-σ₈ models evolve exponentially faster than high-σ₈ models, for a given mass cluster. For example, the number density of Coma-like clusters decreases by a factor of ~10³ from z=0 to z≃0.5 for σ₈=0.5 models, while the decrease is only a factor of ~5 for σ₈≃1. The strong exponential dependence on σ₈ arises because clusters represent rarer density peaks in low-σ₈ models. We show that the evolution rate at z≲1 is insensitive to the density parameter Ω or to the exact shape of the power spectrum. Cluster evolution therefore provides a powerful constraint on σ₈. Using available cluster data to z~0.8, we find σ₈=0.83±0.15. This amplitude implies a bias parameter b≃σ⁻¹₈=1.2±0.2, i.e., a nearly unbiased universe with mass approximately tracing light on large scales. When combined with the present-day cluster abundance normalization, σ₈Ω^0.5≃0.5, the cosmological density parameter can be determined: Ω≃0.3±0.1.

We present a generic algorithm for generating Gaussian random initial conditions for cosmological simulations on periodic rectangular lattices. We show that imposing periodic boundary conditions on the real-space correlator and choosing initial conditions by convolving a white noise random field results in a significantly smaller error than the traditional procedure of using the power spectrum. This convolution picture produces exact correlation functions out to separations of L/2, where L is the box size, which is the maximum theoretically allowed. This method also produces top-hat sphere fluctuations that are exact at radii R≤L/4. It is equivalent to windowing the power spectrum with the simulation volume before discretizing, thus bypassing sparse sampling problems. The mean density perturbation in the volume is no longer constrained to be zero, allowing one to assemble a large simulation using a series of smaller ones. This is especially important for simulations of Lyα systems where small boxes with steep power spectra are routinely used.

We also present an extension of this procedure that generates exact initial conditions for hierarchical grids at negligible cost.

We derive a theoretical relation between R_BLR, the size of the broad emission line region (BLR) of active galactic nuclei (AGNs), and the observed soft X-ray luminosity and spectrum. We show that in addition to the well-known R_BLR~L^1/2 scaling, R_BLR should depend also on the soft X-ray spectral slope, and we derive the expected relation between R_BLR and the X-ray luminosity and spectral index. Applying this relation to calculate a predicted BLR radius for 10 AGNs with reverberation data, we show that including the dependence on the spectrum improves the agreement between the calculated BLR radius and the radius independently determined from reverberation mapping. Similarly, we evaluate an expression for the line width and show that including the dependence on the spectrum significantly improves the agreement between the calculated BLR velocity dispersion and the observed FWHM of the Hβ line.

The theoretical expression for the line width also provides a physical explanation to the anticorrelation between the soft X-ray slope and the emission-line width observed in narrow-line Seyfert galaxies.

Relatively few intensively star-forming galaxies at redshifts of z>2.5 have been found in the Hubble Deep Field (HDF). This has been interpreted to imply a low space density of elliptical galaxies at high z, possibly due to a late (z<2.5) epoch of formation or to dust obscuration of the ellipticals that are forming at z~3. I use Hubble Space Telescope UV (~2300 Å) images of 25 local early-type galaxies to investigate a third option, that ellipticals formed at z>4.5 and were fading passively by 2<z<4.5. Present-day early-type galaxies are faint and centrally concentrated in the UV. If elliptical galaxies formed their stars in a short burst at z>4.5 and have faded passively to their present brightnesses at UV wavelengths, they would generally be below the HDF detection limits in any of its bands at z>2.5. Quiescent z~3 ellipticals, if they exist, should turn up in sufficiently deep IR images.

We present the gamma-ray spectrum of the BL Lacertae object, Markarian 421, above 500 GeV during its most intense recorded TeV flare, on 1996 May 7. The spectrum is well fitted by a power law with an exponent of -2.56±0.07±0.1 (statistical and systematic errors). The spectrum extends above 5 TeV with no evidence for a cutoff, favoring determinations of the extragalactic infrared energy density that do not produce a sharp cutoff at these energies.

BL Lacertae was detected by the EGRET instrument on the Compton Gamma Ray Observatory at the 10.2 σ level with an average flux of (171 ± 42) × 10⁻⁸ photons cm^-2 s^-1, at energies greater than 100 MeV, during the optical outburst of 1997 July. This flux is more than 4 times the previously highest level. Within the July 15-22 observation there was a dramatic factor of 2.5 increase in the gamma-ray flux on July 18.75-19.08, apparently preceding, by several hours, a brief optical flare. The gamma-ray flux decreased to its previous level within 8 hr, and the optical flux decreased to its prior level in less than 2 hr. The gamma-ray photon spectral index of 1.68±0.12 indicates that the spectrum during the 7 day observation was harder than the previous detection.

Resolution of both approaching and receding ejecta in the galactic microquasars makes it possible to measure the flux ratio S_a/S_r of twin ejecta, which contains important information about the nature of the jets. We show that the flux ratio S_a/S_r=8±1 observed from GRS 1915+105 during the prominent 1994 March/April radio flare can be explained in terms of relativistic motion of discrete radio clouds, if one assumes that the twin ejecta are similar but not completely identical, i.e., allowing for some asymmetry between the plasmoids in their speeds of propagation and/or luminosities. The recoil momentum due to asymmetrical ejection of the pair of plasmoids may be comparable to the momentum accumulated in the inner accretion disk. We suggest a possible explanation for the observed anticorrelation between the X-ray and radio flares; it may be the result of drastic structural changes in the inner disk caused by production of powerful jets. The delay between the times of decline of X-rays and the appearance of strong radio flares is explained by the time needed for expansion of radio clouds to become optically thin.

We report RXTE/PCA observations of 4U 1608-52 on March 15, 18, and 22 immediately after the outburst in early 1996. The persistent count rates ranged from 190 to 450 counts s^-1 (1-60 keV). During this period of time, 4U 1608-52 was in the island state. We detected quasi-periodic oscillation (QPO) features in the power density spectra (PDS) at 567-800 Hz on March 15 and 22, with source fractional root mean square (rms) amplitude of 13%-17% and widths of 78-180 Hz. The average rms amplitude of these QPO features is positively correlated with the energy. Our results imply that the neutron star spin frequency is possibly between 300 and 365 Hz.

We observed the low-mass X-ray binary and Z source GX 17+2 with the Rossi X-Ray Timing Explorer during 1997 February 6-8, April 1-4, and July 26-27. The X-ray color-color diagram shows a clear Z track. Two simultaneous kHz quasi-periodic oscillations (QPOs) are present in each observation, whose frequencies are well correlated with the position of the source on the Z track. At the left end of the horizontal branch (HB), only the higher frequency peak is observed, at 645 ± 9 Hz, with an rms amplitude of 5.7% ± 0.5% and an FWHM of 183 ± 35 Hz. When the source moves down the Z track to the upper normal branch, the frequency of the kHz QPO increases to 1087 ± 12 Hz, and the rms amplitude and FWHM decrease by a factor of 2. Farther down the Z track, the QPO becomes undetectable, with rms upper limits typically of 2.0%. Halfway down the HB, a second QPO appears in the power spectra with a frequency of 480 ± 23 Hz. The frequency of this QPO also increases when the source moves along the Z track, up to 781 ± 11 Hz halfway down the normal branch, while the rms amplitude and FWHM stay approximately constant at 2.5% and 70 Hz. The QPO frequency difference is constant at 293.5 ± 7.5 Hz. Simultaneously with the kHz QPOs, we detect HB QPOs (HBOs). The simultaneous presence of HBOs and kHz QPOs excludes the magnetospheric beat-frequency model as the explanation for at least one of these two phenomena.

We report the discovery of an unusual X-ray burst from the direction of the globular cluster M28 using data acquired with ASCA. The burst was recorded by all four ASCA telescopes and displays a fast (≲70 ms) rise followed by an exponential decay (τ=7.5 s) and a steady afterglow that lasts between 800 and 3250 s. The image of the burst is consistent with an ASCA point source and is centered on quiescent X-ray emission from the core of M28. The burst temporal profile is similar to type I bursts emitted by accreting neutron stars of low-mass X-ray binaries (LMXBs). We argue that the burst arises from a LMXB that is located in the core of M28. The burst is unique in two ways: it is intrinsically subluminous, ~0.02L_E, and more importantly it originates from a source whose quiescent luminosity is fainter than that of the known cluster bursters by 3 orders of magnitude. We suggest that this burst is from a highly magnetized neutron star accreting at a low rate. These accreting systems may account for the mysterious low-luminosity X-ray sources in globular clusters.

We have obtained the first classification spectrum and present the first direct spectral classification of the ionizing star of an ultracompact H II region. The ultracompact H II region is G29.96-0.02, a well-studied object with a metallicity value roughly twice that of the Sun. The near-infrared K-band spectrum of the ionizing star exhibits C IV and N III emission and He II absorption, but lines of H I and He I are obliterated by nebular emission. We determine that the star has a spectral type of O5-O7 or possibly O8. We critically evaluate limits on the properties of the star and find that it is compatible with zero-age main-sequence properties only if it is binary and if a significant fraction of the bolometric luminosity can escape from the region. G29.96-0.02 will now be an excellent test case for nebular models, as the properties of the ionizing star are independently constrained.

The difference in formation process between binary stars and planetary systems is reflected in their composition, as well as orbital architecture, particularly in their orbital eccentricity as a function of orbital period. It is suggested here that this difference can be used as an observational criterion to distinguish between brown dwarfs and planets. Application of the orbital criterion suggests that, with three possible exceptions, all of the recently discovered substellar companions may be brown dwarfs and not planets. These criterion may be used as a guide for interpretation of the nature of substellar-mass companions to stars in the future.

Thermal emission spectra from hydrogenated amorphous carbon (HAC) over the 2.5-15 μm range have been obtained at temperatures between 573 and 773 K. These spectra are similar, but not identical, to absorption spectra of HAC samples subjected to the same thermal cycle. A distinct 3.29 μm aromatic CH emission is found in samples heated to temperatures in excess of 723 K. This emission is also observed at lower temperatures in samples that have been thermally cycled to higher temperature. Laboratory emission spectra from HAC are shown to provide a good simulation of 3.2-3.6 μm emission from dust in the extended atmospheres of post-asymptotic giant branch objects. The ratio of 3.4 and 3.3 μm emission in such objects may be a useful indicator of evolutionary status.

The faint extended broad (≥1000 km s^-1) optical emission lines associated with giant H II regions are shown here to be produced in a shell of ISM material smoothly accelerated soon after breakout. Two-dimensional calculations of remnants caused by a strong energy deposit in a low metal abundance ISM are here shown to undergo breakout once encountering a steep density gradient, leading to a fast-moving shell capable of producing the broad and faint emission lines. Energetic sources lead to fast, thick, and hot shells, and when evolving in a low-metallicity ISM, to quasi-adiabatic shells that strongly delay their fragmentation owing to Rayleigh-Taylor instabilities. At the same time, these are smoothly accelerated to reach large distances from the breakout point. The shell acceleration is promoted by the passage of several shocks with small relative speeds, caused by the continuous push exerted by the hot gas that steadily increases its speed to fill the deformed superbubble volume.

This Letter presents the observations of the first two coronal mass ejections (CMEs) obtained with the Ultraviolet Coronagraph Spectrometer of SOHO. Both CMEs were observed at high spectral resolution in the ultraviolet domain. The first event on 1996 June 6-7 was observed in H I Lyα λ1216 and Lyβ λ1026, O VI λλ1032 and 1037, Si XII λλ499 and 521 and imaged within 1.5 and 5 R_☉. The second event on 1996 December 23 was observed in several H I lines and cool lines such as C III λ977, N III λλ990-992, and O V λ630. The analysis of line profiles has allowed us to determine the line-of-sight velocities of the extended corona during a mass ejection. In particular there is evidence for mass motions consistent with untwisting magnetic fields around an erupted flux tube in one of the events and line of sight velocities of 200 km s^-1 in the early phase of the second event presumably related to the expansion of the leading arch of the transient.

Using observations from the Solar Ultraviolet Measurements of Emitted Radiation experiment flown on the Solar and Heliospheric Observatory spacecraft, we have measured Doppler wavelength shifts in the north polar coronal hole in the 1032 and 1038 Å emission lines of O VI and the 1036 and 1037 Å emission lines of C II relative to chromospheric emission lines. These observations were obtained on 1996 November 2 when the north polar coronal hole boundary extended southward to about 750'' (cos θ=0.65). Our measurements indicate the presence of average net redshifts in coronal holes at temperatures of less than 2.9×10⁵ K. Measurements of systematic wavelength shifts in the Ne VIII resonance lines relative to the quiet Sun suggest a transition to average net outflows near 6.3×10⁵ K in the coronal hole.

A coronal mass ejection (CME) observed by LASCO exhibits evidence that its magnetic field geometry is that of a flux rope. The dynamical properties throughout the fields of view of C2 and C3 telescopes are examined. The results are compared with theoretical predictions based on a model of solar flux ropes. It is shown that the LASCO observations are consistent with a two-dimensional projection of a three-dimensional magnetic flux rope with legs that remain connected to the Sun.

We examine spectral properties of the network chromosphere and lower transition region from the SUMER instrument on the SOHO spacecraft, using time-series data sets discussed in an accompanying Letter by Carlsson, Judge, & Wilhelm. The data were obtained early in the mission with no tracking of solar features and so cannot generally be used to examine intrinsic variations in features on timescales in excess of 383 s. Upon examination of the temporal variations and some preliminary power spectrum analysis, we find the following: (1) Transition region lines show more redshift in network regions than in internetwork regions and also a correlation between line intensity brightenings and increased redshift. (2) The internetwork "Ca II grain" phenomenon is not seen in He I λ584 or in lines of Si III and C III. (3) Very rapid changes are seen in the network for transition region lines with no obvious correspondence with the underlying chromosphere. (4) He I λ584 line profiles show very slow time variations. (5) Small-amplitude (2-5 km s⁻¹) coherent oscillations of 5''-10'' scale length and ~130 s period are seen in Doppler shifts of Si III between regions of bright network elements. (6) Essentially all blueshifts or redshifts are substantially less than line widths. We conclude that upward-propagating acoustic shock waves do not contribute significantly to the heating of the lower transition region, and that ionization equilibrium is likely to fail for the interpretation of certain emission lines. The spatial coherency of the Si III velocity oscillations indicates that the quiet Sun's magnetic field topology is more uniform than emission-line intensity data alone might suggest.

We report on the discovery and analysis of a striking neutral sodium gas tail associated with comet C/1995 O1 Hale-Bopp. Sodium D line emission has been observed at heliocentric distance r≤1.4 AU in some long-period comets, and the presence of neutral sodium in the tailward direction of a few bright comets has been noted, but the extent, and in particular the source, has never been clear. Here we describe the first observations and analysis of a neutral sodium gas tail in comet Hale-Bopp, which is entirely different from the previously known ion and dust tails. We show that the observed characteristics of this third type of tail are consistent with it being produced by radiation pressure due to resonance fluorescence of sodium atoms and that the lifetime for photoionization is consistent with recent theoretical calculations.