Table of contents

Repeated imaging observations have been made of NGC 4639 with the HubbleSpaceTelescope between 1995 April and July over an interval of 71 days. Images were obtained on 12 epochs in the F555W band and on two epochs in the F814W band. The galaxy is parent to the prototypical Type Ia supernova SN 1990N. A total of 18 definite and six likely Cepheids were found with periods ranging between 17 and 67 days. The dereddened distance modulus of NGC 4639 derived from the Cepheids that could be measured in both passbands is (m - M)₀ = 32.03 ± 0.22, which corresponds to a distance of 25.5 ± 2.5 Mpc. The apparent magnitudes of SN 1990N at maximum of B_max = 12.70 ± 0.05 and V_max = 12.61 ± 0.05 combined with the Cepheid distance modulus gives M_B(max) = -19.33 ± 0.23 and M_V(max) = -19.42 ± 0.23 for SN 1990N. Combining these with data for five other SNe Ia calibrated in earlier papers of this series and with a preliminary calibration of SN 1989B gives ⟨M_B(max)⟩ = -19.52 ± 0.07, and ⟨M_V(max)⟩ = -19.48 ± 0.07. Combining these mean absolute magnitudes with the Hubble diagrams for well-observed blue SNe Ia in B and V read at remote redshifts well beyond all local velocity anomalies gives a mean Hubble constant of using no second-parameter corrections. Analysis is also given using various formulations of second-order corrections depending on Hubble type of the parent galaxy, color of the SNe Ia, and/or light-curve shape, which show that these uncertain corrections affect the value of H₀ by less than 10%.

Finally, we argue that the short distance scale espoused from both the distance to M100 in the Virgo Cluster and NGC 1365 in the Fornax cluster (Freedman et al.), giving H₀ = 80 ± 17 and H₀ ≈ 70, respectively, have systematic errors in their precepts concerning the distances of the cluster cores, which when corrected give the long scale with H₀ ≈ 55.

The power spectrum (PS) of mass density fluctuations, independent of "biasing," is estimated from the Mark III catalog of peculiar velocities using Bayesian statistics. A parametric model is assumed for the PS, and the free parameters are determined by maximizing the probability of the model given the data. The method has been tested using detailed mock catalogs. It has been applied to generalized CDM models with and without COBE normalization.

The robust result for all the models is a relatively high PS, with P(k)Ω^1.2 = (4.8 ± 1.5) × 10³(h^-1 Mpc)³ at k = 0.1 h Mpc^-1. An extrapolation to smaller scales using the different CDM models yields σ₈Ω^0.6 = 0.88 ± 0.15. The peak is weakly constrained to the range 0.02 ≤ k ≤ 0.06 h Mpc^-1. These results are consistent with a direct computation of the PS. When compared to galaxy-density surveys, the implied values for β (≡ Ω^0.6/b) are of order unity to within 25%.

The parameters of the COBE-normalized, flat CDM model are confined by a 90% likelihood contour of the sort Ω hn^ν = 0.8 ± 0.2, where μ = 1.3 and ν = 3.4, 2.0 for models with and without tensor fluctuations, respectively. For open CDM the powers are μ = 0.95 and ν = 1.4 (no tensor fluctuations). A Γ-shape model free of COBE normalization yields only a weak constraint: Γ = 0.4 ± 0.2.

A formalism that simultaneously searches for monopolar and dipolar peculiar velocities is presented. The formalism is applied to (1) the Mark III Catalog of Galaxy Peculiar Velocities, (2) Lauer & Postman's Abell cluster catalog, and (3) Riess et al.'s Type Ia supernova sample. The emphasis is drawn to the monopolar peculiar velocities, i.e., peculiar Hubble flows, within these samples. The samples show inconsistent peculiar Hubble flows within a depth of ~60 h^-1 Mpc. Beyond a depth of ~80 h^-1 Mpc, the Hubble flows of all samples converge to the global Hubble flow to better than 10% at the 2 σ level. The results are compared with theoretical predictions. They at face value disfavor models that predict smaller peculiar velocities, such as the tilted cold dark matter model. Limitations of the catalogs are discussed, and so are ways to improve the catalogs so that an accurate map of Hubble flows in our local universe can be drawn and be compared with theoretical predictions.

We calculate the number of damped Lyα absorbers expected in various popular cosmological models as a function of redshift and compare our predictions with observed abundances. The Press-Schechter formalism is used to obtain the distribution of halos with circular velocity in different cosmologies, and we calibrate the relation between circular velocity and absorption cross section using detailed gasdynamical simulations of a standard cold dark matter (CDM) model. Because of this calibration, our approach makes more realistic assumptions about the absorption properties of collapsed objects than previous, analytic calculations of the damped Lyα abundance. CDM models with Ω₀ = 1, H₀ = 50 km s^-1 Mpc^-1, baryon density Ω_b = 0.05, and scale-invariant primeval fluctuations reproduce the observed incidence and redshift evolution of damped Lyα absorption to within observational uncertainty, for both COBE normalization (σ₈ = 1.2) and a lower normalization (σ₈ = 0.7) that better matches the observed cluster abundance at z = 0. A tilted (n = 0.8, σ₈ = 0.7) CDM model tends to underproduce absorption, especially at z = 4. With COBE normalization, a CDM model with Ω₀ = 0.4, Ω_Λ = 0.6 gives an acceptable fit to the observed absorption; an open CDM model is marginally acceptable if Ω₀ ≥ 0.4 and is strongly inconsistent with the z = 4 data if Ω₀ = 0.3. Mixed dark matter models tend not to produce sufficient absorption, being roughly comparable to tilted CDM models if Ω_ν = 0.2 and failing drastically if Ω_ν = 0.3.

The effects of radiation force by the cosmic background radiation (CBR) on early-formed objects shortly after the cosmological recombination are explored. In particular, we are interested in the angular momentum transport in massive disks that originate from tidally spun-up density fluctuations. The radiation force by the CBR is calculated by solving the radiative transfer equation and including Thomson scattering, under the assumption that a disk is locally approximated to be a plane-parallel medium in longitudinal motion. As a numerical technique, we employ a variable Eddington factor method. The results show that the efficiency of angular momentum extraction by the CBR decreases exponentially with optical depth even if the radiative diffusion is effective. This implies that the photon-scattering process in moving media proceeds just like the pure absorption process as far as momentum or angular momentum transport is concerned. Because of the present effects, the distributions of the spin probability of early-formed disks could be significantly modified in the presence of the strong CBR. An optically thin disk originating from a large spin parameter (λ > 0.05) would shed angular momentum until it becomes optically thick, resulting in λ ~ 0.05, almost regardless of mass scales of objects and the initial power spectrum. Also, the CBR force likely helps to enhance the effects of shear viscosity, thereby enabling a seed black hole form at z > 10 to account for quasar formation.

I report on a modern empirical realization of the geometric Baade-Wesselink (BW) method aimed at providing accurate distances to Cepheid variable stars. The method is carefully calibrated using high-precision observational data now available from spectroscopic and interferometric techniques. The distance-dependent linear radius R of a Cepheid, derived from the period P according to the period-radius (PR) relation, is matched to the distance-independent angular diameter Φ inferred from the surface brightness-color (SC) correlation using the infrared color (V - K) as brightness indicator. The resulting pulsation parallax is found to be largely insensitive to most of known biasing effects, notably the reddening and metallicity, because of the application of the magnitude-color combination (V, V - K) in deriving stellar angular sizes.

The reliability of the actual BW realization is critically checked using 25 galactic Cepheids in clusters with available high-precision photometry, reddenings and distances by the zero-age main-sequence (ZAMS) fitting approach. I find that 20 of these stars show an average residual (BW - ZAMS) = -(0.01 ± 0.03) mag with a scatter σ = ±0.12 mag, whereas the remaining five stars deviate by up to 5 σ. The fairly good agreement with 20 primary calibrators gives high confidence on the reliability of the method which can provide now an independent route to the problem of cosmic distance scale calibration. According to the overall residual, the actual uncertainty on the absolute galactic distance scale is suggested to be ±0.04 mag, i.e., a factor of 2 smaller than that currently quoted for the Pleiades distance modulus which limits the ZAMS calibration.

The BW approach is then applied to the extragalactic range by including the Magellanic Clouds Cepheids (LMC and SMC) with available high-precision photometry in V and K passbands. An accurate PR relation is recalibrated by a composite fit to galactic and extragalactic Cepheids. The relation shows a dispersion as low as σ = ±0.11 mag nearly close to the scatter affecting BW distances to calibrating galactic Cepheids. It is argued that this scatter is likely due to the asymptotic spread set by the finite width of the instability strip. The BW distances to the LMC and SMC are also determined to be μ₀(LMC) = (18.58 ± 0.024) mag and μ₀(SMC) = (19.00 ± 0.025) mag or d(LMC) = (52.0 ± 0.6) kpc and d(SMC) = (63.2 ± 0.7) kpc with uncertainties not including the contribution due to the absolute distance scale calibration. The Cepheid-based distance to the LMC shows close agreement with the geometric expansion parallax of the SN 1987A in LMC given by d(SN 1987A) = (51.1 ± 1.5) kpc (Panagia et al. 1996).

The BW method is also calibrated by including the Johnson-Cousins color (V - I) relevant in the HubbleSpaceTelescope (HST) observations of Cepheids. By using sets of SMC, LMC, and galactic Cepheids with high-precision V, I, and K photometry, it is shown that the angular sizes predictable by the magnitude-color combination (V, V - I) can be affected by the metallicity. However, the induced effects on BW distances are found to be as small as 0.03 mag up to the metal content of the SMC. Furthermore, it is demonstrated that the current HST approach to the extragalactic distances using a linear combination of period-luminosity (PL) relations in V and I passbands yields the same observational relation as that of the BW realization in the magnitude-color combination (V, V - I). This achievement plays a major role in improving the extragalactic distance scale set by Cepheid data from HST observations. First, all distances spanning from the galactic to extragalactic range can be now sampled by the same BW relation strongly supported by the fundamental results of several observational techniques. Second, all distances will result to be affected by the metallicity as the BW data, i.e., as the SC correlation applied for inferring the Cepheid angular sizes. Third, all distances will suffer from random errors as the almost reddening-free BW data which cancel out the extinction on a star-by-star basis, notably the differential reddening internal to the parent galaxies. In order to show this relevant improvement, the BW distance to the Virgo galaxy M100 is determined to be μ₀(M100) = (31.03 ± 0.06) mag or d(M100) = (16.1 ± 0.5) Mpc with a random error lowered by about a factor of 3 with respect to that derived according to the PL relations by the HST procedure.

The statistics of gravitationally lensed quasars with multiple images in the 01-7'' range have been measured in various surveys. Little is known, however, about lensed-quasar statistics at larger image separations, which probe masses on the scale of galaxy clusters. We extend the results of the HubbleSpaceTelescope (HST) Snapshot Survey for Lensed Quasars to the 7''-50'' range for a subsample of 76 quasars that is free of known selection effects. Using a combination of multicolor photometry and spectroscopy, we show that none of the point sources in the entire field of view of the HST observations of these quasars are lensed images. Large-separation quasar lensing is therefore not common. We carry out a detailed calculation of the expected statistics of large-separation lensing for this quasar sample, incorporating realistic input for the mass profiles and mass function of galaxy clusters. We find that the observational null results are consistent with the expected effect of galaxy clusters, even if these have existed in their present form and number since z ~ 2 (and certainly if they were formed more recently). The rarity of large-separation lensed quasars can rule out some extreme scenarios, e.g., that the mass function of clusters has been severely underestimated or that large mass concentrations that are not associated with galaxies (i.e., "failed" clusters) are common. The rareness of cluster lensing also sets limits on the cosmological constant λ that are independent of limits derived from galaxy lensing. The lensing frequency depends strongly on the central density of clusters. The lensing statistics of larger quasar samples (e.g., the Sloan Digital Sky Survey) can probe the structure, number, and evolution of clusters, as well as the geometry of space.

A planetary microlensing event occurs when a planet perturbs one of the two images created in a point-mass microlensing event, causing a deviation from the standard Paczyński curve. Determination of the two physical parameters that can be extracted from a planetary microlensing event, the planet/star mass ratio q, and the planet/star separation in units of the stellar Einstein ring, y_p, is hampered by several types of degeneracies. There are two distinct and qualitatively different classes of planetary events: major and minor image perturbations. For major image perturbations, there is a potentially crippling continuous degeneracy in q which is of order δ⁻¹_d, where δ_d is the maximum fractional deviation of the planetary perturbation. Since the threshold of detection is expected to be δ_d ~ 5%, this degeneracy in q can be a factor of ~20. For minor image perturbations, the continuous degeneracy in q is considerably less severe, and is typically less than a factor of 4. We show that these degeneracies can be resolved by observations from dedicated telescopes on several continents together with optical/infrared photometry from one of these sites. There also exists a class of discrete degeneracies. These are typically easy to resolve given good temporal coverage of the planetary event. Unambiguous interpretation of planetary microlensing events requires the resolution of both types of degeneracy. We describe the degeneracies in detail and specify the situations in which they are problematic. We also describe how individual planet masses and physical projected separations can be measured.

The formal V/V_max test is generalized to any kind of object distribution. This task is realized by introducing a new statistic, n/n_max, which takes into account the luminosity function of the objects. Our analysis shows that the values of n/n_max are uniformly distributed in the interval [0, 1] and the expectation value of ⟨n/n_max⟩ is ½ in any kind of distribution of the objects if the luminosity function adopted is correct. Besides the ⟨n/n_max⟩ test, we suggest another analysis which could test the uniformity of the distribution of the values of n/n_max. These two kinds of test provide a strict constraint on the selection of luminosity functions.

Statler demonstrated that self-consistent triaxial models with the perfect density law could be constructed for virtually any choice of axis ratios. His experiments are repeated here using triaxial mass models based on Jaffe's density law, which has a central density that diverges as r^-2, similar to what is observed in low-luminosity elliptical galaxies. Most of the boxlike orbits are found to be stochastic in these models. Because timescales for chaotic mixing are generally shorter than a galaxy lifetime in triaxial models with strong cusps, and because fully mixed stochastic orbits have shapes that are poorly suited to reproducing a triaxial figure, only the regular orbits are included when searching for self-consistent solutions. As a result of the restriction to regular orbits, equilibrium solutions are found only for mass models with a modest range of nearly spheroidal or nearly spherical shapes. This result may explain in part the narrow range of elliptical galaxy morphologies and kinematics.

The prospects of detecting extragalactic supernovae (SNs) down to visual magnitudes 23-25 give us hope for observing them in the most distant parts of the universe. Using a Monte Carlo method of stellar population synthesis (the Scenario Machine), we compute, under standard assumptions of stellar evolution, the rates of SNs of various types in a model galaxy and evaluate the SN rates in the universe. The expected cumulative distribution log N - m (number of events - stellar magnitude) is calculated for various SN types and different star formation histories in the universe. The results are also presented in terms of evolution of supernova units with redshifts. Recent observational data on the high-redshift SN Ia rate are in good agreement with our predictions for the relative density of baryons contained in stars at the present time (Ω_* = 0.0057).

We demonstrate a new way to search for blazars by correlating the positions of gamma-ray sources from the Second EGRET Catalog, and polarized radio sources from the National Radio Astronomy Observatory Very Large Array Sky Survey. Using this survey, we first investigated the radio polarization properties of Angel & Stockman blazars and concluded that blazars have a characteristic radio polarization at 1.4 GHz, P_1.4 > 1%. Using this information and the NASA/IPAC Extragalactic Database, we were successful in reconfirming 14 of 18 known EGRET blazars. In addition, 12 potential identifications, eight of which are unique to this paper, were made for unidentified or low-confidence identification EGRET objects. We suggest four new plausible associations for gamma-ray sources previously identified with AGNs in the Second EGRET Catalog. Finally, one low-confidence AGN suggested by Thompson and coworkers is found to be an unlikely identification.

The COSB source 2CG 135+01 has been observed by the EGRET instrument on 10 different occasions during the first ~52 months of the ComptonGammaRayObservatory mission. The source is detected in all but one of the observations. For that one, the exposure was inadequate. The only likely source that is spatially coincident with the gamma-ray position is the radio source GT 0236+610/LS I +61°303. However, there is no compelling evidence for time variations in the gamma-ray emission associated with the radio outbursts from GT 0236+610. Spectral determinations on a timescale of a few days also give no strong evidence for a spectral variation associated with the radio emission of GT 0236+610. Such fluctuations might be expected based on models involving a compact object in an elliptical binary orbit about a massive star. The search for correlations simultaneous with the 8.4 GHz radio outbursts were supported by coordinated observations with the Madrid Deep Space Network during one of the exposures and by Green Bank Interferometer observations on two others. Although there is some possible variability in the gamma-ray flux, it is not clear that it is related to the radio phasing.

This paper presents an analysis of moderately large samples of type 1 and 2 Seyfert galaxies through optical observations and far-infrared IRAS data, also taking into account theoretical color indices derived from dust emission models. The galaxies in the samples cover a rather large interval in far-infrared luminosity, i.e., 7.6 ≤ log (L_IR/L_☉) ≤ 12.6. We show that both types of Seyferts have approximately the same distribution of number of objects with a given L_IR. Galaxies with similar far-infrared color indices α(100, 60) are grouped together, and the corresponding average color indices are interpreted in terms of a simple model in which the observed colors result from the combination of dust directly heated by the active galactic nucleus with a component from the host galaxy represented by the emission of cool dust. On the basis of the average IRAS colors of the derived groups, we show that type 1 and 2 Seyfert galaxies are undistinguishable from each other. From the luminosity ratios L_IR/L_Hα and L_IR/L_{[O III]}, we show that basically the same model can be applied to both types of Seyfert, only allowing for the variation of model conditions: type 2 Seyferts would be like type 1 Seyferts but with the Seyfert nucleus and broad line region more effectively "hidden" by dust.

Two families of models of dusty tori in active galactic nuclei (AGNs; moderately thick and extended versus very thick and compact) are tested against available observations. The confrontation suggests that the former class better explains the infrared (IR) broadband spectra of both broad- and narrow-line AGNs, the anisotropy of the emission deduced by comparing IR properties of Seyfert 1 and 2 nuclei, and the results of IR spectroscopy and those of high spatial resolution observations. There is, however, clear evidence for a broad distribution of optical depths. We also examine the relationship between IR and X-ray emission. The data support a view in which the matter responsible for the X-ray absorption is mostly dust free, lying inside the dust sublimation radius. The consequences of these results for the hard X-ray background as well as IR counts and background are discussed.

A new hydrodynamic model for aperiodic X-ray fluctuations from black hole objects is presented. X-ray fluctuation is thought to originate from instabilities of the accretion disk around a black hole. We previously proposed the cellular automaton model based on the notion of self-organized criticality to describe the open, dissipative nature of a disk undergoing instabilities. Assuming that an avalanche was triggered when the mass density exceeded some critical value, we could reproduce the basic statistical features of X-ray variability with this model. As an independent inquiry, we also calculated the response of an axisymmetric, advection-dominated disk to thermal perturbations, demonstrating that it can easily produce an X-ray shot as is observed in X-ray fluctuations. In this paper, we couple these two different lines of study and calculate the time evolution of an advection-dominated disk with a critical behavior, namely, if surface density exceeds some critical value, we let viscosity increase. As a result, we succeed in producing 1/f-like fluctuations. We also find that the fluctuation properties depend upon the prescription of the critical condition. The present model predicts a constant slope in the decline part of the power spectral density (PSD), while f_break, the frequency separating the flat and decline parts of the PSD, changes according to variations in the size of the advection-dominated portions of the disk.

Previously we identified some short burst events from the BATSE 1B catalog that were consistent with an origin of primordial black hole evaporation at the quark-gluon phase transition. We showed that these events are also consistent with arising from a homogeneous spatial distribution. Recently the PHEBUS group has also indicated that the short bursts are consistent with V/V_max ~ 1/2. In this paper we describe the results of the study of the BATSE 3B catalog shape that confirm the results from BATSE 1B.

We report the classification of 21 new extreme-ultraviolet sources from the recent catalog of Lampton et al. The optical spectra presented identify the objects as 14 active late-type stars (including two double active stars and a possible T Tauri star), three white dwarfs, and six active galactic nuclei (a Seyfert galaxy, the BL Lac object 1ES 1028+511 [=EUVE J1031+508], and four quasi-stellar objects). We have detected Ca II absorption lines in the BL Lac object and measured its redshift. Two of the white dwarfs are unusually massive (M > 1.1 M_☉). Our sample of late-type stars includes five previously known high proper motion objects (EUVE J1004+503, J2244-332A,B, J1802+642, and J1131-346), of which one is the well-known flare star TX PsA (EUVE J2244-332B). We report an unusually high level of activity for the primary component of the TX PsA system (EUVE J2244-332A), which may indicate flare activity. The group of late-type stars is on average almost 3 mag fainter (⟨m⟩ ≈ 13) than the typical member of the ExtremeUltravioletExplorer (EUVE) all-sky survey catalog. All Galactic and extragalactic objects were also detected in the ROSAT Position Sensitive Proportional Counter survey, and most are at the faint limit of the EUVE detectors. These new identifications substantially increase the total number of EUV-selected extragalactic sources.

We numerically investigate both chemical and dynamical evolution of gas-rich galaxy mergers in a self-consistent manner in order to explore the origin of the color-magnitude (C-M) relation of elliptical galaxies. We especially investigate the dependence of the mean stellar metallicity of merger remnants on the rapidity of gas consumption by star formation during merging. We found that galaxy mergers with more (less) rapid gas consumption by star formation show a larger (smaller) degree of metal enrichment. The reason for this dependence can be explained as follows. For a galaxy merger with more rapid star formation, a lesser amount of interstellar gas is tidally stripped away from the system during galaxy merging principally because a greater amount of initial gas has been already converted to a stellar component before the system suffers more severely from the violently varying gravitational potential of galaxy merging. As a result of this, a greater amount of the gas is consequently enriched to a larger extent with metals and is converted to a stellar component during merging. Thus, greater amounts of metals are shared by the stellar component in the remnant of the galaxy merger with more rapid star formation. This result implies that if more luminous elliptical galaxies are formed by galaxy merging with more rapid star formation, the C-M relation can be reproduced at least in a qualitative way. This result furthermore demonstrates that the origin of the C-M relation of elliptical galaxies can be closely associated with the details of the dynamical process of galaxy merging, which depends principally on the rapidity of gas consumption by star formation in merging galaxies.

We present models of the late stages of stellar evolution intended to explain the "UV upturn" phenomenon in elliptical galaxies. Such models are sensitive to values of a number of poorly constrained physical parameters, including metallicity, age, stellar mass loss, helium enrichment, and the distribution of stars on the zero-age horizontal branch (HB). We explore the sensitivity of the results to values of these parameters and reach the following conclusions.

Old, metal-rich galaxies, such as giant ellipticals, naturally develop a UV upturn within a reasonable timescale—less than a Hubble time—without the presence of young stars. The most likely stars to dominate the UV flux of such populations are low-mass, core helium-burning (HB and evolved HB) stars. Metal-poor populations produce a higher ratio of UV-to-V flux, owing to opacity effects, but only metal-rich stars develop a UV upturn, in which the flux increases toward shorter UV wavelengths.

We present the results from a study of the dynamics of the system of globular clusters around M87. After eliminating foreground galactic stars and background galaxies, we end up with a sample of 205 bona fide M87 globular clusters for which we have radial velocities determined from multislit spectra taken with the Low Resolution Imaging Spectrograph on the Keck Telescope. We find that the mean radial velocity of the M87 globular clusters agrees well with that of M87 itself and that the velocity histogram is well represented by a Gaussian distribution. We find evidence for rotation in the globular cluster system. We find that the observed velocity dispersion of the M87 globular cluster system increases with radius from 270 km s^-1 at r = 9 kpc to ≈ 400 km s^-1 at r = 40 kpc. The inferred mass-to-light ratio in solar units increases from 5 at r = 9 kpc to ≈ 30 at r = 40 kpc with M(r) ~ r^1.7. The long-slit optical spectroscopy near the center of M87 and the recent analysis of the ROSAT X-ray data are in good agreement with this analysis near the nucleus and in the outer parts of M87, respectively.

The close correlation between cooling flows and emission-line nebulae in clusters of galaxies has been recognized for over a decade and a half, but the physical reason for this connection remains unclear. Here we present deep optical spectra of the nebula in Abell 2597, one of the nearest strong cooling-flow clusters. These spectra reveal the density, temperature, and metal abundances of the line-emitting gas. The abundances are roughly half-solar, and dust produces an extinction of at least a magnitude in V. The absence of [O III] λ4363 emission rules out shocks as a major ionizing mechanism, and the weakness of He II λ4686 rules out a hard ionizing source, such as an active galactic nucleus or cooling intracluster gas. Hot stars are therefore the best candidate for producing the ionization. However, even the hottest O stars cannot power a nebula as hot as the one we see. Some other nonionizing source of heat appears to contribute a comparable amount of power. We show that the energy flux from a confining medium can become important when the ionization level of a nebula drops to the low levels seen in cooling-flow nebulae. We suggest that this kind of phenomenon, in which energy fluxes from the surrounding medium augment photoelectric heating, might be the common feature underlying the diverse group of objects classified as LINERS.

We describe the detection of neutral gas within the central galaxy of the Centaurus cluster of galaxies, NGC 4696. The detection is in the form of sodium D absorption, against the background stellar emission. The gas is clearly associated with the compact emission filament system and dust lane in the central 20'' of the galaxy; it shares the same velocity field as the emission-line gas, the absorption line strength correlates with the amount of extinction due to dust, and the equivalent width of narrow absorption features is related to the emission-line strength. The quantitative relationship between absorption line strength and extinction is compared with Galactic ratios. The strength of sodium absorption is large for the typical E(B-V) observed, although similar to what is found in external spiral galaxies. Several possibilities may account for this, including larger internal random motions than are present in the Galactic ISM, higher gas-to-dust ratio than Galactic, or a higher neutral fraction of sodium in NGC 4696. The observations imply a common origin for the warm and neutral components, which may be a merger.

We have obtained optical spectrophotometry of 11 H II regions in the polar ring of NGC 2685 (the Helix galaxy) and have used these data to study the physical characteristics of the polar-ring H II regions. The H II regions have normal spectra with no suggestion of unusual density, temperature, extinction, or composition. Semi-empirical calculations yield high oxygen abundance estimates (0.8-1.1 Z_☉) in all H II regions. This, along with the observed (B-V) color, Hα equivalent width, and molecular gas properties argue against the current picture in which polar rings form from tidally captured dwarf irregular galaxies and suggests instead that the rings are long-lived, self-gravitating structures as predicted by some dynamical models. This would allow the time required for multiple generations of star formation and for the retention of the resulting enriched ejecta for inclusion in further generations of star formation.

We calculate the flux and spectrum of γ-rays emitted by a two-temperature advection-dominated accretion flow (ADAF) around a black hole. The γ-rays are from the decay of neutral pions produced through proton-proton collisions. We discuss both thermal and power-law distributions of proton energies and show that the γ-ray spectra in the two cases are very different. We apply the calculations to the γ-ray source, 2EG J1746-2852, detected by EGRET from the direction of the Galactic center. We show that the flux and spectrum of this source are consistent with emission from an ADAF around the supermassive accreting black hole Sgr A* if the proton distribution is a power law. The model uses accretion parameters within the range made likely by other considerations. If this model is correct, it provides evidence for the presence of a two-temperature plasma in Sgr A* and predicts γ-ray fluxes from other accreting black holes that could be observed with more sensitive detectors.

Physical conditions of molecular gas in the first quadrant of the Galaxy are examined through comparison of the CO J = 2-1 data of the Tokyo-Nobeyama Radio Observatory survey with the CO J = 1-0 data of the Columbia survey. A gradient of the CO J = 2-1/J = 1-0 intensity ratio (≡ R_2-1/1-0) with Galactocentric distance is reported. The ratio varies from ≃0.75 at 4 kpc to ≃0.6 at 8 kpc in Galactocentric distance. This confirms the early in-plane results reported by Handa et al.

We classify molecular gas into three categories in terms of R_2-1/1-0 on the basis of a large velocity gradient model calculation. Very high ratio gas (VHRG; R_2-1/1-0 > 1.0) is either dense, warm, and optically thin gas or externally heated, dense gas. High ratio gas (HRG; R_2-1/1-0 = 0.7-1.0) is warm and dense gas with high-excitation temperature of the J = 2-1 transition (T_ex ≳ 10 K), and it is often observed in central regions of giant molecular clouds. Low ratio gas (LRG; R_2-1/1-0 < 0.7) has low-excitation temperature of the J = 2-1 transition (T_ex ≲ 10 K) because of low density or low kinetic temperature, or both, and is often observed in dark clouds and outer envelopes of giant molecular clouds. It is shown that the CO J = 2-1 emission is better characterized as a tracer of dense gas rather than a tracer of warm gas for molecular gas with kinetic temperature higher than 10 K. The observed large-scale decrease in R_2-1/1-0 as a function of Galactocentric distance is ascribed to the fractional decrease of HRG and VHRG from ≃40% near 5 kpc to ≃20% near the solar circle.

The HRG and VHRG are found predominantly along the Sagittarius and Scutum arms, probably in their downstream. This fact and the deficiency of atomic gas compared with molecular gas in the inner Galaxy indicate that physical conditions of interstellar gas are affected by grand-design, nonlinear processes, such as compression by spiral density waves followed by gravitational collapse, and not by dissociation of low-density molecular gas by young stars.

We construct a two-parameter family of models for self-collimated, magnetized outflows from accretion disks. As in previous magnetocentrifugal wind solutions, a flow at zero initial poloidal speed leaves the surface of a disk in Kepler rotation about a central star, and it is accelerated and redirected toward the pole by rotating, helical magnetic fields that thread the disk. At large distances from the disk, the flow streamlines asymptote to wrap around the surfaces of nested cylinders, with velocity v and magnetic field B directed in the axial () and toroidal () directions. In the asymptotic regime, the velocity secularly decreases with cylindrical radius R from the inside to the outside of the flow because successive streamlines originate in the circumstellar disk in successively shallower portions of the stellar potential. In contrast to previous disk wind modeling, we have explicitly implemented the cylindrical asymptotic boundary condition to examine the consequences for flow dynamics. The present solutions are developed within the context of r-self-similar flows, such that v, the density ρ, and B scale with spherical radius r as v ∝ r^-1/2, ρ ∝ r^-q, and B ∝ r^-(1+q)/2; q must be smaller than unity in order to achieve cylindrical collimation. We self-consistently obtain the shapes of magnetic field lines and the θ-dependence of all flow quantities. The solutions are characterized by q together with the ratios R_A/R₁ and R₀/R₁, where for a given streamline, R₀ is the radius of its footpoint in the disk, R_A is the cylindrical radius where the flow makes an Alfvén transition, and R₁ is its final asymptotic cylindrical radius. For given q and R₀/R₁, R_A/R₁ must be found as an eigenvalue such that the Alfvén transition is made smoothly. In the solutions we have found, the asymptotic poloidal speed v_z on any streamline is typically just a few tenths of the Kepler speed ΩR₀ at the corresponding disk footpoint, while the asymptotic rotation speed v_φ may be a few tenths to several tenths of ΩR₀. The asymptotic toroidal Alfvén speed v_A,φ = B_φ/(4πρ)^1/2 is, however, a few times ΩR₀; thus the outflows remain magnetically dominated, never making a fast-MHD transition. We discuss the implications of these models for interpretations of observed optical jets and molecular outflows from young stellar systems, and we suggest that the difficulty of achieving strong collimation in vector velocity simultaneously with a final speed comparable to ΩR₀ argues against isolated jets and in favor of models with broader winds.

Magnetized gas layers in gravitational fields (e.g., accretion disks and galactic disks) can be subject to the Parker instability, an undular mode of the magnetic buoyancy instabilities. By means of a linear stability analysis, we examined the effects of hot, tenuous regions ("coronae") on the growth of the Parker instability in the underlying magnetized gas layers. As an unperturbed state, we consider the magnetized gas layers in static equilibrium. The stratified gas layers are threaded by horizontal magnetic fields in the x-direction. The temperature varies almost discontinuously at the coronal base in the z-direction. The ratio of magnetic pressure to gas pressure, α, is assumed to be constant. Our analysis has confirmed that the presence of a corona reduces the growth rate of the Parker instability and increases the critical wavelength. It is found that the growth of the Parker instability is more sensitive to the height of the coronal base than the temperature ratio between the disk and the corona is. In particular, the Parker instability is stabilized substantially when the coronal base lies below the height of maximum gravitational acceleration. When the wavenumber vector of the perturbation is parallel to the magnetic field (k_y = 0), the growth rates of all modes in the disk are reduced considerably in the limit of the vanishing coronal base height. The first harmonic mode (1h-mode with odd symmetric velocity eigenfunctions with respect to the equatorial plane) is more easily stabilized by coronae than the fundamental mode is (f-mode with even symmetric velocity eigenfunctions). This is because global convective motion across the equatorial plane is allowed for the f-mode even when k_y = 0, whereas it is not allowed for the 1h-mode. For the f-mode, furthermore, we find that the smallest possible γ (critical gamma) against the instability is γ_crit = 1 + α, regardless of the value of k_y. The reason for this is discussed briefly.

The chemistry of developing and collapsing low-mass protostellar cores is followed using a chemical code with a time-varying density. Two evolutionary scenarios are represented, gravitational collapse in the presence of magnetic fields and the slow core growth by accretion near equilibrium. The chemical code includes gas-phase reactions and depletion onto grains with both CO and H₂O ice mantles. We find that various species will selectively deplete from the gas phase at times that correspond to the middle to late stages of dynamical evolution when the densities are highest. These depletions do not depend in detail on the dynamical solution and should exist for any centrally condensed density profile. Sulfur-bearing molecules are particularly sensitive to the density increase: CS, SO, and C₂S show significant depletions both on a strongly bound water mantle and on the weakly bound CO-covered grain surface. In contrast, CO and HCO⁺ show large depletions only for an H₂O grain mantle and remain in the gas phase for models with CO grain mantles. Two species, NH₃ and N₂H⁺, do not deplete from the gas phase for the densities considered in our models. We also find that for very high densities, n_H2>10⁶ cm^-3, depletion becomes important for all molecules. The effects of coupling chemistry and dynamics on the resulting physical evolution are discussed. We compare our results with current high-resolution observations of preprotostellar cores and to more evolved objects and suggest that ratios of the abundances of few species can be used in concert with our models as sensitive discriminators between different stages of core and star formation.

Three hydrogen recombination lines—H90α at 8.9 GHz, H92α at 8.3 GHz, and H109α at 5.0 GHz—have been observed with the Australia Telescope Compact Array toward the 30 Doradus Nebula, the giant H II region in the Large Magellanic Cloud. In this paper, emphasis is placed on the more sensitive H90α observations, which also include the He90α and H113β lines. A spatial resolution of 15'' was obtained with a velocity resolution of 4.2 km s^-1 and an rms noise of 1.6 mJy. Typical line-to-continuum ratios were 0.04 for the H90α and H92α, and 0.03 for the H109α data. The distributions of radial velocity and electron temperature have been determined. The heliocentric velocity of the ionized gas varies between 245 and 290 km s^-1 with a mean value of 262 km s^-1. In some places the line profile is double, suggesting that sections of expanding shells are being observed. The mean (LTE) electron temperature is close to 7700 K over most of the nebula. The intensity ratio of H113β/H90α is close to the expected LTE value of 0.28, implying that the calculated LTE electron temperatures are close to the true electron temperatures; in one area to the north the ratio is enhanced, 0.39 ± 0.03. The helium abundance (Y⁺ = He⁺/H⁺) shows no significant variation across the source, with a mean value of 0.13 ± 0.02.

We present the results of a new set of medium resolution spectroscopic observations and high-resolution coronographic images of the nebula around the Galactic luminous blue variable (LBV) HR Carinae. The observations were carried out at the ESO/NTT (La Silla) in 1995 May and 1996 January. The nebular morphology and kinematics confirm that the nebula around HR Carinae is truly bipolar and very reminiscent of the η Carinae nebula. The previously identified "filaments" outline the edges of two symmetrical expanding bubbles, originating from the star and located, respectively, in the NW and SE quadrants. The small compact inner nebula, a few arcsec in size, previously detected, represents the "waist" of the bipolar distribution. The orientation in the images and the kinematical study have allowed us to define the true orientation of the bubbles, whose major axis lies at an angle of ≃50° with the plane of the sky, at an inclination of approximately 30° on the line of sight. The maximum projected expansion velocity is of the order of ≃100 km s^-1. In the light of these new kinematical data, we revise the dynamical timescale to a younger age of ≃5000 yr.

The nebula around HR Carinae is relatively young and fast, at variance with other well known LBV nebulae such as AG Carinae's. Spectroscopically, the nebula is of low excitation with [O III] absent, and [N II] fairly strong. [Ni II] λ6667 is detected, but only in the inner regions (≤5''). We find that the electron density increases from 400 cm^-3 in the outer regions to more than 10⁴ cm^-3 in the innermost regions. An analysis of the chemical abundances in different regions of the nebula finds that N is overabundant, indicating that the nebula is composed of CNO processed stellar material. We find that the filamentary H II region, seen to the NW of HR Carinae, is at the same distance and is composed of material with typical H II region abundances, and has a morphology that suggests it has been shaped by the wind of HR Carinae.

We report the detection with the proportional counter array (PCA) onboard the RossiX-RayTimingExplorer (RXTE) of 589 Hz oscillations during three type I X-ray bursts from a low-mass X-ray binary located in the Galactic center region. The bursts, which occurred on 1996 May 15.814, June 4.612, and June 19.414 UT, were observed serendipitously during routine monitoring observations of GRO J1744-28 by RXTE. The strongest pulsations were observed from the May 15 burst, reaching an amplitude of 18% of the >8 keV flux in this burst. The pulsation amplitudes in each burst are a strong function of photon energy since no signals were detected in the 1-8 keV band, but there were strong detections above 8 keV. The evolution of the X-ray spectrum through each burst is consistent with modest radius expansion. The pulsations are confined to the postcontraction portion of each burst profile, consistent with the oscillations recently reported in a burst from KS 1731-26 and also some bursts from 4U 1728-34. The location of the burster in the 1° (FWHM) PCA field of view is constrained by comparing the count rates in the five independent detectors of the PCA. This analysis strongly excludes GRO J1744-28 as the source of these bursts but does not yet allow a definitive identification of the source with any known burster in the field of view. However, the derived position strongly overlaps that of MXB 1743-29, identifying this source as a strong candidate for the 589 Hz burster. We argue that the observed oscillations are consistent with rotational modulation of the X-ray brightness.

High Li abundances have been reported in the late-type secondaries of five soft X-ray transients (SXTs): V404 Cyg, A0620-00, GS 2000+25, Nova Mus 1991, and Cen X-4. Since Li is likely to be depleted in stars of this type, the origin of the Li is puzzling. Li has not been seen in similar secondaries of cataclysmic variables, which suggests that the high Li abundance is not due to an anomalous suppression of Li depletion in close binaries. SXTs in the quiescent state have been modeled in terms of hot advection-dominated accretion flows (ADAFs), in which the ions are essentially at virial temperature. At such temperatures, Li production via α-α spallation is possible. We show that ADAFs can produce sufficient Li via spallation to explain the observations in the black hole SXTs V404 Cyg, A0620-00, GS 2000+25, and Nova Mus 1991. If this explanation is correct, it is the first independent confirmation that accretion flows in quiescent SXTs have a two-temperature plasma with ion temperature much higher than electron temperature. Depending on the Li depletion timescale in the secondary, which may range between 10⁷-10⁹ yr, the model requires ~10^-4 to 10^-6 of the accreted mass to be intercepted by the secondary after undergoing Li production and being ejected. In the case of the neutron star SXT, Cen X-4, we can explain the observed Li only if the mass-accretion rate is ~10^-3 times the Eddington rate and if there is enhanced ejection due to a propeller effect. We discuss possible observational tests of this proposal. Li production during outbursts could be quite important and may even dominate over the production during quiescence, but the estimate of the Li yield is uncertain. We calculate the expected luminosity in gamma-ray lines due to the production of excited Li and Be nuclei, but conclude that the line cannot be detected with current instruments.

Vertical structure models are used to investigate the structure of protostellar, α-law, accretion disks. Conditions investigated cover a range of mass fluxes (10^-9 to 10^-5M_☉ yr^-1), viscous efficiencies (α = 10^-2 and 10^-4), and stellar masses (0.5-3 M_☉). Analytic formulae for midplane temperatures, optical depths, and volume and surface densities are derived and are shown to agree well with numerical results. The temperature dependence of the opacity is shown to be the crucial factor in determining radial trends. We also consider the effect on disk structure of illumination from a uniform field of radiation such as might be expected of a system immersed in a molecular cloud core or other star-forming environment: T_amb = 10, 20, and 100 K. Model results are compared to HubbleSpaceTelescope observations of HH30 and the Orion proplyds.

Disk shape is derived in both the Rosseland mean approximation and as viewed at particular wavelengths (λλ = 0.66, 2.2, 60, 100, 350, and 1000 μm). In regions where the opacity is an increasing function of temperature (as in the molecular regions where κ ∝ T²), the disk does not flare, but decreases in relative thickness with radius under both Rosseland mean and single wavelength approximations. The radius at which the disk becomes shadowed from central object illumination depends on radial mass flow and varies from a few tenths to about 5 au over the range of mass fluxes tested. This suggests that most planet formation occurred in environments unheated by stellar radiation. Viewing the system at any single wavelength increases the apparent flaring of the disk but leaves the shadow radius essentially unchanged. External heating further enhances flaring at large radii, but, except under extreme illumination (100 K), the inner disk will shield the planet-forming regions of all but the lowest mass flux disks from radiation originating near the origin such as from the star or from an FU Orionis outburst.

A new model is presented to elucidate the optical-ultraviolet "flickering" variability in cataclysmic variables. In this model, light fluctuations are produced by occasional flarelike events and subsequent avalanche flow in the accretion disk atmospheres. Flares are assumed to be ignited when the mass density exceeds a critical density. The disk then evolves to and stays at a self-organized critical state, in which seemingly chaotic fluctuations can be produced. We calculated fluctuating light curves by cellular-automaton simulations. The resultant power spectral density (PSD) consists of a flat part [S(f) ∝ f⁰, where f is frequency] at lower frequencies and a steep-decline part [S(f) ∝ f^-β, with β ≃ 1-2] at higher frequencies. It is important that the PSD is rather insensitive to wavelength.

Assuming a diamagnetic interaction between a stellar spot-originated localized magnetic field and gas blobs in the accretion disk around a T Tauri star, we show the possibility of ejection of such blobs out of the disk plane. Choosing the interaction radius and the magnetic field parameters in a suitable way gives rise to closed orbits for the ejected blobs. A stream of matter composed of such blobs, ejected on one side of the disk and impacting on the other, can form a hot spot at a fixed position on the disk (in the frame rotating with the star). Such a hot spot, spread somewhat by disk shear before cooling, may be responsible in some cases for the light-curve variations observed in various T Tauri stars over the years. An eclipse-based mechanism due to stellar obscuration of the spot is proposed. With the assumption of high disk inclination angles, it is able to explain many of the puzzling properties of these variations. By varying the field parameters and blob initial conditions, we obtain variations in the apparent angular velocity of the hot spot, producing a constantly changing period or intermittent periodicity disappearance in the models.

Turbulent viscosity is believed to circularize and synchronize binary orbits and to damp stellar oscillations. Older binaries appear to be circular out to longer periods, which indicates that circularization proceeds during the main-sequence phase. It is also believed that when the tidal period is shorter than the turnover time of the largest eddies, turbulent viscosity is partially suppressed. The degree of suppression, however, is disputed. We reexamine both of these beliefs via (1) direct perturbative calculations, linearizing the fluid equations on a turbulent background; and (2) numerical integration of a chaotic dynamical system subject to periodic forcing. We find that dissipation of rapid tides is severely suppressed. Even when we disregard this suppression, we conclude that some mechanism other than turbulent convection circularizes solar-type binaries.

White dwarfs are the remnants of stars of low and intermediate masses on the main sequence. Since they have exhausted all of their nuclear fuel, their evolution is just a gravothermal process. The release of energy only depends on the detailed internal structure and chemical composition and on the properties of the envelope equation of state and opacity; its consequences on the cooling curve (i.e., the luminosity vs. time relationship) depend on the luminosity at which this energy is released.

The internal chemical profile depends on the rate of the ¹²C(α, γ)¹⁶O reaction as well as on the treatment of convection. High reaction rates produce white dwarfs with oxygen-rich cores surrounded by carbon-rich mantles. This reduces the available gravothermal energy and decreases the lifetime of white dwarfs.

In this paper we compute detailed evolutionary models providing chemical profiles for white dwarfs having progenitors in the mass range from 1.0 to 7 M_☉, and we examine the influence of such profiles in the cooling process. The influence of the process of separation of carbon and oxygen during crystallization is decreased as a consequence of the initial stratification, but it is still important and cannot be neglected. As an example, the best fit to the luminosity functions of Liebert et al. and Oswalt et al. gives an age of the disk of 9.3 and 11.0 Gyr, respectively, when this effect is taken into account, and only 8.3 and 10.0 Gyr when it is neglected.

We present K-band spectra (R ~ 525) of 38 northern Galactic WR Stars, of which 16 are WC, 19 are WN, two are WN/WC, and one is WO. The spectra have the expected trend of stronger lines for higher ionization species with earlier spectral subtype. Spectra for the late WC stars can appear to have weak emission lines, an effect due to different amounts of dust dilution in the individual stars. There are also differences in spectral morphology for stars within other subtypes. In general, the spectra for all WC stars earlier than WC9 tend to be quite similar, while the spectra for WN subtypes are more easily differentiated. Several previously unidentified emission lines are seen in the spectra, most notably, features near 2.247 and 2.368 μm in late-type WN stars, and one near 2.222 μm in late-type WC stars. We attribute the 2.247 μm line to a N III transition and argue that it might provide the best method for discriminating between WNL and OIf⁺ stars in the K band.

We investigate the behavior of the 2.11 μm (He I + N III), 2.166 μm (H I + He I + He II), and 2.189 μm (He II) emission lines in WN types and find that these lines provide for accurate discrimination within the sample to within 1 subtype. From this investigation, it appears that the ratio of W_{2.189 μm}/W_{2.11 μm} is sensitive to subtype and shows the least dispersion within subtypes. In addition, we find that the W_{2.189 μm}/W_{2.166 μm} ratio also scales with subtype in a well-behaved manner once it is corrected for contamination of the 2.166 μm line by He II 14-8 and for the presence of an O star companion in binary systems.

We also investigate the behavior of the 2.058 μm (He I), 2.08 μm (C IV), and the 2.11 μm (C III + He I) emission lines in WC types. The ratio of W_{2.08 μm}/W_{2.11 μm} correlates with subtype; however, it is not easy to distinguish between individual subtypes earlier than WC8 by just using this quantity. The dominance of the 2.058 μm line in WC9 types distinguishes this subtype from all other WC subtypes. Two WC9 stars in our sample have nearly featureless spectra due to dust dilution.

It is possible to classify a WR star to within one subtype in the WN sequence based upon the sample in this atlas. The similarity of WC spectra makes it difficult to distinguish among individual subtypes earlier than WC8.

GRO J1744-28 is the first known X-ray source to display bursts, periodic pulsations, and quasi-periodic oscillations. This source may thus provide crucial clues that will lead to an understanding of the differences in the nature of the X-ray variability from various accreting neutron stars. The orbital period is 11.8 days, and the measured mass function of 1.31 × 10^-4M_☉ is one of the smallest among all known binaries. If we assume that the donor star is a low-mass giant transferring matter through the inner Lagrange point, then we can show that its mass is lower than ~0.7 M_☉ and probably closer to 0.25 M_☉. Higher mass, but unevolved, donor stars are shown to be implausible. We also demonstrate that the current He core mass of the donor star lies in the range of 0.20-0.25 M_☉. Thus, this system is most likely in the final stages of losing its hydrogen-rich envelope, with only a small amount of mass remaining in the envelope. If this picture is correct, then GRO J1744-28 may well represent the closest observational link that we have between the low-mass X-ray binaries and recycled binary pulsars in wide orbits.

We have carried out a series of binary evolution calculations and explored, both systematically and via a novel Monte Carlo approach, the range of initial system parameters and input physics that can lead to the binary parameters of the present-day GRO J1744-28 system. The input parameters include both the initial total mass and the core mass of the donor star, the neutron-star mass, the strength of the magnetic braking, the mass-capture fraction, and the specifics of the core mass/radius relation for giants. Through these evolution calculations, we compute probability distributions for the current binary system parameters (i.e., the total mass, core mass, radius, luminosity, and K-band magnitude of the donor star, the neutron star mass, the orbital inclination angle, and the semimajor axis of the binary). Our calculations yield the following values for the GRO J1744-28 system parameters (with 95% confidence limits in parentheses): donor star mass: 0.24 M_☉ (0.2-0.7 M_☉); He core mass of the donor star: 0.22 M_☉ (0.20-0.25 M_☉); neutron-star mass: 1.7 M_☉ (1.39-1.96 M_☉); orbital inclination angle: 18° (7°-22°); semimajor axis: 64 lt-s (60-67 lt-s); radius of the donor star: 6.2 R_☉ (6-9 R_☉); luminosity of donor star: 23 L_☉ (15-49 L_☉); and long-term mass transfer rate at the current epoch: 5 × 10^-10M_☉ yr^-1 (2 × 10^-10 to 5 × 10^-9M_☉ yr^-1).

We deduce that the magnetic field of the underlying neutron star lies in the range of ~1.8 × 10¹¹ G to ~7 × 10¹¹ G, with a most probable value of 2.7 × 10¹¹ G. This is evidently sufficiently strong to funnel the accretion flow onto the magnetic polar caps and suppress the thermonuclear flashes that would otherwise give rise to the type I X-ray bursts observed in most X-ray bursters. We present a simple paradigm for magnetic accreting neutron stars wherein X-ray pulsars, GRO J1744-28, the Rapid Burster, and the type I X-ray bursters may form a continuum of possible behaviors among accreting neutron stars, with the strength of the neutron-star magnetic field serving as the crucial parameter that determines the mode of X-ray variability from a given object.

We present the results of an effort to model recent HST eclipse observations of the C IV 1550 Å resonance line in the high-inclination nova-like cataclysmic variable UX UMa. Assuming this line to be predominantly wind-formed, we are able to roughly reproduce the shapes and strengths of the observed line profiles (both away from and during eclipse) within the framework of a simple kinematic model in which the outflow is described as a rotating, biconical accretion disk wind. Our preferred model also matches most of the detailed behavior of different parts of the C IV line profile (blue wing, line center, and red wing) as a function of orbital phase.

The most important result of our modeling is that it strongly suggests the presence of a high column (N_H ~ 10²³ cm^-2), relatively dense (n_e ≃ 4 × 10¹² cm^-3), and only slowly outflowing (v_poloidal ≪ v_escape) transition region between the disk photosphere and the fast-moving wind. We also find that the outflow from UX UMa is probably only moderately collimated (the outer opening angle of our preferred model is θ_max ≃ 65°). Finally, the rotational disturbance seen in the data is fitted reasonably well by our model, in which the rotational component of motion in the wind is fixed by conserving the specific angular momentum of the outflowing material. The implications of these results for dynamical models of disk-driven winds are discussed.

Ultraviolet spectra obtained with the HubbleSpaceTelescope of the carbon star TX Piscium (HR 9004) are presented, along with analysis providing information on its outer atmosphere, including flow and turbulent velocities, line formation mechanisms, and variations with time. Both thermal (collisionally excited) and fluorescent emission from the chromosphere of the star appear to be formed near the stellar rest velocity, i.e., in a region below that in which the stellar wind is accelerated. Absorption self-reversals in the Mg II emission confirm the presence of an outflowing stellar wind at a mean velocity of about 9-10 km s^-1. Circumstellar absorption features (Mn I and Fe I) overlying the Mg II emission indicate a cool shell expanding at about 5-6 km s^-1 relative to the photosphere. The widths (FWHM) of various emission lines indicate that the chromospheric turbulence is at least 16 km s^-1, but that it may increase with altitude to as much as 34 km s^-1. Three hours of integration on the C II] lines are examined for any signs of variability that might indicate the presence of shocks, but no statistically significant variations are seen. A previous identification (in spectra of UU Aur) of an emission line at 2807 Å, seen only in spectra of carbon stars, as belonging to Fe I multiplet UV45 pumped by the C II] line at 2325 Å is confirmed by the discovery of an absorption feature corresponding exactly to the wavelength of the pumped transition (Fe I UV13) near 2325 Å. Lines from Fe II UV165, previously seen in solar off-limb spectra and in Goddard High-Resolution Spectrograph spectra of α Tau, are clearly present. The normally much stronger Fe II UV32, 62, and 63 multiplets are seen but are weaker relative to both the UV165 lines and the intercombination lines of C II] and Si II] than in α Tau. The weakness of these Fe II lines is indicated both by their absolute flux levels and by their narrow, single-peaked profiles, which are in sharp contrast to the broad, double-peaked profiles seen in oxygen-rich cool giant and supergiant stars. The weakness of the Fe II lines and the presence of the Fe I 2807 Å line suggest that the ionization fraction of iron (Fe II/Fe I) is significantly lower in the outer atmospheres of carbon stars. Fluxes in emission lines of Fe II and Mg II are ≥2-3 times lower than in a 1984 IUE spectrum of TX Psc, confirming that the latter was obtained at an epoch of unusual UV brightness for the star. The Mg II profiles are heavily mutilated by overlying absorption, even more so than in 1984. The TX Psc profiles are very similar to those seen in the carbon star TW Hor but are dramatically different than those in another carbon star, UU Aur, whose lines show violet wing emission out to much shorter wavelengths than in the other two stars.

Because of the apparent difficulty in accelerating previously unaccelerated pickup ions locally at the solar wind termination shock, a model for anomalous cosmic-ray protons is considered in which the initial acceleration of pickup ions to ~10-20 MeV occurs in the inner heliosphere. These accelerated pickup ions are assumed to be accelerated either by propagating interplanetary shocks or by some other, unspecified process. Voyager2 Low-Energy Charged Particle (LECP) observations at low energies are used to normalize their spectra. The ions are then transported to the termination shock, where they are further accelerated to anomalous cosmic-ray energies. We use a well-established transport model to simulate this process. In this model, the two-dimensional cosmic-ray transport equation is solved using the assumed (and normalized) source of energetic particles (~0.01-20 MeV) located at 10 AU. We show that the computed spectra at higher energies (≳100 MeV) are consistent with observed anomalous cosmic-ray fluxes and suggest that interplanetary shocks, especially those in the inner heliosphere, may play an important role in the initial acceleration of pickup ions leading to anomalous cosmic rays.

A mechanism for small-scale generation in magnetically dominated, homogeneous space plasmas is proposed, resulting from transverse collapse (filamentation) of weakly nonlinear, circularly polarized Alfvén waves. It is argued that waves whose frequencies exceed a magnitude of about 10 Hz can significantly contribute to the heating of open coronal holes and to the acceleration of the fast solar wind. In coronal loops, this frequency lower bound reduces to 1 Hz. The importance of Alfvén wave filamentation is also demonstrated in the warm ionized phase of the interstellar medium.

We describe a simple, kinematic model for a dynamo operating in the vicinity of the interface between the convective and radiative portions of the solar interior. The model dynamo resides within a Cartesian domain, partioned into an upper, convective half and lower, radiative half, with the magnetic diffusivity η of the former region (η₂) assumed to exceed that of the latter (η₁). The fluid motions that constitute the α-effect are confined to a thin, horizontal layer located entirely within the convective half of the domain; the vertical shear is nonzero only within a second, nonoverlapping layer contained inside the radiative half of the domain. We derive and solve a dispersion relation that describes horizontally propagating dynamo waves. For sufficiently large values of a parameter analogous to the dynamo number of conventional models, growing modes can be found for any ratio of the upper and lower magnetic diffusivities. However, unlike kinematic models in which the shear and α-effect are uniformly distributed throughout the same volume, the present model has wavelike solutions that grow in time only for a finite range of horizontal wavenumbers.

An additional consequence of the assumed dynamo spatial structure is that the strength of the azimuthal magnetic field at the location of the α-effect layer is reduced relative to the azimuthal field strength at the shear layer. When the jump in η occurs close to the α-effect layer, it is found that over one period of the dynamo's operation, the ratio of the maximum strengths of the azimuthal fields at these two positions can vary as the ratio (η₁/η₂) of the magnetic diffusivities.

Numerical models of interfacedynamos are constructed, and their properties discussed in some detail. These models are extensions in spherical geometry of the Cartesian interface models considered by Parker and in the first paper of this series. The models are cast in the framework of classical mean-field electrodynamics and make use of a realistic solar-like internal differential rotation profile. The magnetic diffusivity is assumed to vary discontinously by orders of magnitude across the core-envelope interface. This allows the buildup of very strong toroidal magnetic fields below the interface, as apparently required by recent models of erupting bipolar magnetic regions.

Distinct dynamo modes powered either by the latitudinal or radial shear can coexist and, under certain conditions, interfere destructively with one another. Hybridmodes, relying on the latitudinal shear both in the envelope and below it, are most easily excited in some portions of parameter space, and represent a class of dynamo solutions distinct from the true interface modes previously investigated in Cartesian geometry. Which mode is preferentially excited depends primarily on the assumed ratio of magnetic diffusivities on either side of the core-envelope interface. For an α-effect having a simple cos θ latitudinal dependency, the interface mode associated with the radial shear below the polar regions of the interface is easier to excite than its equatorial counterpart. In analogy with more conventional dynamo models, interface modes propagate equatorward if the product of the radial shear (∂Ω/∂r) and α-effect coefficient (C_α) is negative, and poleward if that product is positive.

Interface dynamo modes powered by the positive radial shear localized below the core-envelope interface in the equatorial regions can be produced by artificially restricting the α-effect to low latitudes. For negative dynamo number, those modes are globally dipolar, propagate toward the equator, and are characterized by a phase relationship between poloidal and toroidal magnetic field components that is in agreement with observations.

While the models discussed in this paper are linear and kinematic, and consequently rather limited in their predictive power, results obtained so far certainly suggest that interface dynamos represent a very attractive alternative to conventional solar mean-field dynamo models.

A model for solar flares is proposed in which the flare energy is the magnetic energy released when two current-carrying flux loops reconnect to form two new current-carrying flux loops between the original four footpoints. It is assumed that a magnetic flux, ΔΨ, and an electric current, ΔI, are transferred during the reconnection, subject to the constraint that the flux and the current at each footpoint are unchanged. The change in the magnetic energy is separated into parts due to the transfer of current and of "potential" field, and the latter is neglected (although the argument for doing so is a weak one). The current transferred, ΔI, is assumed equal to its maximum possible value, which is the weaker of the two currents in the initial flux loops. The change in magnetic energy is calculated for some specific simple configurations of the spots (footpoints of the loops). Some favorable configurations for energy release are when a positive polarity spot is near a negative polarity spot (so that one of the final loops is very small) and when the two initial loops are at a large angle to each other. The model allows reconnection only between loops with the same handedness (the sign of the helicity ∝I/Ψ) and the handedness is preserved during reconnection. Observational evidence that there is a preferred handedness for coronal magnetic structures suggests that this requirement of the model is automatically satisfied, but specific data on the helicity in active regions is less conclusive. It is energetically favorable for overlapping colinear loops to reconnect to form a longer and a shorter (nested) loop, and it is suggested that very long loops connecting different active regions form through a sequence of such reconnections involving ephemeral flux loops that emerge between the active regions.

Few examples of the creation of a filament channel or filament have ever been documented. In a recent paper, Gaizauskas and coworkers observed the early stages of creation of such a channel and then the formation of a filament in it. The filament channel was born when a new activity complex emerged near an old, decaying bipolar active region. The filament itself then formed after convergence of flux in the channel.

In this paper, force-free models are constructed for two phases of the channel's development. For the early days, the models show that the formation of the filament channel seen in Hα is due to the emerging activity complex. The field lines that give the best comparison to the fibril observations are low-lying and have a strong horizontal component. Later, when the activity complex has matured and a filament has formed between it and the adjacent decaying bipolar region, the models give a good representation of the path of the filament in the channel. It is found that the presence of flat or dipped field lines and of converging flux are necessary but not sufficient conditions for filament formation. Furthermore, the magnetic field lines of the filament itself form a narrow, vertical, sheetlike flux-tube corridor that is flat and low-lying. It connects one particular magnetic source to a sink and is bounded by separatrix surfaces that separate the filament from the old remnant region and most of the newly emerged flux.

The Hanle effect concerns the modification of polarized resonance-line scattering by magnetic fields; thus, it can be used as a diagnostic of stellar magnetic fields. The Hanle effect has been used to determine the field strength and distribution of magnetic structures present in prominences of the Sun. To investigate its potential use in stellar astronomy, the simplified case of an optically thin axisymmetric ring illuminated by a stellar point source is considered. The results are then used to derive the polarization from polar plumes, equatorial disks, and spherical shells. The integrated line polarization is calculated for axisymmetric rings with a variety of magnetic field orientations, and in every case the polarization is proportional to sin²i (where i is the viewing inclination), just as in the zero field case. It is also found that the Hanle effect can significantly alter the integrated line polarization. In some cases the position angle of the polarization in the line can be rotated by 90° relative to the zero field case. We consider the Hanle effect as a possible diagnostic of magnetic fields in stellar winds with prominent ultraviolet and visible resonance lines. For these lines the diagnostic has sensitivity in the range of 1-1000 G. The Zeeman effect is not normally applicable for diagnosing magnetic fields in stellar winds in the subkilogauss range; thus, the Hanle effect should provide an especially useful new method of determining magnetic fields in stars other than the Sun. Possibilities for measuring the fields in early-type stars using ultraviolet observations is discussed.

Theoretical O V electron-density-sensitive emission line ratios for R = I(2s²¹S₀ - 2s2p³P₂)/I(2s²¹S₀ - 2s2p³P₁) are presented. Inspection of Goddard High Resolution Spectrograph HubbleSpaceTelescope spectra of RR Tel reveals the presence of the [O V] 2s²¹S - 2s2p³P₂ line at 1213.80 Å, which is 4.62 ± 0.12 Å away from the 2s²¹S - 2s2p³P₂ intercombination transition at 1218.42 Å, in good agreement with the theoretical prediction of Δλ = 4.54 Å. The resultant value of R = 0.82 ± 0.11 implies a logarithmic electron density, log N_e, of 5.2 ± 0.2 cm^-3, in good agreement with that found from other ions with high electron temperature, such as Ne VI, which also provides support for the identification.

Using a combination of both theoretical and experimentally derived potential curves and dipole transition moments, photodissociation cross sections have been calculated for the SiH⁺ molecule. From the v'' = 0 level of the X¹Σ⁺ ground state, with a threshold energy of 3.256 eV, dissociation occurs through the A¹Π state producing Si⁺ (3p²P) and H atoms. For energies above 9.45 eV, photodissociation through the 2¹Π state yields excited neutral Si (3p²¹D) and protons. Excited Si⁺ (3s3p²²D) and H atoms are produced above 10.08 eV by dissociation through 3¹Σ⁺ and 3¹Π. The 2¹Σ⁺ and 4¹Π states are found to be negligible dissociation channels. Photodissociation cross sections as a function of energy and photodissociation rates in the interstellar radiation field for both diffuse and dense molecular clouds are presented. Photodissociation cross sections from LTE distributions of vibrational and rotational levels of the X¹Σ⁺ state to the A¹Π state are also obtained for temperatures between 1000 and 9000 K. Their possible contribution to a missing opacity source in the near-ultraviolet spectra of red giant stars is discussed.

Erratum to ApJ, 424, 983 (1994)

We calculate the anisotropies in the cosmic microwave background induced by long-wavelength primordial gravitational waves in a universe with negative spatial curvature, such as are produced in the "open inflation" scenario. The impact of these results on the COBE normalization of open models is discussed.

We report on Westerbork 840 MHz and 1.4 and 5 GHz radio observations of the improved Interplanetary Network Wide Field Camera (IPN WFC) error box of the γ-ray burst GRB 970111, between 26.4 hr and 120 days after the event onset. In the ~13 arcmin² area defined by the IPN (BATSE and Ulysses) annulus and the published refined BeppoSAX WFC error box, we detected no steady sources brighter than 0.56 mJy (4 σ) and no varying radio emission, down to 1.0 mJy (4 σ). We also report on B-, V-, R-, and I-band observations of the error box with the 4.2 m William Herschel Telescope at La Palma.

The Hubble Deep Field (HDF) is the deepest set of multicolor optical photometric observations ever undertaken, and it offers a valuable data set with which to study galaxy evolution. Combining the optical WFPC2 data with ground-based near-infrared photometry, we derive photometrically estimated redshifts for HDF galaxies with J < 23.5. We demonstrate that incorporating the near-infrared data reduces the uncertainty in the estimated redshifts by approximately 40% and is required to remove systematic uncertainties within the redshift range 1 < z < 2. Utilizing these photometric redshifts, we determine the evolution of the comoving ultraviolet (2800 Å) luminosity density (presumed to be proportional to the global star formation rate) from a redshift of z = 0.5 to z = 2. We find that the global star formation rate increases rapidly with redshift, rising by a factor of 12 from a redshift of zero to a peak at z ≈ 1.5. For redshifts beyond 1.5, it decreases monotonically. Our measures of the star formation rate are consistent with those found by Lilly et al. from the Canada-France Redshift Survey at z < 1 and by Madau et al. from Lyman break galaxies at z > 2, and they bridge the redshift gap between those two samples. The overall star formation or metal enrichment rate history is consistent with the theoretical models of White and Frenk and the predictions of Pei and Fall based on the evolving H I content of Lyα QSO absorption line systems.

We report the results of a survey for H₂O maser emission in the 6₁₆-5₂₃ transition at 1.35 cm wavelength in 29 active galactic nuclei (AGNs). One new maser was detected. The detection rate among objects with recessional velocities <7000 km s^-1 is consistent with rates reported elsewhere for similarly nearby objects (about one in 15). The new maser lies in the edge-on Seyfert 2 galaxy NGC 3735 (inclination 77°) within 10 km s^-1 of the systemic velocity. No other emission has been identified at velocities within ±500 km s^-1 of the systemic velocity. The maser is coincident with the radio continuum peak of the nucleus at 6 and 3.6 cm wavelengths to within the estimated 1 σ astrometric uncertainty of 01 (15 pc at a distance of 30 Mpc).

It has been known that recent microlensing results toward the bulge imply mass densities that are surprisingly high, given dynamical constraints on the Milky Way mass distribution. We derive the maximum optical depth toward the bulge that may be generated by axisymmetric structures in the Milky Way, and we show that observations are close to surpassing these limits. This result argues in favor of a bar as a source of significantly enhanced microlensing. Several of the bar models in the literature are discussed.

We present results from microwave background observations at the Owens Valley Radio Observatory. These observations, at 14.5 and 32 GHz, are designed to detect intrinsic anisotropy on scales of 7'-22'. After point-source removal, we detect significant emission with temperature spectral index β ≃ -2 toward the north celestial pole (NCP). Comparison of our data with the IRAS 100 μm map of the same fields reveals a strong correlation between this emission and the infrared dust emission. From the lack of detectable Hα emission, we conclude that the signals are consistent either with flat-spectrum synchrotron radiation or with free-free emission from T_e ≳ 10⁶ K gas, probably associated with a large H I feature known as the NCP Loop. Assuming β = -2.2, our data indicate a conversion T_f/I_100μm = 7.5 × 10^-2ν K (MJy sr^-1)^-1.

The detection of such a component suggests that we should be cautious in any assumptions made regarding foregrounds when designing experiments to map the microwave background radiation.

We present a study of the gravitational time delay of arrival of signals from binary pulsar systems with rotating black hole companions. In particular, we investigate the strength of this effect (Shapiro delay) as a function of the inclination, eccentricity, and period of the orbit, as well as the mass and angular momentum of the black hole. This study is based on direct numerical integration of null geodesics in a Kerr background geometry. We find that, for binaries with sufficiently high orbital inclinations (>89°) and compact companion masses greater than 10 M_☉, the effect arising from the rotation of the black hole in the system amounts to a microsecond-level variation of the arrival times of the pulsar pulses. If measurable, this variation could provide a unique signature for the presence of a rotating black hole in a binary pulsar system.

A rarity among supernova, SN 1993J in M81 can be studied with high spatial resolution. Its radio power and distance permit VLBI observations to monitor the expansion of its angular structure. This radio structure was previously revealed to be shell-like and to be undergoing a self-similar expansion at a constant rate. From VLBI observations at wavelengths of 3.6 and 6 cm in the period 6-42 months after explosion, we have discovered that the expansion is decelerating. Our measurement of this deceleration yields estimates of the density profiles of the supernova ejecta and circumstellar material in standard supernova explosion models.

The nebular optical spectra of the unusual Type Ia supernova 1991T are modeled by treating important atomic processes with reliable atomic data and by assuming both Chandrasekhar-mass and sub-Chandrasekhar-mass white dwarf explosion models. Similar to the case of normal Type Ia supernovae, a better agreement between the calculated and observed spectra of SN 1991T is obtained by assuming a sub-Chandrasekhar-mass model than Chandrasekhar-mass models, though the required white dwarf mass of ~1.1 M_☉ is higher for SN 1991T than the ~0.9 M_☉ for normal Type Ia supernovae. This demonstrates that variation in the behavior among Type Ia supernovae can be naturally realized by a range of sub-Chandrasekhar-mass models. The optical emission of SN 1991T is well accounted for by the sub-Chandrasekhar-mass model in both brightness and spectral shape from ~7 to 14 months for a distance of 14 Mpc to the supernova.

We report broadband imaging and spectral observations in the energy range 0.8-12 keV with ASCA toward the H II region complex W3. We find a hard X-ray source, with substantial emission at energies greater than 4 keV, coincident with the W3 core region. The SIS images show that the emission in the W3 core is extended in the east-west direction, similar to the distribution of the compact H II regions. Fitting an optically thin, thermal plasma model to the X-ray spectra from the W3 core region, we find a temperature of 7 × 10⁷ K, an absorbing hydrogen column of 2.1 × 10²² cm^-2, and a luminosity of ≈ 10³³ ergs s^-1 in the 2-10 keV band. Weak line emission from the He-like ion of Fe XXV at 6.7 keV is also detected. If the X-ray emission is typical of Galactic compact H II regions, then they cannot be the primary source of the Fe XXV emission from the Galactic disk.

Advection-dominated, high-temperature, quasi-spherical accretion flow onto a compact object, recently considered by a number of authors, assumes that the dissipation of turbulent energy of the flow heats the ions and that the dissipated energy is advected inward. It is suggested that the efficiency of conversion of accretion energy to radiation can be very much smaller than unity. However, it is likely that the flows have an equipartition magnetic field with the result that dissipation of magnetic energy at a rate comparable to that for the turbulence must occur by ohmic heating. We argue that this heating occurs as a result of plasma instabilities and that the relevant instabilities are current driven in response to the strong electric fields parallel to the magnetic field. We argue further that these instabilities heat predominantly the electrons. We conclude that the efficiency of conversion of accretion energy to radiation can be much smaller than unity only for the unlikely condition that the ohmic heating of the electrons is negligible.

We investigate the behavior of the high-frequency quasi-periodic oscillations (QPOs) in 4U 0614+091, combining timing and spectral analyses of RXTE observations. The energy spectra of the source can be described by a power law (α ~ 2.8) and a blackbody (kT ~ 1.5 keV), with the blackbody accounting for 10%-20% of the total energy flux. We find a robust correlation of the frequency, ν, of the higher frequency QPO near 1 kHz with the flux of the blackbody, F_BB. The slope of this correlation, d log ν/d log F_BB, is 0.27-0.37. The source follows the same relation even in observations separated by several months. The QPO frequency does not have a similarly unique correlation with the total flux or the flux of the power-law component. The rms fraction of the higher frequency QPO rises with energy from 6.8% ± 1.5% (3-5 keV) to 21.3% ± 4.0% (10-12 keV). For the lower frequency QPO, however, it is consistent with a constant value of 5.4% ± 0.9%. The results may be interpreted in terms of a beat-frequency model for the production of the high-frequency QPOs.

We investigate the evolution of molecular abundance in quiescent protoplanetary disks that are presumed to be around weak-lined T Tauri stars. In the region of surface density less than 10² g cm^-2 (distance from the star ≳10 AU in the minimum-mass solar nebula), cosmic rays are barely attenuated even in the midplane of the disk and produce chemically active ions such as He⁺ and H⁺₃. Through reactions with these ions, CO and N₂ are finally transformed into CO₂, NH₃, and HCN. In the region where the temperature is low enough for these products to freeze onto grains, a considerable amount of carbon and nitrogen is locked up in the ice mantle and is depleted from the gas phase in a timescale of ≲3 × 10⁶ yr. Oxidized (CO₂) ice and reduced (NH₃ and hydrocarbon) ice naturally coexist in this part of the disk. The molecular abundance both in the gas phase and in the ice mantle varies significantly with the distance from the central star.

We have measured the proper motions of three molecular hydrogen (v = 1-0 2.121 μm) knots in the Herbig-Haro 1 object using a 4.4 yr baseline (1992-1997). The HH 1F knot, which probably corresponds to the tip of the working surface of HH 1, has a proper motion of 0.19 ± 0.10 (arcsec yr^-1) and a position angle of 315^{+ 29}₋₂₅ deg. This motion is comparable to that determined in the atomic emission lines of Hα and [S II] λλ 6717/31, and confirms that the warm molecular H₂ gas is partaking of the same motion as the atomic/ionic gas.

We report the discovery of a new molecular hydrogen outflow in the Serpens molecular cloud. Narrowband filter images taken in the 2.12 μm v = 1-0 S(1) transition of H₂ and adjacent continuum reveal a series of bright knots of pure line emission apparently emerging to the north-northwest from the embedded source SMM-3 and passing close to the visible star CK-8. Low-resolution H- and K-band spectra of the region show more than a dozen distinct H₂ transitions, whose strength ratios point to shock heating with T_exc ~ 2000 K. Echelle spectra of the S(1) transition with 20 km s^-1 resolution reveal unusual kinematics: the line center velocity increases linearly with distance to the north-northwest from SMM-3 until the bright knots of emission, at which point the velocity begins dropping to a fraction of its maximum value. The molecular hydrogen emission likely arises in limb-brightened bow shocks as a jet from SMM-3 encounters the ambient molecular cloud. This scenario is strengthened by recent HCO⁺ and SiO submillimeter observations of SMM-3, which show an apparent outflow corresponding to the H₂ structures.

On 1996 March 12, during the commissioning phase of the SOHO mission, we obtained observations of the quiet-Sun with the SUMER instrument. The observations were sequences of 15-20 s exposures of ultraviolet emission-line profiles and of the neighboring continua. These data contain signatures of the dynamics of the solar chromosphere that are uniquely useful because of wavelength coverage, moderate signal-to-noise ratios, and image stability.

We focus on data for the internetwork chromosphere. The dominant observed phenomenon is an oscillatory behavior that is analogous to the 3 minute oscillations seen in CaII lines. The oscillations appear to be coherent over 3''-8'' diameter areas. At any time they occur over about 50% of the area studied, and they appear as large perturbations in the intensities of lines and continua. The oscillations are most clearly seen in intensity variations in the ultraviolet (λ > 912 Å) continua, and they are also seen in the intensities and velocities of chromospheric lines of CI, NI, and OI. Intensity brightenings are accompanied by blueshifts of typically 5 km s^-1. Phase differences between continuum and line intensities also indicate the presence of upward propagating waves. The detailed behavior is different between different lines, sometimes showing phase lags. The 3 minute intensity oscillations are occasionally seen in second spectra (CII λ1335) but never in third spectra (CIII and SiIII). Third spectra and HeI λ584 show oscillations in velocity that are not simply related to the 3 minute oscillations. The continuum intensity variations are consistent with recent simulations of chromospheric dynamics (Carlsson and Stein), while the line observations indicate that important ingredients are missing at higher layers in the simulations.

The data show that time variations are crucial for our understanding of the chromosphere itself and for the spectral features formed there—the quiet-Sun's chromosphere is very dynamic and not "quiet." The implications of these data should be considered when planning chromospheric work with instruments such as those on SOHO.

The interaction of f- and p-modes with a slab of vertical magnetic field of sunspot strength is simulated numerically in two spatial dimensions. Both f-modes and p-modes are partially converted to slow magnetoatmospheric gravity (MAG) waves within the magnetic slab because of the strong gravitational stratification of the plasma along the magnetic lines of force. The slow MAG waves propagate away from the conversion layer guided by the magnetic field lines, and the energy they extract from the incident f- and p-modes results in a reduced amplitude for these modes as they exit from the back side of the slab. In addition, the incident p-modes are partially mixed into f-modes of comparable frequency, and therefore larger spherical harmonic degree, when they exit the magnetic flux concentration. These findings have important implications for the interpretation of observations of p-mode absorption by sunspots, both in terms of the successes and failures of this simple numerical simulation viewed in the sunspot seismology context.