The outer gas disk of our Galaxy (and many others) is warped, bending away from the plane defined by the inner disk. The bend begins just outside the solar circle; gas at longitudes ℓ ≃ 80° reaches highest above the plane, while material at ℓ ≃ 260° lies below it. Shortwavelength ripples are superposed. The orbit of a free particle inclined to the galactic disk precesses at a rate which depends on galactocentric radius; warped structures will tend to do the same, winding the warp into a tight spiral. The large-scale galactic warp has no sense of spirality, and is not obviously of recent origin - why then has it survived?

The problem of the persistence of warping in the outer regions of many galactic HI discs is now well known. Such a warp is observed in the Milky Way, where further deviations from a flat plane are seen in the form of a systematic tilt of the gas disc near the centre, with a short-wavelength (approximately 2 kpc) corrugation at intermediate radius.

We formulate and solve the hydrodynamic equations describing an azimuthally symmetric galactic disk as a two-fluid system. The stars and the gas are treated as two different isothermal fluids of different velocity dispersions (CS ≫ Cg), which interact gravitationally with each other. The disk is supported by rotation and random motion. The formulation of the equations closely follows the one-fluid treatment by Toomre (1964). We solve the linearized perturbation equations by the method of modes, and study the stability of the galactic disk against the growth of axisymmetric two-fluid gravitational instabilities.

The paper aims at a demonstration of the principal differences between the oscillation spectra of multi-component systems and a one-component medium. The character of the mutual motions of components appears then to be of importance. Three cases are considered: (1) motionless components in the inertial frame of reference; (2) inertial subsystems moving at constant relative velocities; (3) a rotating n-component system. The oscillation spectra in these three cases have qualitative differences between each other and when compared with those of resting or rotating one-component systems.

The formation and maintenance of spiral structure in galaxies is discussed in terms of spiral modes with a general perception consistent with the hypothesis of quasi-stationary spiral structure. The latter is explained in some detail to give it the original proper perspective and to contrast it with recent studies of non-stationary spiral structures. We again emphasize the possible coexistence of fast-evolving spiral features and slow-evolving spiral grand designs. Spiral modes are described in terms of three categories of propagating waves. The reason is given why isolated non-barred galaxies with spiral grand design are found to be mostly two-armed. Some suggestions are made for future research.

Recently it has been possible to perform computer calculations with sufficiently high resolution to obtain the response of the gas in galaxies to a tightly wound spiral potential due to a stellar density wave. Previously for pitch angles ~ 10° the problem had to be simplified using the tightly wound approximation, with gradients parallel to the arms neglected compared to perpendicular gradients, yielding a quasi-1-dimensional calculation. This paper reports the early results of a 2-dimensional calculation, and compares them with the 1-dimensional results.

Short-wavelength adiabatic oscillations of an infinitesimally thin, rotating and self-gravitating gas disk in dynamical equilibrium are studied. The problem is formulated in terms of the theory of oscillations of gas spheres (i.e. stars) in order to help galactic and stellar seismologists towards their mutual understanding of their rather similar subjects.

The equilibrium and the stability of homogeneous gaseous and uniformly rotating MacLaurin spheroids imbedded in a rigid, homogeneous and spherical stellar halo are considered. Such systems may be useful as crude models of disk-halo galaxies. Explicitly, we have determined the modes of oscillations stemming from the fundamental radial mode of the homogeneous compressible sphere and its non radial “f” (or Kelvin) modes belonging to the second (ℓ=2) and third (ℓ=3) spherical harmonics, together with the first “pressure” and gravity modes p1 and g1 belonging to the first spherical harmonics (ℓ=1). It has been assumed that the oscillations are adiabatic with a ratio of specific heats γ=5/3, and several sequences of spheroids corresponding to various values of the halo mass have been examined.

N-body experiments with stellar disks like those performed by Hohl (1971) and Sellwood (1981) show regular, global, two-armed spiral structures associated with bar formation. The question is whether the spiral is caused by the bar, and if this is true, how the generating mechanism works.

This paper estimates the effect of the stellar velocity dispersion on local and global instabilities in galaxy disks. Measurements of rotation velocities and velocity dispersions are illustrated for the disks of NGC 488 (Illingworth and Kormendy, in preparation), NGC 936 and NGC 1553. These are used to derive the Toomre stability parameter Q, i.e. the ratio of the observed dispersion to that needed for marginal stability against local modes. In both NGC 488 and 1553, Q = 2.5 − 4, depending on the assumed mass-to-light ratio. These values larger than 2 imply that the stellar disks are stable, although the gas in NGC 488 may remain unstable. This is consistent with the observation that both galaxies largely lack coherent spiral structure: NGC 1553 is an SO and NGC 488 a flocculent spiral. The SBO galaxy NGC 936 is more extreme: Q = 5 (estimated error is a factor of two). This is consistent with Sellwood's suggestion that a bar heats the disk until it is too hot to support spiral structure. In contrast to the above galaxies, Toomre has shown that our Galaxy has Q = 1.6 ± 0.5 near the Sun. It is therefore responsive enough to make spiral structure, as observed.

We also explore the stellar kinematics of the disk of NGC 1553 as a function of radius. The exceptionally strong lens component in this galaxy is very hot. At half of the radius of the lens the velocity dispersion appears to be large enough to have an effect even on global stability.

The study of orbits of a test particle in the gravitational field of a model barred galaxy is a first step toward the understanding of the origin of the morphological characterstics observed in real barred galaxies. In this paper we confine our attention to the inner rings. Inner rings are a very common characteristic of barred galaxies. They are narrow, round or slightly elongated along the bar (with typical axial ratios from 0.7 to near 1.0), and of the same size as the bar. A first step to test the old hypothesis that inner rings consist of stars trapped near stable periodic orbits would be a study of particle trapping around periodic orbits encircling the bar. Such a study is contained in the work of several authors (Danby 1965, de Vaucouleurs and Freeman 1972, Michalodimitrakis 1975, Contopoulos and Papayannopoulos 1980, Athanassoula et al. 1983). In the above works the stability of periodic orbits was studied with respect to perturbations which lie on the plane of motion z = 0 (planar stability). To ensure the possibility of formation of rings, a study of stability with respect to perturbations perpendicular to the plane of motion (vertical stability) is necessary. In this paper we investigate the properties of periodic orbits which we believe to be relevant for the inner-ring problem using a sufficiently general model for the galaxy and sets of values for the parameters which cover a wide range of different possible cases. We also study the stability, planar and vertical, with respect to large perturbations in order to estimate the extent of particle trapping. A detailed numerical investigation of three-dimensional periodic orbits will be given in a future paper.

A relatively minor oval distortion in the central region of the Galaxy, turning at a representative angular pattern speed, can excite outgoing waves at the outer Lindblad resonance of that pattern speed. Associated with the density crest of these waves is fast-expanding gas flow. The physical basis of this phenomenon can be understood through a linear analysis. However, to explain the observed expanding velocity in the “3-kpc arm”, the non-linear theory must be used. In our calculations an oval distortion turning at 118 km s-1 kpc-1 with a perturbation of 10% of the mean gravitational field at the outer Lindblad resonance (located at 3 kpc in the present case) can generate an outgoing velocity of 53 km s-1 at the first density crest of the wave (located at 3.6 kpc).

The observed gas distribution and kinematics in the central 4 kpc of our Galaxy show many aspects reminiscent of barred spirals. In this note I describe a collection of (l,v) maps obtained from a gas flow model of a barred spiral.

Most approaches to explaining the long-range order of the spiral arms in galaxies assume that it is induced by the long-range gravitational interaction. However, it is well-known in many fields of physics that long-range order may be induced by short-range interactions. A typical example is magnetism, where the exchange interaction between magnetic spins has a range of only 10 ångströms, yet a bar magnet can be made as large as one likes. Stochastic self-propagating star formation (SSPSF) starts from the point of view of a short-range interaction and examines the spiral structure arising from it (Seiden and Gerola 1982). We assume that the energetic processes of massive stars, stellar winds, ionization-front shocks and supernova shocks, in an OB association or open cluster can induce the creation of a new molecular cloud from cold interstellar atomic hydrogen. In turn this new molecular cloud will begin to form stars that will allow the process to repeat, creating a chain reaction. The differential rotation existing in a spiral galaxy will stretch the aggregation of recently created stars into spiral features.

Galaxies are dissipative systems, and the spatial and time structure of the interstellar medium and young stars is governed by reaction-diffusion equations. The coherent galactic oscillations of star formation self-organized in spiral waves, previously detected by numerical simulations (Seiden, Schulman, Feitzinger, 1982) can be analytically described by the concept of a limit cycle. Analytical work on self-propagating stochastic star formation is also done by Kaufman (1979), Shore (1981, 1982) and Cowie and Rybicki (1982).

We examine how star formation occurs in the Galaxy and come to the following conclusions. (1) The distribution of newly-born stars in the Galaxy depends on the origin of giant-molecular-cloud complexes. For individual complexes, we favor the mechanism of Parker's instability behind galactic shocks. The production of “supercomplexes” may require the mediation of Jeans instability in the interstellar gas. (2) Magnetic fields help to support the clumps of molecular gas making up a complex against gravitational collapse. On a timescale of 107 years, these fields slip by ambipolar diffusion relative to the neutral gas, leading to the formation of dense cloud cores. This timescale is the expected spread in ages of stars born in any clump. (3) When the cores undergo gravitational collapse, they usually give rise to low-mass stars on a timescale of 105 years. (4) What shuts off the accretion flow and determines the mass of a new star is the onset of a powerful stellar wind. The ultimate source of energy for driving this wind in low-mass stars is the release of the energy of differential rotation acquired during the protostellar phase of evolution. The release is triggered by the entire protostar being driven convectively unstable when deuterium burning turns on.

We present a model of the interstellar medium in spiral galaxies. The ISM is simulated by a system of particles, representing gas clouds. It is a probabilistic N-body system in which “cloud” particles orbit ballistically, undergo dissipative collisions with other clouds, and experience velocity-boosting interactions with expanding supernova remnants (SNRs). Associations of protostars may form in clouds following collisions or SNR interactions (thus allowing sequential star formation); these associations take finite times before becoming active and undergoing their own SN events (Roberts and Hausman, 1983; Hausman and Roberts, 1983).

We present further results of our model of the interstellar medium in spiral galaxies (Roberts and Hausman, 1983; Hausman and Roberts, 1983). The ISM is simulated by a system of particles, representing gas clouds, which orbit ballistically in a spiral-perturbed galactic gravitational field, collide inelastically with one another, and receive velocity impulses from expanding supernova remnants. Star formation may be triggered in the clouds by either collisions or SNR interactions.

Abundances in interstellar clouds, as determined from interstellar absorption lines, are discussed first, including abundances in ‘abnormal’ (high-velocity) clouds. HII-region abundances are then discussed and compared to results from the interstellar clouds. The present status of an abundance gradient as determined from HII regions is given. Abundances in planetary nebulae are then given for various categories of nebulae, and compared to HII regions. Finally a short status report on abundances near the galactic center is given.

We report first results of a systematic study of the galactic interstellar nitrogen isotope ratios, based upon measurements of the (J,K) = (1,1) and (2,2) lines of the 14NH3 and 15NH3 molecules. We find significant deviations of the [14N]/[15N3] abundance ratio from the solar-system value (~270), with the most extreme enrichment in the galactic-center region, where [14N]/[15N] ~ 1500. This value is consistent with a nitrogen nuclear-synthesis scheme in which 14N is the main product of normal secondary CNO-burning, whereas 15N is depleted by the same process.