A mechanism for the destruction of rotating stellar bars by dissipative processes accreting mass at the center[1] is tested by means of 3D N-body simulations (N ≈ 2·105)[2,3]. One or two ILR, originated from the growth of central mass concentration, cause the main elongated plane orbit family supporting the bar to be replaced by plane orbit families perpendicular to the bar, which can then no longer be sustained. The gas is represented in a crude way by a small (9%) set of particles contracting their phase-space with an ad-hoc friction law v̇ = −αv, where α is the inverse of a dissipation time scale. Typically, in a few Gyr an otherwise steady bar is transformed into a shorter, faster and weaker triaxial ellipsoid, similar to many bulges of normal spirals. These might be relics of destroyed bars. An example is shown in Fig. 2 where, after 8 Gyr, dissipation has brought 1.3% of the total mass inside 0.2 kpc, increasing the local average density by a factor of 14. The conservative case is represented in Fig. 1. The a/b/c axis ratios change from 1/0.54/0.55 to 1/0.74/0.60 at 1 kpc, and from 1/0.55/0.35 to 1/0.89/0.51 at 5 kpc. With a stronger dissipation (larger α) or equivalently at later time, the complete bar destruction seems unavoidable.

Numerical models permit a detailed study of the dynamics of merging disk galaxies. To first order, a typical galactic encounter is a well-posed problem in collisionless stellar-dynamics, since ≳ 90% of the mass is in the form of stars or dark matter. Mergers between disk/halo galaxies produce objects resembling elliptical galaxies in structure and kinematics. The three major orbit families expected in triaxial ellipsoids are all present in these merger remnants, and the distribution of stars between orbit families reflects the initial orientations of the colliding disks.

Models including gas dynamics, developed in collaboration with L. Hernquist, show how mergers, and perhaps other violent interactions, may fuel nuclear starbursts. In these experiments a large fraction (~ 60%) of the gas distributed throughout the galactic disks winds up in a central cloud with dimensions of a few hundred parsecs, a result of gravitational torques which transfer angular momentum from the gas to stellar bar. Nuclear gas clouds and starbursts observed in merging galaxies such as NGC 520 and Arp 220 may well be generated by this mechanism.

Because of their mutual gravitational interactions, galaxies do not live forever. In extreme collisional events, galaxy pairs can merge into single spheroidal systems. Numerical simulations (Barnes 1988) indeed support the suggestion that the formation of some ellipticals does result from the merging of two disk galaxies (Toomre and Toomre 1972). However, one must then explain the very low molecular-gas content of ellipticals compared to that of spiral galaxies. Could the burst of star formation triggered during the merging exhaust the molecular material of the disk progenitors ? In order to test this hypothesis, data must be collected on the gas content of mergers at various stages of evolution and of “genuine” ellipticals as well. We here report on multi-line 12CO and 13CO observations of a sample of such objects. Our aims were (i) to contribute to a better knowledge of the molecular content of elliptical galaxies, and (ii) to help understand when and how the molecular gas of a merging system is consumed on its way towards an elliptical. The detailed presentation and discussion of the results are to be found in Casoli et al (1988 : Paper I; 1990 : Paper III), Dupraz et al. (1990 : Paper II), and Dupraz and Casoli (1990).

A major argument against forming ellipticals by mergers is that the large velocity dispersions of clusters preclude slow enough encouters that may lead to mergers (Ostriker 1980). We show below that the observed morphology-density relation in clusters (Postman & Geller 1984) is consistent with the formation of cluster ellipticals by mergers, using a simple “chemical evolution” calculation embedded in the standard spherical cosmological infall scenario.

Observations indicate that the frequency of merging galaxies in the present epoch is ≈0.3%, and an extrapolation to past yields a frequency ≈5% (Toomre, 1977, Tremaine, 1980). Loosely bound pairs of galaxies that had separated to great distances in the general cosmic expansion and have lately fallen together again in comet-like plunging ellipses seem to be the most lucrative candidates for mergers. But could all ellipticals be merger remnants ? To study this problem we have determined the frequency of merging galaxies on the basis of the collision theory, by studying many collisions with different collision parameters and progenitor pairs (using the impulsive approximation) and compared the values so obtained with the observational ones. The mean values of the collision parameters favourable for mergers for different types of progenitor pairs are segregated from a statistical study of many collisions and are compared to the corresponding values in dense regions where a galaxy and its nearest neighbour can be visualised to form a loosely bound pair, to determine the corresponding frequencies. The theory and method is described in detail in Chatterjee, 1987, 1990, except that subsequently we have studied disk-disk mergers by using basically the same method.

Using numerical simulations in the context of a simple dissipative model (Abadi et al. 1990) we study the final morphologies and dynamical properties of galaxies under different initial conditions. The protogalaxies are embedded in dynamically relaxed potential wells of dark matter with different initial spin parameters λ and velocity dispersions σ in the dissipative component. We find that both λ and σ control the final morphology of the galaxies. The angular momentum might be acquired by tidal torques, and several sources may account for the velocity dispersion. For instance, under the assumption that supernovae explosions provide kinetic energy to the medium giving σ ∼ 8-10 km/s (Mc Kee & Ostriker 1977) our models suggest that the border line spin parameter between early type and late type morphologies would be approximately λ ∼ 0.04.

Galaxy disks are observed to be young, dynamically cold, and centrifugally supported. These properties suggest that they formed in a quiescent dynamical context, reached their current configuration while still gaseous, and have been relatively undisturbed since formation. This situation can be understood if disks grow by the cooling of a gaseous atmosphere within a quasi-equilibrium dark halo. A simple version of this model can be combined with results from numerical simulations of halo formation to provide quantitative explanations for the Tully-Fisher relation, for the typical sizes of disks, and for their star formation history. A viable model seems to require stellar winds and supernovae to heat the gaseous halos of spirals via a galactic fountain. Such models predict observable X-ray halos, and suggest that continued infall replenishes the gas consumed by star formation in late-type galaxies.

As it was shown in the observations the collisions and mergers of galaxies exert the powerful initiating influence on the star formation on the galactic scales. These processes are accompanied by strong IR emission, OH maser flares and some other manifestation of the galactic activity. As the IR survey depth will increase the number of IR sources on the celestial sphere related to earlier epochs ought to also increase. The possible reason for the change of the IR source statistics is the evolution of the galaxy mass spectrum (GMS).

Groups of galaxies may constitute an excellent laboratory for the study of galaxy interactions. These should naturally be more spectacular in the apparently densest groups, known as “compact groups”, presumably the densest isolated regions in the Universe. The theoretical question that arises is what is the dynamical evolution of a dense group?, while a more down-to-earth question is are compact groups as dense as they appear? I will attempt here to answer both of these questions in turn, basically summarizing two contributions to an earlier conference (Mamon 1990a,b).

We have studied mergers between disk galaxies and smaller companion galaxies by using a 60 000 particle self-gravitating disk in interaction with a point-like perturber. The disk is a stable Mestel disk with an adjustable inert halo, combining to make a flat rotation curve. The hyperbolic, parabolic and circular initial orbits of the perturber have been studied separately (Byrd et al. 1986, Valtaoja 1990, Valtonen et al. 1990). In hyperbolic encounters, the friction may be evaluated by the impulsive approximation by Spitzer (1958). The agreement with experiments is good except for the dependence on the encounter velocity which may deviate somewhat from the analytic theory. In case of circular orbits, we made comparisons with earlier work by Quinn and Goodman (1986) and others by turning off the self-interaction in some of the simulations. In addition, the analytical calculation of dynamical friction by Chandrasekhar (1942) was extended to disk galaxies, and the results were compared with the numerical work. Our results without self-interaction agree well with Quinn and Goodman when we use the same halo/disk mass ratio (Q0 ≅ 1.5) as they did. However, we show that the results are very sensitive to the assumed halo mass. Moreover, the inclusion of self-interaction can change the results considerably in the case of a retrograde satellite. Our analytic theory explains well the merger rate in those situations where self-interaction is not very important, i.e. in case of satellites in direct orbits. The orbital decay time Δt = r (dr/dt)−1 may then be expressed in term of the crossing time in the disk, Tcr = R/v0, where R is the radius of the disk and v0 the constant rotation velocity: Here m1, ms and M are the masses of a star, the satellite and the disk, respectively, and r is the orbital radius of the satellite in its (nearly) circular orbit.

NGC 1052 is a flat-spectrum radio elliptical exhibiting a characteristic low excitation LINER spectrum. Independent evidence, such as extensive and infalling HI gas, and misalignment of the stellar and ionized gas components, suggest a merger has occurred. Optical CCD imaging reveals the presence of dust and a spiral morphology for the Hα emission (see Forbes, Sparks and Macchetto 1990). The X-ray luminosity is consistent with a hot gaseous halo in which a cooling flow is claimed to be operating (Thomas et al. 1986).

In terms of averaged physical quantities, small scale fluctuations yield an additional momentum and energy transfer between the stellar and gaseous component. The nonlinearities lead to dynamical friction (ρs1gg1) and energy loss of the stellar system [1],[2]. Here is the gravitational force and the momentum density, subscripts s and g refer to the stellar and gaseous component, respectively. Since the density fluctuations in the stellar and gaseous component are correlated due to gravitational forces, these transfer functions are in general nonzero. We treat the stellar system collisionless and the ambient gas hydrodynamically including dissipative processes. We computed the effect of fluctuations induced by the pointlike structure of the stars in quasilinear approximation taking stable modes into account only. Thus the resulting transfer rates seem to be lower bounds, since the effect of growing modes, nonlinearities in the wave structure, and fluctuations induced by different mechanisms are not included. For the timescales τb and τE of momentum and energy loss, respectively, we find the relation where I1 corresponds to the ‘Coulomb logarithm’ in the usual scattering theory and I2 is an additional asymmetry measure of the power spectrum [1]. We discussed two very different astrophysical enviroments, where this effect of density fluctuations is of considerable magnitude. First, the systematic motion of a young stellar system, i.e. with low velocity dispersion, in a molecular cloud is slowed down in a timescale τb as short as 106yr [see also Deiss et al., this Vol]. The timescale τE of energy dissipation is of the same order of magnitude. This estimate shows that friction and dissipation terms may not be neglected in the phase of early evolution of stellar systems as done in [4] e.g. — Secondly, the fluctuation theory can be applied to clusters of galaxies as well [3]. In this context the heating of the intracluster medium and a possible mass segregation of the galaxies are the dominating effects. The main uncertainty arises from the badly known mass-to-light ratio of the galaxies.

We have investigated the combined effects of transient spiral waves and giant molecular clouds in heating stellar discs by solving the Fokker-Planck equation with a Monte Carlo simulation. By using obervations of the velocities of local stellar populations we are able to estimate the product of the mean surface density and the mass-weighted mean mass of the clouds and the RMS potential fluctuations associated with the spiral waves. The estimates show good agreement with observations of giant molecular clouds in our galaxy and HI observations in other spiral galaxies. We make predictions of the vertical and radial velocity distributions of two disc populations.

I wish to discuss two basic questions:

1) what do we know from the observations about the relationships between the physical conditions of the Interstellar medium (ISM) in disk galaxies and their past and present star formation rate, heating of dust and acceleration of relativistic cosmic rays ?

2) are the above relations and mechanisms sensitive to the intergalactic environment in rich clusters of galaxies?

I will discuss these questions by reviewing three topics:

a) gas content versus star formation

b) FIR emission versus star formation

c) radio continuum emission versus star formation and by discussing weather and how these topics are related to the environment.

Spiral galaxies belonging to rich clusters are generally found to be deficient in atomic gas; this has been interpreted as being due to the ram pressure from the intracluster X-ray emitting gas or to the tidal interaction between the galaxies. However, the molecular content of spirals does not share these characteristics, at least in the Virgo cluster (Stark et al., 1986; Kenney and Young, 1989).

Galaxies in the Virgo cluster offer a convenient distance independent sample for the study of large scale properties of star formation (SF). While the gas is the raw material for SF, the galaxy mass seems to play an important role. The luminosities in various bands are sensitive to stars of different masses (ages) and as such sample SF over different periods of time. For example, the far-infrared (far-IR) luminosity samples recent (few tens of million years) and massive stars, while the blue luminosity samples stars over about billion years. The H band luminosity represents almost the average SF over the life of the galaxy. Relationships between luminosities and gas contents, can then offer clues to the history of SF in the galaxy and its variation from galaxy to galaxy. In order to be free of richness effects, correlations are studied by normalising with M(G), the dynamic mass within the optical diameter of the galaxy. The mass normalised quantities will be referred to as specific mass and specific luminosity.

It is well known that there are a lot of unidentified absorption lines in the spectra of QSOs. The usual method of identification of these lines based on the standard list of the most strong lines typical for interstellar medium and the corona of our Galaxy has some serious failures and never will lead to the absolute success.

We have carried out numerical simulations of self-gravitating disks in order to find out the effects of galaxy encounters on the stability of disk galaxies. We use the two-dimensional polar-grid code of Miller (1971, 1974), and study a Mestel disk which has a flat rotation curve. Due to a softening parameter and a finite grid size, the disk is stable to small perturbations, mimicing the effects of velocity dispersion at the level of QT = 1.5 − 2 (Toomre 1964). At sufficiently large perturbations the disk develops global instabilities, a bar, spiral arms and mass inflow towards the center of the disk which may be connected to the origin and maintenance of Seyfert activity in the nucleus of the disk (Byrd et al. 1986, 1987, Sundelius et al. 1987, Thomasson et al. 1989). In rich clusters of galaxies the encounters are generally fast and cannot explain why spiral galaxies in rich clusters are radio sources ten times more powerful than spirals outside clusters (Gavazzi and Jaffe 1987). Gavazzi and Jaffe propose that this results from the ram pressure of the gaseous cluster medium. However, we find that the general tidal field of the cluster is able to trigger instability in the disk of the spiral galaxy when the galaxy falls inside about 700 kpc of the cluster center (Byrd and Valtonen 1990). The instability creates mass inflow to the center of the galaxy, cloud-cloud collisions in the disk and bar formation in the central part of the galaxy. Besides initiating activity in the nuclei of these galaxies (see e.g. Venturi et al. 1990 for the effect in the Coma cluster), the instability may enhance star formation and supernova rate, which may finally lead to cleaning of the spiral disks of their gaseous component. At the same time the instability leads to thickening of the stellar disk and possible change of morphology of the galaxy to an S0.

Recent millimeterwave observations of CO emission from ultraluminous infrared galaxies (Lir > 1012 L) and quasars suggest that abundant supplies of molecular gas may be a common property of both types of object. Current published CO detections are summarized, and the possibility of an evolutionary connection between infrared galaxies and quasars is discussed.

FIGURE 1

CO Intensity Distribution