Starburst galaxies are currently forming massive stars at prodigious rates. I discuss the star-formation histories and the shape of the initial mass function, with particular emphasis on the high- and on the low-mass end. The classical Salpeter IMF is consistent with constraints from observations of the most massive stars, irrespective of environmental properties. The situation at the low-mass end is less clear: direct star counts in nearby giant H II regions show stars down to ~1 M⊙, whereas dynamical arguments in some starburst galaxies suggest a deficit of such stars.

The formation of young star clusters in merging galaxies is, by now, well established. The new challenge is to use these young clusters as a tool to address some of the outstanding questions. For example, what fraction of these young clusters become globular clusters? Is this enough to explain the difference in the specific globular cluster frequencies for spirals and ellipticals? What is it about the collision between two gas-rich galaxies that triggers giant molecular clouds to form star clusters? Can the star clusters be used to age date merger remnants and establish a convincing evolutionary connection between merging spirals and elliptical galaxies? This review will focus on the last of these items.

The Antennae galaxies (NGC 4038/39) are a pair of relatively nearby (D ~ 19 Mpc for H0 = 75) gas-rich spirals of comparable mass in the process of merging with extensive dynamical and evolutionary synthesis modeling available. With HST WFPC1 Whitmore & Schweizer (1995) (WS95) detected 738 bright Young Star Clusters (YSCs) around the Antennae. YSCs have been detected in many interacting and starbursting galaxies, but the YSC system in the Antennae is the most populous one, well defining an exponential Luminosity Function (LF) over more than 5 mag above the completeness limit. Problems arising from crowding of the YSCs on a bright and variable galaxy background and from the PSF of WFPC1 may cause blending of YSCs and an overestimation of their effective radii.

Millimeter-wave interferometry has clearly shown the existence of enormous masses (109 – 1010M⊙) of molecular gas concentrated in the nuclear regions (R < 500 pc) of many luminous and ultraluminous infrared galaxies. In these systems, molecular gas is an obvious source of fuel for nuclear starbursts and active galactic nuclei (AGN). In several ultraluminous systems (e.g., Arp 220 and Mrk 231), there now exists CO (2-1) interferometry at ≤ 1″ resolution which reveals for the first time extremely dense, gaseous accretion disks on the scale 50-300 pc. Based on the low velocity dispersion of the molecular gas in the nuclear disks, we believe them to be extremely thin (10-50 pc). In addition, high brightness temperatures in the CO lines (10-20 K) imply that these disks are nearly uniformly filled with a continuous gas distribution, rather than being relatively isolated, self-gravitating GMCs. Although the gas is ‘uniformly’ distributed, the gas densities must be high, > 104 cm−3. When viewed near the plane of the disk, the central AGNs, if they exist, will be totally obscured at optical and near infrared wavelengths. In Mrk 231, our line of sight is probably within 60° of the disk axis, but in Arp 220 the disk is closer to edge-on. In fact, recent near infrared imaging of Arp 220 with the NICMOS camera on the HST reveals totally opaque dust disks embedded within the central star clusters of both nuclei.

The discovery, with HST imaging, of proto–globular cluster candidates in NGC 1275 (Holtzman et al. 1992) was regarded by many as a major success of the merger model for globular cluster formation (e.g. Ashman & Zepf 1992) and has been cited in support of the idea that elliptical galaxies form from the merger of two or more spiral galaxies. A prediction of the Ashman & Zepf model was that newly-formed clusters should be observable in currently or recently merging systems. The NGC 1275 clusters constitute an important test of globular cluster formation models. NGC 1275 is the peculiar cD galaxy at the center of the Perseus cluster. It shows evidence for a merger history and may indeed be undergoing a merger at present. It also has one of the largest known cooling flows.

Luminous starbursts are observed to occur mostly as a result of a collision/merger in gas-rich galaxies, and most luminous infrared galaxies (LIGs) are indeed gas-rich mergers. In order to determine the relationship between the IR and molecular gas properties and the galaxy-galaxy interactions, we study LIG mergers in the intermediate merging process. We have observed nearly 20 LIG mergers and together with the CO data in the literature, we have found a correlation between the CO luminosity, LCO, and the projected separation of merger nuclei, RSep, in > 50 LIG mergers. The correlation suggests the molecular content is decreasing as merging advances and is better established with ~ 40 LIG mergers excluding ultraluminous ones, which resembles more a volume-limited, statistically complete sample of LIG mergers. In addition, an anti-correlation between LIR/LCO (the measure of star formation efficiency, SFE) and RSep is evident. One interpretation is that the molecular gas content of LIG mergers is being rapidly depleted due to the merger-induced starbursts and the increase of SFE as merging progresses.

The rate of star formation is enhanced in many galaxies that show disturbed morphology and/or evidence of interaction. However, the physical explanation of this connection has proved elusive, since it appears that peculiar morphology of a specific type is neither necessary nor sufficient to promote star formation. To explore the relation between star formation rates and galaxy morphology we have selected a sample of galaxies from the Arp-Madore Catalogue of Southern Peculiar Galaxies and Associations (AMC) and the Principle Galaxy Catalogue (PGC), and estimated their star formation rate using the far-infra-red (FIR) power listed in the QMW IRAS Galaxy Catalogue (QMCIGC). There are 933 objects common to all three catalogues, potentially providing the necessary data for the peculiar galaxies.

The main results of this study can be summarized as follows: (a) The HI and CO linewidths are well correlated. Interaction between galaxies has little influence on the HI and CO line broadening. A rapidly rotating nuclear disk in the galaxy could lead to CO line broadening, while the HI line is less affected by the rotating disk. Molecular gas in Markarian galaxies is centrally concentrated. (b) For past and present star formation activity both HI and H2 components of the gas are important. The atomic and molecular gas surface densities are well correlated with blue, FIR, and radio continuum surface brightnesses, but the H2 surface density is better correlated than that of the HI. The two gas phases are also connected. (c) In general, galaxies with UV-excess (Markarian galaxies) are not distinguished by star formation properties from non-UV galaxies, however some second order differences may exist, like the relation between atomic surface density and radio continuum surface brightness.

We obtained HI, radio continuum, and 12CO J = 1 → 0 observations at resolutions of 12″ to 33″ (= 2.9 - 8 kpc), and B, I, J, and K–band images, of the galaxy NGC 2535 and its small starburst companion NGC 2536. NGC 2535 has an ocular (eye-shaped) structure indicative of a recent, close, nonmerging encounter. Our observations reveal widespread high velocity dispersions (30 km s−1) in the HI gas and five clouds with masses of ~ 108M⊙ in the tidal arms of NGC 2535. CO emission was detected at the center and on the tidal tail, but close to the center, of NGC 2535; no CO emission was detected from the companion. NGC 2535 has an intrinsically oval shape to the disk, an extended (R = 48 kpc) HI envelope and an outer elliptically–shaped HI arc that may be a gravitational wake produced by the passage of the companion within or close to the extended HI envelope. The starburst companion, NGC 2536, lies in a 2 × 109M⊙ clump of HI gas at the outer end of the tidal bridge from NGC 2535. A full account our results appears in Kaufman et al. (1997, AJ, 114, 2323).

We report mid-infrared images of several luminous infrared galaxies (LIGs) taken with ISOCAM on the Infrared Space Observatory (ISO). These LIGs were chosen to represent different phases of a merger sequence of galaxy-galaxy interaction with special emphasis on early/intermediate stages of merging. The molecular gas distribution of these LIGs has also been mapped at high spatial resolution (see contribution by Gao et al., this volume). The goal is to do a synoptic study of the evolution of physical conditions in these LIGs along the merger sequence.

We have studied the collisional interaction of disk galaxies using SPH and N-body numerical simulations. We follow the dynamics of a two-component star-gas disk centrifugally balanced by the gravity of a rigid halo. A companion galaxy is modeled self-consistently, and its two-component disk is balanced by the potential of the stellar halo. This model allows us to follow the dynamics of the gas component in the colliding galaxies simultaneously with the dynamics of the collisionless halo of the companion.

To determine metallicity and age distributions of globular clusters (GCs) in distant galaxies–now accessible, e.g. to Keck spectroscopy–it is important to have reliable calibrations of the color-metallicity, color-age, and index-metallicity relations.

Using the latest stellar evolution models, theoretical stellar spectra, and a compilation of observed emission line strengths from Wolf-Rayet (WR) stars, we have constructed evolutionary synthesis models for young starbursts (Schaerer & Vacca 1997; see also Schaerer 1996). We provide detailed predictions of UV and optical emission line strengths for both the WR stellar lines and the major nebular hydrogen and helium emission lines, as a function of several input parameters related to the starburst episode.

We have obtained images of the H2 and Brγ emission lines in the galaxy interacting system NGC 3690 + IC 694. We have also obtained simultaneous H- and K-band spectra for three of its 2μm continuum peaks. The most detectable line emission is concentrated at the continuum peaks. Therefore, the emission lines as well as stellar absorption lines can be used as tracers of the activity in the nuclei themselves. From the strong Brγ and marginal detection of Br10 at the nucleus of IC 694, we derive a large extinction for the fully ionized gas in this nucleus. If we adopt this extinction also for the [Fe II]1.64μm emission, the extinction-corrected [Fe II]1.64μm/Brγ ratio will lie at the higher end of starburst galaxies, and is typical for AGNs or AGN/starburst composites. This might imply that many SNRs are involved in the starburst at this nucleus, unless it includes an AGN. All of our results for Component C, including very little CO absorption in the K band, a large EW(Brγ), a small H2/Brγ ratio, the effective temperature (Teff ≃ 40,000K) derived from HeI 1.70μm/Br10 and HeI 2.06μm/Brγ, are consistent with a very young starburst.

We present recent results of X-ray observations of two luminous infrared galaxies, NGC3690+IC694 (Arp299) and NGC1614 obtained by the Japanese X-ray astronomical satellite ASCA. Both galaxies have quite high infrared luminosity (> 1011L⊙) and strong evidence of merger.

At luminosities above ~ 1011L⊙, infrared galaxies become the dominant population of extragalactic objects in the local Universe (z < 0.5), being more numerous than optically selected starburst and Seyfert galaxies, and QSOs at comparable bolometric luminosity. At the highest luminosities, ultraluminous infrared galaxies (ULIGs: Lir > 1012L⊙), outnumber optically selected QSOs by a factor of ~ 1.5–2. All of the nearest ULIGs (z <0.1) appear to be advanced mergers that are powered by both a circumnuclear starburst and AGN, both of which are fueled by an enormous concentration of molecular gas (~ 1010M⊙) that has been funneled into the merger nucleus. ULIGs may represent a primary stage in the formation of massive black holes and elliptical galaxy cores. The intense circumnuclear starburst that accompanies the ULIG phase may also represent a primary stage in the formation of globular clusters, and the metal enrichment of the intergalactic medium by gas and dust expelled from the nucleus due to the combined forces of supernova explosions and powerful stellar winds.

A review of recent optical and infrared spectroscopic results on luminous infrared galaxies is presented. The main emphasis is on the ultraluminous objects. Possible correlations with infrared luminosity are identified. These results are used to constrain the nature of the dominant energy source in luminous infrared galaxies, and to test whether these objects may represent an evolutionary phase between starburst galaxies and active galactic nuclei.

Ultraluminous infrared galaxies (ULIRGs) form the dominant component of the local galaxy population at the highest luminosities (e.g., Sanders & Mirabel 1996 and references therein). They are major mergers of gas-rich galaxies and are characterized by a molecular interstellar medium (ISM) that is concentrated in the central kpc, where it forms a dynamically significant, or perhaps dominant, component (Scoville et al., 1991). Sanders et al. (1988) proposed that ULIRGs might harbor, or evolve into, dust-enshrouded quasars, formed in their obscured centers.

A unified formation mechanism of nuclear starbursts is presented; all the nuclear starbursts are triggered by binary supermassive black holes made in the final phase of galaxy mergers. Minor mergers cause both nuclear starbursts and hot-spot nuclei while major mergers cause (ultra) luminous infrared galaxies. We discuss the case of Arp 220 in detail.

Is QSO activity sometimes triggered by interactions or mergers of galaxies? Are all QSOs the result of interactions or mergers? Such questions, arising explicitly in the late 1970s, had their roots in the well-known Stoking the Furnace? section of Toomre & Toomre (1972), the Feeding the Monster paper of Gunn (1979), and the early observational evidence that weaker forms of nuclear activity, such as Seyferts, seemed often to be found in interacting systems (e.g., Adams 1977). Reviews of the evidence regarding the relation of interactions and mergers to QSO activity have been given by Stockton (1990) and Heckman (1990). Some of the more important developments since these reviews are: