Star formation in starbursts appears to be biased toward compact clusters, with up to 20% of all stars formed in them. Observations with HST show that many of these clusters have luminosities (-9 > M-V > -16), UBVI colors, and half-light radii (R-eff approximate to 3 pc) consistent with their being young globular clusters (YGC). Although we know little about the long-term stability of the youngest clusters (<20 Myr), compact clusters older than similar to 10 t(cross) (20-40 Myr) are bound gravitationally and very likely YGCs. The present review concentrates on recent progress achieved in dating such clusters and on evidence that mergers of spiral galaxies can produce relatively rich subsystems of YGCs of solar metallicity in the remnants' halos. Studies of such subsystems suggest a scenario in which second-generation globulars form from giant molecular clouds squeezed into collapse by the high-pressure environment of starbursts. These bursts, often driven by mergers, may explain ongoing cluster formation in NGC 4038/39, the young halo globulars found around protoellipticals like NGC 3921 and NGC 7252, and the subpopulations of red metal-rich globulars observed in many giant ellipticals.

The Wide Field Planetary Camera 2 of the Hubble Space Telescope has been used to obtain high-resolution images of NGC 4038/4039 that go roughly 3 mag deeper in V than previous observations made during cycle 2. These new images allow us to measure the luminosity functions (LFs) of clusters and stars over a range of 8 mag (-14 < M-V < -6). To first order, the LF is a power law, with exponent alpha = -2.12 +/- 0.04. However, using a variety of different techniques to decouple the cluster and stellar LFs, which overlap in the range -9 less than or similar to M-V less than or similar to -6, we find an apparent bend in the young cluster LF at approximately M-V = -10.4. Brightward of this magnitude the LF has a power-law exponent alpha = -2.6 +/- 0.2, while faintward the slope is alpha = -1.7 +/- 0.2. The bend corresponds to a mass approximate to 1 x 10(5) M., only slightly lower than the characteristic mass of globular clusters in the Milky Way (approximate to 2 x 10(5) M.). It is currently not feasible to determine the cluster LF fainter than M-V approximate to -8, where individual stars are likely to dominate. The stellar LF in the range -9 < M, < -6 is much steeper, with alpha = -2.9 +/- 0.1, and is dominated by young red and blue supergiants. The star clusters of the Antennae appear slightly resolved, with median effective radii of 4 +/- 1 pc, similar to or perhaps slightly larger than those of globular clusters in our Galaxy. However, the radial extents of some of the very young clusters (ages less than 10 Myr) are much larger than those of old globular clusters (e.g., the outer radius of knot S exceeds 450 pc). This may indicate that the tidal forces from the galaxies have not had time to remove some of the outer stars from the young clusters. A combination of the UBVI colors, H alpha morphology, and Goddard High Resolution Spectrograph (GKRS) spectra enables us to age date the clusters in different regions of the Antennae. Star clusters around the edge of the dust overlap region appear to be the youngest, with ages less than or similar to 5 Myr, while clusters in the western loop appear to be 5-10 Myr old. Many star clusters in the northeastern star formation region appear to be similar to 100 Myr old, with an LF in V that has shifted faintward by similar to 1.0 mag relative to the younger (0-20 Myr) clusters that dominate over most of the rest of the galaxy. A third cluster population consists of intermediate-age clusters (similar to 500 Myr) that probably formed during the initial encounter responsible for ejecting the tails. A handful of old globular clusters from the progenitor galaxies are also identified. Most of these lie around NGC 4039, where the lower background facilitates their detection. Age estimates derived from GHRS spectroscopy yield 3 +/- 1 Myr for knot K (just south of the nucleus of NGC 4038) and 7 +/- 1 Myr for knot S in the western loop, in good agreement with ages derived from the UBVI colors. Effective gas out flow velocities from knots S and K are estimated to be about 25-30 km s(-1), based on the above cluster ages and the sizes of the surrounding H alpha bubbles. However, the measured widths of the interstellar absorption lines suggest dispersion velocities of similar to 400 km s(-1) along the lines of sight to knots S and K.

Gravitational interactions and mergers are shaping and reshaping galaxies throughout the observable Universe. While observations of interacting galaxies at low redshifts yield detailed information about the processes at work, observations at high redshifts suggest that interactions and mergers were much more frequent in the past. Major mergers of nearby disc galaxies form remnants that share many properties with ellipticals and are, in essence, present-day proto-ellipticals. There is also tantalizing evidence that minor mergers of companions may help build bulges in disc galaxies. Gas plays a crucial role in such interactions. Because of its dissipative nature, it tends to get crunched into molecular form, turning into fuel for starbursts and active nuclei. Besides the evidence for ongoing interactions, signatures of past interactions and mergers in galaxies abound: tidal tails and ripples, counter-rotating discs and bulges, polar rings, systems of young globular clusters, and ageing starbursts. Galaxy formation and transformation is clearly a prolonged process, occurring up to the present-day. Overall, the currently available observational evidence points towards Hubble's morphological sequence being mainly a sequence of decreasing merger damage.

We investigate trends of the cold and hot gas content of early-type galaxies with the presence of optical morphological peculiarities, as measured by the fine-structure index Sigma. H I mapping observations from the Literature are used to track the cold gas content, and archival ROSAT Position Sensitive Proportional Counter data are used to quantify the hot gas content. We find that E and SO galaxies with a high incidence of optical peculiarities are exclusively X-ray underluminous and, therefore, deficient in hot gas. In contrast, more relaxed galaxies with little or no signs of optical peculiarities span a wide range of X-ray luminosities. That is, the X-ray excess anticorrelates with Sigma. There appears to be no similar trend of cold! gas content with either fine-structure index or X-ray content. The fact that only apparently relaxed E and SO galaxies are strong X-ray emitters is consistent with the hypothesis that after strong disturbances, such as a merger, hot gas halos build up over a timescale of several gigayears. This is consistent with the expected mass loss from stars.

Globular clusters formed in galactic mergers (e.g., The Antennae) can now be studied at different stages of their evolution. In young merger remnants (e.g., NGC 7252) these "second-generation" globulars appear by the hundreds as young halo clusters of roughly solar metallicity. While at first bluer and much more luminous than old metal-poor globulars, they become redder after 1 - 1.5 Gyr and can then be observed as still overluminous red clusters of intermediate age in perturbed-looking E and SO galaxies (e.g., NGC 1316, 1700, 3610). There is evidence from the color distributions, projected radial distributions, and perhaps also luminosity functions that these clusters eventually assume the properties of red metal-rich globulars observed in many giant ellipticals. Studies of globular clusters in ongoing mergers and young remnants suggest that second-generation globulars form from giant molecular clouds shocked by the rapid pressure increase in the merger-induced starburst. This pressure-induced formation lends credence to Cen's (2001) argument that the general pressure increase during cosmological reionization at z approximate to 7-15 triggered the near-simultaneous formation of the universal population of first-generation metal-poor globulars observed in galaxies of all types.

Collisions and, mergers of gas-rich galaxies trigger bursts of star and cluster formation. Of the thousands of clusters typically formed during a major merger, only the most massive and compact survive for Gigayears as globular clusters (GCs). In less than or equal to 1 Gyr old merger remnants, these 'second-generation' GCs appear by the hundreds as young halo clusters of similar tosolar metallicity. Their likely descendants, metal-rich GCs of intermediate age (2 - 5 Cyr), have recently been found in - 10 E galaxies, where they appear as slightly overluminous red GCs With a still power-law-like luminosity function. Their color and radial distributions suggest that they evolve into the red metal-rich GCs observed in old ellipticals. There is. good evidence that second-generation GCs form from giant molecular clouds shocked by the rapid pressure increase in merger-induced starbursts. This mechanism supports the view that the universal pressure increase during cosmological reionization. may have triggered the formation of the metal-poor globulars observed in galaxies of all types.

The Chandra monitoring observations of "The Antennae'' (NGC 4038/4039) have led to the discovery of a variable, luminous, supersoft source (SSS). This source is detected only at energies below 2 keV and, in 2002 May, reached count rates comparable to those of the nine ultraluminous X-ray sources (ULXs) detected in these galaxies. Spectral fits of the SSS data give acceptable results only for a similar to90 - 100 eV blackbody spectrum with an intrinsic absorption column of N-H similar to (2-3) x 10(21) cm(-2). For a distance of 19 Mpc, the best-fit observed luminosity increases from 1.7 x 10(38) ergs s(-1) in 1999 December to 8.0 x 10(38) ergs s(-1) in 2002 May. The intrinsic, absorption-corrected, best-fit luminosity reaches 1.4 x 10(40) ergs s(-1) in 2002 May. The assumption of unbeamed emission would suggest a black hole of greater than or similar to 100M(.). However, if the emission is blackbody at all times, as suggested by the steep soft spectrum, the radiating area would have to vary by a factor of similar to10(3), inconsistent with gravitational energy release from within a few Schwarzschild radii of a black hole. Viable explanations for the observed properties of the SSS are provided by anisotropic emission from either an accreting nuclear-burning white dwarf or an accreting stellar-mass black hole.

Young massive clusters differ markedly from old globular clusters in featuring extended, rather than tidally truncated, envelopes. Their projected-luminosity profiles are well fit by Elson-Fall-Freeman models with core radii 0.3 pc, r <= 8 pc and power-law envelopes of negative exponent 2.0 <= gamma <= 3.8. These envelopes form within the first few 10(6) yr and last similar to 10(8) -10(9.5) yr, depending on the environment. Many YMCs show clumpy substructure that may accelerate their initial relaxation. The cores of Magellanic-Cloud clusters show universal expansion from r(c) < 1 pc at birth to r(c) = 2 - 3 pc after 10(8) yr, but then seem to evolve along two bifurcating branches in a r, - log(age) diagram. The lower branch can be explained by mass-loss driven core expansion during the first 109 yr, followed by slow core contraction and the onset of core collapse due to evaporation. The upper branch, which shows continued core expansion proportional to logarithmic age, remains unexplained. There is strong evidence for rapid mass segregation in young clusters, yet little evidence for top-heavy IMFs or primordial mass segregation. Finally, YMCs show similar structure throughout the Local Group and as far away as we can resolve them (<= 20 Mpc).