Galaxies are the fundamental building blocks of the observed universe. They are routinely used to trace the large-scale structure of matter. Their internal structure keeps a fossil record of the complex dynamics of both baryons and dark matter in the universe. By now, galactic systems are observed to exist at redshifts at high as $\sim $4--5, and so they serve as laboratories for structure formation at a time when the age of the universe was $< 10\%$ of its current value. For all these reasons, the formation and evolution of galaxies is one of the most important problems in modern cosmology.

Although galaxies were first discovered as islands of light in the dark sky, they served a pivotal role in arguing for the existence of dark matter in the universe. Dark halos are thought to extend well beyond the optically bright core of galaxies. Various methods, such as gravitational lensing or the dynamics of satellites, are currently used to explore the extent of galactic halos beyond the range of several tens of kpc, where HI rotation curves are unavailable. At the same time, numerical simulations have reached the resolution threshold where they can be compared quantitatively to such studies. The comparison between simulations and data on the structure of halos has the potential of unraveling the nature of the dark matter, be it cold, cold+hot, or baryonic dark matter. Complementary information is currently sought after by dark--matter experiments and active microlensing searches for compact objects in the halo of the Milky-Way galaxy.

The luminous baryons are embedded in the core of the dark matter halos. Their dynamical properties raise a number of interesting theoretical puzzles. Although gravity is dominated by the baryons near the galactic center and by the dark matter in the halo, the rotation curve is often flat in-between these regions, implying some dynamical coupling between the two components. Moreover, the total luminosity emitted by the baryons appears to be tightly correlated with the velocity dispersion of the galaxy(the so--called "Tully--Fisher" relation for spirals or the "Faber--Jackson" relation for ellipticals). The tightness and universality of this correlation make it a distance indicator, that can be used to estimate the Hubble constant or the peculiar velocity field in redshift surveys. The above dynamical regularities of galaxies are particularly puzzling in view of the complex gas dynamical processes which are associated with the formation of stars. Recently, new populations of disk galaxies have been discovered. These include galaxies with disks that have two streams of stars rotating in opposite directions, and disks with unusually low surface brightnesses. The ITP workshop could help address the origin of such objects, as well as the process that differentiates between the more conventional spiral and elliptical galaxies, in the context of generic models for structure formation in the universe.

The luminous appearance of galaxies is intimately related to the theory of star formation and evolution. Existing theoretical work on star formation has, by and large, been stimulated by observations of the complex environments of the interstellar medium in our galaxy. Deep imaging and spectroscopy of high redshift galaxies offer additional sources of observational feedback for models of star formation. Recent deep observations with the Hubble Space Telescope and ground--based telescopes are already able to sketch the history of star formation in the universe. In particular, the unprecedented depth of the Hubble Deep Field revealed galaxies out to redshifts $\sim$ 5. Recent spectroscopic observations with the Keck telescope have confirmed that high--redshift galaxies can be efficiently identified by color as UV--deficient, due to strong galactic and intergalactic absorption by hydrogen beyond the Lyman--limit. Complementary information on the history of the galactic gas and its metallicity is drawn from recent