1. INTRODUCTION

Globular star clusters (GCs) are among the oldest radiant objects in the universe. With typical masses ~ 10⁴ - 10⁶ M (corresponding to luminosities of ~ M_V = -5 to -10) and compact sizes (half-light radii of a few pc), they are readily observable in external galaxies. The 15 years since the Annual Review by Harris (1991, "Globular Cluster Systems in Galaxies Beyond the Local Group") have seen a revolution in the field of extragalactic GCs. It is becoming increasingly apparent that GCs provide uniquely powerful diagnostics of fundamental parameters in a wide range of astrophysical processes. Observations of GCs are being used to constrain the star formation and assembly histories of galaxies, nucleosynthetic processes governing chemical evolution, the epoch and homogeneity of cosmic reionization, the role of dark matter in the formation of structure in the early universe, and the distribution of dark matter in present-day galaxies. GCs are valuable tools for theoretical and observational astronomy across a wide range of disciplines from cosmology to stellar spectroscopy.

It is not yet widely recognized outside the GC community that recent advances in GC research provide important constraints on galaxy formation that are complementary to in situ studies of galaxies at medium to high redshift. The theme of this article is the role of GC systems as tracers of galaxy formation and assembly, and one of our primary aims is to emphasize the current and potential links with results from galaxy surveys at high redshift and interpretations from stellar population synthesis, numerical simulations, and semi-analytical modeling. In what follows we will attempt to chronicle the observations that mark recent milestones of achievement and to place them in the wider theoretical and observational context. We will focus most closely on work carried out since about 2000. The preceding period is well-covered by the book of Ashman & Zepf (1998) and the Saas-Fee lectures of Harris (2001). ¹ Among the significant topics not directly covered are young massive clusters (potential "proto-GCs"), X-ray sources in extragalactic GCs, and ultra-compact dwarf galaxies. Neither do we include a comprehensive discussion of the Galactic GC system.

The fundamental premise in what follows is that GCs are good tracers of the star formation histories of spheroids (early-type galaxies, spiral bulges, and halos), in the sense that major star-forming episodes are typically accompanied by significant GC formation. Low-level star formation (e.g., in quiescent galactic disks) tends to produce few, if any, GCs. Since most of the stellar mass in the local universe is in spheroids (~ 75%; Fukugita, Hogan, & Peebles 1998), GCs trace the bulk of the star formation history of the universe. Although the relationships between star formation, GC formation and GC survival are complex and do not necessarily maintain relative proportions under all conditions, this underlying assumption is supported by a number of lines of argument. Massive star clusters appear to form during all major star-forming events, such as those accompanying galaxy-galaxy interactions (e.g., Schweizer 2001). In these situations, the number of new clusters formed scales with the amount of gas involved in the interaction (e.g., Kissler-Patig, Forbes, & Minniti 1998). The cluster formation efficiency (the fraction of star formation in clusters) scales with the star formation rate, at least in spiral galaxies where it can be directly measured at the present epoch (Larsen & Richtler 2000). This may suggest that massive clusters form whenever the star-formation rate is high enough, and that this occurs principally during spheroid formation. Perhaps most importantly, the properties of GCs (especially their metallicities) are correlated with the properties of their host galaxies.

¹ Available online at http://physwww.mcmaster.ca/~harris/Publications/saasfee.ps. Back.