The Formation of Giant Elliptical Galaxies and Their Globular Cluster Systems

The bimodal globular cluster (GC) metallicity distributions of many giantelliptical galaxies are often cited as evidence for the formation of suchgalaxies through mergers involving gas-rich spirals. In such models, the metal-rich GCs are assumed to have formed during the merger process. We explore analternative possibility: that these metal-rich clusters represent the galaxy'sintrinsic GC population and that the metal-poor component of the observed GCmetallicity distribution arises from the capture of GCs from other galaxies,either through mergers or through tidal stripping. Starting with plausibleassumptions for the initial galaxy luminosity function and for the dependenceof GC metallicity on parent galaxy luminosity, we show that the growth of apre-existing seed galaxy through mergers and tidal stripping is accompanied bythe capture of metal-poor GCs whose properties are similar to those which areobserved to surround giant ellipticals. We describe a method of using theobserved number of metal-poor and metal-rich GCs to infer the merger historiesof individual elliptical galaxies, and use this technique to derive limits onthe number of galaxies and total luminosity accreted to date by M49. We arguethat although GC specific frequency is conserved in galaxy mergers, the samemay not be true of tidal stripping by the mean field of the host galaxycluster. Comparisons of model GC metallicity distributions and specificfrequencies to those observed for the well-studied galaxies M49 and M87 showthat it is possible to explain their bimodal GC metallicity distributions anddiscordant specific frequencies without resorting to the formation of new GCsin mergers or by invoking multiple bursts of GC formation.