4.4. Accretion Scenarios
An almost diametrically opposite approach to the problem is taken by Côoté et al. (1998) through an accretion model. Here, the original E galaxy is imagined to form in a single major burst which generates the metal-rich stars and clusters. Then, over time it accretes many smaller galaxies (which each have lower metallicity) and builds up the metal-poor halo and cluster population. They show by numerical simulations that the accreted population can build up the right MDF shape for the clusters and, potentially, match many specific cases.
This model has the advantage of using a process that we are fairly certain is happening to real galaxies; Sagittarius and its disruption by the Milky Way is the best-known example. But this approach also is not without problems. A significant one is that for a typical gE, the number of metal-poor and metal-rich clusters is roughly equal, which implies that roughly half the present-day galaxy must have been accreted material. But then, why is the color of the halo light as red (metal-rich) as it is, and not halfway between the MRC and MPC clusters? In addition, the sheer numbers of accreted dwarf galaxies needed by the model are very large (many hundreds would be required, for example, to build up our test example, NGC 4472), and imply that the great majority of the original dwarf population must now be absorbed by the larger ones, despite their very small collision cross sections. Detailed dynamical simulations of the process are needed to determine whether or not this assumption is realistic. Finally, if many of the dwarfs are dE or irregular types, then their rather low specific frequencies (SN ~ 1-2) would tend to lower the SN of the final giant elliptical rather than keep it at its desired level of 5 or more.
The conditions of this accretion scenario would change noticeably if the incoming dwarfs were highly gaseous, so that new clusters could form in the process and also maintain most of the stars at the necessarily high metallicity that we see in giant ellipticals (see also Hilker 1998 for a similar view). If so, however, then the entire picture once again begins to resemble the standard Searle/Zinn galaxy formation model whereby a large galaxy builds up from many small dwarf-sized gas clouds. The presence or absence of gas is therefore a critical factor in evaluating the success of any formation picture.