Annu. Rev. Astron. Astrophys. 1991. 29: 239-274 Copyright © 1991 by Annual Reviews. All rights reserved

4.2 Mergers?

Many formation scenarios invoke a hierarchical buildup of collapsed clumps to typical galaxy masses (21, 105, 368). There appears to be supporting evidence that (many) ellipticals have formed in a hierarchical way. The presence of shells and kinematically distinct cores, and the low rotation of bright ellipticals are expected in a hierarchical formation scheme. Detailed studies of nearby merger remnants have shown that these systems are remarkably similar to elliptical galaxies (203, 308, 309). Furthermore, as pointed out by Toomre, the present merger rate of spiral galaxies is high enough to have produced a significant fraction of current ellipticals during the lifetime of the universe (332).

Most of the arguments against merging apply to very simple scenarios where, e.g., ellipticals were formed from low-luminosity ellipticals, or from gas-poor spirals (264). There is considerable evidence that the gas content of galaxies was different at early epochs, so that arguments against forming ellipticals from current spirals (including gas) may in fact say more about the epoch at which galaxies formed, then about the mechanism by which they formed (150, 206, 343). The globular cluster frequencies of galaxies may provide interesting constraints (150, 343), but unfortunately, we have a very poor understanding of the formation and evolution of these systems (cf. the completely different formation scenarios discussed in 131 and 183).

Merger simulations are extremely useful because of their predictive power. For this reason they are an excellent paradigm against which current observations can be tested. The simplest way to include gaseous processes in the simulations is by assuming that the gas formed a disk before the merger event, and is (partially) expelled afterwards. This type of scenario is now being investigated by Barnes (15, 16, 17), and detailed comparison of his predictions with observed quantities will be very interesting. However, the main uncertainties of this scenario concern the properties of the galaxies before the merger (especially their gas content), and the question whether galaxies assemble their mass in one big event, or in many, more gentle events. Simulations of galaxy formation (e.g., 57), and observations of high redshift galaxies may provide useful additional information.