Galaxy clusters are old in two distinct respects: first, cluster
galaxies probably collapsed early; second, dynamical processes run
faster in proportion to 
. Thus clusters should contain
remnants of many high-redshift mergers. Archeological evidence from
nearby clusters provides important clues to these mergers.
4.1. Merger Formation
After some controversy, it's generally accepted that elliptical galaxies can be formed by fairly recent mergers of disk galaxies. Support for this position includes:
Studies of proto-elliptical merger remnants like NGC 7252 [56] and models of disk galaxy mergers reproducing such objects [57, 58].
H
line strengths in
some ellipticals indicating recent star formation
[59].
``Fine structures'' in elliptical galaxies correlating with residuals in luminosity-color and luminosity-line strength relations [60, 61].
These results enable us to trace the gradual assimilation of recent
merger remnants into the larger population of field ellipticals. But
such evidence is not available for cluster ellipticals, which seem to
be a more homogeneous population (eg.
[62]). Studies of the
fundamental plane out to z 
0.8 indicate that cluster
ellipticals evolve passively and probably formed the bulk of their
stars at z 
2
[63].
Thus cluster ellipticals are
unlikely to show the signs which betray aging merger remnants in the
field.
Counter-rotating or otherwise decoupled ``cores'' are probably the clearest signs that cluster ellipticals were formed by ancient mergers [64, 65]. High-resolution imaging shows that kinematically distinct nuclear components are usually disks [64, 66]. Such disks typically have high metal abundances [67] and low velocity dispersions [68]. These properties indicate that they formed dissipationally during major mergers [69, 70]; merger simulations producing counter-rotating nuclear gas disks back up this hypothesis [71].
The nature of the mergers which formed cluster ellipticals is unknown; often invoked are highly dissipative encounters of gaseous fragments. But the existence of counter-rotating disks indicates that the penultimate participants can't have been very numerous or very gassy. If many small objects coalesced, the law of averages would make counter-rotation extremely rare. And counter-rotation is unlikely to arise in essentially gaseous mergers since gas flows can't interpenetrate.
Once formed, kinematically distinct disks would be easily disrupted by dissipationless mergers [72]. Thus observations of such structures in cluster galaxies imply that few mergers occur once a cluster has virialized. This is entirely plausible on dynamical grounds since encounters at speeds higher than about twice a galaxy's internal velocity dispersion don't result in mergers [73].