Published online by Cambridge University Press: 30 March 2016
Typical orbital periods for galaxies in rich clusters are observed to be several billion years. Consequently these large systems have had little time to come to equilibrium, and indeed if they formed from small initial density perturbations they must have taken a substantial fraction of the age of the Universe to separate from the Hubble flow and recollapse (Gunn and Gott 1972). Despite the relatively short time available, dynamical effects can cause appreciable evolution both in the distribution of galaxies within clusters and in the observable features of individual galaxies. The overall aspect of a cluster changes as violent relaxation irons out in homogeneities and establishes a relaxed core-halo structure. At the same time collisional relaxation causes massive galaxies to lose energy to lighter objects in an attempt to reach equipartition of kinetic energy, and to spiral towards the cluster centre. These processes are relatively easy to analyse in any particular scheme for cluster formation, and below we concentrate on comparing their predicted effects with observation.