A wide variety of physical mechanisms may be affecting the structure and dynamics of galaxies in clusters. Which of these mechanisms are we certain are operating and which are still just possibilities?
The cause of the morphology-density relation is still unknown. Nearly all the mechanisms listed in Section 2 are still candidates. It is generally believed that a local mechanism is responsible for the effect, but this review argues that global conditions may be more important than local conditions in determining the morphological type of a galaxy. This is based on the fact that the correlation of the morphology with cluster-centric distance is as good as with local galaxy density. The low fraction of ellipticals in compact groups, which have very high local galaxy densities but low binding energy, also suggest that global conditions may be the more fundamental parameter.
There is good evidence that galactic cannibalism is occurring in D and cD galaxies, as well as tentative evidence that the extended envelopes of cD galaxies arise from the accumulated tidal debris of the cluster (whether from galaxy-galaxy interactions or the tidal shear from the mean cluster potential is unknown). HI deficiencies for central cluster galaxies, and smaller HI radii, are clearly established, but whether ram-pressure stripping or gas evaporation is responsible is unknown.
The central stellar velocity dispersions of galaxies do not appear to be greatly affected by the cluster environment, although caution is advised if differences of only a few percent are important (e.g., determination of peculiar motions). Velocity dispersion profiles in cD galaxies appear to rise in their outer regions, suggesting the accretion of material from galactic cannibalism or tidal stripping. There is tentative evidence that rotation curves for spiral galaxies near the centers of clusters are falling, but no clear indication of what mechanism is responsible. While the removal of the dark matter halo by tidal stripping seems like the simplest explanation several other possibilities exist.
As cosmological simulations become more sophisticated they offer the possibility of answering some of the questions posed in this review. In the past, the emphasis has been on cD formation. Several recent simulations are now examining the role of merging halos, and beginning to add dissipational processes to the calculations. These simulations offer the hope that we can begin to address details such as the evolution of morphology, size, and distribution of mass in the not too distant future.
There is no conclusive evidence that any of the forms of late evolution we have discussed is a dominant mechanism for most of the galaxies in the cluster. Many people advocate the idea that most of the interactions happened early in the evolution of the cluster, and the recent interactions have only introduced relatively minor changes since that time.
Some of the key observations that need to be made are: determination of positions and relative velocities for a large sample of both D and cD galaxies (i.e., are they in local enhancements or the dynamical center of the cluster), measurements of intracluster light in a large sample of groups and clusters with different morphologies, velocity dispersion profiles in the outer regions of elliptical galaxies (both stripped galaxies and cD galaxies), and rotation curves for spiral galaxies in groups and clusters. Of particular interest for the sponsor of this workshop is the potential for using the Hubble Space Telescope to see deeper, and hence farther back in time, to an epoch when many of these processes are occurring. As Dressler (1984a) states, we are all "looking forward to looking back".
I would like to thank Vera Rubin, Duncan Forbes, Steve Zepf, Diane Gilmore, Marylin Bell, and Kent Ford who collaborated on several projects reported in this paper. I would especially like to thank Alan Dressler for sending me his computer files containing his measurements for over 6000 galaxies in the 55 clusters he studied and to thank Vera Rubin for permitting me to present Figure 7 prior to publication. Finally, I would like to thank Alan Dressler, Mike Fall, Mike Fitchett, Paul Hickson, Bill Oegerle, and Steve Zepf for commenting on early versions of this paper, and the Astrophysical Journal for allowing the use of the figures which originally appeared in their publication.