CLUSTERS OF GALAXIES, COMPONENT GALAXY CHARACTERISTICS AUGUSTUS OEMLER, JR. A few percent of galaxies are members of rich clusters, containing hundreds of galaxies in volumes a few megaparsecs across. The existence of such clusters was recognized early in this century, and they became important tools in pioneering studies of galaxies by Edwin Hubble, Fritz Zwicky, and others. Because all cluster members are at the same distance from us, one redshift (which determines the distance by Hubble's law) suffices to provide the distance to all. Cluster members are, therefore, the easiest galaxies to study, and much of what we know about galaxies is derived from them. However, what is learned about cluster galaxies can be applied to galaxies in general only if the former are typical. It was soon realized that they are not: Cluster galaxies differ in a number of ways from those which lie outside of rich clusters, in what is often termed the field. Although this fact limits their usefulness as exemplars of the general galaxy population, it provides a most important tool for understanding the processes of galaxy formation and evolution. The differences between cluster and field galaxies are presumably due to the differences in the environments in which they occur. This, by itself is significant, and shows that galaxies have a sensitivity to their surroundings that stars, for example, lack. If one can understand how the environment has shaped cluster galaxies, one will have gone a long way towards understanding the processes governing the formation and evolution of all galaxies. THE GALAXY CONTENT OF CLUSTERS Morphology The most obvious difference between cluster and field galaxies is in their morphological types. The majority of field galaxies are gas-rich spirals and irregulars; gas-poor ellipticals and SOs are outnumbered by about 2 to 1. In contrast, the centers of rich clusters are dominated by the latter types. The few spirals seen are probably chance superpositions: field galaxies or outlying cluster members lying along the line of sight to the cluster center. The shift in galaxy populations is not abrupt: There is a continuum from spiral-dominated to SO- dominated populations as one moves from the most remote field to the densest cluster centers. The SO-to-spiral ratio also depends on the nature of the cluster: populous, dense, symmetric clusters have fewer spirals than do poorer, more irregular clusters. In addition to this dependence on the global cluster properties, it has been shown that galaxy populations are sensitive to the local environment: The ratio of SOs to spirals increases with the number of neighboring galaxies. There is reasonably good evidence for a continual variation in environmental sensitivity as one moves along the Hubble sequence of galaxy types, from ellipticals, through SOs, to spirals and irregulars. As the density of the environment increases, irregulars disappear first, followed by late-type spirals. The abundance of SOs rises as spirals disappear, but only at the highest densities does the abundance of ellipticals begin to increase. This suggests that the effect of environment is to skew the distribution of galaxies along the Hubble sequence towards one or the other end, but such an interpretation is by no means unavoidable. As important as the differences between cluster and fields populations are the similarities: Cluster and field galaxies are both drawn from the same Hubble sequence; cluster ellipticals look approximately like field ellipticals, and cluster Sc galaxies like field Scs. GALAXY LUMINOSITIES If environment influences what kind of galaxies form, it might also influence how bright they are. Furthermore, the luminosity function of ellipticals, at one end of the Hubble sequence, differs markedly from that of irregulars, at the other end. For both these reasons, one might expect the luminosity function of cluster galaxies to differ from that of field galaxies. If such differences exist, however, they are quite subtle. The gross shape and characteristic luminosity of the cluster-galaxy luminosity function is almost indistinguishable from that of field galaxies. The small differences which do exist are attributable to processes involving the brightest few galaxies, as discussed below. THE STRUCTURE OF CLUSTER G~IES In the cores of populous clusters, the average separation between galaxies is less than 10 times their size. One would, therefore, expect frequent close encounters between galaxies, during which their mutual gravitational forces could affect their internal structure. The tidal forces of encounters can remove the outer envelopes of the galaxies. In the rarer event of interpenetrating collisions, the process of dynamical friction can remove enough of the orbital energy of the galaxies to cause them to merge. Direct evidence for tidal perturbation of cluster galaxies is difficult to find. Optical evidence of tidal interactions is much rarer in clusters than in small groups. This is not in itself unexpected. It can be shown that most tidal stripping of galaxy envelopes should have occurred early in the history of the cluster. Furthermore, tidal encounters become less effective when the relative velocities of the galaxies are much higher than the velocities with which their stars orbit within them; in most rich clusters this ratio is of order 2-10. The only clusters in which many tidal encounters can be seen are that rare subset with irregular structure and a large population of spirals. This suggests that these clusters are, in fact, quite young, and have yet to reach equilibrium. In their present state, the relative velocities of neighboring galaxies may be sufficiently low to permit effective tidal encounters. There is, indeed, very little evidence that significant tidal stripping ever occurred in most clusters. A comparison of the structure of elliptical galaxies in clusters and the field shows little sign that the outer envelopes of the former have been tidally truncated. There is somewhat better evidence for the occurrence of mergers, driven by dynamical friction. The brightest member of a cluster, which is always a giant elliptical and usually located in the cluster center, often has a number of satellite galaxies located within its outer envelope. Observations of the velocities and positions of the satellites, and of tidal distortions due to interaction with the brightest cluster member, suggest that some fraction of these satellites may be merging with the central galaxy. Such a process will cause the central galaxy to grow at the expense of the smaller galaxies. There is also some evidence of the results of this process in the luminosity functions of the galaxies in clusters with a dominant central galaxy. Although the dominant cluster member is always an elliptical, it is not always a normal one. W. W. Morgan and his collaborators first pointed out that some are sufficiently unusual to warrant a special class, which they named cD. cD galaxies are characterized by elliptical-like centers on which are superimposed very extended, low surface brightness envelopes. The total size and luminosity of some envelopes is remarkable, with diameters greater than 2 Mpc (10 times that of a large elliptical) and total luminosities 10 times that of the underlying central galaxy. These galaxies are clearly products of the cluster environment in a way unique for galaxies. They only occur as brightest members of clusters. Their envelopes are so extended that the outer parts must be considered part of the cluster as a whole, rather than of the central galaxy. Most telling, the total luminosity of the envelope is a unique function of the total cluster richness. It is most likely that cD galaxies began as normal luminous ellipticals which, because of their mass, sank to the center of their clusters during the period of cluster formation. Once there, they were in a favorable position to accrete material from within the cluster. This material might be smaller galaxies, which were captured by dynamical friction, or it might be tidal debris stripped from other galaxies. Neither possibility is entirely consistent with what is known about cluster galaxies. Material from mergers will be distributed over a much smaller area in the cluster center than that covered by the cD envelopes. Tidal debris would be expected to have a broader distribution, but, as mentioned earlier, there is little evidence that cluster galaxies have lost much of their outer envelopes to tidal stripping. THE CONTENTS OF CLUSTER GOES The difference between the galaxy populations of the field and of clusters is essentially a difference between galaxies which contain significant amounts of gas and young stars and those which do not. This is probably the only difference between the spirals which dominate the field population and the SOs which dominate clusters. The lack of young stars is confirmed by detailed study of the colors and spectral energy distributions of cluster galaxies, which show that most contain only old stars. A similar difference in the gas content is also well documented. This persists when one examines each Hubble type separately. SOs seldom contain much gas, but field SOs contain substantially more than do those in clusters. Those spirals which survive in clusters are much depleted in gas compared to their field counterparts. Interestingly, this gas depletion seems to be confined to the outer parts of a galaxy, beyond the region in which star formation normally occurs. In the inner parts, the gas content appears to be normal, which is why the galaxies are still able to make the new stars which label them as spirals. One cannot find, among the now-quiescent elliptical and SO cluster galaxies, much evidence for recent star formation. However, examination of clusters at earlier epochs shows that clusters contained many more star-forming galaxies in the recent past. Because of the finite velocity of light, distant clusters are observed at earlier times, which allows one to examine the properties of cluster galaxies over about the last third of the age of the universe. The fraction of cluster members whose colors indicate recent star formation has decreased from about 25% five billion years ago to a few percent today. We do not yet know whether these blue galaxies are identical to the spirals seen in nearby clusters. Their distribution within the cluster is similar, and there are some indications that they are disk galaxies, rather than star-forming ellipticals. However, their spectral energy distributions suggest that a significant fraction may have undergone large bursts of star formation shortly before the epoch at which they are observed. This is in contrast to today's spirals, in which star formation appears to have decreased at a slow and steady rate throughout their lifetimes. Large bursts of star formation are rare among galaxies today, except for those which are interacting. These observation suggest a fundamental evolution in the cluster environment, or in the properties of galaxies, over the last third of the age of the universe. WHY ARE CLUSTER GALAXIES DIFFERENT? There are three possible routes by which the population of cluster galaxies might have come to differ from that in the field. 1. Only elliptical and SO galaxies were formed in the cores of clusters. 2. The spiral galaxies which formed in clusters were of a type which exhausted their gas more rapidly than the typical field galaxy; thus evolving, on their own, into SOs. 3. Cluster populations were originally identical to those in the field, but have been transformed by the cluster environment. The existing evidence is contradictory. The discovery of a blue population in distant clusters would seem to rule out possibility 1, but it is possible that those galaxies have faded into obscurity as their starbursts died, leaving an unchanging primordial population of E/SOs. The observation that the dependence of galaxy type on local density extends to regions beyond the cores of rich clusters favors 1 and 2, and suggests that the density of the environment out of which the galaxy formed is the proximate cause of its morphological type. However, models of the formation of clusters indicate that the environment near a galaxy at its bird, is only loosely related to its environment today, and early differences may not have been great enough to affect the morphological type of a forming galaxy. The strong correlation of global cluster Properties and galaxy types favors possibility 3, but effective mechanisms to alter the properties of galaxies have been difficult to find. It has been suggested that interactions between galaxies, and between galaxies and the hot gaseous intracluster medium, could remove the gas from the disks of spirals and transform them into SOs. Unfortunately, the observations of gas in cluster galaxies mentioned earlier show that gas does not get stripped from those regions of the galaxies in which star formation usually occurs. Furthermore, the existence of population differences outside of the cores of clusters cannot be explained easily by stripping mechanisms. It is possible that the evolution of cluster galaxies and their present properties are the result of a number of independent processes, acting in concert over the entire epoch of galaxy formation and evolution, to produce the populations that we observe today. Additional Reading Bahcall, N.A.(1977). Clusters of galaxies. Ann. Rev. Astron. Ap. 15 505. Dressler, A.(1984). The evolution of galaxies in clusters. Ann. Rev. Astron.Ap.22 185. Haynes, M.P.(1988). Morphology and environment. In The Minnesota Lectures on Clusters of Galaxies and Large Scale Structure, Astronomical Society of the Pacific Conference Series No. 3, J.M. Dickey, ed. Astronomical Society of the Pacific, San Francisco. p. 71. Sandage, A.(1990). Properties of galaxies in groups and clusters. In Clusters of Galaxies, W. Oegerle, L. Danly, and M. Fitchett, eds. Cambridge University Press, Cambridge. Strom, S.E. and Strom K.E.(1979). The evolution of disk galaxies. Scientific American 240 (No. 4) p.72. See also Clusters of Galaxies; Galaxies, Formation; Galaxies, Properties in Relation to Environment.