5.7.6. Evolution of cooling flows and active galaxies
One important question is what effect the evolution of clusters has on these cooling flows. An important type of evolution is the merging of subclusters; the observed double clusters (Section 4.4.2) may be systems undergoing this process. Such a merger will probably heat up the gas and disrupt any existing cooling flows (McGlynn and Fabian, 1984). Along these lines, Stewart et al. (1984b) speculate on the possibility that the Coma cluster, with its pair of D galaxies, is the result of such a merger of two subclusters. Perhaps each of the subclusters had a cooling flow, and accretion-driven star formation produced the two D galaxies. The merger would disrupt the flow, which would explain why Coma apparently lacks a cooling flow despite its high X-ray luminosity. The heating of the gas during the merger might explain the unusually high temperature of the intracluster medium in Coma (Section 4.5.1). In this way, accretion could have formed the central galaxies in clusters that do not currently have accretion flows.
Although this review is concerned with clusters, elliptical galaxies in irregular clusters and groups also appear to have considerable quantities of X-ray emitting gas, which may also form cooling flows (Section 5.8.3). These cooling flows are probably powered by stellar mass loss within the galaxies, rather than accretion of intracluster gas. For giant ellipticals, stellar mass loss rates of 1 M / yr are expected. In the past, rates of stellar mass loss in elliptical galaxies were probably higher than they are today (Section 5.10.1) Since cooling flows in cluster centers were probably weaker in the past, this suggests that individual early type galaxies may make an important contribution to the emission from clusters at high redshift (Fabian et al. 1986a).
As was noted above, accreting central galaxies are often strong radio sources, indicating that some small portion of the cooling gas is accreted by the nucleus of the galaxy. Quasars appear to be galaxies with extremely luminous nonthermal nuclei. It is possible that they are also powered by accretion through cooling flows (Sarazin and O'Connell, 1983; Fabian et al., 1986a). It is possible that the evolution of cooling flows in clusters or individual galaxies may help to determine the evolution of active galactic nuclei (Section 5.10). Active galaxies often show strong optical line emission; perhaps the lower excitation line emission seen in some active galaxies may be due to thermal instabilities in cooling flows, the mechanism that produces the line emission in nearby cooling flow galaxies.
In summary, it appears possible that most of the accreted gas in cooling flows is converted into low mass stars, and that this accretion-driven star formation may provide much of the luminous or nonluminous mass of central galaxies in clusters with cooling flows. This process may compete with mergers and tidal effects as a mechanism for the formation of cD galaxies (Section 2.10.1). It may also affect the evolution of active galaxies.