Annu. Rev. Astron. Astrophys. 1982. 20: 547-85
Copyright © 1982 by . All rights reserved

Next Contents Previous

4.1 Cluster Dynamical Evolution and the Growth of cD Galaxies

cD galaxies are the largest and most luminous stellar aggregations. Bautz & Morgan (1970) defined a D galaxy as having "an elliptical-like nucleus surrounded by an extensive envelope," and cD galaxies as "outsanding, large, luminous D galaxies." For the extreme cD galaxy in Abell 1413, Oemler (1976) has shown that a low surface brightness envelope extends to a radius greater than 2 Mpc.

Ostriker & Tremaine (1975) and White (1976b) postulated that cD galaxies grow through galactic cannibalism as dynamical friction causes cluster galaxies to spiral into the centrally located giant, where they are disrupted. Richstone (1976) showed that the envelope, of a cD galaxy could result from galactic halos collisionally stripped from their parent galaxies, which become bound instead by the cluster potential well. Dressler (1979) obtained spectra of the envelope of the giant cD galaxy in Abell 2029, which demonstrated that the velocity dispersion increases with radial distance. This result and the cD's luminosity profile were interpreted through a three-component model composed of a normal elliptical galaxy (Msun / Lsun ~ 10) with a luminous halo (Msun / Lsun ~ 35) at the center of a dark (Msun / Lsun gtapprox 500) cluster-binding super-structure. Dressler argued that the halo material, with its intermediate M/L and velocity dispersion, could be provided by the cannibalized remains of the luminous portions of massive bright galaxies, while the high M/L component could result from the stripped dark halos of cluster galaxies.

Numerical simulations of galaxy mergers suggest that both merging and tidal stripping are most rapid in small groups of galaxies, which are the first relaxed units in hierarchical cluster evolution. Carnevali et al. (1981) studied groups of 10-40 galaxies and found the development of a "merging instability," a runaway of stripping and merging (also see Dekel et al. 1980). Aarseth & Fall (1980) and Carnevali et al. showed that as the cluster relaxes, the frequency of galaxy mergers declines. Collisions and mergers become less likely because galaxy halos containing the bulk of the galaxy mass have been disrupted so that galaxies present smaller cross sections, and because the velocities of encounter are larger, as appropriate for an evolved rich cluster rather than a subcluster.

Bright cluster galaxies will be preferentially cannibalized by the cD, since these massive galaxies will tend to lose energy and settle to the cluster core as the galaxy distribution evolves toward equipartition. In accord with this hypothesis, Sandage & Hardy (1973) found that the absolute magnitude of the first-ranked galaxy is brighter on average by 0.6 mag in Bautz-Morgan Class I clusters than in Class III clusters, and that the second- and third-ranked galaxies are fainter by 0.5 mag in Class I compared to Class III. In addition, the luminosity functions of cD clusters studied by Dressler (1978) show a depletion of bright galaxies. The merger hypothesis also is supported by frequent observations of multiple nuclei in cD galaxies (Hoessel 1980), which Hausman & Ostriker (1978) suggested are the remnants of cannibalized galaxies.

The observations show a range of dynamical evolutionary stages for clusters with X-ray-dominant central galaxies, ranging from the unevolved Virgo cluster and other spiral-rich systems to evolved clusters like Abell 85. The evolutionary phases are distinguished by galaxy properties and cluster dynamical indicators. The observations of Oemler (1976) and Thuan & Romanishin (1981) suggest that the growth of the dominant galaxy may be traced by the development of the extended optical envelope. Figure 4 shows the optical surface brightness profiles for four dominant galaxies and graphically illustrates the increase. The absence of multiple nuclei also may distinguish the most evolved cD galaxies from less evolved systems. In Hoessel's observations, none of the cD galaxies in hot XD clusters were reported to have multiple components. These include Abell 85, Abell 399, Abell 401, Abell 1795, and Abell 2029. However, the brightest galaxy in each of the five XD clusters with cool gas (Abell 400, Abell 1991, Abell 2063, Abell 2199, and Abell 2634) has multiple components. Although the number of clusters is small, this rule, if confirmed, could be explained by the longer time required for less massive galaxies to spiral into the central galaxy in the more evolved, high-temperature clusters, where the more massive galaxies already have been cannibalized and their nuclei disrupted.

As the dynamical evolution of a cluster progresses, subclusters (possibly with coalesced galaxies) will merge. Thus young cD galaxies in spiral-rich clusters may continue to accrete galaxies and galactic debris as the cluster evolves. Cluster galaxies that continue to grow may become giant cD galaxies, while others, depending on their location in the cluster and their velocity with respect to the cluster mean, may grow at much slower rates br even lose their halos). The bright elliptical galaxies in Coma may be examples of galaxies that did not mature into giant cDs.

Figure 4

Figure 4. The surface brightness in mag arcsec-2 is plotted against r1/4 in units of (kpc)1/4 for four cD galaxies studied by Oemler (1976). The inner region of each galaxy has been fitted to a de Vaucouleurs' law (solid line). This figure is adapted from Thuan & Romanishin 1981).

The evolutionary hypothesis implies a spectrum of dominant, central galaxies with increasing optical envelopes. This reduces the usefulness of the distinction between D and cD galaxy classifications. Historically, cD galaxies were selected visually from photographic plates. The optical observations required to quantify the discussion of bright, centrally located cluster galaxies, e.g. spatially resolved spectra and surface brightness profiles, are difficult to obtain. However the X-ray observations permit the determination of the mass distribution in the central regions of clusters and therefore facilitate cluster classification.

Next Contents Previous