Annu. Rev. Astron. Astrophys. 1982. 20: 431-468 Copyright © 1982 by Annual Reviews. All rights reserved

2. THE INFLUENCE OF THE EXTRAGALACTIC ENVIRONMENT

2.1. Cluster Environment

Clusters have long been suspected as sites favorable in some ways to the formation and development of galactic activity. Certainly the high frequency of activity in cluster-centered cD galaxies is well known, as is the likelihood of an intracluster medium (ICM) with a substantial pressure (see Miley 1980 for summary). Generally, and perhaps surprisingly, cluster galaxies other than cDs (which are extremely massive and probably unique to the cluster environment) show little if any preferential propensity for activity over their more Isolated counterparts.

RADIO LUMINOSITY    Bright radio emission is a sure and conveniently observed sign of galactic activity. The probability that a galaxy in a given range of absolute magnitude has a radio luminosity within specified limits is known as the bivariate radio-luminosity function. Auriemma et al. (1977) and Perola et al. (1980) have shown that the bivariate luminosity function of field and rich-cluster galaxies does not differ by more than a factor of two. Adams et al. (1980) reanalyzed the data of Auriemma et al. and reached a slightly different conclusion; Luminous cluster galaxies are stronger radio sources than luminous field galaxies, but for faint galaxies the opposite is true. This may partly reflect the unusual nature of the very luminous dominant-cluster galaxies (the cDs).

Longair & Seldner (1979) have found that the strongest radio galaxies tend to be found in regions intermediate in galaxy richness between Abell clusters and randomly selected fields, whereas the distribution of less luminous radio sources is statistically similar to the distribution of randomly selected galaxies. Longair & Seldner included spiral as well as elliptical galaxies in their "control sample" whereas radio galaxies are generally found in ellipticals and ellipticals are preferentially located in clusters, so their results may be misleading. Stocke (1979) and Longair & Seldner both find that radio sources in dense galactic environments tend to have a more complex morphology than relatively isolated radio sources.

EMISSlON LINE LUMINOSITY    Gisler (1978) conducted a comprehensive survey of a heterogeneous sample of 1316 galaxies for line emission. He concluded that emission-line galaxies (ELGs) are much less common in dense than in groups, associations, and the field. The statistics for the very active ELGs (i.e Seyfert galaxies) are less compelling in Gisler's sample, although a tendency for such galaxies to avoid the rich-cluster environment is apparent. The same conclusion, though not statistically firm, has been drawn by Adams (1977) and van den Bergh (1975). Furthermore Hine & Longair (1979) find that 3CR radio galaxies with strong emission lines behave similarly, and Roberts et al. (1977) have shown that QSOs are less frequently associated with rich Abell clusters than are radio galaxies (which have much weaker emission lines).

The literature is rich in exceptions, however, Seyfert galaxies detected in clusters include NGC 4388 (Phillips & Malin 1981) and perhaps NGC 4235 (Abell et al. 1978) in Virgo, NGC 1365 (Veron et al. 1980) and NGC 1386 (Phillips & Frogel 1980) in Fornax, IC 4329A (Wilson & Penston 1979, Pastoriza 1979) in Klemola 27, IC 1182 = Mrk 298 (Bothun et al. 1981b) in Hercules, and of course, some cD galaxies such as NGC 1275 in Perseus. Hawley & Phillips (1978) find two extremely luminous Seyferts in loose clusters. Stauffer (1981) suggests that the frequency of Seyfert-like ELGs in Virgo is not significantly different than the field. Finally, low-redshift quasars sometimes turn up in clusters (e.g. Wehinger & Wyckoff 1978, Phillips 1980, Stockton 1980, Wyckoff et al. 1980, and Steiner et al. 1981).

These counterexamples notwithstanding, the apparent avoidance of ELGs in rich clusters may be interpreted in several ways. There may be an intrinsically low incidence of activity. Or, active galaxies may not be able to generate observable radiation even if they are active. Gisler (1978) suggested that the sweeping of gas by a dense, hot ICM in rich clusters may explain the low frequency of detectable EGNs in this environment. This conclusion, however, is not fully supported by several lines of evidence. First, an extensive survey of H I in cluster galaxies by Bothun (1981) failed to find any statistically significant sweeping of gas in cluster spirals except at the center of Coma. Second, studies of radio-source morphology and calculations of stripping efficiency by the ICM both indicate only the loosely bound gaseous "halo" of a galaxy can be stripped (e.g. Jones & Owen 1979, Wilkinson et al. 1981). X-ray observations suggest that some cD galaxies are accreting the ICM, rather than losing it (see below). Other nondominant cluster galaxies can also have X-ray halos (Jones et al. 1979, Forman et al. 1979), showing that efficient sweeping is not ubiquitous in clusters.

BL LAC OBJECTS    Butcher et al. (1976) suggested that 3C 66A might be located near the center of a rich cluster at z ~ 0.37. Moreover, NGC 1275, which displays many of the defining characteristics of a BL Lac (Angel & Stockman 1980), is near the center of the Perseus cluster.

However, most BL Lacs are evidently not located in the central regions of rich clusters. Visvanathan & Griersmith (1977) find that AP Lib may be in a poor group of galaxies at z = 0.0487, while Baldwin et al. (1977) reach similar conclusions regarding 1400+162 at z = 0.244. We have examined the locations of six Northern Hemisphere BL Lacs at z < 0.10 with respect to Zwicky and Abell clusters and find that Mrk 421, Mrk 501, 3C 371, and BL Lac lie near the edges of mostly open clusters. On the other hand, Mrk 180 (z = 0.044) and IZw 1757+50 are apparently not in any cluster.

DOMINANT-CLUSTER GALAXIES    The brightest and most massive galaxies known - the cD or dominant-cluster galaxies - are situated at the geometrical and probably gravitational centers of clusters. It has been frequently suggested that cD galaxies are growing by occasionally cannibalizing a smaller cluster member (e.g. Hoessel 1981, McGlynn & Ostriker 1980) and by accreting the nearby hot ICM (Fabian & Nulsen 1977, Cowie et al. 1980, Binney & Cowie 1981, Mushotsky et al. 1981). This infalling matter, particularly the ICM, might be available to provide 10² M yr^-1 of fuel to the nucleus. As discussed later, there is fairly strong evidence for these "cooling accretion flows."

As Burns et al. (1981c) have emphasized, cD and related galaxies are unusually active in the radio compared to other cluster ellipticals, whether located in poor groups or rich clusters. However, given the well-established, strong, and positive correlation between the likelihood of radio emission and galaxy absolute magnitude (see below), it is not clear that these galaxies are unusually active for their absolute magnitude. While the luminous X-ray sources and emission-line nebulosities (Burns et al. 1981a, Jones et al. 1979, Ford & Butcher 1979, Heckman 1981) frequently associated with these superluminous galaxies might be taken as signs of activity, the emission is spatially extended and presumably arises as the hot gas is being steadily accreted onto the galaxy, a process believed to be unique to these objects. Thus these X-ray galaxies should not be considered genuinely active unless their nuclei exhibit the usual symptoms of activity. Nonetheless their central location in the cluster and their large masses contribute to their ability to accrete material for the nucleus.