The energy source of AGNs is most probably a rotating accretion disk about a massive black hole. If the galaxy has a black hole at or near its nucleus, a continual source of mass delivered to it, with nearly zero angular momentum on the scale of the galaxy, is required. Several recent studies agree in confirming earlier ideas that gravitational interaction with a nearby companion or slowly orbiting galaxy is often the mechanism by which this fueling process occurs (Adams 1977; Simkin, Su and Schwarz 1980; Kennicutt and Keel 1984; Dahari 1984). However, very strongly interacting and apparently disrupted galaxies tend not to be Seyferts (Dahari 1985). Evidently in these cases the perturbations or interactions are too strong to bring appreciable amounts of mass to the center with near-zero velocity.
Some small, loose groups of galaxies also contain several AGNs; evidently in such groups the interactions can have the effect of delivering gas close enough to the nucleus to refuel the accretion disk (Kollatschny and Fricke 1985). On the other hand, rich, dense clusters contain practically no AGNs, again indicating that strong or frequent interactions inhibit, rather than promote, the formation of AGNs (Dressler, Thompson and Schechtman 1985).
There is strong evidence for the presence of a black hole of mass M 5 x 106 M in the nucleus of our own Galaxy (Genzel et al. 1985). This is much smaller than the black-hole masses necessary to provide, through their accretion disks, the energies radiated in typical observed AGNs. Though we cannot observe the nucleus of our Galaxy directly in the optical wavelength region, because of the very great extinction in the galactic plane, far infrared and radio measurements confirm that it is only very mildly active. In the nucleus of M 31 there is evidence for a black hole of mass M 107 M (Dressler and Richstone 1988, Kormendy 1988). There is little if any evidence of an AGN in it (Munch 1960; Rubin and Ford 1986). Yet the fact that both these galaxies, whose nuclei we can observe in such great detail, contain small black holes, suggests strongly that there is a continuous range of black hole masses, from QSOs and Seyfert galaxies, probably down through LINERs, through lower and lower luminosity AGN-like nuclei in which similar physical processes are going on, but at a level too faint for us to detect. Probably the simplest working hypothesis is that every galactic nucleus contains a black hole, and that those which have unusually massive black holes, that are unusually well supplied with fuel, are observed as AGNs.
I am very grateful to W.G. Mathews, J.S. Miller, R.A. Shaw, R.W. Goodrich, R.W. Pogge and S. Veilleux for many stimulating discussions of the subjects treated in this review paper, and to W.W. Morgan and R.D. Dreiser for permission to reproduce Figure 1 from their paper. I am also grateful to the National Science Foundation for partial support of this research under Grant AST 86-11457.