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

6.3. Causes of Activity

There are several proposed models that offer appealing but incomplete scenarios for the origin and evolution of galactic activity. Examples include accretion flows, interactions and mergers, gas motions in nonaxisymmetric potentials, and periodic ejection followed by infall (hot gusts). Taking the observations and models as a whole and exercising the prerogative of authors of Annual Review articles to end their papers in unmitigated speculation, we now attempt to develop a coherent view of the origin of galactic activity and, hence, to point out additional areas for future research.

In the conventional picture two necessary (and perhaps sufficient) conditions for the onset of nuclear activity are the presence of a nuclear "monster" and a supply of "food" (Gunn 1979). The monsters are presumed to be long-lived, compact, and presumably massive objects in galactic nuclei.

Do all galaxies harbor monsters? We would argue that the lack of activity in the nuclei of late-type galaxies, despite the apparent abundance of food (as evidenced by the on-going star formation), means that monsters do not live in such galaxies. Instead, the striking incidence of activity in bulge-dominated systems strongly indicates that the core of a bulge is the monster's favored lair and that monsters are fairly common in systems with massive bulges. Moreover, monsters need not all be alike. Some galaxies may have multiple monsters in orbit about one another whose orbital axis can be an important preferred direction.

Activity is more common in luminous (massive) galaxies, although the dependence is stronger for radio sources than for emission-line nuclei (i.e. Seyfert galaxies and quasars). The form of this dependence suggests that monsters are fed more continuously in the more massive galaxies. Dominant-cluster galaxies would be expected to receive an almost continual supply of food from accretion flows or cannibalization of cluster galaxies. Less massive galaxies may erupt intermittently (Bailey & Clube 1978), perhaps suggesting that food for the monster is only occasionally plentiful. Yet, low-level activity is present even between outbursts in at least a third of all early-type galaxies, so the monsters must get frequent snacks.

The role of galaxy rotation in feeding or fostering a monster needs clarification; despite the naive expectation that food can be more easily supplied to the nucleus in a low angular momentum system, the most menacing monsters are seemingly found in galaxies in which rotation is dynamically important. Perhaps a disk component, even though weak, is required to supply food at a large rate. Better statistical data on radio luminosity and galaxy rotation are needed.

The evidence seems to lend support to a number of currently popular schemes to feed the monster. Mergers and interactions may be generally relevant to nuclear activity, since they can obviously supply the monster with fresh food in the form of gas (Gunn 1979) or, if digestible, stars sent inward on plunging orbits. Interactions may also trigger the growth of oval distortions in the galaxy's gravitational potential leading, eventually, to the deposition of gas at the nucleus (Simkin et al. 1980). Finally ellipticals may periodically accrete their own ejecta from previous hot gusts, a process which conceivably could be aided by the "back-pressure" supplied by the gaseous medium found in compact groups of galaxies.

This description of the care and feeding of galactic activity carries with it the usual obligatory statement "More and better observations are needed." Few of the observational underpinnings of this scenario are based on proper, statistically secure results. Many facets of the problem are complicated by its multivariate nature, and the truly important independent variables are yet to be identified.