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

3.3. Absolute Magnitude

RADIO EMISSION     The dependence of radio luminosity on the optical luminosity of radio galaxies was discussed in detail in Section 3.1. To summarize, for radio power 10²⁵ W Hz^-1 at 1.4 GHz, luminous galaxies are far more likely to produce a radio source (probability L_opt^1.5) as discussed by Auriemma et al. (1977), Meier et al. (1979), and Dressel (1981). Below this radio power, more optically luminous galaxies are still more likely to be radio sources, but the dependence is weaker and/or more complex.

Similar patterns are found in other morphological types of galaxies. Dressel (1981) concludes that SO galaxies show a strong correlation between probability of radio emission and absolute magnitude. Unlike E galaxies, the radio emission of SOs tends to be dominated by a compact core. For spiral galaxies, the most comprehensive work is that of Hummel (1981a) who shows that nuclear radio power is directly proportional to the luminosity of the parent galaxy. The same trend holds for total radio emission, but disk emission, which is probably unrelated to galactic activity, frequently dominates the observed radio flux in this case (Hummel 1981a, Gioia et al. 1981). Finally, Heckman (1980b) detected nuclear radio emission in 13 of 20 galaxies with M_B < - 20, and in only 2 of 12 with M_B > - 20.

EMISSlON LINE GALAXIES     The luminosity function of Seyfert galaxies has been investigated by Huchra & Sargent (1973) and Terebizh (1980). Both studies conclude that the probability of finding a Seyfert nucleus increases strongly with absolute magnitude, and that almost all galaxies with M_B < - 23 are Seyferts. However, this result is partly circular; in luminous Seyfert galaxies nuclear light dominates the total luminosity (de Bruyn & Sargent 1978, Heckman et al. 1978, Yee 1981), especially in type 1 Seyferts (Weedman 1977b) for which the nuclear luminosity consists largely of nonstellar continuum. This is supported by Terebizh's data. Type 2 Seyferts (narrow permitted lines), for which the nonstellar continuum is very weak (Koski 1978, de Bruyn & Sargent 1978), show little, if any, correlation between absolute magnitude and the probability of becoming a Seyfert galaxy. In any case, no Seyferts have been identified in a galaxy fainter than M_B ~ - 18. Were such objects common they would have turned up in existing surveys as very low-z quasars.

Heckman (1980b) concluded that LINERs, which are found almost exclusively in early-type galaxies, do not preferentially occur in luminous galaxies. Indeed, some examples occur in faint (M_B ~ - 18.5) galaxies. Stauffer (1981) investigated both LINERs and their high-excitation counterparts (HINERs). LINERs are found to occur in galaxies that are typically ~ 1 mag brighter than galaxies with HINERs. (M_B ~ - 20.5 vs. M_B ~ - 19.5.)

QUASARS     Studies of galaxies surrounding QSOs are made difficult if QSO distances, as we shall now assume, are cosmological. In fact, only recently is there any evidence whatsoever that the fuzz around QSOs consists of a galactic disk or bulge (Morton et al. 1978, Hutchings et al. 1981) or even starlight (Cowie et al. 1981). Absolute magnitudes of the fuzz in several QSOs average M_R ~ - 21.8 ± 0.8 (H₀ = 60 km s^-1 Mpc^-1), consistent with galaxies of typical luminosities (Wyckoff et al. 1981). Interestingly, the magnitudes range from those like cD galaxies in 3C 273 (M_R ~ - 23.2) to a possible unresolved dwarf in PKS 1510-089 (M_R > - 17.4). Aside from the last faint example and perhaps a few other unresolved QSOs discussed by Hutchings et al., these results are consistent with what is known about Seyfert galaxies. Furthermore, looking at the data of Wyckoff et al., there appears to be no correlation between galaxy magnitude and nuclear luminosity, spectrum of the nucleus, or the shape and size of the radio source associated with the QSO.

BL LACS     The spectroscopic studies of Miller et al. (1978) suggest that the host galaxies of BL Lacs are quite luminous. More recent data (Miller 1981) show that the average magnitude of the host is M_R ~ - 22.7 ± 0.6 (H₀ = 60 km s^-1 Mpc^-1), or intermediate between the hosts of QSOs and Seyferts on the one hand and first-ranked cD galaxies in rich clusters on the other. Thus it would seem that BL Lac host galaxies are a relatively luminous population.

In summary, it seems that the probability of activity in galaxies and its degree depend in some way on the luminosity - and, by implication, the mass - of the associated galaxy. This statement is valid for almost every type of galactic activity and morphological class of galaxy. However, with the possible exception of the most luminous galaxies (M_R < - 23.5), most galaxies are not active. Luminosity (mass?), then, is a statistical indication of propensity for activity.