Annu. Rev. Astron. Astrophys. 1995. 33: 581-624
Copyright © 1995 by Annual Reviews Inc. All rights reserved

Next Contents Previous


The search can be divided into three parts.
1. Look for dynamical evidence of central dark masses. In practice, we look for high mass-to-light ratios. If M/L increases toward the center to values that are several times larger than normal, this is a meaningful clue because the range of M/L values in old stellar populations is small. Is there any escape from the conclusion that M/L is high? If not, then we have discovered a ``massive dark object'' (hereafter MDO). It could be a supermassive BH, or it could be a cluster of low-mass stars, brown dwarfs, stellar remnants, or halo dark matter.
2. Once a few MDOs have been found, we need to improve the observations enough so that alternatives to a BH can be confirmed or ruled out. Proof of a BH requires detection of relativistic velocities in orbits at a few Schwarzschild radii. This is not imminent. But in practice, the most plausible alternative to a BH is a cluster of stellar remnants; it would already be important progress if we could rule out such a cluster on physical grounds. When we say that ``we have evidence for a BH'' in some galaxy, we mean we find such arguments persuasive. When we say ``we have found an MDO'', we are emphasizing the uncertainty in arguments against BH alternatives. Step 2 of the BH search is just beginning; we will say little about it here.
3. BH astrophysics requires more than the detection of a few examples. Ultimately, we want to know the mass function and frequency of incidence of BHs in various types of galaxies. This requires statistical surveys. Surveys are challenging because BH detectability depends on galaxy type. For example, BH detection is easier in rotating disk galaxies than in nonrotating giant ellipticals. Type-dependent biases will be a problem. In practice, we make statistical surveys in parallel with step 2. If MDOs turn out not to be BHs, the demographic results will nevertheless be important to our understanding of galaxy nuclei and AGN activity. BH surveys are reviewed in Section 6.

At present, we are nearing the end of step 1. We have dynamical evidence for MDOs in eight objects. Table 1 lists them subjectively in order from most secure to most weak, separately for objects based on stellar and gas dynamics.

But step 1 is not quite finished. For all objects, the MDO discovery observations have been confirmed by at least one other group. Some of these papers improve significantly on the analysis. But none improves much on the discovery resolution. The next iteration in the search - a substantial improvement in spatial resolution - is in progress.

Most of this paper is a review of step 1. How robust is the claim that MDOs have been detected? What kind of evidence should we believe?

Table 1. MDO census

Galaxy Typea D b
MB, bulgec Fradiod
Lradio / e [Lradio (SgrA*)] MBHf
log[MBH g / Mbulge] References h

M31 Sb 0.7 -18.82 0.03 0.3 3 x 107 -3.31 Dressler &
Richstone 88;
Kormendy 88c;
(also M32); Bacon+94
NGC 3115 S0/ 8.4 -19.90 <0.33 <460. 1 x 109 -1.92 Kormendy &
Richstone 92
M32 E 0.7 -15.51 <4. <54. 2 x 106 -2.60 Tonry 84, 87;
Dressler &
Richstone 88;
van der Marel+94b;
Dehnen 95
NGC 4594 Sa/ 9.2 -21.21 ~102 ~2 x 105 5 x 108 -2.99 Kormendy 88d;
Galaxy Sbc 0.0085 -17.65 7 x 102 1. 2 x 106 -3.78 McGinn+89;
Kent 92;
Genzel+ 94a;
Evans & de Zeeuw 94; Haller+95
NGC 3377 E 9.9 -18.80 <0.5 1 x 103 8 x 107 -2.24 Kormendy+95b

NGC 4258 Sbc 7.5 -17.3 1.8 2 x 103 4 x 107 -2.05 Miyoshi+95
M87 E 15.3 -21.42 ~103 ~7 x 106 3 x 109 -2.32 Sargent+78;
van der Marel 94a

a Morphological type (de Vaucouleurs & Pence 1978; de Vaucouleurs et al. 1991).
b Distance (mostly Faber et al. 1995; for NGC 4258: Faber 1995, personal communication). Values are only slightly modified from Lynden-Bell et al. (1988) and Faber et al. (1989). They are based on a large-scale flow field solution, an assumed Virgo Cluster distance of 15.3 Mpc, and a Hubble constant of H0 = 80 km s-1 Mpc-1. They agree well with distances in Tully et al. (1992). BH mass scales as D; log MBH / Mbulge is independent of D. Demographic conclusions are not sensitive to the distance scale: see versions of Figures 1, 2, 3, and 14 for H0 = 50 km s-1 Mpc-1 in Kormendy (1993, 1994).
c B-band bulge absolute magnitude, based mostly on BT in de Vaucouleurs et al. (1991). Exceptions are M31 (VT = 3.28 from Kent 1987b), NGC 3115 (BT = 9.75 from Capaccioli et al. 1987), NGC 4594 (BT = 8.65 from Burkhead 1986), and the Galaxy (MB adapted from Kent et al. 1991). Galactic absorptions are taken from Burstein & Heiles (1984). Assumed bulge-to-total luminosity ratios are: M31: B/T = 0.24 (de Vaucouleurs 1958), NGC 3115: 0.94 (Capaccioli et al. 1987), NGC 4594: 0.93 (Burkhead 1986), NGC 4258: 0.063 (an approximate value near the bottom of the distribution of B/T values for Sbc galaxies, see Simien & de Vaucouleurs 1986).
d Nuclear radio continuum flux. Wavelengths and sources: M31: 3.6 cm, Crane et al. (1992, 1993a); NGC 3115: 6 cm, Fabbiano et al. (1989); M32: 21 cm, Hummel (1980); NGC 4594: 3.6 - 21 cm, van der Kruit (1973), de Bruyn et al. (1976), Hummel (1980); the Galaxy: 6 cm, Zhao et al. (1992); NGC 3377: 6 cm, Fabbiano et al. (1987); NGC 4258: 6 cm, Turner & Ho (1994); M87: 2 cm, 21 cm, Hummel (1980), Biretta et al. (1983). The value for M87 depends on how much of the jet is included at the spatial resolution of the observations.
e Nuclear radio luminosity in units of the luminosity of Sgr A* at the same wavelength. Sgr A* fluxes are 0.5, 0.7, 0.8, and 0.9 Jy at 21 cm, 6 cm, 3.6 cm, and 2 cm, respectively (Zhao et al. 1992).
f MDO mass for the best-fitting isotropic dynamical model.
g Logarithm of MDO mass fraction; Mbulge is the bulge mass determined from MB,bulge and from M/L ratios measured in the BH papers (see note h). For NGC 4258, M/Lr = 4.08 is from Kent (1987a); this disk value is probably a good approximation for the bulge-like component as well.
h MDO references; et al. is abbreviated as + and (e.g.) 1994 as 94. Only papers that measured MBH are included; additional papers that confirm the kinematics are discussed in Sections 4 and 5.

Next Contents Previous