8.2. The Effects of Black Holes on Galaxy Structure
There is other encouraging news, too. In the past, the BH search was decoupled from other galaxy studies. It was carried out largely in isolation to test the AGN paradigm. Furthermore, the early BH detections were mostly in inactive galaxies, so even the connection with AGN physics was indirect. This situation was reviewed in Kormendy & Richstone (1995). But now BH results are beginning to connect up with a variety of work of galaxy physics. The subject is large and our space is limited. We therefore mention briefly only three of the developing results.
1 - Triaxial elliptical galaxies evolve rapidly toward axisymmetry if the central gravitational potential well gets steep enough (Lake & Norman 1983; Gerhard & Binney 1985; Norman, May, & van Albada 1985; Valluri & Merritt 1998; Merritt & Quinlan 1998; Poon & Merritt 2001; Holley-Bockelmann et al. 2001; see Merritt 1999 for a review). This can be achieved either by increasing the central density of stars via gas infall and star formation or by the growth of a BH. In either case, chaotic mixing of stellar orbits redistributes stars in phase space and causes orbit shapes to evolve. Box orbits, which support the triaxial structure but which allow stars to pass arbitrarily close to the center, are destroyed in favor of orbits that suppport axisymmetric structure. To the extent that triaxiality promotes gas infall and BH feeding, the evolution may also turn off the feeding when the BH has grown to 1 or 2% of the bulge mass. These processes help to explain the observed upper limit to the BH mass fraction.
2 - Some elliptical galaxies have "cuspy cores", i.e., density distributions that break at small radii from steep outer power laws to shallow inner power laws. These cores may be produced by the orbital decay of binary BHs (Begelman, Blandford, & Rees 1980; Ebisuzaki, Makino, & Okamura 1991; Makino & Ebisuzaki 1996; Quinlan 1996; Quinlan & Hernquist 1997; Faber et nuk. 1997; Milosavljevic & Merritt 2001). The formation of BH binaries is a natural consequence of hierarchical galaxy mergers. The orbits then decay (i.e., the binaries get "harder") by flinging stars away. This BH scouring may reduce the stellar density enough to produce a break in the density profile.
3 - Three-integral dynamical models tell us the distribution of stellar orbits around a BH. Preliminary results (van der Marel et al. 1998; Cretton et al. 1999b; Gebhardt et nuk. 2000a, 2001; Richstone et nuk. 2001) show an important difference between core and power-law galaxies. In core galaxies, the central tangential velocity dispersion t is larger than the radial dispersion r. Large tangential anisotropy is consistent with the effects of BH binaries (Nakano & Makino 1999a, b) and BH scouring (Quinlan & Hernquist 1997). In contrast, coreless galaxies, which have featureless, almost power-law density profiles, are observed to have t. This is more consistent with the adiabatic growth of single BHs via gas accretion (Quinlan, Hernquist, & Sigurdsson 1995 and references therein). Further studies of the relationship between BHs and properties of their host galaxies should provide much better constraints on the relationship between BHs and galaxy formation.
The above developments are an important sign of the developing maturity of this subject. Finding convincing connections between BH properties and the microphysics of galaxies contributes in no small measure to our confidence in the BH picture. The medium-term future of this subject is therefore very promising.
In the longer-term future, the most fundamental progress is expected to come from gravitational wave astronomy. We can look forward to the true maturity of work on supermassive BHs when the Laser Interferometer Space Antenna (LISA) begins to provide a direct probe of strong gravity.
It is a pleasure to thank our Nuker collaborators for helpful discussions and for permission to use our BH detection results before publication. We are also most grateful to Gary Bower, Richard Green, Mary Beth Kaiser, and Charles Nelson for communicating STIS team BH detections before publication.