|Annu. Rev. Astron. Astrophys. 1989. 27:
Copyright © 1989 by . All rights reserved
A major development in recent years has been the realization that galaxies accrete significant amounts of material in the form of gas and small companions. The next three sections discuss some of the evidence. We begin with observations of distinct dynamical subsystems in galaxy cores.
The first clear example was NGC 5813. Efstathiou et al. (1982) found a core-within-a-core brightness profile in this otherwise normal elliptical (i.e. its core contains a second, smaller core of higher surface brightness). The inner core rotates more rapidly than the outer, and, except for the central measurement, has a smaller velocity dispersion. Kormendy (1984) suggested that these observations are the signature of a merger between a low- and a high-luminosity elliptical. Low-luminosity ellipticals have smaller core radii and higher central surface brightnesses than giant ellipticals. Kormendy showed that the robust core of a small elliptical can survive a merger with a giant elliptical and form a distinct subsystem at the center. He predicted that the rotation axis of the subsystem should be oriented randomly with respect to the main galaxy. In practice, observed orientations may be somewhat nonrandom because merger cross sections depend on encounter geometry. Nevertheless, this provides a test of the merger hypothesis. Also, the nucleus should in general rotate more rapidly than the rest of the galaxy because low-luminosity ellipticals are rapid rotators (Davies et al. 1983). Finally, the Faber-Jackson (1976) relation predicts that the velocity dispersion should in some cases decrease toward the center. These effects are also seen in N-body simulations (e.g. Balcells & Quinn 1988).
A number of galaxies that dramatically show this behavior have now been found. Franx & Illingworth (1988), Jedrzejewski & Schechter (1988a), and Bender (1988b) have found seven elliptical galaxies whose cores are kinematically distinct from the rest of the galaxy. In four cases, the inner and outer parts rotate in opposite directions. This is strong evidence for accretion. Further support is provided by the observation of isophote twists between the two subsystems (Efstathiou et al. 1982, Bender 1988b).
In the new cases, no core-within-a-core structure is seen. This is not surprising. Efstathiou et al. (1982) warned us that their NGC 5813 photometry is not of high quality. It should be checked, especially since dust can counterfeit a core-within-a-core structure. Also, distinct cores are only expected in extreme cases, e.g. when a galaxy like M87 eats one like M32 (Figure 3 in Kormendy 1984). Such events should be rare, because faint ellipticals are rare (Binggeli et al. 1985, Sandage et al. 1985b). Mergers between nearly equal galaxies are not likely to leave a signature in the brightness profile.
The merger interpretation is attractive, but alternatives are possible. For example, if the figure rotation velocity in an elliptical is backward with respect to the streaming velocity of the stars, the sum (which is what we observe) can change sign. However, counterstreaming is difficult to achieve (Vietri 1986, 1988) and does not explain subsystems that rotate at right angles to their galaxies (NGC 4406; Bender 1988b, Franx 1988). This interpretation seems improbable. Another possibility is that the inner and outer parts of a galaxy acquire different angular momenta through tidal torques (Binney 1987, Barnes & Efstathiou 1987). This cannot be excluded, although it is least likely near the center. Like the above authors, we conclude that the observed dynamical subsystems result from mergers. These could be mergers of bulges or ellipticals, or ones involving gas infall and star formation (Section 4). IC 1459 and NGC 5322 may be examples of the latter: Their nuclear subsystems appear to be counterrotating stellar disks (Franx & Illingworth 1988, Bender 1988b).
Only a fraction of all merger remnants can be recognized from observations like the above. The fact that about one third of the ellipticals examined so far show nuclear subsystems (Bender 1988b, Jedrzejewski & Schechter 1988a, b) suggests that mergers affect a significant fraction of galaxies.