The question of rotation in dwarf ellipticals has been addressed only recently. Obviously, the measurement of rotation requires spectroscopy at an even fainter level of surface brightness than measurement of a central . The sparse data we have to date suggest that dwarf ellipticals are not supported by rotation. This excludes at least some ``dS0'' systems, which we have generally lumped together with the dE class. Data are published for two bright dE's in the Virgo cluster, VCC 351 and IC 794 (Bender and Nieto 1990); the three bright dE companions of M31: NGC 205 (Carter and Sadler 1990; Held et al. 1990; Bender et al. 1991), NGC 185 (Bender et al. 1991; Held et al. 1992), and NGC 147 (Bender et al. 1991); and the resolved Fornax system (Paltoglou and Freedman 1987; Mateo et al. 1991). A convenient measure of the dynamical importance of rotation is provided by the anisotropy parameter
where v is the rotational velocity,
is the mean velocity
dispersion, and is the
ellipticity at the radius where v is measured
(Binney 1978;
Kormendy 1982).
If
1, the object can be assumed to be
flattened by rotation; if is significantly smaller than 1, the object must
be supported by velocity anisotropy. The six dwarfs from above all have
< 0.4,
hence they are clearly not
rotation-supported. As mentioned, the dS0's may not follow this trend, e.g.,
UGC 7436, a dS0(5),N system in the Virgo cluster is a
rotating disk galaxy (Bender, private communication). Many dS0's exhibit
disk-like morphologies
(Binggeli and
Cameron 1991,
cf. Sect. 2), but their
kinematic nature has yet to be explored.
Bender and Nieto
(1990) have also measured a number of faint classical
and compact E's in the Virgo cluster, which all seem to be
rotation-supported, in accord with the trend found by
Davies et al. (1983).
An exception to this rule is the very compact galaxy M32 with 0.5
(Tonry 1984;
Tonry 1987;
Dressler and
Richstone 1988),
but M32 is likely
tidally influenced by M31.
The separation between rotating low-luminosity E's and
anisotropic dE's
apparently goes hand in hand with the photometric discontinuity between
E's and dE's described above
(Sect. 2.2). However, a much larger
kinematic sample is needed to test for a real discontinuity in .
Bender and Nieto
(1990)
argue that the specific angular
momentum (J / M) of the dwarfs, unlike that of giant ellipticals,
falls onto the J / M - M relation for rotating ellipticals. The dwarfs
seem to compensate for their slow rotation with a larger effective radius.
Bender and Nieto
(1990)
suggest that dE's gained their anisotropy
from expansion due to winds after the onset
of star formation, but admit that gas accretion and tidal interactions
could be important as well.
Shaya and Tully
(1984)
have suggested that the low angular momentum content of elliptical
galaxies is due to tidal interaction with the collapsing
protocluster. Such a mechanism may also apply to dE galaxies, although
their higher specific angular momentum may be somewhat problematical.