4.2.7. Velocity Fields in Bulges and Ellipticals
Fairly detailed maps of the velocity fields of ellipsoidal components have been published only for the bulges of several disk galaxies. Figure 39 shows such data for the edge-on S0 NGC 3115, from Illingworth and Schechter (1982). Similar observations of NGC 4565, 4594, 5866 and 7814 have been published by Kormendy and Illingworth (1982a, KI). The bulge of M31 has been studied by McElroy (1981), and references therein.
 |
Figure 39. Four one-dimensional cuts
through the velocity field of
the edge-on S0 galaxy NGC 3115, from
Illingworth and
Schechter (1982).
For the upper-right panel the spectrograph slit was oriented parallel
to the disk but was offset from it by z = 20". At this
height the
measurements refer only to the bulge. The major-axis rotation curve is
repeated schematically. The lower-right panel is derived using a slit
oriented parallel to the minor axis and displaced from it by
r = 25".
Each panel shows rotation velocity V (km s-1), velocity
dispersion |
All of the edge-on galaxies studied except NGC 4565 show the
following features. Rotation velocities along the major axis rise
rapidly with increasing radius, sometimes show a maximum in the bulge,
and then level off to constant values at larger radii. Along slit
positions parallel to the disk but offset from it in z, the rotation
curve rises more slowly and more linearly than along the major axis.
The maximum rotation rate is smaller than along the major axis, but is
still quite large
(
100 km
s-1) even several kpc above the plane. No
significant rotation is observed along the minor axis. Most
interesting are the "perpendicular cuts" obtained by orienting the
spectrograph slit parallel to the minor axis but offset from the center
in radius. These show how rotation decreases with increasing height
above the plane. In all five galaxies, the rotation rate decreases
smoothly and continuously with height z. There is no sharp
distinction
between a disk that rotates rapidly and a bulge that rotates slowly.
Comparable data have not been published for elliptical galaxies, but
preliminary results from Davies and Illingworth
(1982b,
see KI, p.475) suggest that they behave similarly.
No detailed study has been made of the implications of the above results for dynamical models or for theories of galaxy formation. In general terms, the velocity fields, like the global rotation measures, are more consistent with dissipational-collapse theories than with dissipationless galaxy formation. The smooth decrease of rotation with height suggests that stars did not form either before or after the collapse but rather simultaneously with the collapse process. Qualitatively, galaxy formation models incorporating simultaneous collapse and star formation produce velocity fields which are similar to the data (e.g., Larson 1975, 1976; see KI and Illingworth and Schechter 1982 for further comparisons with models). However, a much wider range of formation processes is likely to be consistent with the data, including early mergers of partly gaseous protogalaxies (e.g., Silk and Norman 1981), and disk building at late stages by the infall of loosely-bound gas (Gunn 1981). The encouraging point here is that we are rapidly accumulating kinematic data with enough accuracy and detail to allow a meaningful confrontation with developing theories.
Finally, it is interesting to note the special behavior of box-shaped bulges (see section 3.4.1). KI discuss NGC 4565 in some detail, and find that velocities along the r = 27" perpendicular cut are almost constant up to z = 30" (3 kpc for H0 = 50 km s-1 Mpc-1) above the disk (contrast Fig. 39). That is, the bulge rotates cylindrically to large distances above the plane. Similar behavior is seen in the box-shaped bulges of NGC 1381 (Davies and Illingworth 1982a), NGC 7332 (Illingworth and Kormendy 1982) and very probably NGC 128 (Bertola and Capaccioli 1977). As noted in section 3.4.1, this large amount of rotation at high z may be connected with the box-shaped distortion of the isophotes.
(km s-1) and line strength
as a
function of radius (arcsec). Circles
and squares refer to opposite sides of the symmetry axis (the minor
axis for the upper panels, the major axis for the lower ones). Near
the center, where the present photographic spectra are saturated, plus
signs show velocities from