2.2 Global Mechanisms
The difference between local and global mechanisms is not always
clearcut, as demonstrated by the fact that several mechanisms are
included in both our global and local lists. In this review we shall use
the following criteria for differentiating between local and global
mechanisms. If it is necessary to know only the conditions in the
immediate vicinity of a galaxy (e.g., the number density
encompassing the ten nearest galaxies) the mechanism will be called
local. If it is necessary to know the properties of
the cluster as a whole (e.g., the position relative to the
dynamical center of the cluster) the mechanism will be called global. We
should also note that the term global does not imply that the mechanism
is operating throughout the cluster.
- Initial Conditions (hybrid models) - In a hierarchical
picture of galaxy formation (e.g.,
Peebles 1980),
with initial conditions in an isolated protogalaxy determining
the morphological type, it is difficult to understand how elliptical galaxies
preferentially make their way into the higher density regions of the
clusters. This
led to the expectation that "late" evolution by the cluster environment
might play the major role in determining the properties in a cluster
(Dressler 1984a).
Several mechanisms have recently been suggested for providing this connection,
including biased galaxy formation (e.g.,
Davis et al. 1985,
Frenk et al. 1985,
Zurek, Quinn, and
Salmon 1988,
Carlberg and Couchman
1989),
adiabatic perturbation models (e.g.,
Doroshkevich et
al. 1980),
and tidal torque models
(Shaya and Tully 1984).
This review will focus on the possible role of late evolution. However,
we should keep in mind that initial conditions remain a viable explanation for
most of the phenomena we will discuss.
- Tidal Shaking (from mean cluster field) -
Miller (1988)
has argued that a
severe "shaking" of a galaxy as it passes by the high density core of a
cluster may
rearrange the distribution of mass into a more elliptical-like distribution.
- Tidal Stripping (shear from mean cluster field) -
White (1982) and
Merritt (1984)
have stressed that the mean tidal field of the cluster may be more
effective in
stripping stars from the outer regions of galaxies than galaxy-galaxy
interactions.
The stripped material would probably end up in the extended halo of the
central cD
galaxies. Merritt argues that the effect may be so strong during the
initial cluster
collapse that subsequent evolution of the galaxies would be negligible,
since the
cross sections for dynamical friction and galaxy-galaxy interactions
would be much smaller for the haloless galaxies.
- Ram Pressure Stripping and Gas Evaporation - The existence
of diffuse
hot (
108 K)
intracluster gas near the centers of clusters, as evidenced by X-ray
observations, makes it very likely that any galaxy that passes near the
center of a
cluster will lose some or all of its interstellar gas by ram-pressure
stripping
(Gunn and Gott 1972)
or gas evaporation
(Cowie and Songaila
1977).
The evidence for HI deficiencies (see reviews by
Haynes, Giovanelli, and
Chincarini 1984,
or Haynes 1990)
suggest that one or both of these mechanisms is probably
occurring. The big
question is whether one of these mechanisms can also explain the high
fraction of
S0 galaxies in clusters by shutting off the star formation in spiral
galaxies. The fact
that the fraction of spirals decreases at roughly the same rate as the
fraction of S0s
increases (as a function of local projected galaxy density; see
Dressler 1980)
makes this an attractive hypothesis. However, Dressler points out that most S0
galaxies are
actually found in the field where ram-pressure stripping is not
effective. This makes
it very unlikely that spirals are being converted to S0 galaxies via
this mechanism.
- Truncated Star Formation - Several authors advocate the slow
buildup of disks
from an extended gaseous halo, rather than a rapid formation concurrent
with the collapse of the spheroid (e.g.,
Gunn 1982).
Larson, Tinsley, and
Caldwell (1980)
have suggested that the lack of spiral galaxies near the centers of
clusters might be
caused by the removal of these gaseous halo during cluster
collapse. However,
Fall (1983)
has pointed out problems with this hypothesis based on angular momentum
considerations. In short, the slow rotation of the ellipticals implies
the removal of
so much material that it becomes difficult to understand why ellipticals
are more
massive than spirals.
- Cooling Flows -
Fabian and Nulsen (1977),
and
Cowie and Binney (1977)
have
suggested the possibility that radiative cooling in the densest parts of
the hot
intracluster gas may result in significant cooling flows. Observational
evidence that this process may be occurring comes from strong Balmer lines
(Romanishin 1987),
optical filaments
(Heckman 1981),
and X-ray temperature inversions
(Fabian et al. 1981).
Accretion rates of up to a few hundred
M
yr-1 have
been computed by
Fabian, Nulsen, and
Canizares (1982).
If these rates are sustained for an appreciable
fraction of a Hubble time they may explain the extended envelopes seen around
cD galaxies. However, there is no evidence for abnormal color gradients in cD
galaxies
(Schombert 1988)
indicative of continuing star formation.
Fabian, Nulsen, and
Canizares (1982)
have suggested that the accretion may be in the
form of very
low mass star formation (i.e., dark matter), presumably near the
center of the cD
galaxies where the density is highest. Measurements of velocity
dispersions in cD galaxies
(Malumuth and Kirshner
1981,
Mathews 1988)
indicate that the central
values of M/L are normal for elliptical galaxies, so this mechanism is
not likely to be relevant for many clusters.