|Annu. Rev. Astron. Astrophys. 1979. 17:
Copyright © 1979 by . All rights reserved
After reviewing all the evidence, it is our opinion that the case for invisible mass in the Universe is very strong and getting stronger. Particularly encouraging is the fact that the mass-to-light ratio for binaries agrees so well with that for small groups. Furthermore, our detailed knowledge of the mass distribution of the Milky Way and Local Group is reassuringly consistent with the mean properties of galaxies and groups elsewhere. In sum, although such questions as observational errors and membership probabilities are not yet completely resolved, we think it likely that the discovery of invisible matter will endure as one of the major conclusions of modern astronomy.
In addition to the dynamical evidence, there are other indirect indications of dark material in galaxies. The most important of these are the stability analyses of cold, self-gravitating axisymmetric disks (e.g. Ostriker & Peebles 1973, Hohl 1976, Miller 1978), which show them to be susceptible to bar-formation if not stabilized by a hot dynamical component. This hot component may or may not be related to massive envelopes.
Although present data give us little information on the shape of massive envelopes, further study of the outermost hydrogen in spirals may tell us more about this question. For example, the apparent lifetime of warps in many spirals poses severe theoretical difficulties as long as it is assumed that disks are self-gravitating (e.g. Binney 1978, Bosma 1978). This problem would not arise if the warps existed within the potential of a nearly spherical massive envelope. The precession of the warp due to the torque of the disk would then be much smaller, and the warp would be very long-lived. Alternatively, Binney (1978) has suggested that a warp might actually be driven by a triaxial dark halo. Finally, z-motions of H I far from the nucleus can be used to measure the space density of matter in the plane and thus to set limits on the flattening of the envelope.
Despite the general lack of observational evidence on the shapes of massive envelopes, there exists a strong consensus among theorists that they cannot be very flat. It is widely suggested that the large radial extent of the dark material relative to the luminous matter is due to the dissipationless collapse of the invisible matter. If so, a thin disk is unlikely, and we would more plausibly expect a thickened mass distribution spheroidal or triaxial in shape.
Suggestions as to the identity of the unseen matter include massive neutrinos (Cowsik & McClelland 1972, Gunn et al. 1978), faint stars (Ostriker et al. 1974), black holes (Truran & Cameron 1971), and comets (Tinsley & Cameron 1974). Many attempts have been made to detect luminous matter in the halos of edge-on galaxies (e.g. Freeman et al. 1975, Gallagher & Hudson 1976, Hegyi & Gerber 1977, Kormendy & Bruzual 1978, Spinrad et al. 1978). Although faint luminosity has been found in some cases, it can plausibly be identified with the normal spheroidal stellar component.
Further progress in the study of unseen matter will continue to be made by mapping the gravitational potential using all observable test particles. Massive envelopes may well have a significant effect on the shapes and velocities of bridges and tails created in tidal encounters. The embarrassingly short theoretical lifetimes of binary galaxies and compact groups require careful consideration, as does the hypothesized stripping of extended halos and subsequent redistribution of the dark matter during the collapse of dense clusters. Most important, we need to know whether luminosity is a good indicator of mass density over scales greater than a few kiloparsecs. If a sizeable fraction of the mass in the universe is uncorrelated with the visible light, our dynamical analyses might be greatly in error. For example, our basic model for the formation of a group or cluster as a dissipationless collapse of noninteracting mass points might need serious revision. It is to be hoped that the systematic redshift surveys now in progress, coupled with more realistic theoretical simulations of galaxy interactions, will eventually yield definitive answers to these and related questions.
We would like to thank C. Cox and G. McLellan for their aid in preparing the manuscript. We also thank R. Webbink for useful discussions and permission to reference his unpublished work. M. Roberts suggested the term ``massive envelope.'' We are also grateful to S. Wyatt, N. Krumm, I. Iben, A. Faber, and A. Whitford for their critical comments on the original draft.