Annu. Rev. Astron. Astrophys. 1994. 32: 531-590 Copyright © 1994 by Annual Reviews. All rights reserved

2.6. Background Dark Matter

None of the forms of matter discussed above can have the critical density required for the Universe to recollapse: _crit = 3H²₀ / 8 G = 2 × 10^-29 h^-2 g cm^-3, where h = H₀ / (100 km s^-1 Mpc^-1). As discussed later, disk dark matter could only have _d ~ 0.001, while the halo and cluster dark matter could only have _h ~ 0.01-0.1 and _c ~ 0.1-0.2, respectively. However, according to the currently popular inflation theory (Guth 1981), in which the Universe undergoes an exponential expansion phase at some early time, the total density should have almost exactly the critical value ( = 1). [See, however, Ellis (1988) and Ellis et al (1991) for a different point of view.] This would have two possible implications: 1. There is another dark component, or 2. galaxy formation is biased (Kaiser 1984, Dekel & Rees 1987) in the sense that galaxies form preferentially in just a small fraction of the volume of the Universe. Although the second possibility avoids a proliferation of dark matter species. some people now invoke a mixture of hot and cold dark matter anyway (e.g. Taylor & Rowan-Robinson 1992).

In either case, one would expect the mass-to-light ratio to increase as one goes to larger scales, and there is some indication of this from dynamical studies. One can probe the density on scales above 10 Mpc, for example, by analyzing large-scale streaming motions (Dressler et al 1987, Bertschinger & Dekel 1989) or by determining the dipole moment of the IRAS sources (Rowan-Robinson et al 1990). In all these analyses, the inferred density depends on the bias parameter b (dynamical effects depending on the product ^0.6 b^-1). More sophisticated analyses are needed to determine and b separately (Peacock & Dodds 1994, Nusser & Dekel 1993). The IRAS dipole suggests a critical density if the IRAS sources are unbiased (b = 1); however, this conclusion would be erroneous if there was a significant contribution to the dipole from distances beyond 100 Mpc (Scaramella et al 1991). For a recent review of the evidence for and against = 1, see Coles & Ellis (1994).