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

1. INTRODUCTION

The evidence for dark matter, on all scales from star clusters (10⁶ M) to the Universe itself (10²² M), has built up steadily over the past 50 years (Faber & Gallagher 1979, Trimble 1987, Turner 1991, Ashman 1992). Although the strength of the evidence on different scales varies considerably, there is now little doubt that only a small fraction of the mass of the Universe is in visible form. However, we remain uncertain as to the identity of the dark material. Proposed candidates span the entire mass range from 10^-5 eV to 10¹² M, with a dichotomy between those - primary particle physicists - who would like the dark matter to be some sort of elementary particle and those - primarily astrophysicists - who would prefer it to be some sort of astrophysical object. In the first case, the dark matter would have to be nonbaryonic, with the particles being relics from the hot Big Bang; in the second case, it would have to be baryonic, with the dark objects being made out of gas which has been processed into the remnants of what are sometimes termed "Population III" stars.

During the 1970s the dark matter was usually assumed, at least implicitly, to be baryonic (e.g. Ostriker et al 1974, White & Rees 1978), but in the 1980s attention veered towards the nonbaryonic candidates. This was partly because of developments in particle physics, but also because it was realized that there are good cosmological reasons for believing that not everything can be baryonic (Hegyi & Olive 1983, 1986). For a while "hot" dark matter was popular, but soon "cold" dark matter took center stage and many people still regard this as the "standard" model. In the past few years, however, attention has returned to the baryonic candidates - partly because of perceived problems in the cold model and also because there may now be direct evidence for baryonic dark matter. Entire conferences are now devoted to the topic (e.g. Lynden-Bell & Gilmore 1990), and there seems to be a growing realization that there are so many dark matter problems that one probably needs both baryonic and nonbaryonic solutions.

This review focuses almost exclusively on baryonic dark matter. Section 2 presents the observational evidence for dark matter in various contexts; the discussion here is rather brief because it is only necessary to highlight those issues that relate to baryonic dark matter in particular. Section 3 reviews the general arguments for baryonic and nonbaryonic dark matter, concluding - as indicated above - that one probably needs both. Section 4 discusses why one expects baryonic dark matter to form; this involves a brief review of the "Population III" scenario. Section 5 summarizes the constraints on the Population III scenario which come from background light and nucleosynthetic considerations. These limits are essentially as described by Carr et al (1984) but Sections 6 and 7 focus on topics - dynamical effects and lensing effects - that have seen important recent developments. The most plausible candidates seem to be the black hole remnants of high mass stars or low mass objects, so we focus on these candidates in more detail in Sections 8 and 9, respectively. We conclude in Section 10 with a reappraisal of these and other candidates and we assess the prospects of finding or excluding them.