2. THE COSMIC INVENTORY

A useful starting point is the cosmic baryon budget drawn up by Fukugita, Hogan & Peebles (1998), hereinafter FHP, shown in the accompanying table. The total from Big Bang nucleosynthesis (BBNS) adopted here agrees quite well with the amount of intergalactic gas at a red-shift of 2 to 3 deduced from the Lyman forest, but exceeds the present-day stellar (plus cold gas) density by an order of magnitude. ⁽¹⁾

Table 1. Inventory of cosmic baryons and "metals"
Densities expressed as , in units of _crit = 1.54 × 10¹¹ h₇₀² M Mpc^-3

All baryons from BBNS
(D/H = 3 × 10-5 a)	0.04 h70-2
Stars in spheroids	0.0026 h70-1 b
Stars in disks	0.0009 h70-1 b
Total stars	0.0035 h70-1 b
Cluster hot gas	0.0026 h70-1.5 b
Group/field hot gas	0.014 h70-1.5 b (0.004h75-1 in O VI systems c)
Total stars + gas	0.021 h70-1.5 b
Machos + LSB gals	?? b
Z (stars, Z = 0.02 d)	7 × 10-5 h70-1
Z (hot gas, Z = .006)	1.0 × 10-4 h70-1.5 b
	1.2 × 10-4 h70-1.3 e
Yield Z / *	0.051 h70-0.3 ( 3Z!)
Damped Ly- (H I)	0.0015 h70-1 b, f
Ly- forest (H+)	0.04 h70-2 b, g
Gals + DM halos
(M/L = 210 h70)	0.25 b, h
All matter
(fB = .056 h-1.5)	0.37 h70-0.5 b, i
a O'Meara et al 200l; but see also Pettini & Bowen 2001; b Fukugita, Hogan & Peebles 1998; c Tripp, Savage & Jenkins 2000; d Edmunds & Phillipps 1997; e Mushotzky & Loewenstein 1997; f Storrie-Lombardi, Irwin & MacMahon 1996; g Rauch, Miralda-Escudé, Sargent et al 1998; h Bahcall, Lubin & Dorman 1995; i White & Fabian 1995.

FHP pointed out that a dominant and uncertain contribution to the baryon budget comes from intergalactic ionized gas, not readily detectable because of its high temperature and low density. The number which I quote is based on the assumption that the spheroid star-to-gravitational mass ratio and baryon fraction are the same in clusters and the field, an assumption that had also been used previously by Mushotzky & Loewenstein (1997). The resulting total star-plus-gas density is within spitting distance of _B from BBNS, but leaves a significant-looking shortfall which may be made up by some combination of MACHOs and low surface-brightness galaxies; it is not clear that a significant contribution from the latter has been ruled out (cf O'Neil 2000).

¹ The stellar density taken here from FHP is based on B-luminosity density estimates and might be revised upwards by 50 per cent in light of SDSS commissioning data (Blanton et al 2000) or downwards by 20 per cent in light of 2dF red-shifts plus 2MASS K-magnitudes (Cole et al 2000); in either case we are following FHP in assuming the IMF by Gould, Bahcall & Flynn (1996), which has 0.7 times the M / L_V ratio for old stellar populations compared to a Salpeter function with lower cutoff at 0.15 M. Back.