5.1. Comparison with Earlier Estimates
Let us compare our estimates with those by Persic & Salucci (1992), Gnedin & Ostriker (1992), and Bristow & Phillipps (1994). These three groups considered baryons in stars (often classified into those in early and late types of galaxies) and in hot X-ray emitting gas around clusters. The baryon abundance in stars estimated by Gnedin & Ostriker is in good agreement with ours, and that by Persic & Salucci is close to our minimum estimate. On the other hand, the estimate of Bristow & Phillipps is 6 times ours. The same pattern applies to baryons in X-ray emitting gas: Persic & Salucci obtained 1/3 our central value and Bristow & Phillipps found twice our estimate. As a result Persic & Salucci find that the total amount of baryons is substantially less than that expected from nucleosynthesis, while Bristow & Phillipps find the nucleosynthesis value is saturated by stars plus hot X-ray emitting gas alone. Our attempt to take account of a broader range of constraints and states of baryons, including plasma in groups inferred but not directly observed, has led us to conclude that, apart from baryons that may have been sequestered at very high redshift, the bulk are either directly visible or in forms which are well constrained and accounted for with plausible extrapolations.