Because the baryon-to-photon ratio eta_{10} is in some doubt, we drop nucleosynthetic constraints on eta_{10} and fit the three cosmological parameters (h, Omega_M, eta_{10}) to four observational constraints: Hubble parameter h_o = 0.70+-0.15, age of the universe t_o = 14+7-2 Gyr, cluster gas fraction f_o \equiv f_G h^{3/2} = 0.060 +- 0.006, and effective shape parameter Gamma_o = 0.255 +- 0.017. Errors quoted are 1 sigma, and we assume Gaussian statistics. We experiment with a fifth constraint Omega_o = 0.2 +- 0.1 from clusters. We set the tilt parameter n = 1 and the gas enhancement factor Upsilon = 0.9. We consider CDM models (open and Omega_M = 1) and flat LambdaCDM models. We test goodness of fit and draw confidence regions by the Delta chi^2 method. CDM models with Omega_M = 1 (SCDM models) are accepted only because the large error on h_o allows h < 0.5. Baryonic matter plays a significant role in Gamma_o when Omega_M \sim 1. Open CDM models are accepted only for Omega_M \gtrsim 0.4. The combination of the four other constraints with Omega_o = 0.2 +- 0.1 is rejected in CDM models with 98% confidence, suggesting that light may not trace mass. LambdaCDM models give similar results. In all of these models, eta_{10} \gtrsim 6 is favored strongly over eta_{10} \lesssim 2. This suggests that reports of low deuterium abundances on QSO lines of sight may be correct and that observational determinations of primordial 4He may have systematic errors. Plausible variations on n and Upsilon in our models do not change the results much. If we drop or change the crucial Gamma_o constraint, lower values of Omega_M and eta_{10} are permitted. The constraint Gamma_o = 0.15 +- 0.04, derived recently from the IRAS redshift survey, favors Omega_M \approx 0.3 and eta_{10} \approx 5 but does not exclude eta_{10} \approx 2.

Since the baryon-to-photon ratio "eta" is in some doubt at present, we ignore the constraints on eta from big bang nucleosynthesis (BBN) and fit the three key cosmological parameters (h, Omega_M, eta) to four other observational constraints: Hubble parameter, age of the universe, cluster gas (baryon) fraction, and effective shape parameter "Gamma". We consider open and flat CDM models and flat "Lambda"-CDM models, testing goodness of fit and drawing confidence regions by the Delta-chi^2 method. CDM models with Omega_M = 1 (SCDM models) are accepted only because we allow a large error on h, permitting h < 0.5. Open CDM models are accepted only for Omega_M \gsim 0.4. Lambda-CDM models give similar results. In all of these models, large eta (\gsim 6) is favored strongly over small eta, supporting reports of low deuterium abundances on some QSO lines of sight, and suggesting that observational determinations of primordial 4He may be contaminated by systematic errors. Only if we drop the crucial Gamma constraint are much lower values of Omega_M and eta permitted.

Nucleosynthesis in the standard hot big bang cosmology offers a successful account of the production of the light nuclides during the early evolution of the Universe. Consistency among the predicted and observed abundances of D, $^3$He, \hel4 and \li7 leads to restrictive lower and upper bounds to the present density of nucleons. In particular, the upper bound ensures that nucleons cannot account for more than a small fraction $(<0.06 h{^{-2} _{50}})$ of the mass in a critical density (Einstein-de Sitter) Universe. In contrast, x-ray observations of rich clusters of galaxies suggest strongly that baryons (in hot gas) contribute a significant fraction of the total cluster mass $(\ge 0.2 h{^{-3/2} _{50}})$. If, indeed, clusters do provide a ``fair" sample of the mass in the Universe, this ``crisis" forces us to consider other ways of mitigating it, including the politically incorrect possibility that $\Omega < 1$. The options, including magnetic or turbulent pressure, clumping, and non-zero space curvature and/or cosmological constant, are discussed.