Cosmological and astrophysical consequences of a high primordial deuterium abundance

We explore the consequences of the first detection of deuterium in a high-redshift, very metal-poor absorbing cloud complex, D/H = (1. 9-2. 5) x 10^-4^, by Songaila et al. and Carswell et al. , obtained with the Keck telescope. This value reflects closely the primordial abundance ratio provided that the observed spectral features are not due to the corruption of the signal by an errant hydrogen cloud of misfortunate velocity. Assuming that the measured D abundance is free from contamination, the baryon/photon ratio is now confined to the range 1. 3-2 (instead of 3-4 and more), in both the classical and inhomogeneous big bangs. Other light elements (³He, ⁴He, and ⁷Li) are consistent with these figures. The low baryonic density of the universe that ensues leaves no room for baryonic matter in the extended halos of elliptical galaxies, especially if the Hubble parameter is close to 100 km s^-1^ Mpc^-1^. Nonbaryonic matter clearly dominates the gravitating mass of clusters of galaxies. The upper limit of the gas density at high redshift (before bulk galaxy formation) is now consistent with the baryonic one. A massive destruction of deuterium, in the course of the evolution of the galaxy (say, by a factor of 10-25) is required to match the D/H ratio observed in the local interstellar medium. The higher D destruction proposed up to now corresponds to galactic evolutionary models devised by Vangioni-Flam & Audouze (1988) and Vangioni-Flam, Olive, & Prantzos (1994). We discuss the virtues and the limits of this class of models and propose an alternative based on mass loss related to a galactic wind.