2.1. Historical Development

The historical development of BBN is reviewed by [26], [13], [27], [10] and [6]. Here we mention a few of the main events.

The search for the origin of the elements lead to the modern Big Bang theory in the early 1950s. The expansion of the universe was widely accepted when Lemaitre [28] suggested that the universe began in an explosion of a dense unstable ``primeval atom''. By 1938 it was well established that the abundances of the elements were similar in different astronomical locations, and hence potentially of cosmological significance. Gamow [29], [30] asked whether nuclear reactions in the early universe might explain the abundances of the elements. This was the first examination of the physics of a dense expanding early universe, beyond the mathematical description of general relativity, and over the next few years this work developed into the modern big bang theory. Early models started with pure neutrons, and gave final abundances which depended on the unknown the density during BBN. Fermi & Turkevich showed that the lack of stable nuclei with mass 5 and 8 prevents significant production of nuclei more massive than ⁷Li, leaving ⁴He as the most abundant nucleus after H. Starting instead with all possible species, Hayashi [31] first calculated the neutron to proton (n/p) ratio during BBN, and Alpher [32] realized that radiation would dominate the expansion.�� By 1953 [33] the basic physics of BBN was in place. This work lead directly to the prediction of the CMB (e.g. Olive 1999b [7]), it explained the origin of D, and gave abundance predictions for ⁴He similar to those obtained today with more accurate cross-sections.

The predicted abundances have changed little in recent years, following earlier work by Peebles (1964) [39], Hoyle & Tayler (1964) [40], and Wagoner, Fowler & Hoyle (1967) [34]. The accuracy of the theory calculations have been improving, and they remain much more accurate than the measurements. For example, the fraction of the mass of all baryons which is ⁴He, Y_p, is predicted to within Y_p < ± 0.0002 [35]. In a recent update, Burles et al. [6] uses Monte-Carlo realizations of reaction rates to find that the previous estimates of the uncertainties in the abundances for a given were a factor of two too large.