This paper revises the model of Fowler, Greenstein and Hoyle (FGH) for the nucleosynthesis of D, Li, Be and B by high energy particles from the Sun during the early history of the solar system. In this model these nuclei are produced by spallation reactions, mainly on O 16, in metric-sized planetesimals. Large numbers of neutrons are also produced. A fraction of these are thermalized and react by B 10( n, α) Li 7, Li( n, α) H 3, to produce the terrestrial Li and B isotopic ratios. Additional D is produced by H 1( n, γ) on hydrogen retained in the planetesimals as H 2O. The total energy required in high energy particles is about 2 × 10 44 ergs. The nuclear calculations have been generalized in an approximate manner to include a dependence on the duration of the irradiation caused by the long lifetime of Be 10. The first stage of the calculation yields the required spallation yields and time-integrated neutron flux to produce the terrestrial Li, Be and B abundances and isotopic ratios. The required flux is 4 × 10 21 n/cm 2, identical with that obtained by FGH, and does not depend significantly on the choice of irradiation time. The predicted spallation yields depend more strongly on the irradiation time. Those are compared with cloud chamber data for O 16 + 300 MeV neutrons. The predicted low spallation yield for Be 9, which merely reflects its low relative abundance, is consistent with the cloud chamber data. Good agreement is not obtained for the B Li spallation ratio; however, for large values of the irradiation time, this could be due to uncertainties in the parameters. This does not seem possible for short irradiation times, however. Tho nuclear processes are less reasonable if C is present in the planetesimals or if higher energy particles are assumed in order to get appreciable amounts of LiBeB from spallation of SiFe as well as from O 16. The second stage of the calculation yields the required hydrogen concentration, H Si ≈ 1 , and “dilution factor,” F d ≈ 20 (approximately the ratio of unirradiated to irradiated material) to give the terrestrial D H ratio. FGH calculated F d = 10 and H Si = 8 . The amount of water is considerably reduced in the present calculations. Although a loss of O during the formation of the Earth still must be postulated, the amount to be lost is much less than in tho FGH case. We point out that the FGH model is compatible with suggestions that the Moon has a high water content and that biotic material has formed in the carbonaceous chondrites and on the lunar surface. The nucleosynthesis of C 13 in its present terrestrial abundance appears quite feasible; however, in this case the solar C 13 C 12 ratio will be much less than that observed terrestrially. Conflicting experimental results exist on this point at the present time. Observed C 13 C 12 variations in meteorites appear to be due to chemical fractionation. The fact that the isotopic composition of Li, Gd, and K in stone meteorites is identical with that found terrestrially requires that both terrestrial and meteoritic material were subjected to the same particle flux and had the same fraction of material irradiated. This implies that the Earth and the meteorites had a common initial history if the basic features of this model are to be retained. A lunar origin for stone meteorites could very well provide the required astrophysical situation to meet this requirement, whereas an asteroidal origin presents many more difficulties.