Abstract
The probability of gene identity between members of a multigene family was investigated, assuming mutation, random genetic drift, and unequal crossing-over. The theoretical results on the equilibrium and transient behavior of the identity probability were applied to the analyses of sequence variability in the variable regions of immunoglobulins. The probability of amino acid identity between corresponding residue sites at different but homologous immunoglobulin sequences was computed both within and between species from data reported by other workers. The average probability of amino acid identity was obtained for the hypervariable regions and also for the framework regions. From statistical analyses based on population genetics theory it is concluded that somatic mutations cannot be the major cause of hypervariability and that the results can readily be explained by the germ line theory.