Similarities between essential enzymes indicate homology and therefore origin from a smaller number of ancestral genes, but there are also indications that the immediately preceding biochemical system was of about the same complexity as the present one. It is argued that to maintain function with a smaller number of enzymes, earlier enzymes must have been less specific, catalyzing classes of reactions. Lower specificity also resulted in ambiguous translation, each cistron producing a family of related proteins. Though individual protein molecules need not have been less specific, each family as a whole functioned as a catalyst of lower specificity. The number of kinds of amino acids incorporated into proteins may have been larger than at present. The evidence supporting this, some of its implications, and the kinds of additional data that would be useful in such problems are discussed.