The process of transduction of a proto-oncogene involves the loss of intervening sequences, the loss of 3' mRNA terminal sequences, and possible base changes. It is not possible to know how similar the c-rel sequences we studied are to sequences from which v-rel arose. If one of the alleles we studied is identical to progenitor sequences of v-rel, then base changes must have occurred during the process of transduction. Alternatively, there may exist an allele of c-rel which did not need to undergo base changes to give rise to v-rel. The polymorphism of c-rel within a turkey population and between chicken and turkey shows clear evidence in c-rel coding sequences have not been detected within a turkey population. There is an EcoRI site in c-rel from turkey which is not in v-rel. There are unequal length substitutions in c-rel within the turkey population and between chicken and turkey. Unequal length substitutions can be explained if there has been a combination of two events, for example, insertion or deletion with extensive base changes, or two insertions or deletions. Two events occurring in a short evolutionary time span might occur if c-rel is diverging at a relatively fast rate. c-rel is much less conserved across species lines, in nucleic acid hybridization to v-rel (Chen et al. 1981; Wong and Lai 1981) than other proto-oncogenes, consistent with it diverging at a relatively fast rate. Since the evolutionary changes that have been seen for c-rel are in noncoding sequences, it is most likely that v-rel arose by transduction of a somatic mutation generated variant of c-rel or by transduction which involves sequence changes or later mutations.