Primary Actions 273 



precursors or through autocatalysis. We shall return to this problem 

 later. In the present context it is very difficult to see how the nature 

 of a primary product in situ could be analyzed. At this point only 

 indirect evidence from genetics, of the type just discussed, is available. 



Lewis (1951) has used the case, discussed above, of three 

 "pseudoallelic" loci with bithorax eflFect to draw some conclusions 

 concerning primary gene products in situ. The situation is mutatis 

 mutandis similar to Pontecorvo and Roper's case, the main diflFerence 

 being that a morphogenetic effect is involved and that the primary 

 chemical events are purely hypothetical, based completely upon a 

 literal interpretation of Beadle's one gene — one enzyme hypothesis. 

 The derivations are further based upon an unfaltering belief in the 

 classic corpuscular gene, and finally upon the premise that the three 

 pseudoallelic loci are the result of duplication with subsequent change 

 in function of one original gene (as already discussed in the chapter 

 on pseudoallelism ) . Thus a powerful preoccupation with the problem 

 of the origin of new genes by means of duplication is at the basis of 

 the analysis. Like Pontecorvo, Lewis assumes that the genie product 

 in situ is produced in such small amounts or is so limited in its ability 

 to diffuse that the reactions take place independently in one of the 

 homologous chromosomes. The direct products of the genes are 

 assumed to be highly complex molecules. The actions of the pseudo- 

 alleles is then assumed to be a sequential one: if S is the available 

 substrate, and A and B the products of the two duplicated genes +" 

 and +^ the reactions would follow in the order S-^A->B. This 

 seriation is the product of evolution, which Lewis conceives in the 

 following way. The reaction controlled by the old gene is a reversible 

 one: K *=? L. The old gene would be expected to share a specificity 

 for both the substrate K and the product L of the reaction. If a 

 mutation to a new function occurs, the new gene could use both K 

 and L as substrates. If the new gene also utilizes K, it is in com- 

 petition with the old gene for the substrate. But if it uses L, a new 

 product M may result, and the sequential reaction K — > L — > M is 

 established. These progressive steps involving complex molecules 

 may be possible only when the genes in question had a common 

 origin, since unrelated genes would not be similar enough in structure 

 to share a specificity for the same complex molecule. The consequences 

 of this argumentation for the explanation of pseudoallelism are the 

 same as those developed above, following Pontecorvo. 



Recently a new and remarkable example of pseudoallelism was 

 found both by McKendrick and Pontecorvo (1952) and by Lewi§ 



