MANNER OF PRODUCTION OF MUTATIONS 597 



quantitatively and can even, on occasion, involve seemingly quite 

 different characters. The different alleles of one gene commonly arise 

 from a given allele with different frequencies. The seemingly smaller 

 changes of a given gene do not necessarily arise oftener than the large 

 changes so as to tend to form a normal curve, but the more marked change 

 is sometimes the most frequent. 



Among the principles which have attracted the greatest amount of 

 attention is the fact that, in most mutant genes studied with reference to 

 the question, reverse mutations can occur but that the reverse mutation, 

 which must be conceived as ordinarily being a more specific type of 

 biochemical change, is usually much less frec^uent than the direct one. 

 For the same reason, full reversion in some loci requires more than one 

 step. Actually, however, the relative frequency of mutation in two 

 opposite directions does not depend primarily on which direction of muta- 

 tion is from normal to abnormal and which is the reverse but rather on 

 which direction is toward a more " hypomorphic " (or ''amorphic") and 

 which toward a more " hypermorphic " (or "neomorphic") condition (see 

 Chap. 7). It is because most reverse mutations are of the hyper- 

 morphic, or neomorphic, type that they are less frequent than their 

 opposites. Not only the expression of the mutant gene may change 

 when a mutation occurs but also its apparent mutability. In fact, there 

 can even be different alleles having the same expression but different 

 apparent mutabilities (Timofeeff-Ressovsky, 1932a, b; Giles, 1952). 

 Further complications occur when the geneticist is deahng with a group 

 of closely neighboring loci, related by duplication or by position effect. 

 A discussion of all these matters, however, has its place in a treatment 

 primarily concerned with genetics proper rather than with mutagenesis. 



21. CONCLUDING CONSIDERATIONS 



The findings in Sect. 20, especially those dealing with a given locus, 

 illustrate the inadequacy of the concept of "energy levels" by itself 

 (advocated by Timofeeff-Ressovsky, Zimmer, and Delbrlick, 1935; 

 Schrodinger, 1944; Blum, 1951; McElroy and Swanson, 1951) for under- 

 standing mutation and the influence of radiation and other mutagenic 

 and antimutagenic agents on this process. Certainly different genes and 

 different alleles of a gene must, to some extent, have different energy 

 levels and must be separated by different energy peaks which have to be 

 surmounted if a mutation is to occur, as is true in the case of any other 

 chemical reaction. Moreover, there must in some cases be intermediate 

 shallower troughs, representing less stable alleles, lying between the 

 observed ones. However, although these energy differences must form 

 one part of the picture, the process of mutation cannot usually be the 

 simple one of merely raising the amount of energy enough in general, or 



