1250 BIOLOGICAL EFFECTS OF RADIATION 



The essential experiment is one which Muller (112) had undertaken 

 before the calculations discussed in the foregoing were completed and 

 subsequently discontinued. If radiations constitute a sine qua non for 

 the mutation process, when radiations are screened off there should be 

 fewer — or no — mutations. This is a most difficult experiment — the 

 spontaneous mutation being as low as it is. But it is a critical one. 

 For without it (and, under certain not too improbable assumptions even 

 with it) the idea is tenable that many factors are concerned with the 

 stability of genes, but that the "accidents" which constitute a mutation 

 always involve the effect of radiation. 



THE NATURE OF THE MUTATION PROCESS 



The discussion in the preceding section, as in those before it, dealt 

 more or less descriptively with the characteristics of the mutation process. 

 Most of the conclusions as to the "causes" of mutations rest on assump- 

 tions, explicit or implicit, as to the nature of the genes and the mechanism 

 whereby they change. The assumptions fall into two categories: (o) 

 the genes are independent units, and mutations are changes within the 

 units; or (6) the chromosome is a continuum and changes in any portion 

 thereof constitute mutations. A bridge between these is found in the 

 position-effect hypothesis, which permits the expectation that a change 

 in one gene modifies the behavior of its neighbors. 



Under both of these broad categories there is a class of mutation 

 theory which dates back to Bateson's "presence and absence." Either 

 mutations are inactivations or losses, total or partial, of the unit, or 

 they are greater or lesser losses of portions of the chromosome. There 

 is no point here in discussing the details of these hypotheses; much con- 

 troversial blood has been spilled, and yet it would seem that the points at 

 issue have no concern with the experiments which purport to study them. 

 This is particularly evident in the studies on reverse mutations (59, 140, 

 182). These are important, as has been seen, in any attempt to under- 

 stand the question of the direction of mutation. Actually they were 

 undertaken largely in an attempt to show that the effect of X-rays on 

 genes is not to destroy them; that is, to prove that the radiation-induced 

 mutations are not deficiencies. The proof consists of the induction of 

 mutations in both directions at the same locus. This could, however, 

 only amount to proof if it is granted beforehand that mutation is a 

 change within the gene, as Goldschmidt (51) has pointed out. Many 

 mechanisms are conceivable, particularly with the intervention of a 

 position effect (cf. Stadler, 166), whereby reverse mutations could 

 occur and still be due to quantitative changes of some kind. 



One of the most elaborate recent attempts to provide both a structure 

 of the gene and a mechanism of mutation is the hypothesis of subgenes, 

 due to Serebrovsky and Dubinin and their coworkers (1, 2, 35 to 39, 



