RADIATION AND THE STUDY OF MUTATION IN ANIMALS 1221 



The same technique has been used by Patterson and Muller (141) in 

 the study of X-ray-induced mutations. They irradiated females, a 

 large proportion of whose daughters would receive both X-chromosomes 

 from their mother. Among 104 daughters so tested, two lethals were 

 found, and two reverse mutations at the scute locus. In all four cases, 

 the mutation had occurred in only one member of the allelomorphic 

 pair. 



There is another body of data which leads to the same conclusion. 

 When adult males of Drosophila are irradiated, which implies the treat- 

 ment of mature spermatozoa, mosaic mutants appear among the progeny. 

 Muller (105) has termed these "fractional" mutations and has pointed 

 out that they usually involve half of the fly. A fractional mutation must 

 therefore involve the mutation of one of the daughter genes of a given 

 chromosome, without the same mutation in the other — an even more 

 striking example of the strict localization of the mutation process than 

 those previously given. It should be noted that many of the cases of 

 "somatic" mutation found in untreated material probably belong to the 

 same category as the fractionals in treated material, so that the phe- 

 nomenon is common to both induced and spontaneous mutation. The 

 possible interpretations of fractional mutation will be considered later 

 with reference to Moore's (96) data and the possibility of an aftereffect 

 of X-rays. 



From these three lines of evidence, then, it is clear that the mutation 

 process as we know it in Drosophila is the result of a local disturbance, 

 and not a general change to which specific genes respond. Something 

 happens in a given neighborhood, the result being, as we have seen in the 

 previous sections, determined by the neighborhood. 



CHROMOSOME BREAKAGE AND THE PRODUCTION OF MUTATIONS 



In the preceding section it has been shown that the mutation process 

 must be traced to a local disturbance and not to the response of a given 

 gene to some general change in the cell. It is now worth while to examine 

 the proposition from another angle: What are the effects on the production 

 of mutation of other local changes in the chromosome? The simplest of 

 these is chromosome breakage. 



This occurs normally, of course, at every meiosis, in organisms where 

 crossing over is found. Crossing over is not, however, followed by 

 mutation ordinarily, nor is mutation known even to be associated with 

 the ordinary crossing over (Sturtevant, 169). In the case of the frequent 

 "mutations" of the gene Bar in Drosophila melanogaster, Sturtevant has 

 shown (169, 170) that this, a special case, is concerned with unequal 

 crossing over. Two aberrant types result, in the case of the homozygous 

 Bar female: one which has no Bar genes; and one which is an extreme Bar, 

 and contains two Bar genes in one chromosome. The juxtaposition of 



