106 Nature of the Genetic Material 



All tlie older claims for reverse mutation, even those which seemed 

 to be best established, must be looked on with suspicion, because 

 position effect was not known then, and no detailed cytological check 

 was made. An ordinary genetic or cytological check would usually 

 overlook small rearrangements which are not apparent genetically. I 

 might mention the following case: for a long time we carried a mutant 

 of yellow as a stock, in which later the mutant white appeared, and a 

 yw stock was carried and used as any ordinary marker stock. When, 

 for some reason, a cytological check was made (see Goldschmidt, 

 1945a), it turned out that both mutants were position eflFects of small, 

 homozygous viable inversions. If a similar situation had appeared with 

 a reversion to true breeding wild type, probably nobody would have 

 checked the possibility of small rearrangements involving a few bands. 



Though reverse mutations in Drosophila have become very doubt- 

 ful, they play a considerable role in the genetics of Neurospora and 

 in bacteria, where reverse mutation for biochemical requirements 

 seems to be almost as frequent and is even inducible at will. Giles 

 (1951; Giles and Partridge, 1953) has made a special study of such 

 reversals of biochemical (nutritional) mutants, both spontaneous ones 

 and those induced by irradiation with ultraviolet or treatment with 

 nitrogen mustard. In some he could establish, by genetic tests, that 

 the apparent reversal was due to a suppressor locus, which could be 

 transferred to other strains with the same effect. In a number of 

 others ( inositol-dependent strains returning to inositol independence), 

 Giles thinks that he has proved genuine reversal, because the reversals 

 behave in every respect like the it-strain, segregate normally in crosses, 

 do not show ascospore abortion or segregate the mutant after crossing. 

 Furthermore, no changes in crossover values were found. These results 

 seem to exclude gross rearrangements as an explanation. 



However, there are other facts which to Giles favor the explana- 

 tion by reverse mutation but make me still more skeptical. In a group 

 of eight inositol-less mutants, marked differences were found in regard 

 to the frequency of return mutation, varying betw.een the extremes of 

 frequent spontaneous reversal and complete stability. These frequency 

 differences segregate in the asci and are therefore inherited; they 

 seem to be bound to the locus or its neighborhood, as experiments 

 with closely linked markers show. It is furthermore important that 

 X rays are more effective in producing reverse mutants. Giles thinks 

 that he has demonstrated a series of multiple alleles of inositol-less 

 distinguished only by their ability for reverse mutation ( see, however, 

 McChntock's "reverse mutation" in maize). He also claims different 



