RADIATION AND THE STUDY OF MUTATION IN ANIMALS 1235 



lished data is there any information as regards the frequency in which 

 the original mutation occurred, there is no possibility of testing the 

 hypothesis. Harris (72) has indicated that his results on lethals induced 



Table 17. — Viable Mutations from Germ Cells of Different Stages 



d" d^ treated with radium, mated to fresh virgin 9 9 at intervals of seven days, giving 



tkree broods in all. (Hanson and Winkelman, 71) 



Table 18. — Comparison of the Frequency of Viable Sex-linked Mutations 

 IN Drosophila melanogaster, in the Germ Cells of Individuals Treated 



WITH X-RAYS (1325 r) at DIFFERENT StAGES OF DEVELOPMENT 



Data of Moore (96) 



in immature germ cells show that the growth of the gonad in Drosophila 

 depends on an apical-cell mechanism, so that one of the products of 

 division acts as a primordial germ cell. It is easy to see, as indeed 

 Muller has shown (98), how difficult is the determination of the actual 

 mutation rate in germ cells which have yet to divide; even without 

 differential rates of multiplication, the number of mutants recovered is 

 much le.ss than those that occur, and the two percentages may be different, 

 in the observed direction. There is as yet no evidence against the 

 assumption that a gene is as likely to mutate in one tissue as another. 



Timofeeff-Reesovsky (172), and more especially Patterson (130, 

 131, 133), have studied the frequency of mutation in the imaginal disks 

 of Drosophila. Their technique was to irradiate larvae heterozygous for 

 known genes, and to detect in the adults when they emerged mosaics for 

 the characters affected. These mosaics must then be caused either by 

 mutation of the gene in question or by some chromosomal change. In 



