12 



Scientific Proceedings (ioi). 



as possible, in order to be sure that any mutations that might 

 be discovered later had arisen separately, and were not merely 

 sister representatives of one original mutation. By continuing 

 this method of breeding from separate families over five or more 

 generations after the preliminary tests, the sex ratios in 385 

 families were counted. Thirteen of these were found to be 2:1 

 ratios; this is a proportion of one new lethal mutant among each 

 thirty females that are bred. This figure is of a far higher mag- 

 nitude than any which had been anticipated. It should be noted 

 that at the same time as all these lethals arose, no mutations 

 causing ordinary visible character variations were observed. 



The correctness of classification of most of the thirteen lethals 

 was verified by further breeding tests, but there were a few doubt- 

 ful cases, and it was realized that ratios intermediate between 

 1 : 1 and 2 : 1 are sometimes brought about in other ways. Although 

 the possible error due to these cases was not enough to change the 

 order of magnitude of the frequency found, a new set of experi- 

 ments was undertaken in which a still more definite test of lethal 

 factors than the sex ratio was used — namely, the test of linkage 

 to known factors in the X chromosome. The breeding procedure 

 — having preliminary tests, breeding from many separate families, 

 etc. — was the same as before, but instead of using pure wild type 

 flies for the work, the following cross was made in each genera- 

 w e v f 



tion: ^yy p 9 X w e v f cf . In this case a lethal arising in either 



X chromosome makes itself known not only by the 2 : 1 sex ratio, 

 but by the practically total absence of all males containing factors 

 on both sides of the lethal. By noting whether any expected 

 class of males was absent, it could thus be determined whether 

 a lethal was present, and, if so, approximately where it was located 

 in the chromosome. 1,062 families were examined in this way, 

 after the preliminary tests, and twenty lethals were found — a ratio 

 of one in fifty-three. Enough work has been done on them thus 

 far to know that they occurred in at least ten different loci scat- 

 tered along the X chromosome — but this is a bare minimum. 

 Four of the lethals (perhaps five) are more strictly speaking "semi- 

 lethals," as they occasionally allow the male possessing them to 

 live (and then produce some curious morphological effects in him) 



