OF MUTANT CHARACTERS. 285 



genes. Such eclipses are only temporary, hut are serious in tliiit they 

 spoil the usefulness of the mutation as a working tool. 



One of the Fi squat dichaete males was out-crossed to a wild fonuilo 

 and produced four classes of offspring (culture 2993; Sg 21 : D' 20 : S 

 D' 20 : +24). Since this was a back-cross test of the miUe, it is evident 

 that squat is not in the third chromosome. If squat were third- 

 chromosome there would have been no squat dicha'tes. 



At the time that the dichaete test was made a parallel cross to stur was 

 started to test the relation of squat to the second chromosome. Here 

 also in the Fi cultures (2738, 2828) the squat could be distingui.^hed 

 only poorly. 



The back-cross test of the male attempted in this case failed, prol>- 

 ably through sterility. But the answer was obtained, though less 

 surely, by a female test started at the same time. One of the Fi star 



squat females f — — j, when out-crossed to a wild male gave among 



the offspring a few star squats (culture 2827). Since the squat was 

 poor, no accurate records were kept, though it seemed very probable 

 that squat was showing linkage to star. That the eclipse of squat did 

 not mean extinction was proved in the next generation; for a star 

 female and a not-star male, both selected as showing no squat, produced 

 squat offspring (3061). These squats were dominants, not recessives, 

 as proved by out-crossing them to wild flies, whereupon in Fi neiirly 

 half of the flies were squats (3339) . 



OTHER MUTATIONS. 



In culture 3061, just described, the second-chromosome recessive, 

 "narrow" v^ngs was found (February 7, 1916). From the linkage 

 with star which it showed, it had evidently been introduced to the cross 

 through star. 



In culture 3858, which was simply a stock culture of squats whose 

 parents were taken from 3061, there reappeared the second-chromo- 

 some recessive mutant "commas," which had apparently been carrried 

 along in the same chromosome with squat, but being recessive had 

 hitherto no opportunity to show itself. 



LOCUS OF SQUAT. 



It had by now become apparent that squat was second-chromosome, 

 and it was thought advisable to make a rough determination of its 

 locus. Squat was therefore crossed to black i^lexus and a back-cross 

 test of the female made. Black is a control for the middle and plexus for 

 the right end of the second chromosome. No control of the left end 

 was made, since the fairly free crossing-over between star and scjuat 

 observed in some stock cultures had made it evident that the locus of 

 squat is fairly distant from the left end of the chromosome. 



