OF MUTANT CHARACTERS. 



229 



75 and 76). Two or three out-crosses of the pink comma flies were 

 also attempted to secure Pi for a back-cross, but these failed. 



The F2 of the comma by pink gave only 8 commas, all females, among 

 440 flies. Of these 8, 3 were pink commas, so that comma was proved 

 to be not third-chromosome. The fewness of the commas was partly 

 due to the crowding of the mass-cultures, but more to the fact that the 

 character fails to show on all the flies that are genetically commas. 



Table 75. — Pi, comma 9 X pink cf. Fi, wild-type 9 9+ Fi wild-type cf cf. 



The F2 from the cross of comma by vestigial gave 18 females and 

 1 male with comma, but not one of these was vestigial. While these 

 numbers are not large enough to prove that comma is second-chromo- 

 some, the probablity is high that it is. 



Table 76. — Pi, comma 9 X vestigial cf. Fi wild-type 9 9 + Fi 



wild-type cf cf. 



Comma reappeared in the experiments involving "squat" and 

 followed this second-chromosome mutant in distribution in such a 

 way as to make it practically certain that comma is also in the second 

 chromosome. 



VALUATION OF COMMA. 



Aside from carrying on a stock of commas for a few generations and 

 noting that the percentage of commas was seldom above 30, nothing 

 further was done with this mutant because of the poorness of the 

 character. When present, the character comma was perfectly sharp 

 and clear, even when, as was quite often the case, it showed on only 

 one side of the thorax. 



Perfectly definite and trustworthy classifications can be made with 

 characters such as these, but they are very inefficient, since only flies 

 actually showing the character can be considered. The conclusions 

 drawn from such data are not so safe as the classification, since it has 

 been our experience that such characters are particularly sensitive 

 to intensification or repression (in percentage of appearance) by the 

 other genes present in the crosses. 



