356 



CALVIN B. BRIDGES 



the amount of crossing over between cream III and some two 

 other genes whose loci were already known. The best method 

 of finding the amount of crossing over is by the backcross, which 

 consists in testing a multiple heterozygote by the corresponding 

 multiple recessive. The double recessive cream ebony III was 

 made up by breeding the eosin ebony and the cream III flies of 

 Fo (culture 361) to each other. Nearly all of these matings gave 

 only eosin flies in Fi, but one pair gave half of the flies eosin and 

 the other half eosin ebony. These eosin ebony flies were hetero- 

 zygous for the cream gene since their father was homozygous 

 cream. When these F3 eosin ebony flies were bred together a 

 quarter of their offspring (F4) were cream III ebony. One of 



TABLE 14 



The Fi offspring from the outcross of a cream III female to an eosin ebony male 



the resulting cream III ebony males was outcrossed to an eosin 

 female and their Fi eosin daughters were backcrossed to fresh 

 cream III ebony males from the general stock which had mean- 

 while been made up and carried on (table 15). 



As soon as this experiment began to produce results, a great 

 variation in the amount of crossing over in the various cultures 

 was noticed. The ebony stock used throughout this experiment 

 was supposed from previous tests (Sturtevant, '13) to be homo- 

 zygous for a modification of the amount of crossing over in the 

 third chromosome. This modifier (Cm) when in the hetero- 

 zygous form, greatly reduces the amount of crossing over, and 

 it was this kind of result that was expected. Cultures 580 and 

 582 of table 15 gave a low percentage of crossing over, and in the 



