190 



THE SECOND-CHROMOSOME GROUP 



ultimately be extended and may furnish the main basis for the accurate 

 mapping of curved. 



Meanwhile, the purple curved test, while less satisfactory because 

 of the longer interval with the attendant correction necessary for 

 double crossing-over, was more readily handled, and this led to a rapid 

 accumulation of data on the purple curved cross-over value. The 

 black purple curved triple recessive was obtained (May 1913), and the 

 back-cross itself carried out. Thirteen of the FI wild-type daughters 

 from the cross of black purple curved to wild were tested by back- 

 crossing singly to males of the triple form. These same parents were 

 at the end of 10 days transferred to fresh culture-bottles and second 

 broods were raised. The details of the data of these cultures have 

 already been published (Bridges, 1915), and we need repeat here only 

 the totals for the first broods (table 43). 



TABLE 43. Total offspring in the first broods of the black purple curved X 

 wild back-cross (details published J. E. Z., July 1913, p. 8). 



The second broods confirmed on a large scale the fact first brought 

 to light in the purple vestigial back-cross (table 33), that in the second 

 chromosome the amount of crossing-over changes with the age of the 

 mother (see Bridges, 1915). 



The numerical relations in the first broods of this experiment con- 

 firmed the position of curved as already mapped. A map of the second 

 chromosome was constructed on the basis of the data then on hand 

 (October 1913), and was as follows: 



b p, v g c 



0.0 5.3 17.5 25.0 



The black purple curved back-cross was carried out in only one of 

 the four possible ways, and is therefore unbalanced. However, the 

 results showed that inviability was negligible. What little deviation 

 there was from expectation can be attributed to curved, which still 

 appeared to the extent of 97 flies for every 100 expected. 



ALTERNATED BACK-CROSSES. 



In cases where only one type of back-cross is to be made, the poorest 

 type is that in which all the mutants are together, as was the case in 

 the black purple curved X wild experiment just cited. The flies 

 having the most mutant characters are relatively the least viable, and 

 this type of cross furnishes the highest proportion of such individuals. 



