OF MUTANT CHARACTERS. 183 



was entirely due to inviability. Slight differences of this order, but 

 many of them undoubtedly significant, are continually appearing in 

 our work. Other known causes of linkage variation besides inviability 

 are: differences in the age of parents (Bridges, 1915) or of the tempera- 

 tures at which the experiments are conducted (Plough, 1917), or 

 mutant "cross-over" genes (Sturtevant, Muller, and Bridges), and 

 probably to several other internal and external factors not yet analyzed. 

 The best that can be done in correction is to calculate mean values 

 from as many experiments as possible where none of the recognized 

 causes of variation are especially active, and thus obtain a sort of 

 composite picture of the "normal" condition. 



THE VARIATION IN CROSSING-OVER WITH AGE. 



The reason for raising second broods in these experiments was to 

 obtain more offspring from each female and thus secure a more trust- 

 worthy index of the genetic behavior of individual. This practice was 

 extended to all the work at this time, and was continued until a compar- 

 ison of the cross-over values of the first and second broods brought out 

 a remarkable relation in the cases involving the second chromosome. 

 There was found to be a change in the amount of crossing-over, so 

 that both in the totals for each experiment and in a great majority of 

 the individual cultures the cross-over value had fallen significantly. 

 Equally surprising was the fact that there was no such change in the 

 case of the first chromosome, and this added another proof of the 

 distinctness of our linkage groups — that is, of the individuality of the 

 chromosome involved. The first case in which this decrease for the 

 second chromosome was clearly seen was that of the back-cross tests 

 of the purple vestigial linkage given in tables 33 and 34. Of the 8 

 females whose tests are given in table 33, seven showed a decrease in 

 percentage of crossing-over and only one (F) showed an increase, 

 which, however, was smaller in amount than the smallest of the 

 decreases. In the complementary case "repulsion" (table 34) all 6 

 females showed a decided drop. The totals likewise reflected this 

 same change; the decreases were 2.9 and 4.9 units respectively. The 

 cross-over value calculated from the balanced second broods was 8.0, 

 a decrease of 3.8 units, or, compared with the corresponding cross- 

 over value (11.8) from the balanced first broods, a 32 per cent decrease 

 from the normal amount. Many other experiments have confirmed 

 the fact of change in crossing-over frequency with the age of the 

 mother, and a partial analysis has been made. 



THE LOCUS OF PURPLE-A TWO-POINT MAP. 



The repetition of the purple vestigial back-crosses was not carried 

 out until the summer of 1913; meanwhile considerable progress had 

 been made with, the mapping of the second chromosome. The test 



■ I 



