156 



HAROLD H. PLOUGH 



The difference in the percentage of crossingover is so striking 

 as to leave no doubt that it is caused by the change in tem- 

 perature at which the F x females were hatched. These females 

 are offspring of the same females, and as larvae eat practically 

 the same food, but develop at different temperatures and there- 

 fore the interval before hatching is shorter or longer as the case 

 may be. The X 2 values for the control with either heat or cold 

 are so great that there is no possibility at all that the differences 

 can be due to random sampling, and it is very obvious that these 

 are greatly in excess of the X 2 values found in an earlier para- 

 graph for the greatest observed difference between controls. 

 The increase in the percentage of crossingover for both B and C 

 is roughly 100 per cent for the first class, or shorter region of the 

 chromosome and 33 per cent for the longer or second region. The 

 fact that longer regions, which permit of greater double cross- 

 ingover tending to obscure the actual results, show smaller per- 

 centages of observed increase, is brought out very clearly in 

 later experiments. The most striking thing about the result 

 as a whole is that both a temperature considerably above and 

 one considerably below the normal give approximately the same 

 result — an increase in the amount of crossingover, or conversely 

 a reduction in the strength of linkage. 



The increase caused by heat was confirmed with the different 

 stock containing the mutant characters — black-purple-vestigial. 

 In this case the temperature used for the test was 31.5°C, 

 slightly higher than in the previous experiment. The results 

 as shown by the backcross were as follows: 



The control value for the black-purple region is here milch 

 larger than usual, perhaps due to the presence of some modifying 



