STUDIES ON HIGH AND LOW NON-DISJUNCTION 



91 



scute eosin daughters were crossed to wild males from stock and I 

 proceeded as in the foregoing cases. As it may he seen from tahle 9, 

 there occurred in this way 2159 zygotes with a percentage of excep- 

 tions of 12,83 + 0,72 %. And here the part common to the two A':s 

 and inherited from tlie original high non-disjiinctional A':s was the 

 whole chromosome with the exception of a piece to the left of a point 

 between scute and eosin plus a piece to the right of a point between 

 tan and forked. 



TABLE 8. 



Offspring of scute non-disjiinctional daughters from 463 table 6 and 



their exceptioncd female descendents when mated by various males. 



I think thus that the inference may be made that the lowering of 

 the percentage of exceptions in these cases is due to the absence of 

 some parts of the X-chromosome. But as a matter of fact this lowering 

 in all three cases is not great enough to reach the limits of error of the 

 ordinary low percentage, but reaches only a percentage between this 

 low percentage and the high percentage typical of the eosin line of 

 high non-disjunction. On the other hand if we compare the percentage 

 of exceptions from tables 7 and 8 their difference is 2,35 and this dif- 

 ference divided by the mean error of the difference is 2,6i. In the 

 same way the difference between the values from tables 7 and 9 is 3,55 

 and the ratio of the difference and the mean error of the difference 

 is 4,08. Finally the corresponding values for the comparison of the 

 figures from tables 8 and 9 are 5,9n and 5,07. It is therefore not possible 



