STUDIKS ON HIGH AND LOW NON-DIS.IL'NCTION 99 



no such common piece exists in which however we have a percentage 

 which is much higher than the ordinary low percentage even if we 

 pay regard to the error of sampHng. Thus in the cultures recorded 

 in tal)le 11 the exceptional females are homozygous for a piece of the 

 A' which lies to the right of cut hut in table 15 they are homozygous 

 for a fragment lying to the left of cut. From the first of these tables 

 we find, however, the percentage 15,58+ l,oi and from the second the 

 percentage 8,02 db 0,8o. Likewise in table 13 the exceptional females are 

 homozygous with respect to a piece of X which has its left hand end to 

 the right of vermilion and in table 16 the part of Ä' has its right hand 

 end to the left of vermilion but in the first of these tables the 

 percentage of exceptions is 13,28 ib I,«« and in the second one the per- 

 centage is 9,51 rb 0,90, 



It seems thus that there probably can not be any genes in the A' 

 which are responsible for the high percentage of exceptions in the 

 eosin line of high non-disjunction. Of course, however, since in the 

 experiments I have not made use of any genes to the left of scute or 

 to the right of forked it may have happened that in all the cases I 

 had merely got flies homozygous for pieces of X lying to the left of 

 scute or to the right of forked. But this is something so improbable 

 that I think we may neglect this possibility. But even if it were so, we 

 have not explained the great variability of the exception percentage 

 arrived at in the experiments of the foregoing chapter. 



Indeed if we arrange the different percentages of exceptions from 

 the cultures where the exceptional females have been homozygous for 

 various parts of the A:s in order of their magnitude then we have 

 a series beginning with 6,93 + 0,75 (table 8) and ending with 15,5,s + l,iii 

 (table 11). But how is this to be explained? There is really one ex- 

 planation which fits the figures fairly well, but I wish from the be- 

 ginning to point out that this explanation must be looked upon only 

 as an alternative one until data are available to a greater extent. 



The explanation is as follows, there are no genes which are res- 

 ponsible for the different percentages of exceptions, but it is the length 

 of the part of the A per se with respect to which the exceptional fe- 

 males are homozygous which regulates the percentage of exceptions, in 

 such a way that when this part is long then the percentage is high and 

 when this part is short then we have a low percentage of exceptions. 



When I started the experiments I believed that there should be 

 genes in the X responsible for the high percentages and by this process 

 of fragmenting the X and then uniting the different parts I expected 



