STUDIES ON I1IÜH AND LOW NON-DISJl NCTION 103 



between the length of the used part of the X and the percentage of ex- 

 ceptions calcuhited three different vakies for the whole h'ngtli of the 

 X-chromosome viz. 76,o, 85,5 and 85,7 units. Now — so far as I am 

 aware — the whole known length of the X-chromosome is about 71 

 units and if we rememl^er that all the calculations are only approxi- 

 mate and that the length of the X-chromosome may be longer than the 

 part of it that we know to day then it must be said that the calculated 

 values agree very well with the value 71. 



Here it may be the place to make a short summary of the reasons 

 for the supposition that it is the length of the part of the X with 

 respect to which the exceptional females are homozygous which de- 

 termines the percentage of exceptions: 



1. Data are secured which show that XXT-females homozygous 

 for the whole X produce the high percentage (about 22 %) of excep- 

 tions and which indicates that neither the autosomes nor the Y-chro- 

 mosomes have any influence upon this percentage. 



2. XXy-females being homozygous with respect to different parts 

 of the X produce percentages of exceptions all lying between the ordi- 

 nary low percentage (4,3 %) and the high percentage of the eosin line 

 (about 22 %). 



3. In a number of different experiments XXY-females which have 

 been homozygous for parts of the X having no points in common have 

 produced percentages of exceptions which coincide neither with the 

 ordinary low percentage (4,3 %) nor with the high percentage (al)out 

 22 % ) even if we consider the error of sampling. This makes the 

 possibility of sex-linked genes for the percentage of exceptions at least 

 highly improbable since it has been concluded from other experiments 

 that if such genes existed they should have been recessive. 



4. X'XF-females which are homozygous with respect to a longer 

 piece of the X produce really a higher percent of exceptions than 

 do XX'^y-females which are homozygous for a shorter piece of 

 the X. 



5. The approximate relation y = 4,3 + 0,2 z between the percent 

 of exceptions (y) and the length (z) of the part of X with respect to 

 which the XXT-females are homozygous fits all the experiments very 

 well. 



6. The total length of the X-chromosome calculated by means of 

 this relation is in good agreement with the known length of this chro- 

 mosome. 



It must be pointed out that the essential point is not that there is 



