No. 524] CHROMOSOMES AND HEREDITY 



493 



thenogenetic female for producing males. Tlie facts 

 make out a strong case in favor of the view that we have 

 probably found the mechanism by means of which sex 

 is determined. 



When we try to analyze the results, however, I mean 

 when we try to make clear to ourselves how the acces- 

 sory determines sex, we fail to make good a consistent 

 story of the process. 



If we assume, as Wilson and I have done, that the re- 

 sult is purely quantitative in that the female develops 

 because the egg fertilized by the female producing sperm 

 contains one more chromosome than the egg that be- 

 comes a male; when we make this assumption, we seem 

 to leave unexplained how sex is determined in a large 

 number of cases when the odd chromosome has a 

 partner of equal size. 



On the other hand, if we assume that the accessory 



then the mate of the accessory, or one of the correspond- 

 ing chromosomes in the female, must be male-producing. 

 To make this mechanism "go" we must assume select- 

 ive fertilization; for which at present there is no evi- 

 dence. 



I shall try to indicate in the barest outline the further 

 analysis of the two statements just made. 



When the accessorv has no mate, as in the examples 

 just given, we have the problem of the three X chro- 

 mosomes. The following situation then develops. 



(A) The three X's are identical as everything we 

 know about them indicates. Their position on the reduc- 

 tion spindle both in the male and female is so far as we 

 know fortuitous. It follows then the female results when 

 two X's meet in the same egg, and a male when only one 

 X is present. This is the simplest explanation yet found 

 that is strictly in conformity with observed relations for 

 this class. It encounters five difficulties: first, it does 

 not seem to apply when the accessory has a mate of 

 equal size, if that mate be another X (see below); 

 second, sex in hermaphroditic forms is not apparent 



