538 Edmund B. Wilson. 



any direct relationship between these two classes of chromosomes. 

 Montgomery has called attention to the fact that the ra-chro- 

 mosomes vary greatly in size in different species, graduating down 

 to excessively minute forms (such as those occurring in Archi- 

 merus.) It is evident that these chromosomes have undergone 

 a symmetrical reduction which, if continued, might lead to the 

 disappearance of both; and such a process, if repeated, would 

 lead in the history of a species to a progressive and parallel 

 reduction of the number in both sexes. When these facts are 

 compared with those presented by the idiochromosomes the 

 thought can hardly be avoided that the reduction of the m-chro- 

 mosomes may be correlated with a corresponding change that is 

 taking place equally in both sexes; while the reduction of the 

 small idiochromosome may represent a change that is taking place 

 more rapidly in one sex than in the other, or affects one sex only. 

 4. How the foregoing conclusions and suggestions regarding 

 the idiochromosomes and heterotropic chromosomes will square 

 with McClung's hypothesis ('02, 2) and my own similar sug- 

 gestion ('05) that these bodies may be in some way concerned 

 with sex-determination, does not yet clearly appear from the 

 known data; but there are some considerations that are too 

 interesting in this connection to be ignored. If the heterotropic 

 chromosome- be a univalent body the conclusion is unavoidable 

 (since the spermatogonial number is odd) that in the production 

 of males, the number of chromosomes contributed by the two 

 germ-cells cannot be the same. To this extent the facts -har- 

 monize with the view of McClung; but further consideration 

 gives reason to doubt some of the more specific features of his 

 hypothesis. The presence of the heterotropic chromosome in 

 the male by no means proves that it is of paternal origin in fer- 

 tilization, still less that it is specifically the male sex-determinant 

 indeed, I believe the facts point in the opposite direction. In 

 Anasa, for example, where the spermatozoa possess either ten 

 or eleven chromosomes, offspring (males) having twenty-one 

 would be produced by the fertilization of an egg having ten chro- 

 mosomes by a spermatozoon having eleven (as McClung would 

 assume); but the same result would follow from the fertilization 

 of an egg having eleven by a spermatozoon having ten. I believe 

 the second of these alternatives to be the more probable one for 

 the following reasons: According to my view, the heterotropic 



