326 HERMANN J. MULLER 



It will be seen that the above grouping of genes leaves one pair 

 of autosomes (presumably the small ones) and the Y-chromosome 

 without any genes to correspond. As regards the Y, it seems at 

 first sight surprising that no mutations have been found involving 

 genes which follow it in their distribution (i.e., genes which are 

 always transmitted from father to son) , for it is of about the same 

 length as the X, in which over thirty genes are known. This fact 

 would force us to one of three conclusions: (1) Either the 

 genes connected with the Y-chromosome for some reason do not 

 mutate; or (2) these mutations are all recessive to dominant normal 

 allelomorphs present in X (in spite of the fact that mutations in X 

 are not dominated, conversely, by allelomorphs in Y); or (3), as 

 Mr. C. W. Metz first suggested, genes are degenerate or entirely 

 absent in the Y-chromosome. The first conclusion is a priori 

 unlikely; the third may now be supported by a number of consid- 

 erations, which it will be of interest to discuss briefly in this con- 

 nection, before proceeding to our account of the fourth chromo- 

 some. 



In the first place, the Y-chromosome is known to vary greatly 

 in size and number in closely related species of animals. Secondly, 

 as Mr. C. B. Bridges has recently shown, a female occasionally, ow- 

 ing to an abnormal reduction division ('non-disjunction'), receives 

 a Y-chromosome in addition to the two X's, yet such a female is in- 

 distinguishable from the ordinary form, which contains no Y. The 

 Y-chromosome therefore either contains no genes or else only genes 

 which are allelomorphic to those in X, but never dominant to them. 

 This is proved also by the fact that mosaic flies sometimes develop, 

 of which a part of the body is female but another part is male, owing 

 to the accidental loss of one of the X-chromosomes in an embry- 

 onic cell-division. These male parts must have an X, but no Y, 

 and yet they are indistinguishable from corresponding parts on 

 real males which contain a Y. Furthermore, Mr. C.B. Bridges has 

 obtained (again by 'non-disjunction') males which must have two 

 Y's, yet these also show no peculiarities. It is therefore certain 

 that Y either contains no genes at all which have an effect upon 

 the individual or it contains genes allelomorphic to those in X. 

 It can be proved, however, that if it contains allelomorphs to the 



