328 HERMANN J. MULLER 



This is because most losses of genes, if they occurred, would be 

 recessive (i.e., one 'dose' of a factor usually has the same effect as 

 two), as indicated by the fact that the one-X part of a mosaic is 

 like the XX part in respect to all characters affected by X except 

 the sex characters (provided the two X's of the XX part are alike) . 

 Now, if, as seems likely, recessive losses are more apt to occur 

 than recessive additions of genes (since one 'dose' usually has the 

 same effect as two) the Y-chromosome will gradually become 

 functionless. On a priori grounds, we should expect such a fate, in 

 any species, for the sex-chromosome peculiar to the heterozygous 

 sex, if crossing-over never occurs between it and its homologue.* 



These considerations will also explain the size difference which 

 often exists between X and Y, the apparent unimportance of super- 

 numerary Y-chromosomes in development, when these occur, and 

 other irregularities of the Y-chromosome. Hitherto it has been a 

 mystery why the difference merely in the sex factor or factors con- 

 tained in the sex chromosomes should often be correlated with such 

 a large difference between the two chromosomes, whereas differ- 

 ences in respect to other factors did not involve any visible size 

 differences in the chromosomes concerned, which presumably 

 contain hundreds of genes. 



There is, therefore, all things considered, no cause for surprise in 

 the fact that no mutations have occurred involving a group of genes 

 transmitted only from father to son, and thus following in their 

 distribution the Y-chromosome. The chief gap, if it may be so 

 termed, then remaining in the parallelism between the configura- 

 tion of the chromosomes and the distribution of genes in Droso- 

 phila has been due to the fact that no genes were found in a fourth 

 independent group (a third independent non-sex-linked group) 

 to correspond with the fact that there are three pairs of autosomes. 



. ^ Conversely, where this chromosome appears degenerate or different genetic- 

 ally from its homologue (in addition to the difference in the sex factors) we should 

 expect to find no crossing-over between the two sex-chromosomes in the heterozy- 

 gous sex, i.e., complete, not partial, sex-linkage. This argument applies to an 

 known cases of sex-linkage, for in all these cases a recessive mutant factor call 

 manifest itself in the heterozygous sex, proving that the sex-chromosome peculiar 

 to the heterozygous sex is different from the other sex-chromosomes in that it 

 contains no dominant normal allelomorph of the mutant factor. 



