Genie Control of Development 357 



compensation appears here. This leads to the question how autosomal 

 mutant loci behave in the two sexes. If dosage alone counts, the sexes 

 should be alike. This is frequently so, but frequently, also, the sexes 

 react differently. In a series of mutants producing a plexation of wing 

 veins in Drosophila, which I studied in detail ( Goldschmidt, 1945a), 

 the females were always one grade more extreme than the males of 

 the same constitution. All these facts show that the phenotype in the 

 two sexes may or may not be dependent upon dosage, and that all 

 the different cases should be explained on the same basis (which we 

 shall present below). 



Haldane (1932) had already pointed out a somewhat different 

 concept: namely, that sex-linked genes may be of such a high potency 

 of action that even the simplex condition may reach the physio- 

 logically possible maximum effect. This condition in the wild type may 

 be the result of selection in the past. The ideas of Stern and Muller 

 follow the same general trend of thinking, but they attribute dosage 

 compensation to modifying genes in the X-chromosome, accumulated 

 by selection. Of course this would also apply to the wild-type condi- 

 tion, assuming that it has an adaptive advantage. The situation for 

 mutants would only reveal what the wild-type behavior would be. 

 There would be a corollary to this: the many sex-linked mutants which 

 do not show dosage compensation would have to be regarded as some- 

 how independent of the dosage modifiers, which does not add to the 

 plausibility of the theory. Muller (1950Z?) has since gone into the 

 problem with more detailed work. He studied mutant loci showing 

 dosage compensation in one dose only in the female, or two doses in 

 the male, obtained through deficiencies and duplications, and found 

 that one dose in the female makes the expression less extreme even 

 than in the male, while two doses in the male produce a higher 

 expression than in the normal female. He even maintains that by 

 photometric methods the same effects can be demonstrated for the 

 wild-type eye in Drosophila, which always looks the same to simple 

 inspection. This would mean that the action according to dosage is 

 present but is compensated for by a modifier system, which must be 

 located in the X-chromosome (as first concluded by Stern, 1929, for 

 the bobbed alleles). Furthermore, whereas many loci in the X-chro- 

 mosome show such behavior, each must have such a modifier system 

 or share in one. It is obvious that this statistical interpretation is based 

 upon the general ideas of Fisher on selection of modifier systems in 

 the heterozygote. 



MuUer does realize that there is a possibility of an alternative 



