352 



The Journal of Heredity 



possibility — that the two apparent al- 

 ternatives are as a matter of fact 

 combined — that from the outset the 

 excess of one of the sex-determining 

 substances causes a change in meta- 

 bolism, but that it also produces a 

 specific substance promoting localized 

 growth in one or the other set of sex- 

 ual characters. 



In higher animals, with gonads 

 producing an endocrine secretion, the 

 probability is that the first alterna- 

 tive is true, or quite possibly the third. 

 For animals without a gondial hor- 

 mone, however, as insects have been 

 shown to be, the probability is that 

 the second alternative, or again pos- 

 sibly some form of the third, is true. 



We should remember that in various 

 low forms, e. g., the oyster.'" seasonal 

 and age differences are associated with 

 change of sex, that in other organisms, 

 e. g.. Hydra and certain plants, 

 sex appears to be wholly or partially 

 determined by external conditions. We 

 can compare regulation of sex by 

 environment, whether internal or ex- 

 ternal, to the regulation of dimor- 

 phism by environment, which occurs 

 in certain plants. In Polygonum am- 

 phibiun, for instance, and in Ranun- 

 culus aquatilis, it is well known that 

 the plant may occur in two mam 

 forms, the aquatic and the terrestial. 

 The assumption of one or the other of 

 these forms is, however, not deter- 

 mined genetically, but by the envir- 

 onment of the plant. Its genetic 

 constitution is so constructed that when 

 the plant grows in water it will em- 

 bark on a series of reactions leading 

 to the water form, and z'icc versa, 

 when it is on land. Here the influ- 

 ence on general metabolism is pri- 

 mary, the specific form-difference is 

 determined secondarily as a result of 

 the ])rimary influence. 



Finally in the Diptera, the group of 

 insects to which the flies belong, we 

 have fairly definite proof that the 

 germinal factors responsible for sex- 

 determination, or at least some of 

 them, are of a nature which can hardly 

 be supposed to exert specific effects 



upon the development of sex-characters, 

 but can readily be supposed to act on 

 general metabolism. In a very remark- 

 able recent paper Bridges summarizes 

 his work upon triploid stocks of Dro- 

 sophila." It is impossible to go into the 

 evidence he presents, but a few points 

 important to our general argument must 

 be mentioned. In the first place, he 

 shows conclusively that sex-determina- 

 tion in Drosophila is not, as was at 

 one time held, a question of the abso- 

 lute amount of some factor contained in 

 the X-chromosome, so that two doses 

 of X invariably produce females, one 

 dose invariably males. Rather it is 

 a matter of a relation between the 

 amount of some factor carried in the 

 X-chromosome and of some other fac- 

 tor or factors carried in the rest of the 

 chromosomes. 



His triploid stock usually carries 

 three instead of the normal two sets 

 of autosomes, together with either one, 

 two, or three X-chromosomes. Let us 

 call a single set of autosomes A. Then 

 it was found that a chromosome for- 

 mula 3X :3A gave normal femaleness, 

 just as does 2X :2A in normal diploid 

 stock. IX :2A in diploid stocks gives 

 normal maleness. IX :3A in triploid 

 stocks gives sterile "super-males" ; 

 while 2X :3A — an intermediate ratio — 

 gives intersexual individuals combin- 

 ing male and female characters ; 3X :2A 

 individuals, discovered in other ex- 

 periments, were sterile "super-females." 

 Thus we have a state of affairs per- 

 fectly consonant with Goldschmidt's 

 ideas. However, Bridges was able to 

 push the matter still further in one 

 direction. Owing to the fact that the 

 sm.all fourth chromosome, though 

 sometimes present in triplicate like 

 the rest of the autosomes, was at 

 other times only present in duplicate, 

 he was aSle to show that there was 

 not a single male-determining factor, 

 but at least two (probably many) — 

 one (or one set) lodged in the fourth 

 chromosome, the rest in the other 

 autosomes. The set in the fourth 

 chromosome appears to exert a dis- 

 proportionately strong effect. Ob- 



