August i8, 192 i] 



NATURE 



781 



distribution of sex is only a symbolical way of 

 representing what actually happens when the 

 mechanism of the X-chromosomes is set to work ; 

 or, as we put it occasionally, both sets of facts 

 express the same thing in different language. 



Recent work has proved the correctness of such 

 assumptions. We need mention only that in the 

 fly Drosophila, where breeding work showed the 

 male to be the heterozygous sex, cytological in- 

 vestigation also demonstrated the existence of an 

 X-Y group in the male (Morgan and collabor- 

 ators) ; in moths, where genetic proof exists that 

 the female is the heterozygous sex, the existence 

 of an odd X-chromosome in the female was con- 

 clusively shown (Seiler). But what we may regard 

 as final proof was furnished by Bridges when he 

 analysed cases in which unexpected genetical be- 

 haviour of sex-linked characters was shown to 

 be explicable on the assumption of a non-disjunc- 

 tion of sex-chromosomes during the meiotic 

 division, and when he was able to add cytological 

 evidence of such an event to the genetic proofs. 

 Thus we are led to believe that the mechanism 

 of the distribution of the two sexes among the 

 offspring is perfectly known ; it is furnished by the 

 distribution during meiotic division of the sex- 

 chromosomes, carrying, among other factors, the 

 sex-differentiators. We are confident that the 

 little opposition which is still encountered occa- 

 sionally will soon vanish before the weight of facts 

 in favour of such conclusions. 



A knowledge of the mechanism at work is a 

 safe basis from which we may attack the second 

 part of the problem of sex and so find an answer 

 to the question : How does the one-X— two-X 

 mechanism act physiologically in order to secure 

 the differentiation of one or the other sex? The 

 first attack upon this problem has been made by 

 analysing a phenomenon which we have termed 

 "intersexuality," and the main line of the facts 

 and the analysis in question are given below. 



The work was done with the gipsy-moth, in 

 which the female is the heterogametic sex and 

 the mechanism of the distribution of sex is per- 

 fectly normal. The phenomenon of intersexuality 

 occurs, then, as breeding experiments show, with- 

 out any disturbance of this mechanism. Inter- 

 sexes — i.e. individuals which show definite mix- 

 tures of the characters of both sexes, and, as a 

 whole, appear to occupy a definite position 

 between the two sexes — are produced regularly 

 and at will in crosses between different geo- 

 graphic races of the gipsy-moth. If, for example, 

 a female of the Japanese race from Tokyo is 

 crossed with a South European male, all the off- 

 spring are normal ; in the reciprocal cross, however, 

 all males are normal, but all would-be females 

 intersexual. Or, again, if we cross a female of a 

 Japanese race from Hokkaido with a male from 

 Fukuoka, all the offspring are normal, but in the 

 reciprocal cross all females are normal and all 

 would-be males intersexual. 



If we fix our attention, for the sake of sim- 

 plicity, only on the intersexual females — i.e. inter- 

 NO. 2703, VOL. 107] 



sexes with the factorial and chromosomal con^ 

 stitution of a female— we may state that the 

 majority of the different races belong to one of 

 two categories — first, what may conveniently be 

 termed weak races ; and secondly, strong races, 

 which are those the males of which, if. 

 crossed with the female of a weak race, produce 

 normal males and intersexual females. In testing 

 the different strong races at our disposal in crosses 

 with females of any particular weak race, we find 

 among the strong races a graded series according 

 to "strength." The males of one strong race 

 produce with the weak female a low type of inter- 

 sexuality, individuals which exhibit only slight 

 addition of maleness to their female constitution. 

 Another strong race produces with the same race 

 of females a higher type of intersexuality ; still 

 another may produce a high grade of intersexual 

 females ; while a fourth may finally trans- 

 form all would-be females into males, which 

 cannot be distinguished (except by breeding tests) 

 from genetic males. If we test the different weak 

 races by crossing their females with any particular 

 race of strong males, we find again a series of 

 degrees of weakness as shown by the lower or 

 higher type of resulting intersexuality. From such 

 experiments it follows that female intersexuality 

 is produced if a female of a weak race is crossed 

 with a male of a strong race ; further, that the 

 grade of intersexuality depends upon two vari- 

 ables — viz. the relative degrees of weakness and 

 strength of the parental races ; in other words, Lt 

 depends upon a quantitative relation of what we 

 have termed weakness and strength. 



By applying breeding tests it was shown further 

 that strength follows in inheritance the distribu- 

 tion of the X-chromosomes or the sex-factor. 

 Strength must therefore be regarded as a property 

 of the well-known Mendelian sex-factor located in 

 the X-chromosome. What we have termed weak- 

 ness, iiowever, is inherited purely maternally* • 

 This may mean that it is transmitted within the 

 protoplasm or the Y-chromosome, and in any event 

 it must be equally present in every e.^^. All these 

 facts show clearly' that an explanation on ordinary 

 Mendelian lines is not possible. Something has to 

 be added to ordinary Mendelian symbolism in order 

 to account for the facts, and this addition is the 

 assumption that the factors in question are pos- 

 sessed of a definite valency which acts in a quanti- 

 tative way. 



The X-chromosome contains the factor for 

 maleness, whereas the factor for femaleness is 

 inherited maternally. The quantity of the latter 

 is constant for each ^^^^ whereas the quantity of 

 the former is double in the male (XX), single 

 in the female (X). If there exists such a 

 normal relation that the one male quantity is less 

 efficient than the female quantity, while two male 

 quantities act more strongly than the constant 

 female quantity, and, further, if it be assumed 

 that the higher quantity controls sexual differ- 

 entiation, it is obvious why normally one or the 

 other sex is produced, although each ^^^ might, 



