Maech 5, 1909] 



SCIENCE 



399 



shared with the female. Can we reconcile 

 these facts with the idea that the female is a 

 male plus something else ? I think so, but we 

 must concede also the possibility that the male 

 may possess certain qualities not merely not 

 manifested by the female, but even not pos- 

 sessed by it. I would ofEer the suggestion that 

 we have a mechanism suitable for the trans- 

 mission of characters exclusively male in the 

 T-element described by Wilson, the " synaptic 

 mate " of the X-element, which takes the place 

 in the gamete of a lacking X-element, and 

 which would not be borne by a gamete pos- 

 sessing that element. If the primary differ- 

 ence between male and female is a defect in 

 the male, the lack of something present in the 

 female, that very defect would constitute a 

 likely place in the germ-cell for new structures 

 to find lodgement, which, behaving as the " syn- 

 aptic mate," the material counterpart of the 

 X-element would pass only into gametes lack- 

 ing X, and so would produce structures pe- 

 culiar to the male, and unrepresented in the 

 female. 



If this idea should prove to be correct, then 

 we should have to revise the generalization to 

 which Wilson gives expression " that so far as 

 the eggs are concerned (and also those sper- 

 matozoa that contain the X-element) . . . 

 every gamete contains factors capable of pro- 

 ducing both the male and female characters, 

 and that this is also true of all the zygotes." 

 If the T-element should prove to be the basis 

 of characters purely male, then such characters 

 would not be represented at all in gametes 

 containing X, and cases like that described by 

 Darwin, in which the hen-pheasant transmits 

 to its hybrid male offspring in crosses char- 

 acters of the male of its own species, could 

 have but one interpretation, viz., that the hen- 

 pheasant produces gametes lacking the X-ele- 

 ment, as well as those which possess it. In 

 other words, the hen-pheasant would seem to 

 be a sex-heterozygote and so to fall in the same 

 category of cases as the moth, Ahraxas gros- 

 sulariata, category B already mentioned. If 

 so, the male pheasant should be incapable of 

 transmitting in crosses characters peculiar to 

 the female pheasant, if such exist. 



This line of thought emphasizes the impor- 



tance of reciprocal crosses in unraveling the 

 mysteries of sex-inheritance and of the inherit- 

 ance of secondary sexual characters. If the 

 two categories of cases A and B really exist, 

 there should be this difference between them. 

 In A the male may transmit recessive charac- 

 ters peculiar to the female, but the reverse 

 relation does not hold. In B, the female may 

 transmit recessive characters peculiar to the 

 male, but the reverse relation does not hold. 



Further, there should be a difference in the 

 two categories of cases in the Mendelian na- 

 ture of fixed sexually dimorphic conditions. 

 In category A, male secondary characters must 

 be dominant in order to be fixable, i. e., they 

 must be represented in the T-element by some- 

 thing not found in the X-element, but which 

 will manifest itself even in the presence of the 

 X-element. In category B, male secondary 

 characters must be recessive in order to be fix- 

 able, i. e., they must have their basis in the 

 absence from T of some element present in X, 

 which absence will not be manifested if even 

 a single X-element is present. For example, 

 in Abraxas the pale lugens character is mani- 

 festly a defect character, due to lack of some- 

 thing found in grossulariata individuals, L 

 being recessive to G. The gametic coupling 

 of the female character with the lugens char- 

 acter, whenever a doubly differential cell divi- 

 sion occurs, is doubtless due to the fact that 

 the grossulariata character acts as the "syn- 

 aptic mate " to the X-element, leaving absence 

 of G (i. e., L) associated with X. If in this 

 cell-division G were associated with X, instead 

 of with T, then it would be possible to produce 

 a stable sexually dimorphic race, with Lc?s 

 and G?s, but the relation being what it is, no 

 stable race can be formed in which the two 

 sexes are G and L, respectively, but only races 

 purely G or purely L in both sexes. 



On the hypothesis suggested in this paper, 

 accordingly, we can account for the fact that 

 secondary sexual characters are more common 

 in the male, if not its exclusive possession, 

 even though the male is, as compared with the 

 female, a defect race, or regressive variation. 

 Transference to the female of characters origi- 

 nally possessed by the male alone could be ac- 

 counted for by the duplication of the T-element 



