INHERITANCE 191 



mean that sex is irrevocably fixed at fertilisation ; like all 

 other genes those which ordinarily determine sex require 

 appropriate external and internal conditions in which to 

 operate; and we shall return to this question in considering 

 sex- differentiation as part of the physiology of development. 



It is interesting to note how nicely balanced the genetic 

 factors influencing sex- differentiation may be. This is well 

 seen in experiments of Goldschmidt on the gypsy moth, 

 Lymantria. Individuals from the same local races of this 

 widely distributed form when bred among themselves produce 

 a normal sex ratio ; when individuals of different local races are 

 crossed the relations of the sexes among the offspring may be 

 abnormal. If females from a European race are crossed with 

 males from a Japanese race, the offspring are normal males 

 together with females showing a num.ber of modifications in 

 the direction of maleness ; the would-be females are intersexual . 

 Reciprocally, when a Japanese female is mated to a European 

 male the F.i generation is normal but a definite proportion of 

 the males in the F.2 are intersexual, i.e. show modifications in 

 the direction of femaleness. Varying grades of intersexuality 

 characterise the results of crossing individuals from different 

 local races ; the grade from the cross between any two given 

 races is always the same. In the extreme case all the individuals 

 of such a cross may be of one sex, but half of these on being 

 bred back to a parent stock can be shown to have the genetic 

 constitution of the alternate sex. By making two assumptions 

 Goldschmidt has brought into line the results of a very large 

 number of such racial crosses : (i) in addition to the genes 

 for maleness for which the female has the constitution Mw 

 and the male MM (as in Abraxas where the same type of 

 sex-linked inheritance occurs) there is something dependent 

 on the constitution of the tggy transmitted therefore through 

 the female parent only, that mxodifies the degree of maleness, 

 and is denoted by the symbol F. Thus a female MmF forms 

 gametes MF and rriF while the male MMF forms gametes M 

 only ; (2) it|is^/urther assumed that in different local races the 

 efficiency of M and F respectively to influence differentiation 

 in the direction of maleness or femaleness are quantitatively 



