640 THE PHYSIOLOGY OF REPRODUCTION 



males (1) produce offspring which are of both sexes, but all of 

 the grossulariata variety, the converse cross (6) yields grossu- 

 lariata males and lacticolor females. It is concluded, therefore, 

 that the wild and presumably pure grossulariata females are 

 heterozygous for sex, femaleness being dominant and maleness 

 a homozygous recessive character. All the females are be- 

 lieved to produce male-bearing and female-bearing gametes in 

 equal numbers, whereas all the males appear to produce only 

 male-bearing spermatozoa. According to this view, in gameto- 

 genesis there is a repulsion between the female determinant and 

 the grossulariata determinant, so that each gamete can be the 

 bearer of one or other of these two characters, but not of both. 

 The results obtained by Miss Durham in her experiments 

 on cinnamon canaries are expUcable on a similar hypothesis. 

 When a cinnamon male was mated with a green female, the 

 female offspring were cinnamon and the males green ; but 

 when a cinnamon female was paired with a green male, all the 

 offspring of both sexes were green.-^ Where, however, a green 

 cock of the second generation (the F^ generation produced by 

 crossing) was mated with a cinnamon hen, both green and 

 cinnamon birds of both sexes were produced ; but when a green 

 cock of the second (F^^) generation was crossed with a green hen 

 the resulting male birds were all green, but the females were of 

 both types. A more complex case of a like kind has been 

 brought to light by Bateson and Punnett in their investigation 

 on the heredity of the black pigmentation of the silky fowl in 

 its crosses with brown Leghorns and other fowls with light 

 shanks. Here two allelomorphic characters, in addition to the 

 two sex determinants, are concerned, but Bateson and Punnett 

 state that the facts point very clearly to some such solution 

 as that indicated by Doncaster's experiments with Abraxas. 

 They suggest further that whereas in Vertebrates it is probable 

 that the female is heterozygous as regards sex (the production 

 of secondary male characters in castrated females supporting 

 this view), the work of Potts and Smith (see p. 308) on parasitic 

 castration in Crustacea points to the converse conclusion that 

 in these animals the male is heterozygous, assuming definite 



' Durham, Report to the Evolution Oommittee of the Royal Society, IV., 

 London, 1908. Of. p. 637. 



