260 Dr. E. A. Cockayne on 



specimens the male wings are considerably larger than the 

 female, contrasting strongly in this respect with the second- 

 ary somatic hermaphrodites, which form the subject of 

 this paper. 



The fact that the somatic characters are almost in- 

 variably the same on both sides in these gynandromorphs 

 from Royston is important, because it disproves the view 

 that either fertilisation of an ovum by two spermatozoa 

 or fertilisation of one half of a prematurely divided ovum 

 gives rise to the condition of gynandromorphism ; though 

 both suggestions have been advanced at one time or 

 another to explain ordinary halved gynandromorphism. 

 An irregular division of chromosomes at the first cleavage 

 of the fertilised ovum may account for the halved examples 

 of gynandromorphous coridon, but will not account for 

 bilateral or crossed forms. In these far rarer specimens 

 an irregular division probably took place earlier, perhaps 

 at the second maturation division before the fertilisation 

 of the ovum. In all cases there is probably some ante- 

 cedent abnormality of the chromosome. 



In the majority of lepidopterous insects the female is 

 a heterozygous dominant for sex, the male a homozygous 

 recessive, and A. coridon is probably no exception. The 

 occurrence of aberrations peculiar to the female, such as 

 abs. tithonus, Meig. (syngrapha, Kef.), and semisyngrapha, 

 Tutt, supports this hypothesis. In the simpler cases of 

 this kind the spermatozoa always have sex chromosomes 

 of one kind, but the ova are of two kinds : one with a 

 large sex chromosome, which on fertilisation produces a 

 female, the other with a small sex chromosome, which 

 produces a male. If we suppose that some part of the 

 large sex chromosome is lost either before or after fertilisa- 

 tion, femaleness may be incompletely dominant. A some- 

 what similar condition may exist in the case of the chromo- 

 somes bearing the secondary sexual characters, and loss 

 of part of the dominant chromosome (that for female 

 secondary sexual characters) may cause an imperfect 

 dominance of female over male secondary sexual char- 

 acters. If the dominance becomes much reduced, the 

 male element may become visible in the form of numerous 

 blue scales and androconia. If the margin is narrow, it 

 may only be visible in the form of a few blue scales and 

 androconia; or, where there is almost a perfect balance 

 between the male and female elements, only one or two 



