84 



THE ORIGIN OF GYNANDROMORPHS. 



(3) On the elimination hypothesis (text-fig. 68, 3), an egg with 

 a W fertilized by a sperm (Z-bearing) should produce a female (ZW). 

 This gives no chance to produce a male side (ZZ) by ordinary elimina- 

 tion. If by somatic non-disjunction (ZZ or no Z) it is not evident that 

 the no-Z part would develop, and if it did, why it should be plain. 



(3a) If an egg with a Z had been fertilized by a Z sperm (text-fig. 

 68, Sa), a male (ZZ) would result from which, by elimination of a Z 

 chromosome bearing the white gene, would produce female parts that 

 are zebra and male parts that are also zebra, which is contrary to the 

 actual conditions in the gynandromorph. If the other chromosome 

 should be eliminated, viz, the one bearing the zebra gene, then the 

 male part would be zebra and the female part would be plain. 



It is evident that this case can not be explained in any of these 

 ways, even though it be assumed that the color-factors are carried by 

 the sex chromosomes. And if we do treat the color-factors as sex- 



Text-figure 68. 



linked, then they can not be the same zebra-white pair of factors 

 described by Toyama in other crosses which are clearly not sex- 

 linked. To apply the above view tactily takes for granted that the 

 zebra-white pair is not the same pair referred to in other crosses. If it 

 is not, then we are not obliged to assume that zebra is dominant to plain. 

 If plain is assumed to be dominant over zebra, the gynandromorph can 

 be accounted for by Boveri's hypothesis or by elimination. Possibly one 

 might try to find an excuse for such an evasion by pointing out that 

 Toyama states that the two gynandromorphs appeared in a cross 

 between a striped French race with yellow cocoons and a common 

 Japanese race with white cocoons, and that this is not the same cross 

 as that which he described in the body of his paper, where he states 

 that the striped race had white cocoons. On the other hand, both 

 Coutagne and Kellogg, according to Tanaka, have found that striped 

 is dominant to plain, and although I can not find that they have 

 made exactly the same cross as that which yielded the gynandromorph, 

 nevertheless the cumulative evidence is strongly in favor of the view 

 that zebra is both dominant and not sex-linked. It is clear, then, that 



