Julian S. Huxley 269 



and polar body would then contain one X-chromosome. This hypothesis 

 of differential maturation is of importance, as it involves the subordi- 

 nation of the chromosome-mechanism to other agencies. 



As moths are of the ZW, ZZ type, differential maturation is a 

 theoretical possibility. Since the first draft of this paper was written, 

 however, I have not only had access to Goldschmidt's latest published 

 statement of his results ('19), but the same author has most kindly sent 

 me the proofs of a forthcoming paper ('20) which supersedes all his 

 earlier work. I will not do more here than to say that the sex-ratios 

 there obtained give further very strong evidence in favour of the 

 possibility we have already indicated for Kuschakewitsch's frogs — viz. 

 that zygotes with the chromosome-constitution of one sex may be 

 transformed into functional individuals of the other sex. Not only that, 

 but such animals, both ZW males and ^-^ females, have been bred from, 

 and, though Goldschmidt himself realizes the need for further data, all 

 the results can be interpreted on this hypothesis, while it is difficult to 

 account for them on any other. In other words, it appears that in these 

 moths, abnormal sex-ratios are first produced by overriding chromosome 

 constitution, and that sex-ratios abnormal in the opposite direction are 

 produced in the next generation, exactly as in Boulenger's fish. Harrison's 

 results ('19) on moths of the sub-family Bistoninae are in many ways 

 similar, but are complicated by being for the most part got from species- 

 or genus-crosses. They too are strong evidence for the possibility of 

 individuals of the chromosome-constitution of one sex being trans- 

 formed into functional individuals of the other sex, although this has 

 not been tested out by controlled breeding experiments with such trans- 

 formed individuals. 



Mention should also be made of Keilin and Nuttall's intersexual 

 lice ('19). Here again we have a graded series from " hermaphrodites " 

 (intersexes) which are preponderatingly of one sex through those with 

 equal development of male and female characters, to those preponderat- 

 ingly of the other sex. Whether, however, this is due to overriding of 

 chromosome-constitution cannot be said until the fuller genetic and 

 cytological analysis which is now in progress has been made. 



Let us now return to our fish. If we suppose that some external 

 factor upset the normal sex-ratio during the first period, we have a 

 provisional explanation of the facts. This factor must have worked 

 in the opposite direction from late fertilization in frogs, converting 

 animals of male constitution into somatic females. Further, in this case 

 not all the males were thus converted. , This is, however, paralleled in 



