THE FACTORS WHICH DETERMINE SEX 693 



{Triton alpestris). He found that if the males were starved severely 

 at a time when spermatogenesis should be active, the development of 

 the secondary sexual characters was arrested, the animals remainiiig 

 more or less "neuter," as in winter. In the following spring the 

 testes were seen to be replaced by bands of fatty tissue, and the 

 secondary characters did not appear. Two such males which were 

 abundantly fed in the next winter partially assumed the coloration 

 of the female. One was dissected on 11th January, and contained 

 only the fatty bodies which had replaced the testes. Another was 

 kept until 8th April, by which time it had become entirely female in 

 appearance. Each fatty body now contained an ovary with oocytes, 

 and there was also an oviduct on either side. Moreover, it fostered 

 the fertile eggs of a female with which it had paired some time before. 



Crew^ has described a number of frogs with abnormal repro- 

 ductive systems, and adduced evidence that these, or some of these, 

 were chromosomally females (XX in composition) which, instead of 

 developing into normal females, became transformed into " somatic " 

 males by the action of some factor or combination of factors which 

 had overridden the chromosome constitution. He points out that 

 the mating of such individuals, functioning as males, must disturb 

 the sex -ratio of the next generation, and that such an interpretation 

 may explain thg unusual sex-ratios recorded by various observers. 



Julian Huxley^ had already suggested such a possibility, and 

 applied it statistically to the case of the " millions fish " (Girardinus 

 pceciloides) for which the sex-ratio had been found to be abnormal, 

 those bred in the Zoological Gardens, as recorded by Boulenger, 

 being produced in the ratio 3 ? : 1 <?, but subsequently changing to 

 2 $ : 3 (?, and then^ later to 1 ? : (?, which last ratio continued for some 

 years and as long as the fish were bred at the Gardens. Huxley 

 has shown that these-^abnormal ratios can be explained on the 

 assumption of an overriding of the chromosome constitution by 

 means of external influences, the theory as applied to this special 

 case being worked out in some detail. 



Boring and Pearl ^ have described hermaphrodite fowls with 

 gonads and external characters as well as sex behaviour in process 



^ Crew, " A Description of Certain Abnormalities of the Reproductive Organs 

 found in Progs, etc.," Proc. Roy. Physiol. Soc. Edin., vol. xx., 1921. See above, 

 p. 663, where the work of Witschi, etc., is referred to ; also footnote, p. 700. 



2 Huxley, " Note on Alternating Preponderance of Males and Females in 

 Fish and its Possible Significance," Jo^or. of Genetics, vol. x., 1920. See also 

 " Eecent Advances in the Biological Theory of Sex," Jour. Roy. Soc. of Arts, 

 vol. Ixx., (January and February) 1922. 



3 Boring and Pearl, "Sex Studies," XI., Jour, of Exp. Zool., vol. xxv., 1918. 

 The authors say that the amount of luteal tissue in the ovary is precisely 

 correlated with the degree of external somatic femaleness. Elsewhere they 

 record luteal cells in the testes of hen-feathered males. See also Crew, Vet. 

 Jour., vol. Ixxix., 1922. * 



