SECTIONAL TRANSACTIONS.— D. 333 



and associated structures were elucidated. The branchio- visceral skeleton has an 

 abbreviated development, but in the adult is quite Ranid. A spiraculum (atriopore) 

 is never developed. Finally, the affinities of the genus are discussed. 



Prof. J. E. DuERDEN. — The Zoology of the Fleece of Sheep. 



An intensive study of the fleece of the merino has called for a comparison of the 

 hairy covering of sheep, goats and mammals generally, and has revealed many facts 

 of high zoological interest. Evolutionary series in different directions are disclosed, 

 and many illustrations of recapitulation as well as of physiological adaptation. An 

 interpretation of the nature of the winter pelage of animals is forthcoming, and 

 numerous stages in the phenomenon of shedding. 



South Africa is peculiarly favourable for a study of this character on account of 

 the large variety of distinctive types of sheep and goats. The highly specialised 

 wooUed merino may be placed at one extreme, the black-head, primitively coated 

 Persian at the other, the latter with its extraordinary fat-rump to be compared with 

 the long, fat-tailed, indigenous Afrikander ; also the highly lustrous ringlets of the 

 Angora goat are to be contrasted with the curly pipes of the Karakul. 



Four morphologically distinctive types of fibres are concerned, known in wool 

 terminology as kemp, wool, heterotypes and gare. These are represented in very 

 different proportions in different types and in different ontogenetic stages. The 

 role of the heterotypes, a fibre coarse and kempy above and woolly below, is elaborated, 

 and furnishes a marked instance of convergent evolution. 



Miss Ruth C. Bamber (Mrs. Bisbee). — Segregation. 



In the course of experiments carried out by Miss Herdman and myself on colour 

 inheritance in cats, facts have come to light which do not seem to be consistent with 

 the theory of the indivisibility of the gene. 



Yellow and black in cats behave as a pair of allelomorphs. The heterozygote is 

 tortoise-shell, but because of sex-linkage, it occurs normally only in the female. In 

 a tortoise-shell female the black and yellow segregate. But all yellow cats examined 

 by us have a few scattered black hairs, and this black does not segregate from the 

 yellow. It seems as though the yellow is contaminated by black ; but contamination 

 would involve the sphtting of the factor for black — and the gene is generally believed 

 to be indivisible. 



Further, in the course of our experiments, a yellow female appeared which had a 

 few very small spots of black. Breeding experiments proved that this black did not 

 segregate from the yellow. Again, the idea of contamination suggests itself. 



Other explanations are possible, however, in both these cases, especially as the 

 exact relationship between black and yeUow is open to question. It is just possible 

 that one underlies the other. 



Later, however, we studied the inheritance of black and white in cats. Self white 

 is dominant to black. But we find that most black cats, probably all, have a few 

 white hairs. Here the relationship between the colours is known, and black, the 

 recessive, should certainly be free from any trace of the dominant white, if the gene 

 for white is indivisible. 



Bateson pointed out that the members of any series of multiple allelomorphs differ 

 from each other quantitatively only and argued that this was strong evidence for the 

 fractionation of a factor. Goldschmidt's experiments on the Gipsy moth have led 

 him to regard the members of a series of multiple allelomorphs as due to different 

 quantities of the same factor and he, therefore, postulates the divisibDity of the gene. 

 Several other workers have made similar suggestions. 



Our own observations, especially in regard to the impurity of a recessive, seem 

 strongly to suggest fractionation of a factor. 



The hypothesis of the indivisibUity of the gene was built on the foundation of 

 observed dean segregation of characters. Mendel's experiments did show such 

 segregation. Other workers have looked for MendeUan proportions and, finding them, 

 have emphasised Mendel's law — and with it the idea of clean segregation with the 

 resulting purity of the gamete. 



But this purity of the gamete seems, on closer investigation, to be relative rather 

 than absolute. Recessive black in cats is not pure from the dominant white. Other 

 cases are also known. In guinea pigs the recessive albino is not pure from the 



