MKXUKMAN EXPERIMENTS WITH r<J\VI,S. 247 



of F, vvliicli 1 have <le.sciilMi<l is not tlie .same jis the white of the 

 original Silky. Bateson states that the adnlt Silky is pure white 

 as a rule, hut the down of the chickens has some colour usually 

 though perhaps not always. The colour in this case is buff, not 

 1)1 ue or black, and it occurs on the sides of the head and on the 

 rump. This again has nothing to do with the result I have 

 described, for the colour in my F^ whites was not oVjserved in the 

 chicks, or in the mature plumage when first developed, but 

 appeared later, and gradually increa.sed in intensity. I have 

 not examined pure Silky chicks, as I have not reai-ed any. 



It is, I think, certain tliat in this particular case the recessive 

 character derived from a cross is not pure, and this naturally 

 suggests that no recessives derived from crosses are perfectly pure. 

 The determinants of these characters in the hybrid individuals 

 have been mixed in the F, individuals with the determinants of 

 other characters. Whatever the determinants may be, whether 

 s[)ecial chemical compounds such as ferments, or parts of chromo- 

 somes, it is possible that the mixing in F, is such that each 

 determinant is affected, tainted as it were by the other. If this 

 were the case it would tend to throw doubt on the whole 

 conception of species and varieties reared on the Mendelian 

 phenomena. It may be that segregation occurs not because the 

 determinants or factors in the gametes are indivisible units, but 

 l)ecause the determinants of certain paii-s of characters are too 

 different to mix completely, while others less different would 

 show more mixture. In any case the lecessive white in my 

 experiment is not pure, and it is improbaVjle that it is the only 

 case of its kind. I hope next sea.son to l>reed from these whites 

 and study the heredity of the colour character in the F, 

 generation. 



Pi-of. Punnett, in conversation with me on this subject, made 

 two different suggestions in explanation of the cfjlour in the 

 white recessive. The first wa.s that the Silky is not a perfectly 

 white breed, but normally carries a little yellow colour, which is 

 usually present in the cocks but not in the hens. As the stock 

 of Silkies from which my cross came wei-e all hens, and I ha\ e 

 no experience of Silky cocks, I cannot disprove this suggestion, 

 but can only point out that it does not agree mth the account of 

 Silkies given by Davenport, nor with that given in Wright's 

 ' Book of Poultry,' which says the plumage should be pure white. 

 The second suggestion was that the yellow or yellow to red colour 

 of the Bankiva might be a distinct factor from that of the black 

 ]>igment, and therefoie be segi-egated independently. This seems 

 to me an untenalde hypothesis, as, if it were true, we ought to get 

 wholly Vjlack and wholly yellow or red birds in F^, and no such 

 c^ses have occurred. ^loreover, the F„ bird No. 1 in my experi- 

 ment has produced one or two single feftthei"s on the back which 

 contain >)lack pigment as well as yellow. I have plucked out one 

 such feather anrl preserved it. 



