MENDELIAN EXPERIMENTS WITH FOAVLS. 251 



as a reticulum of pigment cells in the connective tissue. On the 

 peritoneum tlie pigmentation was slight but distinct, not only on 

 the outer wall but also over the intestines, for example over the 

 gizzard ; the surface of the testes was also black over the 

 anterior three fourths of the organs, while the postei-ior fourth 

 was wliite. 



The next thing to be done was to make a thorough examination 

 of a normal specimen of Gcdlas hanklva in order to compare it 

 with the recessive just described. Foi'tanately, the cock of the 

 original cross being now rather old and blind of one eye, was 

 available for the purpose and was placed at my disposal and killed 

 and plucked, I found a little pigment in the subcutaneous tissue 

 about the ankle-joint posteriorly, but none in the periosteum : 

 this pigment was merely a slight extension of the pigment on 

 the inner surface of the skin, which extended down tlie tarso- 

 nietatarsus as in the F., recessive. In both cases the shanks 

 appeared externally olive-green. In the Bankiva there was no 

 pigmentation anywiiere else, either in the subcutaneous tissue of 

 the abdomen or of the wing-joints, nor in the peritoneum, and 

 the surface of the testes was white. 



It is certain, then, that the pigmentation present in the F., 

 recessive is not derived from the Bankiva, but must have come 

 from the Silky, with the exception of that of the shanks, which 

 occurs in both breeds. The next question is whether the recessives 

 are all pigmented to a similar degree or whether some are desti- 

 tute of pigmentation. Of such recessives, as seen in the list given 

 in Part I. of this paper, there are three. One of these. No. 2 

 of the first bi'ood, is still alive, being kept for breeding ; so far as 

 can be seen it is in the same condition as the one which was 

 killed, as it shows the dark colour round the cloaca. The third, 

 No. 4 of the first brood, also a male, was killed befoi-el thought of 

 looking for slight degrees of pigmentation, and therefore I can say 

 nothing about it. In any case, we have the fact that one of the 

 recessives is not pure, and therefore segregation is not complete 

 and perfect in the gametes, as the Mendelian theory assumes. 

 It may perhaps be suggested that slightly pigmented F„'s ai-e not 

 recessives at all but heterozygotes, but this in the ordinary 

 sense is impossible, because svich birds do not occur in F^, in which 

 all are hetei'ozygotes. The only possible conclusion is that there 

 is a tendency to segregation, but that the recessive character 

 when it separates is no longer pui-e but is, at any rate in some 

 individuals, modified by a slight degree of the opposite chaivacter ; 

 or if we adopt the pi-esence and absence theory, we may say that 

 the positive character is not entirely absent from the recessive 

 individuals, but that distinct traces of it are present in them. 



It is necessary now to make a detailed compai-ison between my 

 experiment and those described by Bateson and Punnett in the 

 paper to which reference has already been made. It will be 

 seen, that the inheritance of the pigmentation of the Silky in 

 different crosses is exceedingly curious and complicated. Bateson 



