64 The Walnut Comb [ch. 



It was not a little surprising to see so striking and 

 characteristic a structure as the walnut or Malay comb 

 appear with strict regularity as the product of two such 

 dissimilar parents as the rose and the pea. Not only is 

 the general appearance of the walnut quite distinct from 

 these, but the presence of the hairs constitutes a feature of 

 absolute difference, for no hairs are found on combs of the 

 usual types. Were the walnut, rose, pea, and single combs 

 found as characteristics of wild birds no naturalist would 

 hesitate to regard them as four distinct specific characters ; 

 and even as the special properties of domesticated birds I 

 suppose they would by many be regarded as evidence ol 

 long-continued selection. Nevertheless, as will be seen, 

 these four forms stand to each other in a simple genetic 

 relation and the fact suggests wide possibilities in regard 

 to many hitherto unexceptionable differences ** of specific 

 value " recognized among animals and plants. 



The interpretation of the facts was at first by no means 

 easy, and I am sorry to have been responsible for the 

 promulgation of a quite erroneous suggestion regarding 

 them. Further knowledge of kindred phenomena has, 

 however, made the elucidation of this case now perfectly 

 clear and simple. 



To return to the experimental results. Having found 

 that rose x pea gives walnut, the next thing to be done 

 was to test the genetic properties of birds thus produced. 

 This was done in two ways (i) by breeding the walnuts 

 together, (2) by breeding them with singles. In what 

 follows, the names may be abbreviated thus : R, rose ; 

 P, pea ; RP, walnut ; and 5*, single. 



Experiment showed that RP x RP gave an F^ family 

 RP, R, P, and S. The appearance of S, which was not 

 known to have been put in, is not at first sight intelligible. 

 Repeated trials proved that the ratio in which these combs 

 appeared was 



gRP : sR : 2,^ : iS. 



It was further proved by experiment that the R birds 

 were either pure R or contained the recessive S, but gave 

 no more P or RP ; that the P birds similarly could only 

 give P, or P and S ; while the 6^ were all pure to that 

 character. 



