INHERITANCE OF PLUMAGE-COLOR. 13 



Table 3 shows that when two imperfect crests are mated the off- 

 spring have imperfect crests (table 3, A). This indicates that absence 

 of occipital feathering is recessive (R) to its presence. When perfectly 

 feathered individuals that are probably hybrids between dominants (D) 

 and recessives (R), i.e. DR's, are mated with imperfectly feathered con- 

 sorts (RR) an equal number (16 : 17) of perfectly crested and imperfectly 

 crested are thrown (table 3, C, ). When two heterozygous individuals 

 are mated there result about 75 per cent perfectly crested offspring to 

 25 per cent imperfectly crested actually 11:4 (table 3, B, /?). When 

 a pure dominant is mated to a pure recessive all offspring are perfectly 

 crested (table 3, C, /?). When both parents have a perfect crest (being, 

 therefore, DD or DR) all offspring have the perfect crest (table 3, B, a). 

 Altogether it appears that, in crested birds, absence of feathers on the 

 occiput is recessive to their presence. Consequently we have in the 

 crest two pairs of allelomorphs : crest C and absence of crest c: occipital 

 feathering and absence of occipital feathering (baldness) o. Then CO 

 is the zygotic formula for perfectly crested, Co for imperfectly crested, 

 cO for the ordinary plain-head, while a fourth possible combination, co, 

 should occur in one-sixteenth of the offspring in F 2 and should be 

 partly bald plain-heads. This combination I seem not yet to have 

 acquired nor have I seen it mentioned ; yet it is to be expected. 

 Nevertheless it is possible that baldness is coupled with crest. 



The Mendelian nature of the inheritance of baldness sufficiently 

 explains the view of some canary breeders cited at the beginning of 

 this section. For two crested parents may throw from 100 per cent 

 to 25 per cent bald offspring according as they are RR or DR in 

 respect to baldness. But, since all crestless parents are DD or DR 

 in respect to baldness, a crested bird mated with a crestless can give 

 at most only 50 per cent bald-headed and may give none at all. 



Series II. THE INHERITANCE OF PLUMAGE-COLOR. 

 STATEMENT OF PROBLEM. 



We have seen that the original canary was "green" and that out 

 of such a pigmented canary there was evolved, over two centuries ago, 

 a yellow race. Although the historical evidence is incomplete it favors 

 the view that the yellow form arose suddenly, as a sport. Similarly, 

 within recent years, in the Australian grass parakeet, or budgerigar, a 

 yellow variety has arisen under domestication. The interpretation of 

 these color changes must rest on the facts of chemical physiology. 

 According to Krukenberg (1882, 21) there is no green pigment in the 

 wild canary, but the green is due to a yellow and a dark pigment. The 

 yellow pigment is of a fatty nature, is easily extracted by boiling 



