90 PROBLEMS OF GENETICS 



nearly resembling the spontaneous origin of dominant factors, 

 and I cannot see such pedigrees as these without recalling Vir- 

 chow's aphorism that every variation owes its origin to some 

 pathological accident. In the evolution of domestic poultry, 

 if Gallus bankiva be indeed the parent form of all our breeds, 

 at least some half dozen new factors must have been added 

 during the process. In bankiva there is, for example, no factor 

 for rose comb, pea comb, barring on the feathers, or for the 

 various dominant types of dark plumage. Whence came all 

 these? It is, I think, by no means impossible that some other 

 wild species now extinct did take part in the constitution of 

 domestic poultry. It seems indeed to me improbable that the 

 heavy breeds descend from bankiva. Both in regard to domestic 

 races of fowls, pigeons, and some other forms, the belief in origin 

 within the period of human civilization from one simple primi- 

 tive wild type seems on a balance of probabilities insecurely 

 founded, but allowing something for multiplicity of origin we 

 still fall far short of the requisite total of factors. Elements 

 exist in our domesticated breeds which we may feel with con- 

 fidence have come in since their captivity began. Such ele- 

 ments in fowls are dominant whiteness, extra toe, feathered 

 leg, frizzling, etc., so that even hypothetical extension of the 

 range of origin is only a slight alleviation of the difficulty. 



Somehow or other, therefore, we must recognize that dominant 

 factors do arise. Whether they are created by internal change, 

 or whether, as seems to me not wholly beyond possibility, they 

 obtain entrance from without, there is no evidence to show. 

 If they were proved to enter from without, like pathogenic 

 organisms, we should have to account for the extraordinary 

 fact that they are distributed with fair constancy to half the 

 gametes of the heterozygote. 



In proportion as the nature of dominants grows more clear 

 so does it become increasingly difficult to make any plausible 

 suggestion as to their possible derivation. On the other hand 

 the origin of a recessive variety by the loss of a factor is a process 

 so readily imagined that our wonder is rather that the phenom- 

 enon is not observed far more often. Some slip in the accurate 



