Permanence and Evolution. 59 



races, had they been found wild, would have had 

 separate genera established for their reception ; 

 which is perhaps true, but shows only in what a 

 chaotic state the classification of birds is, never- 

 theless he thinks the evidence of common de- 

 scent stronger in this case than in that of any 

 other creature. His grounds are chiefly these 

 six : 



ist. If the principal breeds of domestic 

 pigeon have not descended from one stock (the 

 rock dove, Columba livia, comprising three sub- 

 races : C. livia proper, the European rock dove ; 

 C. intermedia, the Indian rock dove ; and 

 C. affinis, the chequered wild pigeon, and one 

 or two others equally slightly different), they 

 must have descended from at least eight or nine 

 aboriginal forms very distinct, otherwise the more 

 distinct domestic races could never have been 

 produced. If so, they must either now exist 

 wild somewhere or be extinct. The former, he 

 says, is so unlikely that it may be put aside ; 



