172 MUTATION 



and in so far as their inheritance was witnessed 

 each refused to blend when mated with a dissim- 

 ilar form. For example, the pea X single gives a 

 pea comb which in the next generation yields sin- 

 gle and pea; cerebral hernia and no hernia give 

 no hernia in the first generation, but hernia again 

 in the second; taillessness may follow the Men- 

 dehan formula, polydactyhsm approaches it, and 

 the color varieties illustrate it strikingly. In a 

 word, while quantitative variations are never ab- 

 sent in poultry, the sudden appearance and dis- 

 appearance of full-fledged characters is most 

 striking. Mutation as thus defined presents to 

 the breeder as a common phenomenon. But, say 

 the neo-Darwinists, your mutations are of a 

 teratological sort and have nothing to do with 

 species as we find them in nature. In reply I 

 admit, first, that under domestication many mu- 

 tations are preserved by man that would perish 

 in nature. It is quite hkely that mutations occur 

 almost as frequently in nature as under domesti- 

 cation, but the unfavorable new forms are apt 

 to suffer early elimination. There remains, how- 

 ever, a host of characters that are not detrimental 

 to the individual, and such are. not necessarily 

 ehminated. They are teratological only in the 

 sense that they are novel to the species, but they 

 are of the same order as many of the specific dif- 

 ferentiee of feral species. Take, for instance, the 

 passerine birds — ^what are some of their striking 

 qualitative characters? We find crossed bill 



