284 Darduin, and after Darwin. 



while maintaining that all specific characters must 

 necessarily be useful, maintains at the same time 

 that any number of varietal characters on the one 

 hand, and a good half of generic characters on 

 the other, are probably useless. Thus he contra- 

 dicts his argument from the " constancy of specific 

 characters" (seeing that generic characters are still 

 more constant), as later on we saw that he contra- 

 dicts his deductive generalization touching their 

 necessary utility, by giving a non-utilitarian ex- 

 planation of whole multitudes of specific characters. 

 I need not, however, again go over the ground so 

 recently traversed ; but will conclude by once more 

 recurring to the only explanation which I have 

 been able to devise of the otherwise inexplicable 

 fact, that in regard to this subject so many natural- 

 ists still continue to entangle themselves in the 

 meshes of absurdity and contradiction. 



The only conceivable explanation is, that these 

 naturalists have not yet wholly divested themselves 

 of the special creation theory. Although professing 

 to have discarded the belief that "species" are 

 " definite entities," differing in kind from " varieties '^ 

 on the one hand and from "genera" on the other, 

 these writers are still imbued with a vague survival 

 of that belief. They well know it to belong to the 

 very essence of their new theory that "species" 

 are but " pronounced varieties,' or. should we prefer 

 it. "incipient genera"; but still they cannot alto- 

 gether Cbcape the pre-Darwinian conception of species 

 as organic units, whose single mode of origin need 

 not extend to other taxonomic groups, and whose 



