164 ESSAY ON CLASSIFICATION. Part I. 



without entering here into a discussion respecting the possibility or practicabiHty 

 of setting aside this difficulty by introducing the consideration of the limited fer- 

 tility of the progeny of individuals of different species, I will only remark, 

 that as long as it is not proved that all the varieties of dogs, and of any others 

 of our domesticated annuals, and of our cultivated plants, are respectively derived 

 from one unmixed species, and as long as doubts can be entertained respecting 

 the common origin of all races of men from one common stock, it is not logical 

 to admit that sexual connection resulting even in fertile offspring is a trustworthy 

 evidence of specific identity. 



To justify this assertion, I would only ask, where is the unprejudiced naturalist 

 who in our days would dare to maintain : 1st, that it is proved that all the 

 domesticated varieties of sheep, of goats, of bulls, of llamas, of horses, of dogs, 

 of fowls, etc., are respectively derived from one common stock ; 2d, that the 

 supposition that these varieties have originated from the complete amalgamation of 

 several primitively distinct species is out of the question ; and 3d, that varieties 

 imported from distant countries and not before brought together, such as the 

 Shanghae fowl, for instance, do not completely mingle ? Where is the physiologist 

 who can conscientiously affirm that the limits of the fertility between distinct 

 species are ascertained with sufficient accuracy to make it a test of sjDecific identity? 

 And who can say that the distinctive characters of fertile hybrids and of unmixed 

 breeds are sufficiently obvious to enable anybody to point out the primitive feat- 

 ures of all oiu- domesticated animals, or of all our cultivated plants ? As long 

 as this cannot be done, as long as the common origin of all races of men, and 

 of the different animals and plants mentioned above, is not proved, while their 

 fertility with one another is a fact which has been daily demonstrated for thou- 

 sands of years, as long as large numbers of animals are hermaphrodites, never 

 requiring a connection with other individuals to midtiply their species, as long as 

 there are others which multiply in various ways without sexual intercourse, it is 

 not justifiable to assume that those animals and plants are unmixed species, and 

 that sexual fecundity is the criterion of specific identity. Moreover, this test can 

 hardly ever have any practical value in most cases of the highest scientific inter- 

 est. It is never resorted to, and, as far as I know, has never been applied with 

 satisfactory results to settle any doubtful case. It has never assisted any anxious 

 and conscientious naturalist in investigating the degree of relationship between 

 closely alHed animals or plants living in distant regions or in disconnected geo- 

 graphical areas. It will never contribute to the solution of any of those difficult 

 cases of seeming difierence or identity between extinct animals and plants found 

 in difierent geological formations. In all critical cases, requiring the most minute 

 accuracy and precision, it is discarded as unsafe, and of necessity questionable. 

 Accm'ate science must do Avithout it, and the sooner it is altogether discarded, the 



