164 LARWINIANA. 



tinction not shared bj genera and other groups which 

 were not supposed to be genealogically connected. 

 How a derivative hypothesis would modify this view, 

 in assigning to species only a temporary fixity, is ob- 

 vious. Yet, if naturalists adopt that hypothesis, they 

 will still retain Jussieu's definition, which leaves un- 

 touched the question as to how and when the " peren- 

 nial successions " were established. The practical ques- 

 tion will only be. How much difference between two 

 sets of individuals entitles them to rank under distinct 

 species? and that is the practical question now, on 

 whatever theory. The theoretical question is — as 

 stated at the beginning of this article — whether these 

 specific lines were always as distinct as now. 



Mr. Agassiz has " lost no opportunity of urging 

 the idea that, while species have no material existence, 

 they yet exist as categories of thought in the same way 

 [and only in the same way] as genera, families, orders, 

 classes," etc. He 



"has taken the ground that all the natural divisions in the ani- 

 mal kingdom are primarily distinct, founded upon different 

 categories of characters, and that all exist in the same way, 

 that is, as categories of thought, embodied in individual living 

 forms. I have attempted to show that branches in the animal 

 kingdom are founded upon different plans of structure, and for 

 that very reason have embraced from the beginning representa- 

 tives between which there could be no community of origin ; 

 that classes are founded upon different modes of execution of 

 these plans, and therefore they also embrace representatives 

 which could have no community of origin ; that orders repre- 

 sent the different degrees of complication in the mode of execu- 

 tion of each class, and therefore embrace representatives which" 

 could not have a community of origin any more than the mem- 

 bers of different classes or branches ; that families are founded 



