LEADING GROUPS OF EXISTING SYSTEMS 181 



fess at the same time that I have not yet been able to discover the 

 principle which obtains in the limitation of their respective subdivi- 

 sions.^^ All I can say is that all the different categories considered 

 above, upon which branches, classes, orders, families, genera, and 

 species are founded, have their degrees, and upon these degrees sub- 

 classes, sub-orders, sub-families, and sub-genera have been established. 

 For the present these subdivisions must be left to arbitrary estima- 

 tions, and we shall have to deal with them as well as we can, as long 

 as the principles which regulate these degrees in the different kinds 

 of groups are not ascertained. I hope nevertheless that such arbitrary 

 estimations are forever removed from our science, as far as the catego- 

 ries themselves are concerned. 



Thus far, inequality of weight seems to be the standard of the in- 

 ternal valuation of each kind of group; and this inequality extends 

 to all groups, for even within the branches there are classes more 

 closely related among themselves than others: Polypi and Acalephs, 

 for instance, stand nearer to one another than to Echinoderms; Crus- 

 tacea and Insects are more closely allied to one another than to 

 Worms, etc. Upon such degrees of relationship between the classes, 

 within their respective branches, the so-called sub-types have been 

 founded, and these differences have occasionally been exaggerated 

 so far as to give rise to the establishment of distinct branches. Upon 

 similar relations between the branches, sub-kingdoms have also been 

 distinguished, but I hardly think that such far-fetched combinations 

 can be considered as natural groups; they seem to me rather the ex- 

 pression of a relation arising from the weight of their whole organiza- 

 tion, as compared with that of other groups, than the expression of a 

 definite relationship. 



^ Professor James D. Dana has thrown out some valuable suggestions upon this point 

 in his review of my Contributions. [In this note to the 1859 edition Agassiz refers to 

 Dana's review of the Essay as it appeared in the first volume of the Contributions. See 

 American Journal of Science, XXV (2d ser., 1858), 12&-128, 202-216, 321-341, especially 

 p. 333.] 



