CLASSIFICATION. ig 



rical law, that certain structures, not necessarily or usually 

 connected together by any visible link, invariably occur in 

 association with one another, and never occur apart, — so far, 

 at any rate, as human observation goes. 



Thus, all animals which possess two condyles on the occi- 

 pital bone, and possess non-nucleated red blood-corpuscles, 

 suckle their young. Why an animal with only one condyle 

 on its occipital bone should not suckle its young we do not 

 know, and perhaps we shall at some future time find mam- 

 mary glands associated with a single occipital condyle. Again, 

 the feet are cleft in all animals which ruminate, but not in 

 any other. In other cases the correlation is even more appa- 

 rently lawless, and is even amusing. Thus all, or almost all, 

 cats which are entirely white and have blue eyes, are at the 

 same time deaf. With regard to these and similar gene- 

 ralisations we must, however, bear in mind the following 

 three points :— 



1. The various parts of the organisation of any animal are 

 so closely interconnected, and so mutually dependent upon 

 one another, both in their growth and development, that the 

 characters of each must be in some relation to the characters 

 of all the rest, whether this be obviously the case or not. 



2. It is rarely possible to assign any reason for correlations 

 of structure, though they are certainly in no case accidental. 



3. The law is a purely empirical one, and expresses nothing 

 more than the result of experience; so that structures which 

 we now only know as occifrring in association, may ultimatelv 

 be found dissociated, and conjoined with other structures of a 

 different character. 



9. Classification. 



Classification is the arrangement of a number of diverse 

 objects into larger or smaller groups, according as they ex- 

 hibit more or less likeness to one another. The excellence of 

 any given classification will depend upon the nature of the 

 points which arS taken as determining the resemblance. Sys- 

 tems of classification, in which the groups are founded upon 

 mere external and superficial points of similarity, though 

 often useful in the earlier stages of science, are always found 

 in the long-run to be inaccurate. It is needless, in fact, to 

 point out that many living beings, the structure of which is 

 fundamentally different, may nevertheless present such an 

 amount of adaptive external resemblance to one another, that 

 they would be grouped together in any "artificial" classifi- 

 cation. Thus, to take a single example, the whale, by its 



