CHAP. XXII.] CLASSIFICATION. 279 



Before going on witli tlie subject of this chapter, I wish 

 to remove a possible source of confusion. I have stated 

 that qiiestions of classification are not mere cpiestions of 

 words and names, but have to do with realities. Thus 

 the assertions that the whale is not a fish, but a mammal 

 having the external form of a fish, and that a cirrhipede is 

 not a molluscan, but a crustacean which has put on the 

 form of a molluscan, are assertions not merely concerning 

 the names which naturalists have agreed to use, but con- 

 cerning the real nature and afBnities of those animals. 

 These are cases of members of one group being disguised 

 in the likeness of another, and in such cases we know 

 nothing unless we know the real affinities of the species 

 under review, and consequently its true classification. But Questions 

 there are other questions of classification which are little cation* 

 more than questions of words. It is, for instance, very ^liich are 

 difficult to say where the series of fishes ends, and that of verbal. 

 Perennibranchiate Batrachians begins ; it is very difficult 

 to say whether the lepidosiren is a fish or a Batrachian. Position of 

 But it is as needless as it is difficult to decide such a siren°" 

 question. The two groups run into each other, and there 

 is so complete a gradation of intermediate forms, that the 

 line can scarcely be drawn between them except arbitrarily. 

 it is scarcely a metaphor to say that the lepidosiren is a 

 fish which we have caught in the act of acquiring lungs and 

 transforming itself into a Batrachian. But MammaUa and 

 fishes, or Crustacea and MoUusca, are not groups that run 

 into each other ; the whale has no tendency to become a 

 fish, nor has the cirrhipede any tendency to become a 

 molluscan. 



It needs no proof that the value in classification of any The value 

 character depends, not on the importance of that character chaKicter 

 to the life of the organism, but altogether on the extent in classifi- 



cation 



to which it is so correlated with other characters as to be depends 

 an index to the general nature and affinities of the °° ]'^^J^S 



^ .an index 



organism. Thus, on the one hand, as I have mentioned to others, 

 already, the presence or absence of wings in some groups 

 of beetles is very inconstant, both as between otherwise 

 similar species and between individuals of the same species ; 



