48 STRUCTURAL AND SYSTEMATIC ZOOLOGY 



of the same value in every tribe. A natural classification 

 must rest on those prevailing characters which are the 

 most constant.^ And such a classification can not be 

 linear. It is impossible to arrange all animal forms 

 from the sponge to man in a single line, like the steps 

 of a ladder, according to rank. Nature passes in so 

 many ways from one type to another, and so multiplied 

 are the relations between animals, that one series is out 

 of the question. There is a number of series, and 

 series within series, sometimes proceeding in parallel 

 lines, but more often divergent. The animals arrange 

 themselves in radiating groups, each group being con- 

 nected, not with two groups merely, one above and the 

 other, below, but with several. Life has been likened to 

 a great tree with countless branches spreading widely 

 from a common trunk, and deriving their origin from a 

 common root ; branches bearing all manner of flowers, 

 every fashion of leaves, and all kinds of fruit, and these 

 for every use. 



The groups into which we are able to cast the various 

 forms of animal development are very unequal and dis- 

 similar. We must remember that a genus, order, or 

 class is not of the same value throughout the kingdom. 

 Moreover, each division is allied to others in different 

 degrees the distance between any two being the 

 measure of that affinity. The lines between some are 

 sharp and clear, between others indefinite. Like the 

 islands of an archipelago, some groups merge into one 

 another through connecting reefs, others are sharply 

 separated by unfathomable seas, yet all have one com- 

 mon basis. Links have been found revealing a relation- 

 ship, near or distant, even between animals whose forms 

 are very unlike. There are fishes {Dipnoi) with some 

 amphibian characters, and fishlike amphibians (Ajcolotl). 

 The extinct ichthyosaurus was a lizard with fish charac- 



