14 GENERAL SURVEY OF THE ANIMAL KINGDOM. 



Note on Classification. 



We naturally group together in the mind those impressions which are 

 like one another. In this lies the beginning of all classification, 

 whether that of the child, the savage, or the zoologist. ' For there are 

 many possible classifications, varying according to their purpose, 

 according to the points of similarity which have been selected as 

 important. Thus we may classify animals according to their habitats or 

 their diet, without taking any thought of their structure. 



But a strictly zoological classification is one which seeks to show the 

 natural relationships of animals, to group together those which resemble 

 one another in their real nature or structure. It must, therefore, be 

 based on the results of comparative anatomy — technically speaking, on 

 " homologies," or real resemblances of structure. Whales must not be 

 ranked with fishes, nor bats with birds. 



To a classification based on structural resemblances, two corrobora- 

 tions are necessary, from embryology and from palaeontology. On the 

 one hand, the development of the forms in question must be studied : 

 thus no one dreamed that a Tunicate was a Vertebrate until its life- 

 history was worked out. On the other hand, the past history must be 

 inquired into : thus the affinity between Birds and Reptiles is confirmed 

 by a knowledge of the extinct forms. 



In classification it is convenient to recognise certain grades or degrees 

 of resemblance, which are spoken of as species, genera, families, orders, 

 classes, and so on. 



To give an illustration, all the tigers are said to form the species 

 Felis tigris, of the genus Felis, in the family Felidas, in the order 

 Carnivora, within the class Mammalia. The resemblances of all tigers 

 are exceedingly close ; well marked, but not so close, are the resem- 

 blances between tigers, lions, jaguars, pumas, cats, etc., which form the 

 genus Felis ; broader still are the resemblances between all members of 

 the cat family Felidse ; still wider those between cats, dogs, bears, and 

 seals, which form the order Carnivora ; and lastly, there are the general re- 

 semblances of structure which bind Mammals together in contrast to Birds 

 or Reptiles, though all are included in the series or phylum Vertebrata. 



It must be understood that the real things are the individual animals, 

 and that a species is a subjective conception within which we include all 

 those individuals who resemble one another so closely that we feel we 

 need a specific name applicable to them all. And as resemblances 

 which seem important to one naturalist may seem trivial to others, there 

 are often wide differences of opinion as to the number of species which 

 a genus contains. In a handful of small shells the "splitters" may 

 recognise 20 species, while the ' ' slumpers " see only 3. Thus Haeckel 

 says of calcareous sponges that, as the naturalist likes to look at the 

 problem, there are 3 species, or 21, or 289, or 591 ! 



But while no rigid definition can be given of a species, seeing that the 

 conception is one of practical convenience and purely relative, there are 

 certain common-sense considerations to be borne in mind : — 



I . No naturalist now believes, as Linnaeus did, in the fixity of species ; 

 we believe, on the contrary, that one form has given rise to another.' 

 At the same time, the common characteristic on the strength of which 



