CLASSIFICATION 257 



illustration, a genus may be compared to a number of human 

 families related to one another in the first degree of cousin- 

 ship. A zoological family may be compared to groups of first 

 cousins related to one another in more distant degrees of 

 cousinship. An order, again, to groups of distant cousins 

 related by remoter ties of tribal kinship. A class may be 

 compared to a number of tribes forming a nation, and a 

 phylum to a number of nations forming a great race, like the 

 Aryan or Mongolian race. The complete genealogy of a 

 tribe would show the descent of all the families composing 

 it from a common ancestor, the founder of that tribe, and 

 it would show also the various degrees of kinship not only 

 between the surviving families of the tribe but between all 

 the families existing at any given time in the past. Similarly, 

 the complete genealogy of a zoological order would show the 

 descent of all the species contained in it from a common 

 ancestor, and would show the degrees of kinship of all 

 existing species and of all pre-existing species at any given 

 period of time. But it need hardly be said that such a 

 complete genealogy does not exist and never will exist. 

 There is no continuous record. Zoologists can only infer 

 the degrees of affinity of species by their greater or less 

 structural resemblance to one another, aided by the very 

 imperfect and fragmentary evidence of descent furnished by 

 extinct forms whose remains have been preserved in a fossil 

 state. Any system of classification, therefore, depends, in 

 the first place, on the extent of our knowledge of the structural 

 differences which obtain among species, and in the second 

 place, upon the judgment shown by the frarner of the system 

 in grouping together or separating species according to their 

 greater or less resemblances and differences. Consequently 

 there is no such thing as a final scheme of classification of 

 any group. Systems must change with the increase of our 

 knowledge and according to the varying judgments of the 

 authors who frame them. It also follows that there is more 

 certainty about the affinities of species and genera than of 

 the higher divisions, for it is easier to recognise near than 

 remote relationships. In fact, the limits assigned to classes, 

 orders, and families are somewhat arbitrary, and these divisions 

 in different phyla of the animal kingdom do not imply equal 

 degrees of affinity. Thus the classes in one phylum may 



Digitized by Microsoft® 



