14 INDIAN FRESH-WATER FISHES. 



forms which differed from each other ages ago should 

 have thrown off convergent branches, so that the 

 descendants of each, becoming more and more like 

 each other, should now be undistinguishable apart. 

 They would still, nevertheless, be distinct species, and 

 would neither consort nor breed together. However, 

 though theoretically imaginable, such cases do not 

 practically occur ; the developments of nature are in- 

 finite, and, however closely some forms may resemble 

 others, there is always some point or other, if we 

 only look close enough, which serves to distinguish 

 each species apart. 



A difficulty somewhat similar does, however, occur 

 practically in dealing with groups. Groups perpetually 

 inosculate with one another, and species may be found 

 which seem in certain respects more nearly allied to 

 other groups than the one to which they really belong. 

 In these cases we have to discriminate between real 

 and apparent affinities, and hence arise frequently 

 differences of opinion between different systematists. 



The more perfect our classification becomes, and the 

 closer we follow nature in grouping such species as we 

 are acquainted with, by so much the less are we 

 troubled with wayward forms which seem to belong 

 neither to one genus or another, but to possess affinity 

 with two or more. 



It must be remembered that we have at best but 

 scanty material with which to build up the scheme of 

 nature. We can observe the species of to-day, but 

 many links are wanting, many forms must have died 

 out to make room for the present ones. 



