158 The Classification of Insects 



Butterflies and Moths may be easily enough distin- 

 guished from Cockroaches, but some of them show 

 a remarkable likeness to Caddis-flies. This difficulty 

 in classification is just what we should expect, if, as 

 we believe is the case, all insects are really related to 

 each other. We are able to draw hard and fast lines 

 between some groups because in the course of ages 

 they have diverged far from their common stock. But 

 the more nearly related insects are, the more difficult 

 do we find it to make them fit into the divisions that 

 we have made for their reception. Could a naturalist 

 have before Jiim all the insects which live or ever 

 have lived on the earth, his divisions would entirely 

 break down, and his system of Species, Genera, 

 Families and Orders would become lost in a well- 

 nigh indefinable gradation of characters. The object 

 of the classifier is so to arrange the insects which he 

 does know that they may fall into their right positions 

 with regard to these vanished races that must once 

 have formed the connecting links between them. 

 The truly natural classification of insects is that 

 which will place them on their right branches of 

 the great tree of animal life. The labours of the 

 classifier are usually presented in the form of lists 

 or catalogues ; and no linear arrangement — genera, 

 families, orders, following one after the other — can 

 exactly express the true relationships between the 

 groups, which must necessarily show affinities in 

 various directions. But in striving after a natural 

 system, making use in his work of all that he 

 can learn of the inner and outer structure and 

 development of the creatures which he classifies, 

 the systematist is doing his part to read the history 

 of insects in the widest sense of the term. And 

 he will, if wise, hold with a light grasp to his 

 systematic divisions, recognising that the increased 



