270 THE PANORtoiD COMPLEX (Introduction), 



of the supposedly ancestral Order, and, indeed, suggests, in more 

 than one place, that the whole of the Lepidoptera are descended 

 from this still existing and abundant genus of Caddis-flies. 

 Throughout the work, the characters of the Lepidoptera are 

 subjected to a searching scrutiny and criticism, but the same 

 tests are not applied to the Trichoptera, which are, nevertheless, 

 an equally important factor in the problem [3, i\. 



Tt has, for long, seemed to me that a problem of this kind, 

 attacked in this manner, i.e., by intensive study of the internal 

 differences within the Order in question, without an equally 

 intensive study of the differences existing within the Oiders 

 supposedly ancestral to it, could not possibly yield a complete 

 and accurate solution. The alternative, chosen by some few 

 authors in recent years, has been to pass in review the sup- 

 posedly archaic characters of related Orders, and to attempt to 

 derive from this evidence some idea as to the standing of one 

 Order to another. As an example of this, I need only mention 

 again Handlirsch's famous attempt to give us a complete Phylo- 

 geny of the whole of the Orders of Insects, as well as a Phylo- 

 geny of the Class Insecta itself(2). 



It will be at once admitted that this second method is the 

 right one. It has, however, one obvious disadvantage. In 

 order to carry it out successfully, the author should be equally 

 expert upon all the Orders that he reviews. But no man could 

 hope to attain the knowledge in half-a-dozen or more Orders 

 that can be attained in one by a life-study of it. Hence we see 

 that, while the outlook of such an author will be a broader one 

 than that of the expert in one Order only, yet the evidence 

 brought forward cannot all be accepted at the same face- value; 

 and the chances of misinterpretations at vital points is evidently 

 very much greater. 



As an example of this, we might contrast the treatment of 

 the Phylogeny of the Lepidoptera by Meyrick on the one hand, 

 and by Handlirsch, on the other. First of all, Meyrick confines 

 himself almost entirely to wing-venation. By an exceedingly 

 full analysis of the different types within the Order, he arrives 

 at the conclusion that the homonomously -winged Jugatse repre- 



