Introduction 15 



based upon long study and extensive material, may however, when 

 published, afford an independent basis for classification. 



Until that time comes we have no system of classification, as already 

 stated, based on larval characters; but alleged resemblances in the 

 larvse have frequently been used to support relationships based pri- 

 marily on adult characters; and if such resemblances are, at least in 

 part, cases of convergence, even such may be hazardous. 



It may be added that McGillivray's key to Coleopterous larvae,' 

 though excellent for the period in which it was prepared, now requires 

 considerable modification. 



I have now given an account of the principal changes that have 

 been proposed in the Leconte system. But it is a bare sketch of their 

 salient points. A complete synopsis and argument may be found in 

 Dr. Gahan's paper from which I have already quoted. This should 

 be studied by every one interested in the subject of family classification. 

 It is, however, a critical paper and points out the weak points in other 

 systems without constructing a new one. The treatment of the Coleop- 

 tera by Brues and Melander,- incorporates many of the ideas which 

 I have endeavored to repeat, but gives no clue to the sequence in which 

 the families should be arranged. So that we are left to choose between 

 the rival continental authors, but with the guidance of Dr. Gahan's 

 impartial criticism and of Sharp and Muu-'s work on genitalia. 



System adopted for Check List. 



In this way, balancing one argument against another, I am led to 

 believe that a division of the Coleoptera into two sub-orders is estab- 

 ' lished ; and that possibly the sub-division of the sub-order Polyphaga 

 into several series, approaching the rank of sub-orders, is at present 

 the best course to pursue. The definition of some of these series, and 

 consequently the inclusion or exclusion of certain families, remains doubt- 

 ful; but for many purposes such definition is practically accomplished 

 by the families included. The two sub-orders would be separated as 

 follows : 



Outer lobe of maxillae palpiform; first visible ventral segment divided; hind wings with 

 cross- veins; pleural sutures of prothorax present; antennae never serrate, clavate or lamel- 

 ate; tarsi 5-iointed (except in the genus Hydroporus); larvae generally campodeaform, with 

 egs, tarsus with one or two claws, sometimes blattoid Adephaga. 



' New York .State Museum Bulletin (iS, 1903, pp. 288-294. 

 - Key to the Families of North .\merican Insects, 1915. 



