202 DEVELOPMENT OF THE WING-VENATIOX OF ODONATA, 



ings took place, even amongst this one portion of the Suborder ? 

 The origin of the tribe Agrionini is doubtful, for their acute 

 quadrilateral might yet be proved to be the reduced remnant of 

 an originally weak Anisopterid triangle-formation. Finally, 

 standing out clearly from all the rest as the most recent offshoot 

 from the Anisopterid line, we see the EjnopJdehia-Synlestes-Lestes 

 line of descent, which branched off from that line not very far 

 from the beginnings of the GomphinfF, and whose sharply angu. 

 lated quadrilateral is almost certainly an Anisopterid remnant. 



Were it not for the problematical position of the Agrionirii, 

 which form the greater number of the Zygoptera, and must 

 therefore, be satisfactorily placed in any classification that is to 

 be of any use, we might see the dividing line in the form of the 

 quadrilateral. Our two main groups would then be the Zygop- 

 tera Rectangularia, with more or less regular quadrilateral, and 

 the Zygoptera Acuta with the distal angle of the quadrilatera 

 sharply acute. But this division is most probably scarcely less 

 unnatural than the Selysian. It would, therefore, be unwise to 

 press for its recognition as a natural basis for classification. On 

 the other hand, just as the terms Calopterygidre and AgrionidcF, 

 though indicating admittedly unnatural groups, do carry a certain 

 value as the names of groups possessing a single character common 

 to every member of each (at least, with few exceptions) — and 

 are, therefore, of great value to students not very familiar with 

 the intricacies of Odonate wing- venation— so also, I would urge 

 a loose acceptance of the terms proposed, on the ground that 

 they may prove of very great benefit in subsequent phylogenetic 

 discussions. If that be recognised, then, for instance, whenever 

 we mention the term Zygoptera Acuta, everyone will recognise 

 that it refers to all Zygoptera with acute-angled quadrilaterals; 

 and similarly for Zygoptera Rectangularia. 



The time is not yet ripe for making more than a beginning at 

 the difficult problem in hand. That beginning I now make by 

 selecting the Epiophlebia-Synlestes-Lestes line of descent as a 

 single pure line, descending from the main Anisopterid stock at 

 a later period than any of the other Zygoptera. These, I con- 

 sider to be worthy of family-rank at least; hence I propose, for 



