the Superim' Wing of the Hymenoptera. 441 



Evaniidae, as regarded by Mr, Bradley, would doubtless 

 form one of the archaic types. 



In the following catalogue of the wing cells and nervures 

 I have attempted to synonymize the names applied to the 

 various divisions by all the authors with whom I am myself 

 familiar, and I will at once say that my sole object has 

 been to arrive at a simple system, which is equally applic- 

 able to all sections of the Order. That I have neglected 

 the works of many authors upon Aculeata and Tenthredi- 

 nidae I am fully aware, and my excuse is that I am 

 entirely persuaded that the simple venational structure 

 was, as one would in the nature of things expect to be the 

 case, the earliest form, and that from which the complex 

 neuration of the sawflies on the one hand and bees on the 

 other have sprung. Jurine attempted to bring the whole 

 Order into line by tracing the venational development 

 from his Fsilus (nee Galesus, Curtis), through Chalcis, 

 Codrus, Omalus, and the Aculeates, up to Hylotoma, of 

 which " the wing is filled with its greatest complement of 

 areolets," I am of opinion we shall arrive at a more 

 natural system by tracing two branches, or better three, 

 for the links with the sawflies appear quite lost, and these 

 correspond peculiarly exactly with the Entomophaga (ex- 

 cluding Phytaphaga) and Aculeata of Westwood, or, as 

 modern authors are pleased to term them, Iclmeumonidea 

 and Vespoidea. 



My conception of the ancestral wing of all Hymeimptera 

 is a compound of those of Bracon and Fhygadeuon. A 

 comparison of these will reveal the actual couformity of 

 their structure, though the nervures are so differently dis- 

 posed that at first sight they appear entirely distinct. In 

 the former, venation is entirely obsolete on the disc, failing 

 to connect the cubital cells, inter se, in any way ; and, in 

 common with all Braconids, the second recurrent nervure 

 is wanting. In the latter it is again the cubital nervure 

 which is deficient, but this time at its base ; it is con- 

 tinuous with the first recurrent nervure, its point of origin 

 being indicated by the " nervelet " or " ramellus " only. 

 This combination will give us the following neuration, 

 traceable more or less in its entirety throughout the Order, 

 as will be seen by the preceding wings, which are num- 

 bered in conformity with the following diagram. Addi- 

 tional nervures occur in the Tenthrediriid.ae, and the 

 number of cubital cells is multiplied in the Aculeata, of 



