BY R. J. TILLYARD. 117 



there is left only the distal secondary forking of Cui into Cuj^ and 

 Cuiij, the point of bifurcation being properly termed the i^pcondary 

 cubital fork {ciif). 



Now the resemblance between the condition of the cubitus in 

 the forewings of these smaller specialised Hemprohiidce^ and in 

 the forewings of the Myrmelcontid(f, is too striking to be ignored. 

 It suggested to me at once that the veins which I had named 

 Cuj and Cu.2 in the Myrmflpontidcf were in reality only Cuj^ and 

 Cu]b respectively, and that, if the true Cu" still existed, it must 

 be sought for near the extreme base of the wing. 1 therefore 

 examined a considerable number of MyvmpUontidw^ and found 

 that, in many cases, there could be detected, close to the base of 

 the wing, a semi-opaque connecting-vein, not unlike an oblique 

 vein, (though not so well defined) running across from Cu^ to 1 A. 

 This appeared to me to be the remains of Cu.^. With a view to 

 proving this point, I decided to examine the pupal tiacheation 

 of some of the older Mynndpontidce, belonging to the suV)family 

 Dendroleontince, in the imagines of which this supposed remnant 

 of Cu.2 appeared to be much more evident than in the genus 

 Myrmeleon itself. 



Most of the larva" of the Dcndroleontince do not make pit-falls, 

 but must be sought for either in loose sand, or in rubble or debris 

 collected at the bases of large trees. During the winter of 1916, 

 larvte of several genera of this type were collected around Sydney. 

 A number of these spun up in the following September. At the 

 end of a fortnight, three likely-looking cocoons were opened, but 

 were found still to contain larvae. They were then placed under 

 close observation until pupation took place. The first pupa ex- 

 amined proved to be very recently formed; so that, although a 

 good dissection of the forewing was obtained, the structure was 

 so exceedingly delicate that all the air passed out of the tracheae 

 before either a photomicrograph or a camera-lucida drawing could 

 be obtained. With the second dissection, I decided to take no 

 risks, but removed it at once to the microscope-stage, and made 

 a camera-lucida drawing of it. Both the pupal tracheation and 

 the imaginal venation showed up very clearly, and the result was 

 9 



