BY R. .1. TILLYARD. 197 



of Archipanorpa magtiifica. The pterostigma is shaded in, as in 

 the figures of Archi^mnorpd. A comparison of the two forms 

 enables us at once to fix the limits of the main veins in A^^chi- 

 jxino7ya, and to name the principal cells of the wing. The steps 

 of the determination may be briefly outlined as follows : — 



(1) The radius (R) is a strongly-])uilt, highly convex vein 

 running through the pterostigma (pt). Therefore I searched 

 first of all for the pterostigma, which was easily located, both 

 from its smoothness compared with the rest of the wing-mem- 

 brane, and because of the pterostigmatic furroiv (ptf) bordering 

 it posteriorly. Having found it, the radius was at once deter- 

 mined. There were then two other veins ending up on the 

 anterior wing-border between the radius and the base of the 

 wing; whereas, in Panorpa^ as in all recent insects, there is only 

 one (the subcosta, Sc). It became clear, therefore, that Archi- 

 panorpa possessed not only a subcostal vein, Sc, but also an 

 archaic costal vein, C, separated from the costal or anterior 

 border, as in some of the larger Carboniferous fossils (e.g., the 

 Meganeuridce of the Order Protodonata). 



(2) Turning next to the area of the wing below R, in specimen 

 No. I06« (hind wing), the cubitus, Cuj, is easily determined as a 

 strongly convex, obliquely-running vein, ending about half-way 

 along the posterior border. Between this and R, basally, there 

 lie two main stems, of which the upper must be the radial sector, 

 Ks, (its point of union with R is not preserved in the fossil), and 

 the lower must be the media, M. 



(3) The first forking of Rs is preserved in the fossil. Here 

 Rs divides into R2+S above, and R4+5 below. The cell enclosed 

 between these two veins basally is the discoidal ceU(dc), R2+3 

 again forks into Ro and R^, R4+5 into R^ and Rr,, and both are 

 sessile upon dc, as in the two Trichopterous wings already dealt 

 with. The area enclosed between Ro and R.. would be, in Tri- 

 choptera, the first apical fork {Ai A in Text-figs. 2-4). Rut, in 

 most Mecoptera, R.^ forks again into Rga and R^b (Text-fig. 7). 

 This occurs in Archipanorpa: but, above and beyond this, R.>ij 

 forks a second time into R2b and R2b'. Moreover, K., which 

 remains unforked in both Trichoptera and Mecoptera, in Archi- 



