BY R. J. TILLYARD. 681 



straight — actually witli a very s'ight double-curving — below and 

 parallel to C and Sc. Tlie question of the formation of the 

 radial sectm' (Rs) must be dealt with under the discussion of 

 the nodus. 



The media (M) divides at the arculus into the concave and 

 weak Ml +3, and the stronger, convex vein, M^. At a point a little 

 beyond the distal end of the triangle, M1+3 divides into M142 

 above, and M3 below, both weakly formed and concave. Imme- 

 diately afterwards, M1+2 gives off, posteriorly, a strong concave 

 vein, corresponding with the bridge-vein of recent Anisoptera. 

 This vein, for the reasons given below in my discussion of the 

 nodus, I have definitely decided to call the Zyyojjferid sector (Ms). 

 It runs parallel to, and a little below, M2. Mi +2 divides into 

 M, and M._, just before these two veins are crossed by the vein 

 descending from the subnodus, as Woodward's figure accurately 

 shows. Below Ms, there is a slight indication of the formation 

 of a radial supplement (Rspl). M4 runs from the arculus to the 

 upper distal angle of the triangle, and, from there on, curves 

 gently downwards, beneath and subparallel to Ms, to reach the 

 posterior wing-border far beyond the le\el of the nodus. Beneath 

 M4, a very long, median supplement (Mspl) is developed, lying 

 slightly concave to M^, but not linked up with the triangle (as 

 would appear from Woodward's figure). 



The cubitus (Cu) is a fairly strong, concave vein, running, with 

 a very slight double-curving, below the arculus, to join the 

 triangle at its proximal angle. There it turns sharply down 

 wards, at right angles to its previous course, forming the proximal 

 side of the triangle itself. Its further course is unfortunately 

 lost, owing to the fracture below the triangle, already mentioned. 

 But, at the distal side of this fracture, it is easy to pick up again 

 a strongly concave vein descending to the wing-border almost 

 directly under the nodus. This must certainly be Cuj ; but where 

 Cuo is, or whether it is present at all, is not so easy to determine. 



The analis (A) is a very strongly-marked, convex vein, running 

 in a gentle curve (concave to the costa) from the base to the 

 proximal angle of the triangle, where it meets Cu, I can find 



4y 



