62 THE WINGS [CH. 



the same tendency has set in in the Agrionine genus Hemiphlebia 

 (fig. 22). Here of course the arculus has been complete right 

 through the evolution of the genus. But the slanting of the 

 arculus has gone on until it came into line with M 4 , and then the 

 distal side of the quadrilateral tended to continue the same line. 

 The final stage of elimination of the basal side has only so far been 

 achieved in the fore-wing of the male, and occasionally in the 

 female also. 



The Anal Field (Plate II, figs. A, B). 



(i) Anisoptera. We can only briefly trace the many changes 

 in this important area of the wing. Owing to the basal fusion of 

 A with Cu (due to the same causes as the fusion of R with M), 

 there was necessarily also a more distal point of departure of A 

 from Cu. This point is the analogue of the arculus, and is repre- 

 sented in the imaginal wing by the anal-crossing, Ac. This usually 

 only appears as a straight cross- vein, often indistinguishable, except 

 perhaps for a slight slant, from a number of others descending from 

 Cu + A, or Cu. 



Just as M, after leaving R, divides into a main stem and three 

 branches, so exactly does A on leaving Cu. But, as the area to be 

 supplied by these branches differs greatly from that supplied by M, 

 both in size and shape, we find them correspondingly shortened 

 and directed posteriad rather than distad. 



The main stem of A, after leaving Ac, is represented by A lt 

 Normally this continues on to meet Cu at, or just below, the 

 posterior angle of the triangle. By a slight re-arrangement of one 

 of the cross-veins above A lt the area known as the sub-triangle or 

 internal triangle is formed (fig. 20, st). In the fore- wing of 

 Libellulidae this may become very highly specialized, by means 

 of a sharp bend in A v At the same time, the cross-vein forming 

 the original basal side of st becomes elongated to continue the line 

 of the proximal part of A, thus giving us the complicated structure 

 seen in the fore-wing of Hemicordulia (Plate II, fig. B), Tramea 

 (fig. 136), and many other genera. This is clearly correlated with 

 transverse elongation of the triangle. In the hind-wing, the sub- 

 triangle is usually absent. 



