60 . THE WINGS [CH. 



supra-triangle is correspondingly long and narrow (fig. 136). But 

 in the hind-wing, the broadening of the basal portion of the wing 

 had an opposite effect. The triangle was not only stretched 

 longitudinally (somewhat after the manner seen in the Aeschninae) 

 but it was drawn towards the wing-base. Finally it took up a 

 stable position with its basal side at the level of the arculus, so 

 that supra-triangle and triangle became coterminous in length 

 (fig. 20^; cf. fig. 136). 



Let us now return to the Zygoptera, in which the discoidal cell 

 has remained an undivided quadrilateral. Original differences in 

 the levels of the arculus, and the bifurcation of Cu, account for 

 the many variations in the length of the quadrilateral still to be 

 seen. (Contrast Pseudophaea, fig. 137, with RhinocypJia, fig. 138, 

 or Calopteryx, fig. 177.) The cross- veins early disappeared in this 

 suborder, except in a number of Calopterygidae, where the original 

 close venation of the wing has persisted up to recent times. Another 

 early tendency was the thickening of a short basal portion of Cu 2 

 in line with the distal side of the quadrilateral . As the anal bridge 

 (Ab) ended up on Cu 2 just at this same point, a special distal 

 portion of the cubito-anal space became cut off between Ac and 

 Cu z , forming a quadrangular area sometimes called the sub- 

 quadrangle (sq in fig. 20 c'). 



We can see an archaic type of quadrilateral still extant in the 

 hind-wing of Epiophlebia (fig. 141). From this type, a further 

 narrowing of the wing-base leads us to the Lestine-type (fig. 20 /" ; 

 cf. fig. 142). Here the distal side, together with its prolongation 

 along Cu 2 , is no longer transverse to the wing- axis, but becomes 

 strongly slanted, while the costal side of the quadrilateral is 

 correspondingly shortened. By a similar method, the sharply- 

 pointed quadrilateral of the Agrioninae arose from the older form 

 seen in the fossil Phenacolestes (fig. 164) and in many recent 

 Megapodagrioninae. The Lestine and the Agrionine-types are 

 closely parallel ; but the former seems to have been in existence 

 for a very long time, whereas the latter is a very recent develop- 

 ment. 



Quite a different tendency has operated in the case of the 

 Epallaginae and Calopteryginae, resulting in a gradual loss of the 

 irregularity of the quadrilateral, and the formation of an exact 



