50 THE WINGS [CH. 



then dominant family Meganeuridae occurred both in the Upper 

 Carboniferous and in the Permian. In the Carboniferous genus 

 Meganeura (fig. 156), R in the fore-wing had two very distinct 

 branches situated close together; but in the hind- wing the same 

 vein was quite unbranched. In the closely allied genus Mega- 

 neurula, R was unbranched in both wings. But in the Permian 

 genus Typus (= Tupus), R was found to be branched in both 

 wings, in the same manner as in present-day Anisoptera, the 

 presence of both bridge and oblique vein having been satisfactorily 

 demonstrated [150]. In other words, in the family Meganeuridae, 

 the condition of R was variable and only of generic value. Un- 

 fortunately there is a large gap between the Permian Protodonata 

 and the earliest true Dragonflies, which appear in the Lias. Hence 

 the further course of this development is not capable of definite 

 palaeontological proof. It is sufficient, however, to remark that, 

 in the Lias, no forms with the intermediate condition of R (as 

 seen in Meganeura) have been found. Every Liassic genus either 

 has R once-branched in both wings, or unbranched in both. In 

 other words, in the Lias, the dichotomy between Anisoptera and 

 Zygoptera was complete, through the dying-out of intermediate 

 forms. 



As regards the vein M, we find that, in Anisoptera from the 

 Lias onwards, it was constant, and gave off only three branches. 

 In Zygoptera however, during the same period, M has always had 

 four branches, one of which (Ms) played the part of the missing 

 branch of R. In the Heterophlebiinae (fig. 159) the two most 

 distal branches of M (M 1 _ 2 ) came off together by a common stalk. 

 Owing to the big gap thus left between M 1-2 and M 3 , we find here 

 the development of a "long bridge" backwards from M z towards 

 the main stem of M. This formation parallels that of the shorter 

 bridge of Anisoptera. Our recent Lestidae are the direct descen- 

 dants of forms such as the Heterophlebiinae. 



A third principal difference between the two suborders lies 

 in the evolution of the discoidal cell. It has, 1 think, been too 

 strongly insisted on that this cell is highly differentiated only in 

 the Anisoptera, its condition in the Zygoptera being considered 

 much more primitive. The structure of this cell is of such interest 

 that we have dealt with it in a separate section (p. 56). 



