BY R. J. TILLYARD. 221 



these, R is unbranched. As a contrast to this, in the Anisoptera, 

 I find Rs present as a distinct sector in the very earliest stages 

 of Hemicordulia and uEschna of ^vhicll it is possible to obtain a 

 photograph. Dr. Ris also finds it present in a very early stage 

 of Libellula depi-essa, and Needham shows the same condition 

 for all ontogenetic stages of Gomphus descriptiis. Thus, the 

 ontogenetic evidence points to an originally unbranched radius 

 for all Zygoptera, and an originally branched radius for all 

 Anisoptera. 



3. Palceontological Evidence. 



In the Frotodonata (Upper Carboniferous and Permian), the 

 family Meganeuridce offers important evidence. The best-known 

 species, Megdueura monyi Brongniart, has R unbranched in the 

 hind wing, but with two distinct branches in the fore wing. A 

 smaller species, Meganeurula selysi Brong., has R unbranched in 

 both wings A Permian species, Typ)us {Tupus) permiaiius 

 Sellards,has R bratichediw both wings, and shows clear evidence of 

 having possessed the bridge and oblique vein of the Anisoptera. 

 Thus, it is fair to conclude that, in the Protodonata, forms that 

 were otherwise closely allied had a variable condition of the 

 radius, which might be branched or unbranched in both wings, 

 or branched in one wing and unbranched in the other. 



In the Odonata of the Lias and Upper Jurassic, we find 

 definite Anisoptera with, the radius branched, the oblique vein 

 and bridge present. We also find a few undoubted Zygoptera 

 {Calopterygidce) with the radius unbranched. There remain the 

 group which Handlirsch named Anisozygoptera^ and which he 

 claimed to be intermediate between true Anisoptera and Zygop- 

 tera. A careful examination of these forms shows : — 



(a). That the Heterophlebiince have an unbranched radius, with 

 a long bridge developed as in the Synlestince. 



(b). That all the other forms hava an unbranched radius and 

 no bridge. 



Intermediate forms, like Meganeura monyi, are not found in 

 the Mesozoic strata. Every form so far known can be, therefore, 

 definitely classed as Anisopter'ons or Zygopterons, according to 

 the condition of its radius. . - 



