BY R. J. TILLYARD. 203 



them, the family-name Lestidcp., with the following definition and 

 subdivisions : — 



Family Lestid^e. 

 Quadrilaterals of all four wings, but especially of forewings, 

 with their distal angle very acute. Rs leaving M., considerably 

 distad from subnodus. A long bridge-vein developed backwards 

 from Rs towards the bifurcation of M^.o from M.,. 



Excluding fossil forms, three subfamilies may now l)e proposed 

 under the above heading : — 



/Bridge connecting with M^ o 1. 



I Bridge connecting with M„ 2. 



/'More than two antenodals, of which two are thickened. 

 M.^ departing from snbnodus. Quadrilateral of hindwing 

 much less acute than that of forewing. General facies 



of insect distinctly Gomphine in appearance 



Subfamily 1. Epiophhhiince. 



Ool}' two antenodals (rarely three). Slender petiolate wings. 

 M„ departing more or less distad from nodus. General 



facies of normal Zygopterid form Subfamily 2. Lestince. 



'Only two antenodals (rarely three). Slender petiolate wings. 



M„ departing far distad from subnodus Subfamily 3. Synlestiiue. 



More robust forms with many antenodals 



[Fossils only . Subfamily Heterophhbiiiue]. 



Of these, the subfamily Bpiophlebiitice contains the single genus 

 Epiophlehia Calveit {= Palceophlebia of Selys). The subfamily 

 Heterophlebiiufc contains only the single fossil genus Hetero- 

 phlehia, which has not yet been fully studied. The subfamil}^ 

 SynlestincH contains the genus SynUstes, Chlorolestes, and pos- 

 sibly other additions from the ranks of our present tribe Poday- 

 rionini. The subfamily Lestinca remains exactly as at present 

 recognised, with Archilestes as its most archaic genus (M^ fairly 

 close to subnodus) and Aitstrolestes as its most reduced, and pi-o- 

 bably most csenogenetic development. 



There can be little doubt that the tangle of forms left over, 

 after the extraction of the homogeneous Lestidce, will eventually 

 be unravelled by scientific treatment. But the problem calls for 

 not months but years of research. The point of division must 

 be looked for still in the behaviour of Rs, since there are pro- 

 bably, in the Calopteiyyidce, a number of aichaic forms in which 



