600 ON THE GENUS DIPHLEBIA, 



two gizzards of FeUiJura (iniantea Leach, described by me in 

 a former paper, we should be content, I think, to regard 

 variations in the actual number of teeth on a fold as purely 

 individualistic. 



The type of longitudinal fold exhibited in these gizzards is, 

 I believe, found in other Calopteryyid larvae whose gizzaixls 

 have been examined. A somewhat similar longitudinal ar- 

 rangement exists in the Gomphine larvae; but, in these, there 

 are only four folds, each containing a great many teeth, and 

 shewing as many as three or four abreast on the upper por- 

 tions. In the Aiiisopfera, it is certainly true that the larval 

 gizzards determine the separation of the main groups, there 

 being so far four main types known, viz., the Lihelhdid 

 (Libellulino-Corduline), ^'Esch^iine, Gomphine, and Petaln- 

 rine. If these are really four coordinate groups (a position 

 which, I believe, could be easily maintained, and strengthened 

 by many cogent arguments), we should seek to apply the same 

 test to our admittedly unsatisfactory Zyyopterous classifica- 

 tion. I should like, therefore, to put it upon record now, 

 though the elaboration of the facts must be left to later 

 papers, that I have found the Diphlebia-iovm of gizzard to 

 exist also in Argiolestes and Isosticta amongst the Agrionidce, 

 though the gizzards of the legions Agrion and Lesfes of de 

 Selys are absolutely different. These facts suggest that the 

 present Agrionidce are a collection of strongly asthenogenetic 

 forms descended along severed main lines from more abun- 

 dantly nervured insects, of which the present-da}' Caloptery- 

 gidce may represent, fairly closely, various stages of descent. 

 In this manner, we can trace, from a common ancestor, in 

 descending asthenogenetric order, the single group-line /^/y;/'- 

 lehia — Argiolestes — ' Isosticta, thoMgh. we cannot perhaps regard 

 any single one of these three genera as lying in the direct line 

 of descent of the group ; that direct line having most pro- 

 bably been lost. It may be added that the main structure 

 and habits of the larvae of these three genera, their caudal 

 gills and their labia in particular, agree with this view. Wo 



