36 



THE CANADIAN EiNTOMOLOGIST 



cubitus with two or more branches, in most genera showing a primitive anteriorly- 

 arching type of branching, and at least three complete anal veins; in the hind- 

 wing, the outer margin with a single complete convex contour, without any 

 re-entrant break^or angle at the distal end of Cus (this character is unique for 



Fig. 1. — Venation of Stenoperla prasina (Newm.), family Eustheniidae, (New Zealand), 

 genus Eusthenia itself has broader wings ancf still denser venation. 



The 



the family), the radial sector branched, and the archedictyon completely pre- 

 served on the anal fan, as well as on the rest of the wing. (Text Fig. 1*). Other 

 archaic characters are the form of the tarsal joints, the wide separation of the 

 front coxae, the presence of strong functional mandibles and a clearly visible 

 clypeus and labrum. 



Separated from these by clear characters, but still very archaic, are the 

 genera Austroperla Needham and Tasmanoperla n. g. (type Eusthenia diver sipes, 

 n. sp.), which differ from the true Eustheniidae in having shorter cerci, somewhat 

 more reduced but still primitive venation, at least a partial fusion of M3+4 with 

 Cuia in the hindwing, no archedictyon on the anal fan, and a distinct re-entrant 

 angle on the outer margin at the distal end of Cu2. (Text Fig. 3). In order 

 to test the validity of this family, I sought carefully for larvae of a different 

 type from those of Eustheniidae, and finally discovered in Tasmania a long 

 cylindrical larva without any gills at all, and with rather short cerci, from 

 which I reared a new species of Tasmatioperla, not yet described. Later on, I 

 also obtained a closely similar type of larva from New Zealand, and reared 

 from it Austroperla cyrene Newm. Thus the formation of this new family was 

 fully justified. 



Numerous species occur in which a considerable measure of specialization 

 has set in, coupled with the retention of a number of archaic characters. These 

 are all closely related to the genera Leptoperla of Newman (1839) and Gripopteryx 

 of Pictet (1841). They therefore form the family Leptoperlidae, this name re- 

 placing the name Gripopterygidae used by Enderlein to include not only these 

 insects, but also the Eustheniidae and Austroperlidae of this paper. Parenthet- 

 ically, it may be remarked that Enderlein, in forming his family, entirely ignored 



I '^-   



*An excellent figure of the wings of Eusthenia spectabilis Gray is to be seen in Comstock's 

 "The Wings of Insects," (1918), p. 247, Fig. 246. 



