222 THE CANADIAN ENTOMOLOGIST. 



by well-developed setae, to Ne^noura by venation, shows that the differ- 

 ences between these groups are not of more than tribal value. If it is 

 desired to have two sub-families, 1 think that the Pterofiarchii should be 

 opposed to all other Perlidce. However, I think that four tribes better 

 express the relationships of the groups. 



These four tribes may be tabulated as follows : 



1. Well developed setae present 2. 



Setae obscure or absent ; one branched vein from the anal cell of fore 



wings ; no series of cross-veins in anal region ; never greenish or 

 yellowish species Nemourini. 



2. A series of cross-veins in anal region of fore wings ; anterior coxae 



approximate Pteronarcini. 



Rarely a series of cross-veins in anal region ; anterior coxse widely 

 separate 3. 



3. Anal cell with two simple or one branched vein from below ; radial 



sector generally forked beyond or at anastomosis; median vein 



usually running parallel to radius in base of fore wings Perlini. 



Anal cell with but one unbranched vein from below; radial sector 

 usually not forked beyond nor at anastomosis ; median vein running 

 into radius much before base : Capnifit. 



The genus PerlifieUa, like Chloroperla of authors, contains some 

 discordant species. The type of Perlinella^ P. trivittata, has some cross- 

 veins in the anal field, a character of Pteronai'cys, The Perla placida^ 

 Hag., and some allied forms, differ from Perlinella in lacking such cross- 

 veins, and are more allied to Perla. They differ from that genus, not only 

 in venation, which is allied to Perlinella^ but also in having much longer 

 setre, with very elongate joints. Therefore I place P. placida as the type 

 of a new genus^ Perlesta, distinguished by the characters given in the 

 table below. The Perlinella frontalis^ Bks., differs from Perlinella and 

 Perlesta in having a forked vein from the anal cell, instead of two simple 

 veins ; it is therefore related to Alloperla. It differs from Alloperla by 

 having a series of cubital cross-veins in the hind wings, and by the two- 

 branched radial sector of fore wings. It therefore belongs to new genus 

 Paraperla. 



All the species originally placed in Chloroperla are now placed by 

 European authorities in Isopteryx, a later genus ; therefore Chloroperla 

 rej)laces Isopieryx. 



The genera of Perlidce, now known to occur in the United States and 

 Canada, may be distinguished by the following tables of the four tribes. A 



