I 



Manclicster Memoirs, Vol. l.xi. (1917), No. *Z. 5 



In two places something of the kind seems present. The cerci- 

 likc termination of the abdomen may or may not be actual 

 cerci, or the appearance may be caused by a surface film of 

 the matrix being- broken away at this point. 



Affinities. — The salient facts of structure which need to be 

 borne in mind in seeking to identify the insect are well marked, 

 and distinctive, notwithstanding its fragmentary structure. I 

 have already summarised the details of wing structure, and 

 they need not therefore be repeated. The general details are 

 as follows :—I"Iead small and bearing well-developed antennae, 

 and powerful mandibles, thorax large and much elongated, 

 especially the prothorax; legs well developed, wings long and 

 delicate; abdomen long. 



This assemblage of characters is not unlike those of certain 

 of the Proto-orthoptera, and it may be that better preserved 

 examples will definitely settle the question of Proto-orthopteron 

 affinities. Should the insect belong to this class, it will prob- 

 ably be found closely related to the family Sfanioderidee, or 

 that of the Sthenar o podiel cc . 



With the family Spanioderidie, the specimen agrees in its 

 elongated prothorax, and legs not adapted for jumping. It is 

 still more closely in agreement with SthenaropodidiC , in which 

 the head and thorax are elongated, the legs slender, with the 

 hind pairs the longest, and the wings have a broadly rounded 

 apex. In this family also, the outer branch of the median 

 unites with the inner branch of the radial sector, and separates 

 again. The cubitus gives off many branches to the inner wing 

 margin. Transverse veins are well developed. Whether the 

 radial sector and the median unite in the Commentry specimen 

 which we are considering is not actually demonstrable, owing 

 to its fragmentary condition. Yet a close study of the vein 

 fragments shows that it is quite likely, and we have assumed 

 such to be the case. Before we, however, class the specimen 

 as belonging to^ the Sthenaropodidce, it may be well to point 

 out that this insect, and certain others classed with the Proto- 

 orthoptera, may possibly yet jirove to be neuropterids, some- 

 what nearly related to the family Perlidec. Brongniart has 

 already shown (Insectes Fossiles, p. 407, 1893) that some justi- 

 fication exists for the supposition that forms allied to. the 

 Perlick-e existed in Coal Measure times, and has descril>ed a 

 wing under the name of Protoperla Westuwodi. The existing 

 forms of Pcrlidce have a well-developed and elongated thorax; 

 four membranous wings, of which the hinder are the largest, 

 and well-developed and widely separated legs. The venation 

 of the wing in such a living form as Pteronarcys frigida is but 

 little removed from that of the Commentry insect, whilst the 

 latter also shows a close resemblance to the wing described 

 under the name of Protoperla Westurood-i, by Brongniart. It 

 differs from the latter in that the sub-costa is clearly attached 

 at its distal end to the radius, and does not reach the costal 



