BY R. J. TILLYARD. 455 



soon dividing into two strong branches, which diverge at a sharp angle; the 

 upper branch, M2a, runs straight along the wing to end up just below the apex, 

 close to MiA, which is a long sector formed between Mi and M2; the lower 

 branch. Mat, runs obliquely just above Ms. The common stem of Ms and M3 

 soon divides at a very acute angle into Ms, which runs straight on, obliciuely 

 across the wing, and Ms, which arches so as to end up about half-way along 

 the posterior border, far from Ms, but very close to M4. M4 is a slightly arched, 

 unbranched, convex vein, rather weakly formed distally. The cubitus is a simple, 

 concave vein, corresponding with Cm of recent Odonata, slightly curved below 

 the arculus, and then running almost straight on to end up at about the middle 

 of the posterior border, a little before M4. Anal crossing present as a weal; 

 cubito-anal veinlet, situated at the end of the petiole, and marking the origin 

 of lA from Cu; lA itself runs between Cu and the posterior border, and ends 

 up not far from Cu. Both Cu and lA become weak and somewhat zig-zagged 

 distally. 



Handlirsch formed the new Suborder Archizygoptera to include the single 

 peculiar and highly problematical genus Protomyrmeleon Geinitz, represented by 

 a single species, P. brunonis Geinitz, from the Upper Lias of Dobbertin in Meck- 

 lenburg. This fossil was placed by him in the family Protomyrmeleontidae, the 

 only family of the Suborder. The present fossil agrees with Protomyrmeleon in 

 the very striking characters of the shortened Se, unformed nodus, peculiar deve- 

 lopment of MiA, two-branched M2 and simple cubitus, but appears to differ in 

 the bas' not being petiolate, M arising separate from R, so that no arculus is 

 formed, separation of the base of Ms from Ms, and entire absence of lA. (It 

 should be noted that Handlirsch's naming of all the veins after Mi is incorrect, 

 Ills M2 being actually MiA; his Rs, M2a; his Ms,M2b; his M4, Ms; and his Cui, Cu2 

 and lA being Ms, M4 and Cu respectively). It would appear highly probable 

 that the true base of Protomyrmeleon has not been preserved, including the 

 petiole (if present), the portion of M fused with R, and the true arculus. This 

 misled Handlirsch in naming the veins. As drawn by him in Plate xlii., fig. 14 

 of his Atlas to "Die Fossilen Insekten," there is, in any case, no true lA in this 

 genus. 



Genus Triassageion, n.g. (Plate lii., fig. 31; Test-fig. 77.) 



To the characters given for the family we may add the following for the 

 genus : — Postnodals about twenty-four, the basal ones mostly continuous with the 

 cross-veins below them, the distal ones not so. Pterostigma short, about twice as 

 long as wide. Ri, Mi, Mia and M2a all ending up close together at or near apex 

 of wing. MiA arises as a weak zig-zag vein from near base of M2, and runs 

 very close above M2 at first, but gradually diverges until, below the pterostigma, 

 it runs as a straight vein about half-way between Mi and M2u. The wide tri- 

 angular spaces between the two branches of M2 and also between Ms and Ms 

 are filled with irregular cellules, without any supplements. Only one row of 

 cellules between lA and the posterior border of the wing. 



Genotype, Triassagrion australiense, n.sp. (Upper Triassic, Ipswich, 



Q.). 



It is useless to try to compare this genus closely with any existing Zygoptera, 

 owing to the great difference in the structure of the arculus, the entire absence 

 of the discoidal cell, the primitive condition of Cu and lA, and the branched con- 

 dition of M2. I would, however, call attention to certain resemblances which it 

 bears to the forewing of the genus Chorismagrion Morton. This latter genus, 

 found in North Queensland at the present day, has the arculus open basally in 





