456 MESOzoic INSECTS or Queensland, ix., 



the forewing; Sc is very short; M1-3 gives off the common stem of M3 and Ms 

 at the subnodus, and Ma far beyond it, with Mia forming a well developed vein 

 between Mi and Ms; Cu, except for current usage, might well be interpreted as 

 a simple vein, lA arising separately out of the posterior margin just beyond the 

 end of the petiole, and connected with Cu above it by the anal crossing. Again, 

 if we look at the genus Hemiphlebia, which also has the arculus incomplete 

 basally in the forewing, we see that the same interpretation of Cu and lA is 

 the obvious one, and that lA actually arises from the cubito-anal veinlet as it 

 does in Triassagrion, with a small cross-vein connecting it with the end of the 

 petiole as in that genus. This latter character, being unique in present-day 

 Zygoptera, is a very significant one. I therefore suggest the probability of our 

 modern Zygoptera having arisen from some such form as Triassagrion by the 

 following changes : — 



(1) Formation of a complete nodus by strengthening of the subnodal cross- 

 veins between end of Sc and R, and between R and M1-3. 



(2) Formation of the strong distal side of the still open discoidal cell, by 

 change of direction of the first cross-vein between M4 and Cu. 



(3) A further bending of Cu below arculus, correlated with (2). 



(4) Cross-vein situated below distal angle of discoidal cell becomes strong 

 and oblique, and, in the nymphal wing, carries a trachea which captures lA and 

 attaches it to Cu. 



(5) Shortening and simplification of MiA. 



(6) Reduction of M2 to a simple vein. 



(7) Approximation of M3 towards Ms. 



Definite proof of the origin of modern Zygoptera from such a type as 

 Triassagrion cannot be given with the present state of our knowledge. It is 

 more probable that a considerable number of archaic types ancestral to various 

 groups of the true Zygoptera were already in existence in the Upper Trias. 

 One, indeed, we already know in the genus Triassolestes, related to Epiophlebia. 

 We can only add that the recent studies of Professor C. H. Kennedy on the 

 penes of Zygoptera strongly indicate the probability of forms such as Hemi- 

 phlebia and the Megapodagrionidae being the oldest existing Zygoptera, and 

 that this result, startling as it appears to be, would be quite in harmony with 

 the evidence of our Upper Triassic fossils. 



Triassagrion australiense, n.sp. (Plate lii., fig. 31; Text-fig. 77.) 



An almost complete wing, probably a forewing. Total length, 21 mm; 

 Greatest breadth, 4.5 mm. The wing is the reverse of a left wing, as is shown 

 by Ri being concave and Mi convex in the impression. 



The wing is complete except for the following missing parts : — Portions of 

 the costal area broken away (o)" before the end of Sc, (6) in two places between 

 Sc and pterostigma, the second of these being a deep triangular break reaching 

 across Ri, as shown in Plate lii., fig. 31, and (c) from pterostigma to near apex; 

 in this last case, the two posterior angles of the pterostigma are visible, and 

 also the whole of the straight and slightly thickened base along Ri, so that the 

 stigma itself can easily be restored in its entirety. The basal piece of Cu up 

 to beginning of arculus is very faintly preserved, and has been restored back- 

 wards to base, in Text-fig. 77, along the line faintly indicated in the fossil. The 

 posterior margin of the petiole is absent, but a clear indication of the cubito- 

 anal crossing and the beginning of lA below it can be seen; most of the course 

 of lA is very faint indeed. Between the origins of Ms and M2, the wing has 

 slightly buckled by lying above a hard, convex object, probably a fruit 



