684 MESOZOIC INSECTS OF QUEENSLAND, ii., 



of the triangular region, the fracture already mentioned cuts us 

 off from completing our survey. Woodward, in his figure, traces 

 the veins, called by me Aj and Ao, across this fracture, together 

 with two interpolated supplements. This is quite unjustifiable, 

 particularly as he makes both these veins, which, before crossing 

 the fracture, are strongly convex, pass, on reaching the other 

 side, into veins which are quite as strongly concave, and which, 

 therefore, if they are parts of the main veins at all, and not mere 

 supplements, must belong to Cu. The only ridgings that can be 

 seen running across the area of bare rock forming the fracture 

 are shown by dotted lines in Plate xliii., fig. 11. As these are 

 continued far beyond the posterior border of the wing, and as 

 similar, though not quite so plain, ridgings can be made out in 

 the area beyond the costal margin of the wing, I can only con- 

 clude that they are part of the structure of the fracture, or of 

 the grain of the rock, and have nothing whatever to do with the 

 wing itself. 



The triangle itself is certainly the most remarkable structure 

 of the kind known in the hind wing of any Anisopterous Dragon- 

 iiv, owing to its intense narrowing in a transverse plane. Its 

 shape, taken in conjunction with its broad and somewhat quad- 

 rangular subtriangle, at once suggests a comparison with the 

 same structures in the forewings of the most highly-developed 

 recent Libellulidw. The two formations are certainly parallel; 

 but, as we cannot trace out the course of the cubitus and its 

 branches with certainty in our fossil, it does not seem possible to 

 institute a detailed comparison. 



Text-fig. 10 shows the actual shape of the triangle and sub- 

 triangle. In Woodward's figure, the shape of the triangle is not 

 quite correctly indicated, and it is made to appear widei- ante- 

 riorly than it really is. Further, the long, distal side is not so 

 curved as in Woodward's figure; and the short, upper side is 

 absolutely straight, with both Cu and A meeting it together at 

 its extreme, proximal end. The cross- veins in the triangle itself 

 are all simple and subparallel, and are more numerous than would 

 appear from AVood ward's figure. One of these cross-veins, the 



