liY R. J. TiLLYARb. 689 



We may summarise the chief points of interest in this region 

 of the wing as follows: — 



(1) The nodus is fully formed, but weak. 



(2) The subnodus is also complete. 



(3) The apparent prolongation of Sc through the nodus is of 

 the same nature as that seen in Telephlebia^ i.e., due to the 

 secondary formation of a supplement between C and R, by align- 

 ment of the boundaries of numerous cellules of the dense mesh- 

 work to form a sahcostal iHtj}j)lHmeut^ 8cspl; but, in the fossil, this 

 Scspl is completely formed, whereas, in Telephlebia, it is of only 

 about two cellules' length. 



(4) Supplements similar to 8cspl, and due to the same cause, 

 are formed proximally to the nodus, between C and fSc, and 

 distally from the nodus, between C and 8cspl; but the latter of 

 these does not reach back proximally to the nodus (see Plate 

 xliii., fig. 11), 



(5) Rs arises from R directly under the subnodus. It is a 

 weak vein, running obliquely to M, crossing Mj and M., not far 

 distad from their point of union, and then dropping towards Ms 

 as the incomplete, oblique vein O. 



It will be seen that, if this fossil gives us the right interpreta- 

 tion of the structure of Rs and Ms, the distinction between the 

 Anisoptera and Zygoptera in this region almost disappears. The 

 only difference would be that, in the former, trachea Rs has 

 become strengthened by capturing Ms, so that the basal part of 

 the latter (the bridge) has become aborted; while, in the latter, 

 it continues to weaken, and fails ever to cross Mi^o- 



Classification of the Fossil Wing. 



As long as we decide to keep to the old classification of the 

 Anisoptera, in which only two families are recognised (the 

 ^EschaidcB and the Libellulidce), it seems to me that we cannot 

 grant full family-rank to the ^Eschnidiidie. These latter have 

 the triangles of the fore- and hind wings alike (both being 

 elongated transversely to the wing-axis), and the antenodals of 

 the first and second series not corresponding. Hence they are, 



