54 THE WINGS [CH. 



forms like Aeschnidium (fig. 163) and the recent Telephlebia 

 (fig. 19 F) seemed to shew the subcosta running through and 

 beyond the nodus. 



The recent discovery of the fossil Mesophlebia (fig. 160) is a 

 great step in advance. Though it may not be true that all recent 

 Odonate nodi have been actually evolved by passing through a 

 stage resembling that of Mesophlebia, yet it seems very clear that 

 this stage gives us a very simple solution of the problem, and one 

 which seems very likely to have been the actual evolutionary 

 solution in the case of the Anisoptera, and probably of most 

 Zygoptera as well. 



In order to follow the steps of the evolutionary process, it is 

 essential to bear in mind that the so-called costa of the Dragonfly 

 wing is not really a main vein formed along a preceding trachea, 

 but merely the strengthened anterior border of the wing. Thus 

 its structure is quite different from that of a main vein. When 

 carefully examined it is seen to be regularly ribbed by a series of 

 spines or teeth along its outer border. It seems to be clear that 

 this formation is due to the action of separate hypoderm cells, 

 and that the portion of the costa carrying one such tooth is the 

 product of a single pair of cells. Thus there is no inherent difficulty 

 in understanding how & joint, such as the nodus, could be evolved 

 on the costa, though this difficulty would be hard to overcome if 

 we were dealing with a main vein of uniform structure. Granted 

 then that the nature of the costa is such that the joint can be 

 easily formed when required, Mesophlebia shews us definitely how 

 it was formed. It is clear from this fossil that the nodus consists 

 of two distinct parts only, viz. the part proximal to the actual 

 joint, and the part distal to it. The former is derived from a 

 strengthening of the original fusion between the costa and the end 

 of the subcosta. The latter is formed simply by specialization 

 of the cross-vein lying distally next to the ending of the subcosta, 

 strengthened by a thickening of the costa itself just distally to the 

 joint. I shall call this cross-vein the nodal cross-vein (nc). The 

 nodus of recent Dragonflies differs from that of Mesophlebia only 

 in two points, (1) by a closer union of parts around the joint, (2) by 

 the alteration in direction of the lower half of the nodal cross-vein. 

 This latter change has been brought about by the need of 



