BY R. J. TILL YARD. 443 



Legion v. Agrion. 



In this legion, we find an exceedingly uniform type of venation, 

 which may certainly be claimed as the most highly specialised in 

 the family. The general result of this line of development has 

 been similar, both in its signal success, and in the type of insect 

 produced, to the more archaic Lestes. By the strengthening of 

 the quadrilatei-al, which assumed, as in Lestes, an irregular form, 

 by the early fixation of the second antenodal at the level of the 

 arculus, and by the stability in position of the median sector, the 

 leoional type became a great success without either having to 

 sacrifice the value of the sectors of the triangle as parts of the 

 wing-scheme or being in any way driven to extremes of petiola- 

 lation. Though a distinct advance on the Lestes wing-form, j'^et 

 this Ayrion-tjTpe of wing must be of considerable antiquit}^, and 

 I regard those membeivs of the legion which show most aberration 

 fi'om that type, not as archaic remnants of a still older wing-type, 

 but i-ather as the result of further asthenogenesis brought about 

 by special conditions. For instance, the genus Agriocnemis is not 

 to be regarded as more archaic than the type-forms, merely because 

 its quadrilateral is more regular and the position of the second 

 antenodal not in line with the arculus; but rather, these must 

 be considered as the inevitable efi'ects of asthenogenetic reduction 

 beyond the type-stage. Proceeding even further along the same 

 lines, we come to the remarkable genus HemijMehia, in which 

 the asthenogenesis is so complete that, in the forewing, the cross- 

 vein forming the lower half of the arculus, i.e., the basal side of 

 the quadrilateral, has been entirely dispensed with, and the vein 

 M4 thereby considerably strengthened and straightened out. 

 This may be fairly claimed as the most advanced asthenogenetic 

 type of Odonate wing yet developed. [In this connection it is 

 interesting to note that, in the female of Hemiphlehia, this cross- 

 vein is still occasionally present in the forewing, indicating how 

 very recently this asthenogenetic development began]. 



Owing to the uniform venation of this legion, it is difficult to 

 choose suitable venational generic characters. If, however, we 

 study the various stages of the asthenogenetic process that are 

 still evident, we may obtain reliable criteria for the subdivisij3*i ;"?r" ^ 



> ^^«^ V^^ 

 t I BR A R Y 31 



