628 CAUDAL GILLS OF ZYOOPTEIUP LARV.t:, 



order. But there are certain conclusions in which the evidence 

 from the structure of the Caudal Gills reinforces that already 

 gained from the study of other larval and imaginal organs: and 

 these, I think, may by now he considered as reasonably proved 

 They are as follows : - 



(1) Of the three families, the Calopterygidcf are the oldest. 

 Within the Calo})lerygidfe, the EpaJhigiuce on the one hand, and 

 the ThorliLCP. on the otlier, remain primitive; but the Caloptery- 

 ifiuce must be reckoned as considerably more specialised. 



(2) The Lestidce appear as a very distinct family, apparently 

 derived from ancestors having close affinity with Calopterygidce. 

 If we could obtain the larva of Epiophlehia^ it is probable that 

 some valuable evidence would be forthcoming as to the more 

 exact origin of this interesting family. The evidence of the 

 Caudal Gills, in any case, reinforces thai of the venation on the 

 point that the LestidcH are most certainly distinct from the 

 AgrionidiP. 



(3) The Agrionidd' are the dominant family, and contain the 

 most highly specialised types of the Suborder. But the Megapod- 

 agr'wuiiiiE reujain still archaic, and far removed from the higher 

 forms. The larval development of Argiolestes is specialised 

 along a line quite unique within the Odonata. The two-jointed 

 gills offer strong evidence in favour c»f the origin of the Froto- 

 ueuriiuH and AgrionimH from a common ancestor, whose larval 

 form must have closely resembled that still extant in the 

 former subfamily. As the P I atyci lemituH {a.hout whose larvye 

 little is known) clearly connect the ProloneuriiKt' on the one 

 hand, and the Ayrioniiue on the other, with the archaic Mega 

 podayrioniiuf^,, it is to be hoped that the study of their larvie 

 will soon be undertaken. 



Apart from the Phylogeny of the Gill-System as a wliole, it 

 would be advisable to state brietly the stages through which the 

 different internal structures of the gills have passed : — 



(1; The Blood-Syatern : — Ov\g\i\2i\\y Si simple extension of the 

 htemoccele into the caudal process, this early became differ- 

 entiated by the closing-off of a narrow canal for the afferent 

 blood-flow. We have already given reasons for supposing that 



1 



