724 A NEW SPECIES OF iSAi\' AOPHLEBJA, 



matiou of unij suitable bracing or strengthening of the wing-area, 

 either on Aeschaine or Libellidine lines. 



I do not contend that it is necessary to assume even that the 

 present-day " quadrilaterals " in the Libellulidce, ever possessed 

 ancestors with a fully-formed three-sided triangle. I think, rather, 

 that their ancestors were the laggards in this race for the perfect 

 triangle, and that they attained only a certain measure of success 

 in that direction, without gaining a position of equilibrium. Need- 

 ham has shown (without emphasising the point) by his excellent 

 diagrams of the gradual formation of the triangle, what an enor- 

 mous stress must be thrown on to those origmally weak cross- 

 veins, which are finally called upon to play the part of strong sides 

 to the triangle. Anything less than complete success, in this diffi- 

 cult piece of evolution, must surely have stood self -condemned, and 

 either exists still as an unstable form (e.g., Synthemis) or em- 

 barked on the backward path of asthenogenesis, in which, since the 

 aim is no longer to j^roduce the ideal flying wing, the reduction 

 back to the weaker form would rapidly proceed. 



• The argument for the Single-Development Theory of the Anal 

 Loops may be briefly put as follows : — All present-day hind-wings 

 of the Anisoptera were developed from an original anisopterous 

 type, which had a hind-wing broad enough to contain all the cell- 

 material necessary for present-day developments. By various ar- 

 rangements of the basal cells and the anal and cubital branch- 

 veins traversing them, all the present-day loops (and probably 

 other kinds now lost) were developed. Thus arose, with varying 

 degrees of final success, the Gompltine, Petalurine, Chlorogom- 

 phine, Aeschnine, Macromian, and Libelluline loops, in all their 

 forms and variations. Of these, the most recent and most success- 

 ful is the Libelluline loop, whose origin may probably be sought 

 near to the point from which Chlorogomphus sprang, and from 

 which that genus app/;ars to have diverged but little. 



Besides these menogeneiic types, at various points along the 

 evolutionary route, unsuccessful competitors gave up the race, and 

 adopted, as a means for their preservation, the process of asthe- 

 nogenesis, so successfully carried out long before by the main 



