170 DEVELOPMENT OF THE WING-VENATION OP ODONATA, 



opment of the strong anal triangle in the male, or the corres- 

 ponding rounded portion of the wing in the female, A3 tends to 

 migrate from A^.g across Ac, to become attached to A4 very close 

 to Ac(Fig.2). But, in the forewing, it still remains attached to 



-^1-3- 



Let us now study the four branches of A from the base out- 

 wards Firstly, A4 is a short, weak trachea, only reaching back 

 from Ac to about half-way towards the wing-base. In many 

 forms, it comes downwards at its end. Towards the posterior 

 edge of the true wing-base, the line of A4 parallel to Cu -f A is con- 

 tinued by the formation of the secoiidary anal vein A', which 

 becomes attached to the wing-base of the imaginal wing. Hence 

 we see that A' is a bridge-vein continuing the weak trachea A4 

 back to the wing-base; its distal portion is formed about A4 itself, 

 but its basal portion is not formed about any important tracheal 

 branch at all. It is comparable, therefore, to the " bridge-vein '' 

 connecting Rs backwards to M^.o, which occurs in all Anisoptera. 



Secondly, A3 is a fairly strong trachea descending either (pri- 

 marily) from A1.3 (Fig.l), or (secondarily) from Ac or A4(Fig.2), 

 towards the posterior border of the wing. About it, in the 

 imaginal wdng, the distal side of the anal triangle is formed in 

 the male(Fig.3). In the female, it is usually less regressed 

 towards the base, and generally descends straight from Ac, 

 though individual variations, both basad and distad from Ac, are 

 sometimes seen. It gives rise, in the female imago, to a descend- 

 ing vein of less importance than the distal side of the triangle in 

 the male, but corresponding to it. The fate of this branch, in 

 both sexes of the Libellidinoi, is of interest, and will be followed 

 out later. 



Thirdly, An is a strong branch descending sharply from Aj.j 

 somewhat distad from Ac, and roughly parallel to Ao. In the 

 uEschnince (and, indeed, in all forms in which a " loop " is de- 

 veloped) it forms the proximal or basal side of the "anal loop" 

 (Al, Fig.3) 



Fourthly, Aj is a strong trachea which continues distally 

 towards Cuo, which it meets quite close under the point of bifur- 

 cation of Cua from Cn^. It then fuses with Cuo for some dis- 



