no. 1331. 



DRAGON-FLY WING VENATION— NEEDHAM. 



721 



deflected portion of the cubitus (the inner side of the triangle) has 

 jbeen elongated, producing the very narrowly elongate triangle of 

 iLibellula (Plate XLVllI, tig. 3) and its allies. 



' B. The recession of the triangle in the hind wijuj. — The Ijcliavior of 

 ithe triangle in the hind wing has consisted siinpl}' in its retraction 

 through the successive position shown diagraniaticalh' in fig. 16, and 

 ^illustrated in such genera as Agrionoptera (Plate XLIV, fig 1), 

 [Raphi.Kinia (Plate XLIV, fig. 3), TJracis^ (Plate L, fig. 2), and PantaJa 

 ((Plate L, fig. 3). Doubtless this recession of the triangle of the hind 

 Jwing could only be advantageous in connection with the developments 

 (described under the next heading and discussed again farther on. 

 \ The anal area. — In the Odonata there is no anal furrow or suture, 

 as in most other insects, sharply delimiting an anal field. The anal 

 ivein, except in a few fossil forms {TstpJdehia^ fig. 31, etc.), is solidly 

 joined to the cubitus, as already shown. We now ])ass to notice the 

 i distribution of the anal 

 i branches in the small area 

 back of the cubital deflec- 

 i tion. 



Pro])al)ly in the early 

 Odonatii the anal tracheal 

 trunk occupied the posi- 

 tion of the anal vein, well 

 back of the cubitus. On- 

 togeny (see figs. 1 and 2) 

 and analogy with insects 

 of other orders would l)oth 

 support this view. But in such recent forms as I have had for study 

 this anal trachea is closely approximated to the cubitus beyond the 

 base, and, midway to the triangle, descends to the lexel of the anal 

 vein, and sends branches both proximal ly and distally, about which 

 that vein is formed (fig. 2, and Plate XXXI, figs. 1-3). Even in 

 the thick wings of nymphs of Anisoptera the contorted position and 

 decreasing size of the anal trachea indicate that, like the costal trachea 

 of the opposite wing margin, it is sufiering from lack of room; or, per- 

 haps, outstripped by the intxn-vening traciiea' that are l)etter situated 

 in relation to air supply. At least we find, especially in thin and 

 narrow wing cases, the anal trachea dwindles, and loses successive!}'' 

 all its )n-anches, which then become incorporated into the cubital air 

 system. The best developed anal trachea we have seen is that of Anax 

 (Plate XXXI, fig. 3), in which it will be noticed that the terminal branch, 

 ^1, is continued l)eyond its ]K)sition of fusion with the cubitus in a 

 recurrent positioti. In the hind wing of (Tonij>Inix{V\\\iQ. XXXI, tig. 1) 

 it will be ol)served that this recurrent portion is attached to vein <'i(.,\ 

 -4i, ai)parently, wanting; yl.,, strongly developed. In the tore wing 

 of the same nymph ^1.. has also been transferred to the cubitus, being 



Fk;. li). — Diagram repkesentini; the recession or the 



TRIANGLE IN THE HIND WINGS OF THE LlBELLULID.K, 1, 

 '1, 3, AND 4, SUCCESSIVE STAGES. 



