o. 1331. 



DRAGON-FLY WING VENATION— NEEDHAM. 



717 



Fig. 11.— Diagram illustrating the behavior of the 

 quadrangle in the cai.opterygid.e. solid lines, a 

 somewhat primitive quadrangle; line of dots, the 



QUADRANGLE OF HcUrrhuV. LINE OF DASHES, THAT OF 



Rhinocypha; arrows indicate the opposite course 

 of rotation. compare with plates li and lii. 



In Agrionida? (Plates LllI and LIV) the quadrangle is undivided, the 

 iniddle cross vein ])eing absent. In the Calopterygida? (Plate LII) the 

 middle cross vein is occasionally wanting, as in AnhopJiHi'd and Kpal- 

 \lage, but in general the quadrangle is elongated and contains numerous 

 [extra cross veins, among which the identity of the typical one is lost. 

 The quadrangle is gener- 

 ally rectangular in this R+M ^-.^w ^ . " ■ R 



family, but in Ilefxrinn 

 (Plate LI, tig. 4) it i^ wid- 

 ened distally and convex an- 

 teriorly; in Thore (tig. 35) 

 and its allies, exactly the 

 , reverse. These facts are 

 [illumined when one sees 

 I what has been the behavior 

 \ of the cross vein which 

 I terminates the quadrangle 

 1 in this family. Vein C^i/^, 

 * separating from vein Ou^ at a right angle, and as suddenly bending 

 I again distally, sets off a transverse basal portion which is in direct line 

 I with this cross vein. The two thus joined rotate together about the 

 1 hind angle of the triangle as an axial point, while the two forms of 

 quadrangle described above are developing. In Iletxrina (Plate LI, 

 I tig. 4) and Lais the medial end of the 



cross vein has proceeded distally, while 

 the base of Cuo^ has been retracted; in 

 Thorc (tig. 35), Rhinocypha^ Lrstcx^ 

 etc., the reverse rotation has taken 

 place, as illustrated in the acconqjany- 

 ing diagram (tig. 11). 



Tlte triangle of the Anisoptera. — Re- 

 turning now to the suborder Anisop- 

 tera, and to the triangle as an individ- 

 ual feature of the wing, we may follow 

 with the aid of tig. 12 the changes that 

 have taken place in it, bringing it 

 from the condition of an ordinary rec- 

 tano'ular cell to its present estate. It 

 will l)e convenient to begin with a 

 triangle hardly more generalized than 

 that of the fore wing of Tetrathemis^ 

 and to follow the shiftings of its parts severally. 



1. The angulation of the cnhitm. — At the dividing cross vein of the 

 quadrangle, the cubitus tends in ail Anisoptera to form an angle, 

 which pu.shes triangle and supertriangle out of line, thus destroying 

 the unity of the quadrangle, in making of it two elements of wing 



:^^^ 



Fig. 12.— Diagram setting forth the 

 behavior of the triangle in the sub- 

 ORDER ANISOPTERA. The heavy LINES 

 BOUND A SOMEWHAT PRIJIITIVE TRIANGLE. 

 1,2, 3, 4, AND 5 ARE STAGES IN THE DESCENT 

 OF THE UPPER CROSS VEIN. 1', 2', 3', AND 

 4' REPRESENT SUCCESSIVE STAGES IN THE 

 RETRACTION OF THE CUBITUS AT THE TRI- 

 ANGLE. 1", 2", 3", AND 4" REPRESENT 

 STAGES IN THE RETRACTION OF THE BASE 



OF VEIN C»2. 1'", 2'". 3'", AND 4'" reprp:- 



SENT STAGES IN THE ASCENT OF THE VEIN 

 CH] up THE OUTER SIDE OF THE TKl AN(; LK. 



