NO. 1331. DRAGON-FLY WING VENATION— NEEDHAM. 747 



distally by an oblique vein. It differs utterly, however, in the form 

 of the quadrano'le of the hind wing-, in the remoteness of the nodus 

 from the arculus, and in the relation of the arculus to the hypertro- 

 phied antenodal cross veins 



It agrees with generalized Gomphines in the last-named particular 

 (compare with Plate XXXV, tig. 3), also in the general relations of 

 bridge and oblique vein, and relations between arculus and nodus; 

 but its quadrangles are undivided, and the hind margin of both wings 

 is utterl}" unlike all known Anisoptera. In having a quadrangle that 

 is obliquely placed, narrowing distally in the fore wing and widening 

 distally in the hind wing and undivided in l)oth, it stands entirely 

 alone. 



De Selys pointed out when he described the species that it showed 

 striking reseml)lance to the fossil genus Ileterophlchia — a genus, unfor- 

 tunately, still insufficiently known. The quadrangle of the fore wing 

 is, in fact, practically identical; and other parts of the wing, so far as 

 known (whether there were hypertrophied antenodals is uncertain), are 

 similar. But the quadrangle of the hind wing in IleteropJdc'hki is 

 very different. The cubitus, instead of being declined before the 



Fig. 36.— Hind wing of Cora incana Hagen. 



i quadrangle as in Palseophleb'ta^ is bent at the middle of its posterior 

 ' side, and from that point springs the dividing cross vein, as in Anisop- 

 tera; and the cross vein is declined so far that it rests against the 

 upper end of the terminal cross vein of the quadrangle. There is yet 

 another point of difference, of perhaps even greater importance. The 

 : bridge at its inner end is directed toward and attached to vein J/j in 

 ' Ileterophlehia, whereas it turns the other way and attaches to vein 

 |J/i+2in Palxoplilebia and in all living Odonata. (Although in the 

 more generalized Gomphina? it is apparentlv symmetrically forked 

 and not turned either way, whenever the fork happens to be un.sym- 

 , metrical the leaning is seen to be toward the anterior side.) This is a 

 difference of kind that is not to bo passed over lightly. 

 [ HeterojMeUa shares this peculiarity with one other genus, Tarso- 

 \phlehia, likewise fossil, and it is a strong bond of union between them, 

 iof more importance than their rather striking differences, these being 

 mainly differences of degree. Tar sop! deb la" has the quadrangles 



a Libellula panneuitziana Goppert belongs in Tarsophlebia, as will be seen by com- 

 i paring Assman's figure of the type (Zeitschr. fiir Entomologie, I (n. s.), 1870, pi. i, 

 [fig. 11) with fig. 3 of Hagen'a plate cited herewith. Heterophkbia jucunda Hagen is 

 f not a Heterophh-hia at all, nor even nearly related thereto. 



