K0.1331. DRAGON-FLY WING VENATION— NEEDHAM. 726 



cross veins are reduced so as to leave those that remain in such position 

 that^ach seems to bear its individual share in the stress upon the wing. 

 Instead of a dozen, nioi-e or less, of inconstant cross veins in the space 

 between veins B^ and 31^ between nodus and stio-ma, as in yeurothemis, 

 there are but three, and these three are constant, and so for other 

 parts. In and about the triangle of Nearotheinin arc many veinlets 

 which have been sacrificed to make the triangle itself stronger in 

 Tetragoneiiria. An actual count of the cells in a hind wing of JVeuro- 

 them/s gives the number 2,695; in a hind wing of Tetragoneurta^ 265, 

 the latter wing being at the same time a little larger. Neurothemis 

 has far outrun Tetragoneurla in all those adjustments of parts in the 

 region of the arculus, already described, as characteristic of the spe- 

 cialization of the Libellulida?; but Tetragoneuria^ having attained a 

 fair measure of mechanically advantageous arrangement of parts, has 

 attained success by disposing of its strength-giving wing material 

 where it is most effective. Tetragoneiiria is vastly superior in flight " — 

 is, indeed, one of the fleetest and most agile of winged creatures. 



CKt)fSS VEINS. 



The vein-building substance of which we speak is of course hypo- 

 dermis. The insect wing is essentially but a flat evagination of the 

 bod}' wall, with a few trachea gi'own out into it. During early devel- 

 opment the hypodermis of the Aving does not differ in an}' essential 

 respect from that of other parts of the body. As elsewhere, it con- 

 sists of a single laj'er of cells which secrete a protecting external 

 layer of chitine. At the time of transformation, when the hypodermis 

 of the two walls of the wing sac is bound together by fused internal 

 processes, blood is forced out into the wing, greatly extending it lat- 

 erally. The hypodermis is thus spread out in a very thin layer. As 

 soon as a definitive layer of chitine is deposited, the hypodermal cells 

 (which, of all cells known to me, possess the greatest capacity for 

 speed}' and extensive shifting and readjustment) begin to be segregated 

 into groups along the lines of the veins that are to l)e, and there 

 deposit additional chitine, which differentiates veins from meni])rane. 



Doubtless in the earliest insect wings the segregation of the hypo- 

 dermal cells was such as to give a membrane crowded with somewhat cir- 

 cular areoles, such as we find in the expanded lateral margins of the 

 pronotum of the Tingitida\ in the tegminaof the Locustidiv, in ahnost 

 the entire wing of the fossil dragon fly ^'Esch)ii<Jh(iii^ and in the wider 

 spaces of the wing of NeirrofJtern/x. The principal veins first would 

 become strongly marked by the accumulation of the hypodermal cells 

 about the trachea?. Cross veins would emerge from the meshwork, as 

 they seem to be emerging in the wing of y^eurothemts^ in the spaces 

 between veins R^ and J/^, between J/j and J/"^, etc., by the dropping 



« I venture this unqualified statement without having seen Neurothemis fly. 



