THE WINGS OF EPHEMERIDA 221 



media, which forking at the base of the wing plays the part of two principal 

 veins; vein IMi+o being concave and vein M3+4, convex. 



The chief branches of the principal veins, counting media as two veins, 

 are of the same nature as the principal veins, except in the case of the radial 

 sector; but an alternation of convex and concave veins has been attained 

 by the development of intercalary \^eins, which in each case differ from the 

 veins between which they are situated. In the area of vein Mi ^2, where 

 the supposed radial sector is interpolated, it requires two intercalary veins 

 to complete the alternation, veins IMi and IR^. 



In addition to the intercalary veins enumerated in the table many 

 other intercalary veins are developed at the margin of the wing in a more or 

 less irregu'ar manner; but wherever a second intercalary vein extends far 

 into the disk of the wing it is accompanied by a third, one being convex, the 

 other concave, except in the anal area where intercalary veins are more of 

 the nature of braces, like cross-veins. 



Correlated with the development of a triangular form of wing, which 

 involves an expanding of its outer margin, is the fact that the secondary 

 longitudinal veins are all added distally in the May-flies. But the method 

 of development of these veins appears to be radically different from what it 

 is in the Neuroptera. There the accessor}^ longitudinal veins are preceded 

 by trachea?, which arise as fine twigs at the tips of older trachese, and which 

 in the course of phylogenetic development branch off" from the parent 

 trachea farther and farther from the margin of the wing, thus making 

 room for the development of other twigs. Here, in the May-flies, the 

 secondary longitudinal veins are evidently thickened folds, each of which 

 arose more or less nearly midway between other veins, with which, at first, 

 it had no connection. 



A fact of prime importance in the study of the homologies of the wing- 

 veins of May-flies is that the corrugations of the wing are the most per- 

 sistent features of it. Hence the most important criterion for determining 

 the homology of a vein is whether it is a concave or a convex one. This is 

 especially true of the hind wings where a variable number of the veins have 

 been lost. 



The stiffening of the costal margin of the wing by the formation of a 

 subcostal furrow has been attained in most of the orders of insects; and in 

 several of them the formation of folds has extended, to a greater or less 

 degree, to other parts of the wing. But as a rule, the latter method of 

 specialization has not been the most important one in perfecting the wing. 

 In the Odonata it has been carried farther than elsewhere except in the 

 Ephemerida. But in the Odonata it has been supplemented by other 

 methods of specialization, with the result that an exceedingly efficient organ 

 of flight has been developed in that order; while in the Ephemerida the 

 cephalization of the flight-function and the corrugating of the wings have 



