Vol. 1.] Wood'WortJi. — Wiug Veins of Insects. 95 



nodus, and the approximation of the first and second posteriors 

 at the other connective, strengthen this view. 



The Odonata, a group rich in problems of venation, are full 

 of interest, but we shall have to confine ourselves to the con- 

 sideration of a single one— the development of the structure 

 known as the triangle. This structure constitutes the sole con- 

 stant difference between the venations of the two divisions of 

 the Odonata. There are almost innumerable points of differ- 

 ence between the venation of typical representatives of these 

 groups, as for instance between Agrion and Esrhna ; but within 

 each group there is such diversity that, if we take up one by 

 one each point of difference and trace it out through the whole 

 series, the differences vanish. 



The triangle is found among the Odonata only in the 

 Anisoptera, and is, indeed, wanting in all other insect wings. 

 All writers have, I believe, uniformly, and probably cor- 

 rectly, homologized all of the veins in the neighborhood of 

 the triangle with the corresponding veins in the Zygoptera; 

 none have, however, attempted to trace the method of the 

 origin of the triangle. The only discussion of its genesis is 

 the brief one by Comstock and Needham ('98 99, p. 908). I 

 can agree with them as to the nature of the bounding veins — 

 the cubitus and two cross veins — which they were the first to 

 point out, but am not so ready to agree as to the course of 

 development which they aim to indicate by their diagrams. 

 There is no necessity, I think, for conceiving the triangle to 

 have been, as they assume, originally the whole of a quadri- 

 lateral cell. Moreover, their figure 65 simply provides 

 for different positions of the triangle, by variations in the 

 angle made by the "inner cross vein" with the basal portion 

 of the cubitus and in that made by the bend of the cubitus; 

 it does not, nor does the wing of any of the Anisoptera, show 

 any quadrilateral cell, nor any transition between a quadrilat- 

 eral and a triangular cell. This view assumes that by the 

 approximation of the corresponding ends of two cross veins 

 till they meet, a quadrilateral area is converted into a trian- 

 gular one. My view is that the triangle is formed by the 

 division of a quadrilateral cell into two triangular cells, one 

 of which is the triangle. How, in my opinion, this has been 

 brought about I will explain directly. 



I would particularly object to their interpretation of the 



