332 Annals Entomological Society of America [Vol. XV, 



THE FORKING OF THE VEINS. 



Subcosta is so completely fused with costa and radius that a discus- 

 sion of its outer course can lead to no certain conclusion. I have 

 postulated a forking beyond the hinge, to account for one of the obscure 

 cells at this point in the Carabidse and the double thickening of the 

 stigma in Hydrous. The fading out of the trachea of Ri in those forms 

 that have one at all, shows that the terminal portion of Ri has 

 been captured by Sc2 (a common occurrence). In the more specialized 

 forms Sc2 is one of the tracheae that persists, and with the disappearance 

 of true veins in the apical part of the wing, takes an oblique course 

 toward the apex regardless of what traces of veins remain (Fig. 5, 6a; 

 but compare Fig. 6). The tracheae of M and 2d A in certain forms behave 

 similarly. 



Radius. — Ri is a vein that tends strongly to weaken and disappear, 

 its terminal portion fusing with Sc. In several Coleoptera there is a 

 plain anterior branch of the radial stem which continues in the common 

 cavity of Sc and R beyond the point at which the main radial trachea 

 leaves it (Fig. 2) . There is no reason to doubt that this is Ri, and that 

 the main trachea, beyond the bifurcation, is the stem of Rg. 



Beyond this point tracheation fails to give evidence, as the sub- 

 divisions of Rs are unstable in all the forms yet studied. The interpre- 

 tation laid out on the hypothetical plan (Fig. 1) is based on the assump- 

 tions, first, that the apex of the wing in a low holometabolous insect is to 

 be sought in the neighborhood of R3, and second, on the plain connection 

 of the stub here identified as R4+5 across to the stem of Rg in such forms 

 as Tetracha (Fig. 13). The tenninal veining here introduced into the 

 hypothetical plan is nowhere so plain as in Hydrous, but well-marked 

 traces survive, not only in the other large Hydrophilidte, but in the 

 Lamellicorns as well. This group of veins might likewise be interpreted 

 as Mi+2, but the strong tendency to reduction in the median and cubital 

 systems of the Neuroptera would suggest a similar interpretation here. 



In the Polyphaga the base of Rg is atrophied, leaving the outer part 

 as an apparent backward-projecting spur — the radial recurrent (Rr). 

 The second radial cross- vein, on the atrophy of a segment of Rg crossing 

 the main folds, swings into this portion of Rs, and is usually reckoned 

 with it as a portion of the radial recurrent, which would then be desig- 

 nated according to the usual terminology as Rs & 2d r. 



Media. — Media has become two-branched in practically all the 

 known Neuroptera, and in many is reduced to a single stem, forking 

 only at a point corresponding to the nearly veinless apical region of the 

 Coleoptera. The presence of but a single medial trachea in all the 

 forms studied, save for unstable terminal branching, would suggest 

 here that no branches need be sought save in the terminal portion of 

 the wing. I have marked as medials those terminal veins that seem to 

 connect most closely with the medial stem. The main trachea in the 

 Adephaga, runs out in the vein here marked M4 (Fig. 3) ; in the Poly- 

 phaga swinging into the same vein with Cu toward the margin. The 

 two veins that survive in the generality of forms are marked Mi and 



