334 Annals Entomological Society of America [Vol. XV, 



obviously fused with the first branch of 3dA, forming a pointed wedge- 

 cell, and leaving only two terminal branches below the second anal 

 furrow (compare Figs. 32 and 37). 



On the basal side of the wedge-cell there is a short oblique vein 

 running from the stem of 2d A to the upper branch of 3d A. Whether 

 this is a fourth branch of 2d A, which has permanently joined 3dAi, 

 or a cross vein, is not clear. Occasionally it contains a trachea arising 

 from 2d A, but tracheae in other crossveins are not entirely unknown. 

 In the pupa of Tenebrio and Dytiscus it is more definitely transverse 

 than in the imago, a significant point. 



Third anal. — The third anal vein forks once near the base of the 

 wing as a rule. Sometimes both branches have a trachea, sometimes 

 only the lower. The upper is connected to 2d A by two transverse 

 veins, enclosing the wedge-cell between them. In a few forms the 

 third anal is simple, presumably by the atrophy of its upper branch, 

 which is broken, for instance, in the Lamellicorns (Fig. 62). 



Fourth anal. — The anterior branch of the fourth anal vein is present 

 in all save markedly reduced species, and contains a trachea in all 

 species with fairly complete tracheation. A second branch is present 

 in most Adephaga, running along the inner margin of the wing, and in 

 a few a third, running back and stiffening the alula. This is the super- 

 ficial interpretation; it is not impossible that a detailed comparison 

 with the Neuropteroids will result in a different interpretation. In 

 particular the vein here considered 3d A2, may possibly belong in 

 fact to 4th A. 



CROSS-VEINS. 



It is evident that the Coleoptera are descended from a form with a 

 considerable number of cross- veins, which were tending at least to 

 take definite positions. Assuming that they were not wholly definite, 

 the survival of certain ones was doubtless determined by the folding, 

 which necessitated a more complete cross-bracing than in the 

 Trichoptera and Lepidoptera, for instance. The humeral has already 

 been taken up. Discussion of other cross-veins in the costal region 

 would be useless, as the longitudinal veins are almost completely 

 fused. Comparison with the Neuroptera would prepare one for a 

 large number of such veins, but the few forms in which C and Sc are 

 separate (e. g., Cupes, Fig. 12) show no sign of them. 



Between Ri and Rg there is plain evidence of several cross-veins. 

 For the region basad of the pivot-fold, the Hydrophilidae, and especially 

 the primitive Elaters, give the best evidence. They plainly show two 

 cross- veins, one variable in position, but well before the fold of the 

 wing, the other stiffening the edge of the first fold, being the outer 

 part of the vein marked Rs+r in Fig. 30. The latter is obscurely 

 indicated in. several Adephaga, but there is no trace, apparently, of the 

 first, the radial fork coming just before the hinge. 



Beyond the hinge the Hydrophilidse show only faint traces of veins, 

 and most Polyphaga none at all; but the Adephaga have two well- 

 developed cross-veins, which may be called 3d r and 4th r. The 



