84 Annals Entomological Society of America [Vol. VI, 



But in the Membracidae several points not compatable with 

 this natural method of reduction presented difficulties. In the 

 first place, both branches showed constant and unmistakable 

 signs of further subdivision at their tips, which would not be 

 likely to be true of the cephalic branch if it were Ri. Moreover, 

 the vein between the cephalic branch and subcosta was often 

 seen to be preceded by a trachea. Again and again in mounts 

 of different species this area contained a trachea which was 

 evidently a branch from the cephalic branch of radius. If this 

 were true, this most anterior branch should be Ri. But Ri 

 normally leaves the main stem proximad of the division of the 

 radial sector, while this branch seemingly pulls off from one 

 half of the sector itself, and this demanded an explanation which 

 was not immediately forthcoming. 



The solution was first found in the wings of Vandnzea 

 arguata and later this peculiar condition (Fig. 9) was verified 

 in other genera. The trachea representing Ri, as will be seen 

 from the figure, is weak and apparently greatly reduced. It 

 leaves the main stem in the normal postion, but runs in close 

 juxtaposition to the radial sector beyond the point at which the 

 latter branches. Here it turns cephalad and runs across to 

 subcosta where it again turns outward and closely parallels sub- 

 costa for some distance in its course toward the tip of the wing. 

 The sharp turns made by the trachea in following this course 

 (Fig. 10) are remarkable, and in the veins which enclose this 

 region of the wing, the bridge from radial sector to subcosta 

 (Fig. 11) gives every appearance of a cross-vein. 



While this interesting behavior of Ri is unusual, and per- 

 •haps peculiar to the Membracidae, it only illustrates another of 

 the vagaries of which radius is capable. In fact, throughout 

 the Hemiptera, radius seems to be most unreliable, and Ri 

 capable of the most peculiar performances, being, according to 

 Miss Patch^^ "the least stable of the hemipterous wing veins". 

 It has been shown in the Cicadidse^*^ that Ri has been crowded by 

 subcosta until its trachea coalesces for its entire length with 

 radial sector and its anterior branch. In the Pentatomidae^^ 

 also, it has been supplanted by subcosta and is entirely 



17. Annals of the Entomological vSociety of America, 1909, Vol. 2, p. 119. 



18. Wings of Insects, p. 245. 



19. Wings of Insects, p. 250. 



