NO. 143S. \VJ.\<uS OF THE TENTHREDIKOIDEA—MacGILLIVRAY. 585 



opposite the point of orii'-in of media, and in Diywura (iig. 63), where 

 it is distinctly l)eyond the radial end of this cross-vein. An interest- 

 ino- nioditieation is found in the subfamily Tenthredinime (tigs. 50-58), 

 where R+M is bent at a prominent angle at the point where it is 

 joined by Scj, indicating a condition more closely related to the Xye- 

 lida? and L3dida? than is found in the other members of the family 

 Tenthredinida?. Sc^ is best preserved in those genera where the medio- 

 cubital cross-vein and the stem of M3+^ are strongly divergent behind. 

 In fact, there seems to be a direct correlation between the divergence 

 of these veins, the widening of the area between costa and Se+R+M, 

 and the preservation of the free part of Scj. This is especially marked 

 in the subfamily Nematinie (tig. iSS). The free part of Sci is entirel}" 

 wanting in the Cephida^ (tigs. 93-96), Oryssida3 (fig. 97), Cimbicina3 

 (figs. 59-60), and numei'ous genera of other groups as Lahidarge (fig. 

 78), FhyUotoma (fig. 5-i), Harpljfhorm, and BlaMlcotoma (fig. 44); while 

 in many genera, as Stroz/ihoceros (fig. 50), Dolerus (fig. 49), and Jlaero- 

 phya (fig. 57) there is a marked thinning out of the costal half of the 

 vein, while in still other genera as AllanUis and Athalia^ there is only 

 a slight projection on the front margin of Sc+R + M. 



If the record has been correctly interpreted, the free part of Scj, 

 as represented in the genus Dineura (fig. 63), has undergone a double 

 migration. First, from near the stigma to the condition found in 

 Xijrla (fig. 35), as shown by the wings of the Xyelidse and Lydida?, and 

 second, after the coalescence of the stem of subcosta with R+M, a 

 remigration toward the stigma has resulted. " 



THE RADIAL AREA. 



Radius divides into R^ and the radial sector just before the stigma, 

 of which the vein R^ forms the hind margin, and beyond the stigma 

 extends along just within the wing margin to or beyond the apex 

 of R3. 



There is only one family of Hymenoptera, the Xyelida? (figs. 31-35), 

 in which all the branches of radius are present. In all the families 

 except the one named the entire free part of R2 has been obliterated. 



In the genus Macroxyela (fig. 33) R.^ arises about midwa}' between 

 the radial cross-vein and the origin of the free part of the vein R^; in 

 Manoxyela (fig. 34) it usually arises near the apex of the cell Rj, though 

 in the same species it may in some specimens arise from the cell R5 

 and in others be interstitial with the free part of the vein R5, and in 

 Xyela (fig. 35), although it normally arises from the cell R^, yet it is 

 sometimes interstitial with the free part of the vein R5. 



The radial cross-vein is situated between the vein R^ and the stem 

 of the radial sector, dividing the cell R^ into two parts. Within cer- 

 tain limits it is fairly constant in its position. In every case, so far 

 as 1 have observed, its anterior end is joined to about the middle of 



