NO. 1438. WTNGS OF TJIE TENTHREDINOIDEA—MacGILLIVRAY. 639 



wing margin by coming into line with M^, the medial cross-vein, and 

 the longitudinal part of M,. The medio-cubital cross-vein is longi- 

 tudinal in the Incaliina?, Lobocerinee, and Pergintv, and transverse in 

 the Acordulecerini^? and Ptervgophorinje. The free part of M^ is 

 longitudinal in the Pterygophorinaj and only about one-third the 

 length of the free part of M^, while in the Lobocerinw and Perginaj 

 the free part of M^ is transverse and two or three times as long as the 

 free part of M3. This is due to the migration of the basal end of the 

 medial cross-vein along the free part of Mg. The free part of the 

 second anal vein is wanting throughout the group, due to the great 

 reduction of the anal lobe of the wing, while the second anal furrow 

 and the axillary incision are near the first anal vein. The loss of the 

 second anal is tlie consummation of a variation that has appeared 

 sporadically in other subfamilies. 



THE CELL R5 GROUP. 



It has been pointed out in the case of the generalized Tenthredin- 

 oidea that the radial cross-vein always ends in the cell R^. In the 

 families Xiphydriida? (tig. 8.5), Siricida^ (iig"'^- 86-91), Megalodontida3 

 (tig. 92), and Cephidiv (tigs. 93-96) this cross-v^ein, with rare excep- 

 tions, also ends in the cell R.. In the family Oryssidae (fig. 97) this 

 cross-vein is apparently wanting, but, as was shown above, the cross- 

 vein is present and is represented by the transverse vein at the base 

 of Rg. The only moditication in the course of the cross-vein is that 

 its posterior end has migrated toward the apex of the wing, so that it 

 is alvva3"s oblique to R3+^ instead of being perpendicular. 



The interrelation of the radio-medial cross-vein, the base of the 

 radial sector, and the base of the media is a prominent characteristic 

 of this group of families. In the Xiph3"driidje there is only a slight 

 departure from the arrangement of parts existing in the typical wing, 

 the base of the radial sector and the base of media being parallel, and 

 the cross- vein perpendicular to them. In the other families, however, 

 the posterior end of the cross-vein has swung around at such an angle 

 as to form an apparently continuous vein with a part of the radial 

 sector, while the basal part of the sector extends transversely between 

 the cross- vein and the stigma like a cross- vein. As a result of this 

 change in the direction of the veins, the cells R, 1st Rj, and 2d Ri + R2 

 are arranged in a row. 



The position of the medio-cubital cross-vein in those families in 

 which the posterior end of the radial cross- vein ends in the cell R5 is 

 also of especial interest. In the Xiphydriida? this cross-vein occupies 

 practically the same position that it does in the tj'pical wing. The 

 Oryssidffi show a stage slightly more advanced than that of the Xiph}'- 

 driida?. The cross- vein is longer than the transverse part of media, 

 vvhich has been brought about by a combined migration of the anterior 

 Proc. N. ]\L vol. xxix— 05 42 



