640 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxix. 



end of the cross-vein alon^ media, and b^- a further coalescence of 

 media with radius until it is ahiiost opposite the anterior end of the 

 cross-vein. A similar condition is found in certain Cephidfe, the 

 cross -vein being- about three times as long as the transverse part of 

 media. In this family the modification has been a migration of the 

 anterior end of the cross-vein along media until in certain species, as 

 Ce})hus py gin sens (fig. 96) it arises in the angle between Sc+R+M and 

 media. In the Siricida? and Megalodontids the modification has been 

 in an entirely different direction. The cross-vein and the transverse 

 part of media in these families are subequal in length, while the origin 

 of media is either opposite or beyond the anterior end of the cross- 

 vein and never before it, as it is in all the forms previously described. 

 The manner in which this arrangement of parts has arisen can be best 

 understood if a stud}^ be made first of the condition found in the 

 Xiphydriidee (fig. 85). In this family the cross-vein and the trans- 

 verse part of media are subequal in length, just as in the families 

 named above. The cross-vein is distinctly bowed on the side toward 

 the base of the wing. This bow in the cross-vein has been preserved 

 in practicall}^ all the Siricida3 (figs. 86-91). Now, if the cross-vein 

 maintain this same form and position, and the point of separation of 

 media from radius be gradually changed, moving toward the apex of 

 the wing by the coalescence of media more and more with the radius 

 until it is opposite or beyond the anterior end of the cross vein, 

 exactly the same condition will be had as is found in the Siricidae. The 

 Megalodontidffi (fig. 92) differ only in that the coalescence has pro- 

 ceeded farther, the transverse part of media being distinctly inclined 

 toward the base of the wing, and the cross-vein is straight instead of 

 being bowed. 



The only other possible solution of the arrangement of veins in the 

 the stigmatal area of the Siricid^ would be that starting with a wing 

 like that of Cephus pygmseus^ the base of media had migrated along 

 the cross-vein until near its middle, and that at some later time the 

 anterior end of the combined cross-vein and media had migrated along 

 radius toward the apex of the wing. This would give exactly the 

 same result that has been explained above in another way. That this 

 latter explanation can not be the correct one is proven by the relation 

 of these veins in the Xyelidfe, Lydidse, Cephidge, and the Tenthredi- 

 nidjB. It has been shown that in the first three of these families the 

 tendency is for the progressive coalescence of media with radius, and 

 coordinated with this a progressive migration of the medio-cubital 

 cross-vein from a position near the apex of the cell R to the point of 

 separation of media from radius. That the tendency is not for media 

 to migrate along the cross-vein when the cross-vein reaches the angle 

 between radius and media, as has been shown in the Tenthredinidje, 

 but instead that the cross-vein continues its migration toward the base 



