594 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxix. 



anal veins. The Cephida? (tig. 11, c) are an interesting example of 

 how far the reduction of this area may be carried; in this family the 

 emargination of the base of the third anal is entirely wanting, the free 

 part of the second anal is perpendicular to the other veins and opposite 

 the medio-cubital cross-vein, while the wing area behind the third anal 

 is so greatly reduced that this vein in some species practically forms 

 the hind margin of the wing. 



The genera Derecyrta^ Brachi/.vlpJms^ and IToiiov.ria (tig. 11, d), of the 

 family Xiphydriidse are described as having the free part of the sec- 

 ond anal vein present and the third anal vein united with 1st A+2d A 

 at the contraction of the third anal vein. The tirst two genera are 

 unknown to me in nature, but the species of these genera, figured by 

 Westwood " and Kirby, show the contraction of the anal cells of the 

 same type as found in Xiphydrla (tig. 85). In Xiphydria the contrac- 

 tion is much deeper and the third anal approximates more closely the 

 1st A+2d A than in any of the other genera of the Tenthredinoidea. 

 In Koncnvla walshii^' the contraction is still deeper, yet the two veins 

 do not come into actual contact with each other. Xlphydria and 

 KonowUi are of interest in showing successive stages of the anterior 

 migration of the third anal at the contraction and to prove that at 

 least one method of the moditication of the anal cells is by the anas- 

 tomosing of the veins at this point. 



The anal area has been reduced in two very different ways; tirst, by 

 the anastomosis of the third anal with the first and second at the contrac- 

 tion in the second anal cell; second, b}" the shortening of the free part 

 of the second anal until the third anal comes in contact with the com- 

 bined first and second anals. Thus it will be seen that in both cases 

 the reduction is due to anastamosis, but that it takes place at a ditierent 

 point and in a different way. 



AVith the exception of the genera of the family Oryssidsv (tig. 97), 

 the only place where the anal area undergoes any reduction at all is in 

 the family Tenthredinidw, and even here the great majority of the 

 genera fall under the tirst class. As to whether the third anal anas- 

 tomoses with the combined first and second anals before or after the 

 atrophy of the free part of the second anal, it is impossible to tell. 

 If we base our conclusions on the Xiphydriidse, the natural supposition 

 would be that it took place before the atrophy of the free part of the 

 second anal; but, from a careful study of this area, I have been led to 

 conclude, because of the difference in the stages within the different 

 families, that these modifications hav^e arisen independently within 



ffj. O. Westwood. Thesaurus Ent. Oxoniensis. 1S74. W. F. Kirby. List 

 Hymen. Brit. Mus., Tenth, and Siric. I, 18cS2. 



'^I am indebted to Mr. J. Chester Bradley for an opportunity to see a specimen ot 

 this species l)elonj^ing to the U. 8. National Museum Collections. The generic refer- 

 ence was made bv Dr. W. H. Ashmeao. 



