82 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 39. 



the coxal subdivision of the episternum (eps^) the " mesosternum," 

 while the part of the upper plate above the groove (z) he calls the 

 "epimerum" and the part below the "episternum." A person who 

 has studied ants alone may be excused for making such an interpreta- 

 tion as this, but, in the light of a comparative study of all the Hymen- 

 optera, the writer can not see how the sutures (r) and (z) can be any- 

 thing other than secondary grooves in the mesepisternum. The 

 writer has not observed a metapleurum in the ants constructed as in 

 Emery's figure of StreUognathus. In Leptogonys and other Ponerines 

 examined the metapleurum (PZ3 and pl^) is very indefinitely demarked 

 from the propodeum {IT), and the metapostnotum (PN^) is not distinct 

 from either. Emery calls the lower part of the metapleurum the 

 " metasternum " while in the upper part he finds both a metepisternum 

 and a metepimerum. He makes a very curious use of the word "par- 

 apternum " which he applies to the anterior subdivision of the meso- 

 scutellum (scl^). The writer has shown elsewhere (1910, footnote a, 

 p. 20) that Audouin's paraptere is a little plate in the pleurum before 

 the base of the wing (see p. 47). In Myrmica piriformis Emery calls 

 what is apparently a subdivision of the metanotum the "metaparap- 

 terum." The writer feels confident that Emery's interpretations of 

 the thoracic parts of ants are due to a deficient study of other Hymen- 

 opteran families leading up to them from the Tenthredinoidea and 

 Siricoidea, and that his homologies must appear erroneous to anyone 

 who will ground his morphological ideas on the thoracic structure of 

 these generalized forms. 



5. WINGS, THEIR VENATION AND ARTICULATION. 



A comprehensive study of the wings is beyond the scope of the 

 present paper, but there are some interesting points brought out in 

 a study of the evolution of their basal parts. The Hymenopteran 

 venation is so different from that of all other insects that any scheme 

 of homology with the other orders involving the branches of the veins 

 is purely speculative. The Comstock-Needham system of nomen- 

 clature as applied to the front wing of Sirex flavicornis is shown by 

 figure 74. It assumes that the. fourth and fifth branches of the 

 radius (R^ and ^5) have been bent back toward the posterior edge 

 of the wing and fused with the neighboring branches of the media 

 and that the third and fourth branches of the media (i/3 and M^) 

 have been likewise turned back and united with the cubitus (Cu), 

 while this last vein fuses with the first anal (lA). If all the terminal 

 branches of the veins in this wing were to be designated according to 

 the veins that unite in their formation, they would have to be given, 

 in many cases, names entirely too long for practical purposes. For 

 this reason Hymenopteran systematists have not commonly adopted 

 the Comstock-Needham nomenclature, but continue to use that of 

 Cresson (1887). Figure 76 shows the front wing of an Ichneumonid, 



