1910[ Ris, the Identity of Tiro Odonata Fossils. 103 



1. None of the antenodals is thickened. This, indeed, could be an 

 error in reproduction. It is evident, that the first one or two anteno- 

 dals are not figured at all ; so the first thickened antenodal would be 

 lost. The second of the represented antenodals could possibly be 

 thickened, although it is not so figured. So this argument carries little 

 weight. 



2. The bridge is of Libellulid form. The Gomphine bridge shows at 

 its proximal end a nervule towards M 8 , and all Gomphirice known to me 

 have supplementary bridge-cross-nervules. This argument, too, it not 

 of great moment, as the nervules in question could be omitted by the 

 author of the design. 



:s. No Gomphirice are known to me, in which the distance between 

 arculus and triangle is so great as that figured. It is perhaps greater 

 in let mux than in any other living Gomphine, but still considerably 

 less than in the figure. I think this argument is of very considerable 

 weight, the point specified being not easily subject to error in drawing. 



4. The distance between the subnodus and the first nervule be- 

 tween B and M, is too great for a Gomphine, but regular for a Libel- 

 lulid. Another argument of only relative importance, as a nervule 

 might bt omitted. 



5. There is a very distinct radial supplement, parallel to Its. This 

 is a well-known feature among TAhelluUdCB, indistinct only in some of 

 the most primitive genera. No Gomphine wing is known to me. where 

 a supplement parallel to Rs is developed. Some (loniphimi may have 

 a nervure for which the term radial supplement might be claimed (as 

 instanced in Xeedham. Proc. U. S. Nat. Mus., 26, 1903; pi. 34. fig. 3— 

 Gomphidia) ; but it is in these cases a nervure branching off from Es 

 and running to the wing-edge in a direction divergent from Es. This 

 seems to me an argument of sufficient importance to exclude the 

 Gomphine hypothesis for Scudder's wing. 



&.— The Corduline supposition— allied to Aeschnosoma. 



2, 3, 4, and 5, as above, are all in favor of a Libellulid. 



6. None of the Lihelliiluiii is known to have a triangle like Scud- 

 der's figure: such a triangle is to be found in Neurococordulia, Platy- 

 cordulia, and Aeschnosoma. I have before me a specimen of Aeschno- 

 soma from Surinam, possibly fordpula, where even a trace of the frac- 

 tured distal side of the triangle may be observed. So it is reasonable 

 to look for the allies of our species (Stenogomphus ca-rletoni) rather 



