16 SCUDDER ON THE DEVONIAN 



reference there would seem to obscure its real isolation. In fact there seems to be not 

 only no family of insects into which it can be placed, but even no sub-order living or 

 extinct, into which it would naturally foil. There is no known insect in which five par- 

 allel and distant nervures follow the course of the costal margin, and of which only two 

 arise from the same root ; and so far as my observations have gone, I have found no neu- 

 ropterousinsect (to which of living groups this is plainly the most neai'ly allied), in which 

 the externomedian vein is the first extensively branched vein, and in which at the same 

 time, the upper branch of this vein is simple. In Ephemeridae (to which group one 

 would most naturally compare it from its general appearance), the externomedian vein, as 

 already stated, is always compomid, and its upper stem is always forked. In this insect 

 on the contraiy, the Uf)per stem is simple (which is the more remarkable from the forked 

 character of the scapular, alwaj's simple in Ephemeridae) and the lower forked, its 

 branches being superior and herein differing remarkably from ordinary types. 



Gerephemera then is not only further removed from modern Ephemeridae than is Plat- 

 ephemera, bvit can be even less closely affiliated with Platephemera than the latter with 

 modern Ephemeridae. It has, nevertheless, some distinctive points in common Avith it. 

 Such are its great size and the probable great expanse of the internomedian area, the dif- 

 fering character of the net-work above and below the uppermost externomedian branch, 

 the polygonal nature of the mesh-work caused by the cross- venation (in common with 

 many other old insects), and the somewhat xuiiform character of that network next to 

 and away from the boi'der. In common with modern Ephemeridae, but in distinction 

 from most other insects, must be mentioned the common feature of intercalary nervures, 

 which here, as in Platephemera, are never free at their origin. 



As points of special distinction from Platephemera may be mentioned the broad area 

 given to the veins above the externomedian vein, the forking of the scapular vein, its 

 course at the bottom of a deep and broad sulcus, the occurrence of a straight intercalary 

 in the scapular-externomedian interspace, the entire structure of the externomedian vein 

 (differing altogether from Platephemera) and the elongated slender form of the wing, 

 which resembles much more closely Dictyoneura and Haplopldebium. 



From these latter genera again, to which we should perhaps consider it most closely 

 allied, this insect differs remarkably in the structure not only of the externomedian vein, 

 but in the wide extent of the wings above that vein, and the number of nervures which 

 fill it. It would appear also to differ in the character of the reticulation above the exter- 

 nomedian vein, a matter of less importance, but in which it agrees with Platephemera. 

 The difierence in tlie frame work of the wing, however, is so great and so deep seated, 

 that there can be no doubt of at least its family distinction from all known types. 

 Whether or no it is worthy of being classed as subordinally distinct, I leave to future 

 discussion. But in allusion- to the apparent fact that the peculiar nature of its neuration 

 has not le'ft its mark on modern types, I propose to call the family group in which it 

 should be placed Atocina.^ It will be sufficiently distinguished from other ancient types 

 (as from modern) by the forking of the scapular vein, the course of the externo- 

 median, its distant removal from the costal margin, and its peculiar division. 



This insect and Xenoneura come from the lowest of the Lancaster Shales which furnish 

 insect remains, called 2)lant bed No. 2, by Professor Hartt. 



, ' From the Greek azoxnq. 



