130 TERTIARY INSECTS OF NORTH AMERICA. 



probably long and slender, furnished with spine-like hairs as long as the 

 breadth of the femora. The abdomen was moderately slender, rather longer 

 than the wings ; its ninth and tenth segments a little enlarged in the 2, the 

 tenth half ($), or three-quarters (,?), as long as the ninth, and the eighth half 

 as long again (?), or twice as long (<?), as the ninth, and a little more than 

 half as long as the seventh. The anal appendages were as long as the tenth 

 segment, rounded triangular (?) or quadrate (r?). 



The species of Dysagrion found at Green River may be separated by 

 the characters drawn from the neuration of the wing in the following table : 



TabJe of the species of Dysayrion. 



Pterostigma at least four times as long as broad; quadrilateral longer than broad ; middle of the area 

 occupied by supplementary veins between the principal and subuodal sectors rilled with quad- 

 rilateral cells. 

 Pterostigma surmounting four cells ; quadrilateral nearly as broad at base as at apex. 



D. fretlericii. 

 Pterostigma surmounting three cells ; quadrilateral nearly twice as broad at apex as at base. 



2. D. lakesii. 



Ptcrostigma only three times as long as broad ; quadrilateral slightly broader than long ; middle of the 

 area between the principal and subnodal sectors tilled with pentagonal cells. . .3. D. packardii. 



1. DYSAGRION FREDERICK. 



PI. 6, Figs. 2, 5, G, 9, 10, 14, 17. 



Dysagrion fredericii Scudd., Bull. U. S. Geol. Geogr. Surv. Terr., IV, 534-537, 775 (1878). 



Several specimens of various parts of the body with wings were found 

 by Mr. F. C. Bowditch and myself in the Green River shales, in a railway 

 cutting by the river bank beyond Green River station. The most important 

 are a neai-ly perfect wing and its reverse, which preserve all the important 

 points of the neuration. A single antecubital appears to be present, nearer 

 the nodus than the arculus; the principal sector, like the short sector (sec- 

 tor brevis). bends slightly upward just as it reaches the arculus; the cel- 

 lules in the discoidal area are half as broad again as long, yet the breadth 

 of the wing is such that the broadest part of the postcostal space, between 

 the nodus and the middle of the wing, is more than half as broad as the rest 

 of the wing at that point. The quadrilateral is subquadrate, about half as 

 long again as broad, its upper and lower margins subparallel and its lower 

 outer angle about sixty degrees; pterostigma four times as long as broad, a 

 little dilated, oblique both within and without, but especially pointed above 

 on the outer side, touching the costal margin throughout. The wing is 

 wholly hyaline, excepting the infumated pterostigma, which is bordered by 



