October, igij Bulletin of the Brooklyn Entomological Society 93 



Clayton, Rabun Co., is near the North and South Carolina lines, 

 and is surrounded by mountains rising to 3,500 feet and is itself at 

 an altitude of 1,800 feet. 



Stone Mt. is in Decatur Co., 16 miles from Atlanta. 



Offerman is in Pierce Co., with conditions very similar to Black- 

 shear. I collected along the Satilla River, about 4 miles from 

 the station. 



Williamsonia, a New Genus of Dragonflies from North America. 



By Wm. T. Davis, New Brighton, Staten Island, N. Y. 



While examining the collection of dragonflies in the American 

 Museum of Natural History I came upon a female example of a 

 species unknown to me. It bore the label "Paterson, N. J., May 

 4," and had been collected and presented to the Museum by 

 Mr. John A. Grossbeck. 



Reference to the "Key of North American Genera of Cordulinas" 

 on page 484 of Dr. Needham's "Aquatic Insects in the Adiron- 

 dacks" placed the specimen in division "a" having "Veins M4 and 

 Cul in the fore wing parallel or a little divergent apically, the 

 number of rows of cells between them increasing toward the 

 margin of the wing." The only genus placed in this section is 

 Neurocordulia to which belongs A'', obsoleta Say and N. yamaskaren- 

 sis Provancher. In these insects, however, the triangles and 

 subtriangles of the fore wings are often divided into three cells by 

 cross veins, the triangles of the hind wings also have each a cross 

 \^ein, while in the New Jerse}^ specimen all the triangles of both 

 wings are open, that is without any cross veins. Neurocordulia 

 has many cells in the space beyond the anal loop, while the speci- 

 men in question has but a few and these arranged in three rows. 



In a foot note on page 484 already referred to, Dr. Needham says 

 in reference to section "a" of his table, "One species, the little 

 Cordulia lintneri of Hagen, may seem to belong in this section of 

 the table, though of course not in the genus Neurocordulia; it is 

 al^o a synthetic type, lacking the special corduline features of 

 venation, which I take to be (1) the approximation of veins M4 

 and Cul, and (2) the general reduction of cross veins; it shows 

 strong libelluline affinities in the conformation of the anal loop and 



