152 Journal New York Entomological Society. [Vol. xxiii. 



more difficult of appreciation and of definition. Although, as Casey- 

 has stated, the species are abundantly distinct, lurida is given as a 

 variety of obscnra. In looking over recently two collections in which 

 the two species were hopelessly confused, I have struck what I 

 believe to be the clue to the matter. The usual analytical tables 

 separate the species by the character " thorax widened from base, 

 obscnra," etc. While this character is true of obscnra, in a large 

 series of lurida many will be found to which it also applies and it is 

 just these luridas with intergrading thorax which are wrongfully 

 identified as obscura and which cause true lurida to be regarded as a 

 variety. 



I have found, however, other characters by which the species may 

 be separated and venture to give these in the following table in 

 order of their importance. 



A (a) Hind coxal plate notched at insertion of femur and with distinct tooth 

 on outer side of notch. 



(6) Lateral margin of thorax viewed from beneath and sidewide present- 

 ing a polished carina for at least one-half of length from base. 



(c) Food plant persimmon obscura. 



AA (a) Hind coxal plate not or very feebly notched and with no tooth. 



(&) Lateral margin or thorax viewed from beneath and sidewise present- 

 ing merely a trace of polished carina, not over one-fifth of length 

 from base. 



(c) Food plant hickory lurida. 



In separating the closely allied species of Dicerca, I believe more 

 attention should be paid to the food plants. In Virginia, the species 

 of the divaricata group which feeds on sugar maple is markedly dif- 

 ferent from the species which feeds on peach. 



I have found difficulty in separating by analytical tables Chalco- 

 phora libcrta and its allies from Cli. fortis. In the small series in 

 my collection, I note that the prosternal spine in libcrta and its allies 

 has two longitudinal parallel sulci, while in fortis there is but one 

 sulcus. If this holds for a larger series, it would serve to separate 

 the two without having to depend upon the ratio of length to breadth 

 of elytra, a character not easy to determine with certainty. — Wirt 

 Robinson. 



A Small Collection of Odonata from Atlin, British Columbia. — 

 Mr. E. M. Anderson, of the Provincial Museum, Victoria, B. C, has 



