168 Mr. R. I. Pocock on a New and Natural Grouping 



Orplmoecus. And if one may be permitted to guess, I would 

 suggest that the first-named will perhaps fall into the Ornitho- 

 ctonidaa and the L^st three into the Selenocosmiidee. 



No organ resembling either of those mentioned above is 

 found in any genus that I have examined of the following 

 Neotropical groups : — Aviculariea?, Eurypelmate^, Tliera- 

 phosea3, and Homoeommateas ; nor yet in the African genera 

 above enumerated. The South-African Harpactira, how- 

 ever, has a peculiar arrangement of hairs on the outer surface 

 of the mandible, there being a dense scopula above and a 

 naked space below it ; but upon this naked area there is one 

 or more curved rows of hairs which have evidently been 

 derived from the fringe which borders the lower edge of the 

 mandible. 



Family Selenocosmiidse, nov. 



With the exception of Baploclastus, which will, perhaps, 

 have to be separated from it, this new family is tolerably 

 compact. It is characterized by the possession of a stridu- 

 lating-organ, composed of a series of thickened rods, upon 

 the inner surface of the coxa of the palp, and of a corre- 

 sponding series of spiniform hairs or of spicules upon the 

 lower half of the external surface of the mandible. The 

 fovea on the carapace is generally small, linear, transverse or 

 crescentic, with the concavity forwards (it is larger in i^sa/mo- 

 pwus). The legs are either without spines or with merely a 

 few short ones at the apex of the protarsi or tibia3. In the 

 male, so far as is known, there is no spur on the tibia of the 

 anterior legs. This, at least, is the case in the male of 

 Foecilot/ieria, Chilohrachijs, Musagetes, Phlogius, and Seleno- 

 cosmia. 



Yet, although thus intimately allied, the genera with which 

 I am acquainted fall readily into the following sections : — 



A. Poecilotherin. 



B. Chilobrachys, Musagetes. 



C. Selenocusmia, Phlogius. 



D. Coremiocnemis, Lyrognatlius, Selenotgpus. 



E. Psalmopceus. 



The mutual affinities of these sections are at present puzzling, 

 and it is impossible to represent them accurately in a dichoto- 

 mous synoptical table. It seems to me that A is related to 

 B, B to C, C to D, C to E, and E to A. I venture therefore 

 to attempt to express this roundabout relationship by the 

 oUowing diagram : — 



