166 ^Ir. E. 1. Pocock on a Neio and JS^alural Groiqying 



sections of his subfamily Aviculariinse, namely the Phlogieaj, 

 Selenocosmiege, and the Poecilotherieag, which may be tabu- 

 lated in the following order : — 



1. rHi,OGiK,"E. — *Orp/i»(rcu.f, Luzon; Chilohrachyx^ Ceylon; PMoijius, 



ludo- and Austro-Malaya; Coremiocneinis, I'iuaug. 



2. Selknocosmik-^. — *Lo.romp/ialia, Phonei/usa (-^vn. JInrpaxotherid), 



JlyMerocraU's, Harpactira, *PeUnobius, Ethiopian Refrion ; *En- 

 cyocrates, Madagascar ; *Ephehopu^, loc. ? ; Selenocosmia, Indo- 

 Malaya; *Lampropelma, Ins. Sangir; * Haplopclma, Borneo; 

 *Ciiri<)paciopus, Tenasserini ; Haploclastus, S. India : and, possibly, 

 Ontothymu!!, Pinang. 



3. PcECiT.OTHEBiE-E. — Pcccilot/terkt, S. India, Ceylon ; Scodra, W, Africa. 



It is to be supposed that the above were regarded as natural 

 groups ; and since the Poecilotherieffl were treated quite apart 

 from the rest of the Old-World genera and in connexion with 

 the Neotropical group of Aviculariege, we may conclude that 

 these two grou))s were looked upon as related to each other, 

 and that the Indian genus Posciiotheria was not regarded as 

 having any near relationship with other Oriental forms. 

 j\loreover, the above-given arrangement of the genera indi- 

 cates, I take it, that the Oriental genera of Selenocosmieae 

 are more nearly allied to the African genera of the same 

 group than they are to the Oriental Phiogieai. But, after 

 working over some of the material of the Aviculariidie con- 

 tained in the British Museum, the conclusion has been forced 

 upon me, tirstly, that Simon's sections are largely artificial, 

 and, secondly, that the genera referred to them may be other- 

 wise grouped, so as to form assemblages which may be 

 regarded as natural, inasmuch as they agree, (1) in the 

 possession of constant characters, and (2) in their geographical 

 distribution. 



In the first place it may be stated that none of the Oriental 

 genera appear to me to be especially related to the Ethiopian, 

 or, to put it more accurately, the genera composing either of 

 the two groups into which the Oriental genera fall are more 

 nearly related inter se than any one of them is to any Ethio- 

 pian genus known to me. This conclusion serves at once to 

 separate off the Ethiopian genera, which I do not propose to 

 deal further with here, except to state that I provisionally 

 divide them into the following three sections : — 



* Genera that ai'e unliLowu to me in nature are murked with an 

 asterisk. 



