MB. G. BENTHAH OK OBCHIDE^E. 305 



in two series ; but the pollinary appendage is very abundant and 

 the masses of both series are adnate to it upwards or nearly 

 their whole length, the smaller series is deficient only in Antho- 

 gonium. The genera mostly belong to the Indo-Australian region ; 

 but Cliysis and Bletia itself are tropical American with one 

 Chinese and Japanese species of the latter, and one species of 

 Phaius has been found in tropical Africa. 



The three genera Acanthephippium, Phaius, and Bletia have 

 been admirably worked out and illustrated in Blame's great 

 work on Javan Orchids, and my own observations lead in 

 every respect to confirm his conclusions, except in regard to 

 the North-American Bletia aphylla, Nutt., which I have above 

 mentioned as an anomalous Liparidea allied to Corallorhiza under 

 Bafinesque's name Hexalectris. The Chinese and Japanese 

 Bletia hyacinthina and the Indian and Malayan Phaius albus 

 and its allies, differ from the rest of the subtribe in their 

 racemes terminating leafy stems ; and Eeichenbach has proposed 

 to separate them generically, the former under the name of Ble- 

 tilla, the latter under that of Thunia, and to transfer them to 

 Lindley's tribe of Arethusese on the supposition that the pollen 

 was entirely granular. But Blume has well shown that this 

 is a mistake. In the early stage the pollen-masses are quite 

 continuous with, and scarcely distinguished from, the mass of 

 pollinary matter forming the appendage, and in withered flowers 

 a considerable quantity of granular pollen will sometimes re- 

 main in or about the anther-case ; but about the time of dehis- 

 cence I have seen the eight pollen-masses quite distinct, and 

 as waxy as in other Bletiese, in dried specimens of Bletia hya- 

 cinthina and Phaius albus and in living specimens of Phaius 

 Benscmiee (Bot. Mag. t. 5694). I therefore follow Blume in 

 reuniting them with Bletia and Phaius respectively, dividing 

 Bletia into two sections, Eubletia and Bletilla, and Phaius into 

 four, Puphaius, Gastrorchis (including Pesomeria, Lindl), Thunia, 

 and Limatodes. I should exclude, however, from the latter the 

 L. gracilis and L. rosea, Lindl., which are both referrible rather 

 to Calanthe. 



Chysis, Lindl., is a tropical-American genus of half a dozen 

 species, in which the granular mass or pollinary appendage is re- 

 markably large, sometimes almost enclosing the pollen-masses. 

 Kephelaphyllum, Blume (Cytheris, Lindl.), contains four species 

 from the Indo-Malayan region, and is well illustrated by Blume. 



