MB. a. BENTHAM ON OBCHIDE^l. 301 



pressed, as in Bulbophyllum, but almost connate into a globular 

 mass. In Parish's Moulmeyn collection is a specimen very 

 closely resembling the Nepal species ; but the pollen, according 

 to Parish's drawing, though similar in shape, has become attached 

 to a stipes with a gland precisely as in Vandese. Mr. Parish's 

 analytical drawings of Moulmeyn Orchids are so exceedingly 

 accurate and so generally taken from living specimens, that 

 Eeichenbach, in describing the collection for the Linnean Trans- 

 actions (vol. xxx.), accepted his representation of the pollina- 

 rium of Monomeria without hesitation, and had it copied in the 

 plate in which he figured the plant as a genus of Vandea?, very 

 unlike any one yet known in that tribe. Unfortunately neither 

 Parish's nor Lindley's specimens afford the means of verifying 

 the point; but in two flowers ready to open, taken from a spe- 

 cimen I had from "Wallich, I find the pollen exactly as drawn by 

 Lindley without the stipes and gland. I cannot help thinking, 

 therefore, that the pollen figured by Parish had become acci- 

 dentally attached to some extraneous body mistaken for the 

 stipes, a conjecture somewhat confirmed by the very exceptional 

 manner in which the pollen appears attached to the supposed 

 stipes, which, moreover, does not correspond in shape with that of 

 the rostellum, from which it would have been detached. 



Dendrochilum, Blume, reduced to his first section, certainly in 

 some measure connects Liparideae (to which I have above referred 

 Blume's second section) with the Dendrobieoe allied to Bulbo- 

 phyllum ; for it has the small flowers in a slender raceme of the 

 former subtribe, with flowering leafless scapes issuing from the 

 stem-like caudex distinct from the unifoliate pseudobulbs, and 

 the labellum articulate on the basal projection of the column as 

 in Bulbophyllum. The pollen-masses are in some measure inter- 

 mediate between those of the two subtribes, but appear to be 

 rather nearer to those of Bulbophyllum than of Liparis. 



JPanisea, Lindl., reduced to the two original species, P. parvi- 

 Jlora and P. reflexa, with the habit of Bulbophyllum reptans, 

 appears also in its flowers to be much nearer to that genus than 

 to Coelogyne, of which Lindley had at first made it a section. 

 But P. apiculata and P. unijlora, afterwards added by Lindley, 

 have not the peculiar labellum of Panisea, and appear in all 

 respects to be true species of Coelogyne, to which, indeed, Pveichen- 

 bach has already transferred the P. apiculata. I cannot, how- 

 ever, concur with him in referring Coelogyne bilamellata, Lindl., 



