MB. Q. BENTHAM ON OECHIDE^!. 309 



of the pollen-masses hardens into a disk-like gland contiguous to 

 the rostellum ; but the perfectly 2-celled anther and the eight 

 distinct pollen-masses tapering into points or caudicles are totally 

 unlike what is observed in Vandeae ; and although the viscum 

 forming the gland may be, as in many of the preceding genera, 

 an exudation of the rostellum, it does not appear to me to be a 

 scale or lamina detaching itself from the upper surface of the 

 rostellum, as in true Vandese, although it might, perhaps, be 

 compared to the gland of Ophrydeae. Calanthe therefore appears 

 to me to be referrible, as well iu character as in habit, to Ccelo- 

 gynese and not to Vandese. The species have been distributed by 

 Lindley into two series, according as the spur of the labellum is 

 elongated, or short or quite obsolete, but the distinction is vague 

 and not confirmed by habit. Some species, however, such as G. 

 densiflora, Lindl., and G. curculigoides, Lindl., are well marked by 

 their dense racemes and scarcely spreading perianths; and, again, 

 three species, C. gracilis, Lindl., without any spur, G. brevicornis, 

 Lindl., with a short spur, and C. densiflora, Lindl., with a long 

 one, have the column exceptionally produced above the tube 

 which it forms by its connection with the labellum. 



The facility with which Limatodes rosea, Lindl., can be made to 

 hybridize with Calanthe vestita has been given as an instance of 

 ready hybridization between two distinct genera ; but the fact 

 appears to be that L. rosea itself has all the characters of Calanthe 

 and not of Ccelogyne, and is indeed a species very nearly allied in 

 every respect to C. vestita. 



Arundina, Blume, is a well-characterized Indo-Malayan genus 

 of about five species, and I quite agree with Eeichenbach in 

 uniting with it Lindley's Dilochia. This genus was distinguished 

 by the lateral lobes of the clinandrium bearing perfect anthers ; 

 but Blume has shown that this anomaly is not constant, and it 

 has been occasionally observed as an accidental anomaly in a few 

 flowers of two or three very different genera of Orchidese. 



Elleanthus, Presl {Evelyna, Poepp.), is a natural and well- 

 characterized tropical-American genus of nearly 50 species, 

 mostly with short dense terminal heads or spikes with closely 

 imbricate bracts, but one or two species, with a longer loose inflo- 

 rescence, assume almost the habit of Neuvoiedia. 



Subtribe 8. Stenoglosse2e. — The genera which I have collected 

 in this subtribe are mostly small-flowered epiphytes, which have 

 been generally classed either as sections of Epidendrum or as 



