MB. Q. BENTHAM ON OBCHTDEjE. 317 



perianth, with a mentum instead of a spurred labellum, being so 

 very different. Lissochilus, E. Br., including Ilypodematiiim, A. 

 Rich., is a purely African genus, tropical or southern, of about 

 twenty-five species, differing from the genuine EuJophias with 

 leafless flowering- scapes chiefly in the large petals, always broader 

 and more coloured than sepals an approach to which character 

 may be observed in Eulopltia herlacea, Lindl., and a few allied 

 Asiatic species. Galeandra, Lindl., comprises half a dozen Ame- 

 rican species, with the inflorescence terminating the leafy stems, 

 as in the two or three abnormal species of South-African Eulo- 

 pliias, from which they differ chiefly in the broad funnel-shaped 

 spur of the labellum, and in the gland or scale of the pollinarium 

 not at all or scarcely produced into a stipes. The three genera, 

 however, of Eulophiea? might almost equally well be considered 

 sections of one genus. 



Subtribc 2. Cvmbidiej:. — The plants of this subtribe have 

 generally the habit nearly of Eulophiene and of some Cyrtopodieae, 

 but have no spur to the labellum nor mentum to the perianth. 

 They are terrestrial, or more or less epiphytical ; the leaves, 

 usually large and plicate or many-nerved, are often borne on 

 pseudobulbs ; the racemes, simple or rarely branched, are on 

 leafless scapes, or in three genera terminate the leafy steins. The 

 pollen-masses are sessile on the scale-like gland of the pollina- 

 rium, rarely produced into a short single or double stipes. The 

 eleven genera are all from the Old World, one only of the some- 

 what auomalous ones {Poly si achy a) being also represented in 

 America. 



Cymbidium itself, including Iridorchis, Blume, has about thirty 

 species, chiefly Asiatic or Australian; the two African species 

 referred to it by Lindley have been shown by Harvey to belong 

 to Eulopliia-, but the C. Sandersonii, Harv., from South Africa, 

 and an allied species from tropical Africa, appear to be true Cym- 

 bidia. The pollen-masses in this and the two following genera 

 have usually within the anthers a granular appendage, like that 

 of Epidendrum, attached to their base and lying along their outer 

 edge; but in Epidendrum, the anther-cells being parallel, the 

 appendages of the two celk are contiguous and parallel, readily 

 uniting in a single lamina ascending from the base ; whereas in 

 Cifmbidium, the anther-cells being divaricate, the appendages are 

 also divaricate, united by their bases only into a single transverse 

 linear lamina, attached by the centre, and, after dehiscence, placed 



