ME. G. BENTHAM Otf ORCIIIDE/E. 315 



is so closely allied in every respect to Cattleya, that one has great 

 hesitation in accepting the technical distinction of the eight pollen- 

 masses in two series instead of the single series of four, especially 

 as hybrids are so readily produced in cultivation in which the 

 number of pollen-masses is variable. It was thought at first that 

 the character was accompanied by a difference in habit ; but this 

 has not been since confirmed, for several species, first described 

 and figured as Cattleyas, on account of their habit, have been sub- 

 sequently found to have the eight pollen-masses of Lcelia (Bot. 

 Reg. t. 1172 and 1838, t. 2 ; Bot. Mag. t. 365G, 3711, 3910, 4700). 

 SchomlurgJcia, Lindl., about twelve species, and Sophronitis, 

 Lindl., four or five Brazilian species, call for no special remark on 

 the present occasion. 



Tribe 2. Vandejs. 



This vast tropical tribe remains, with very few exceptions, as 

 circumscribed by Lindley. The plants are mostly epiphytical 

 and frequently pseudobulbous. The inflorescence is generally 

 lateral, although in a few genera it terminates leafy stems. The 

 essential character resides in the anther and poll i nary apparatus. 

 The anther, operculate and deciduous as in Epidendrea?, is more 

 closely incumbent on the rostellum, and when mature usually 

 appears one-celled or nearly so ; the two cells, perhaps always 

 distinct in the young bud, become confluent when open, or are 

 only separated by a partially raised liue in the anther-case, and in 

 all probability are always divaricate, not parallel as in Epiden- 

 drea?. The pollen of each cell is a waxy, globular, ovoid or oblong 

 mass, not tapering to a point or caudicie, and either entire or 

 more or less divided into two by a trausverse or oblique furrow or 

 section. Thus the pollen-masses, when four in the anther, are fore 

 and aft in each pair, not all four collateral and parallel, as in most 

 Epidendrese. "When the flower opens, the anther-case is very 

 readily detached, leaving the two pollen-masses or pairs of pollen- 

 masses belonging to the two cells separately and firmly attached 

 to a scale or plate which becomes detached from the back of the 

 rostellum. This scale or plate, with the attached pollen-masses, 

 constitutes the pollinarium. When it is small and thick, and 

 almost wholly viscid, it assumes the appearance of a gland, and is 

 so denominated; but in many cases it takes the shape of a broad 

 scale, or is prolonged into a linear or oblong, single or sometimes 

 double, stipes, bearing the pollen-masses at the further end ; but 



