of Ra.f^esm from Manilla. 383 



completely envelop the perianth; they are marked by pro- 

 minent veins, precisely as in R. Arnoldi ; the tube of the pe- 

 rianth originates on a line with the central row of bractea? be- 

 low the two interior rows, and although in the bud at its up- 

 per part it is undivided, yet the lines of its divisions, when 

 expanded, are clearly discernible. The interior of these di- 

 ^dsions of the perianth is marked by tubercles of various forms, 

 as in the other species. 



The column has a convex disc, surrounded by a raised 

 edge; on the surface of this column are eleven processes, 

 rather more than one-eighth of an inch in height, differing 

 from each other shghtiy in size and form, the summits of 

 which are entire and hispid, the hairs much resembling pis- 

 tillary projections. One of these processes is in the centre, 

 the other ten arranged around it at about an equal distance 

 between it and the raised edge. 



The anthers, which are of the same form, with pores and 

 cells like those of the other species described, are ten in num- 

 ber, and are also suspended from the under side of the upper 

 edge of the column, in open cavities formed in the lower part 

 or base of it ; both edges of the open part of these cavities are 

 covered with hairs resembling those on the tips of the pro- 

 cesses on the disc, and that part of the tube of the perianth 

 opposite to these openings is studded with thick, capillary 

 hairs, each terminated by what is apparently a glandular knob. 



DoAvn the centre of the column are lines, evidently bundles 

 of vascular tissue, which pass through the substance of the 

 cup into the root of the Cissus ; all the rest of the interior is 

 cellular. 



I could not perceive any very distinct appearances in the 

 bud of an annular process at the mouth of the tube of the 

 perianth, although it is not improbable, from various marks, 

 that such a ring may be developed when the flower is open. 



There is no appearance, in any of these three specimens, of 

 the cavities exhibited in the figure of R. Patma which con- 

 tain the spores; on this part of the structure of Rafftesia, 

 therefore, these specimens from Manilla do not throw any 

 further light. They are probably male flowers. Of R. Hors- 

 fieldii, which, when expanded, is only three inches in diameter, 

 I have not seen any description. 



I close this paper with the following comparisons of the two 

 species described, and of that which I call, at present, R. Ma- 

 nillana. 



R. Arnoldi. Bud, before expansion, one foot in diameter, sessile on 

 root of Cissus angustifoUa, the under side of its base reticulate : disc 

 of column convex ; processes on surface forty to sixtj', close together. 



