142 BOTANY OF CONGO.
probably of ]\ladagascar, and both are remarkable in having the caiulex dichotomous, or repeatedly divided.
As the Palm on the banks of the Congo was seen in fruit only, it is not difficult to account for Professor Smith's referring it rather to Hypha^ne than to Corypha ; Gsertner having described the embryo of the latter as at the base of the fruit, probably, however, from having inverted it, as he appears to have done in Elseis. It is at least certain that in Corijijlia Taller (^ of the continent of India, which is very nearly allied to C. umbracuhfera, the embryo is situ- ated at the apex, as in Hyphsene.
The journal also notices a species of Raphia, which is probably BapMa vinifera of M. de Beauvois/^ the Sa(/U8 Falma-]miiis of Gsertner.
The collection contains fronds similar to those of Calamus secmidiflorus of M. de Beauvois,^ which was also found at Sierra Leone by Professor Afzelius ; and a male spadix very nearly resembling that of JElate sylvestris of India.
The Cocoa Nut was not observed in any part of the course of the river.
Only five species of Palms appear therefore to have been seen on the banks of the Congo. On the whole continent of Africa thirteen species, including those from Congo, have been found ; which belong to genera either confined to this continent and its islands, or existing also in India, but none of which have yet been observed in America, unless perhaps Elseis, if Alfonsia oleifera of Humboldt should prove to be a distinct species of that genus.
CYPERACE^;. In the collection there are thirty-two species belonging to this order, which forms therefore about one eighteenth of the Phsenogamous plants. This is very different from what has been considered its equinoctial proportion, but is intermediate to that of the northern part of New Holland, where, from my own materials, it seems to be as 1 : 14 ; and of India, in which according to Dr. Roxburgh's Flora it is about 1:25.
^ Boxb. Coromand. o, tahh. 255 et 25G.
= Tlore cVOicare 1, ;;. 75, tahh. 44, 45, et 46.
'^ Op citat. ],i?. 15, tabb. 9, et 10,
�� �
