FLORA NIGRITIANA. 379 
or many-celled ovary, nor the drupaceous or baceate fruit 
appear to be sufficiently absolute distinctions to separate them 
as tribes. The whole might take the name of Coffeew, and be 
divided into four or five subtribes, chiefly according to the 
estivation of the corolla* and insertion of the ovules, viz.: 
Vangueriee, (including Morindee and Canthium), with a valvate 
estivation and pendulous ovules ; Guettardee, with an imbricate 
estivation and pendulous ovules ; Jzoree, with an imbricate sesti- 
vation and laterally attached ovules; Psychotriee, (including 
Cephaélidee), with a valvate æstivation and erect ovules. Possibly 
a fifth might be inserted between Guettardee and Ixorea, with 
an imbricate zstivation, like in those two tribes, but differing 
from Guettardee in the baccate, not drupaceous fruit, and from 
Izoree in the ovules suspended from the apex, or nearly so, but 
I am not acquainted with the fruits of a sufficient number of 
Species of Chomelia, Chiococca, Kraussia, &c., to ascertain 
whether they can be really so separated from Guettardee even as 
a subtribe. I am doubtful also whether the few genera with 
an imbricate zestivation, and ovules erect or ascending, should 
be reckoned among /zoree, or form an intermediate subtribe 
between them and Psychotriee. They are chiefly South Ameri- 
can, and require further examination. 
l. Sarcocephalus esculentus, Sab—DC. Prod. 4. p. 868.— 
Sierra Leone, Vogel, Don ; Senegal and Guinea. 
The fruit of this plant appears to me to be improperly 
deseribed as an agglomeration of berries. The syncarpium, 
especially the upper portion, consisting chiefly of the epigynous 
discs and summits of the calyxes, is indeed fleshy ; but the 
divisions, formed of the combined parietes of adjoining fruits, 
are scarcely so ; the two cells of each fruit have no pulp, and 
the dissepiment which separates them is membranous. The 
Placente are adnate: the seeds (as yet unripe in the fruit I 
* The imbricately aestivated corolle of Rubiacee are usually convolute 
as in Apocyne@ ; but since it is frequently difficult, if not impossible to as- 
Certain whether it is so constantly, I have preferred using the more eye 
term imbricated (of which the convolute is a modification) in opposition to 
valvate, S 
* 
