220 VEGETATION OF THE FEEJEE ISLANDS, 
ficiently distinct from both to authorise the considering that 
species as forming a genus, for which De Candolle’s sectional 
name Rhodomyrtus may be adopted. I find the ovary one- 
celled, with three thick fleshy double placentae, projecting 
to the centre, but without cohering. The margins of each 
placenta, as they reach the centre of the ovary, turn in- 
wards and bear each a single row of densely superposed 
ovules, whilst from the centre (or as it were the midrib) of 
each placenta a spurious dissepiment projects slightly into 
the cavity so as to separate the two rows of ovules, without 
however reaching half-way to the centre. The ovary itself 
in its young state is easily separable from the calyx which 
encloses it. The Rhodomyrtus tomentosa is the only species 
I am acquainted with, and occurs frequently in collections 
from tropical Asia. It is Cuming’s 1253 from the Philip- 
pine Islands and 2264 from Malacca. 
Nelitris Urvillei, DC. Tobie Island, Mr. Barclay. 
Ofthis genus, besides Barclay's specimens, I have ee 
mined Cuming's n. 801, 821 and 1824 from the Philippine 
. Islands and 2271 from Malacca, and East Indian specimens - 
from Roxburgh, Wallich and others. The ovary always ap- 
pears 8 or 10-celled, although strictly speaking it be tetra- 
merous or pentamerous. The placentation is in fact analo- 
gous to that of Rhodomyrtus, but the placente, instead of 2 
meeting only in the centre, cohere together; and the imper- e 
fect dissepiments of Rhodomyrtus, also reach the centre  — — 
Nelitris, and are there united with the common axis, so as t0 — — 
divide each cell into two. Of the above mentioned specimens - 
Roxburgh’s, Wallich's and Cuming's n. 2271, have the 
flowers and ovary tetramerous, and belong to N. paniculata, 
Lindl. Barclay’s and Cuming’s 821 are pentamerous and j 
agree with De Candolle's character of Nv Urvillei ; Cuming's 
n. 801 is allied to N. Jambosella, but has certainly, like all 
the others, several ovules in each cell ; and Cuming’s 1824 is 
probably a new species allied to N. Urvillei. a 
Eugenia vismioides DC? Prod. 3, 267. New Ireland, | 
Mr. Hinds, Mr. Barclay. 2 
