FLORA NIGRITIANA. 363 
lato-venosa, axillis. venarum subtus fasciculum pilorum fo- 
ventibus, ezterum glabra v. juniora minute puberula. Stipule 
lineares, foliaceze, 2 lin. long, valde caduce. Spice in 
axillis superioribus cum pedunculo 3-5-pollicares, jam infra 
medium interrupte floriferz, floribus arcte sessilibus, infe- 
rioribus dissitis, superioribus approximatis, singulis subglo- 
bosis fere 5 lin. diametro. Bractee 3, fusco-membranaceze, 
una exteriore (bractea subtendens), duæ (bracteolz) interiores 
opposite. Sepala 3 lin. longa, lato-ovata, obtusissima, basi 
breviter connata, margine ciliolata, membranacea, venulosa, 
dorso puberula. Petala consistentia et pube sepalis similia 
sed duplo majora, basi angustata et distincta, circa fructum 
persistentia et globoso-conniventia. Glandule breves, late, 
cum petalis discum hypogynum et subperigynum margi- 
nantes. Stamina intra petala fasciculatim inserta, swpissime 
ad quoddam petalum sex, rarius 5 ; filamenta filiformia, pilis 
longis patentibus barbata, petalis zquilonga v. vix longiora. 
Ovarium dense hirsutum. Styli exserti, ultra medium pilis 
longis barbati, apicem versus fere glabri. 
The almost entirely free ovary, and the number of petals, 
as well as that of the glands and of the bundles of stamens, 
being double that of the calycine segments, sufficiently dis- 
Unguish this genus from Blackwellia and Homalium, inde- 
pendently of the remarkable size and form of the flowers. 
R. Brown mentions a Madagascar genus with a free ovary, but 
without any further indications to show how far it differs from 
ours, 
Two other Homalinee have been described from W. Tropical 
Africa; Homalium angustifolium, Sm., from Sierra Leone, (of 
which the true genus is doubtful), and Byrsanthus Brownet, 
Guillem., from Senegambia. The generic name of the latter 
has been changed by Endlicher to Anetia, on account of the 
Pre-existence of Presls Byrsanthus among Lobeliacee ; but if, 
as is suspected, with every probability of correctness, by A. De 
Candolle, the latter should prove to be really not different 
from Siphocampylos, the name of Byrsanthus should be re- 
fained for Guillemin's plant. 
