“ five or six feet high, covered with a brown bark, and 
‘‘-diyides upwards into many small branches, which are 
** oarnished with spearshaped leaves, about four inches long 
“ and a little more than one broad; these are smooth, of a 
“lieht green, and have many horizontal veins, running 
“ from the midrib to the sides, and are placed opposite by 
“pairs. From the wings of the leaves, towards the upper 
“ part of the branches, are produced the flowers, standing 
“upon branching footstalks, each sustaining four or five 
“ flowers, whose tubes are swelling at the base, just above 
«the cup, but contract upwards to the mouth, where the 
“ petal is cut into five broad segments which spread. flat; 
“ they are white.” 
This species differs in regard to one of the features enu- 
merated in the present generic character, by having stamens 
which stand above the mouth of the tube of the corolla. 
It may not be useless to subjoin a version of the character 
by which Mr. Brown has circumscribed the natural order 
he has detached, under the title of ArocinEm, from that of 
his AscLEPIADE®; which orders previously formed the Apo- 
cINE of Jussieu. ‘ 
Calyx five-cleft, persistent. Corolla monopetalous, in- 
ferior, regular, 3-lobed, (before expansion) imbricate, deci- 
duous. Stamens placed upon the corolla, alternate with 
the segments of the limb. Filaments distinct. Anthers bi- 
locular, bursting longitudinally. Pollen granular, applied 
immediately to the stigma (not by the intervention of a par- 
ticular process, as in the AscreprapE®). Germens 2, or 1 
and bilocular, in most manyseeded. Styles 2 or 1. Stigma 
1. Fruit follicular, capsular, drupaceous, or berried, double 
or single. Seeds generally furnished with an albumen. Em- 
bryo foliaceous. Plumule inconspicuous. Trees or shrubs, 
often milky. Leaves opposite, sometimes verticillate 
(whorled), seldom scattered, quite entire, generally fur- 
nished with either cidie or glands between the petioles. Jn- 
florescence subcorymbose. | 
We learn from the same accurate observer of natural 
affinities, that this subdivision forms a truly natural group, 
with perhaps the exception of one or two genera; but 
which if even these were excluded is hardly to be technically 
defined: though easily distinguished by the economy of the 
anthers and the stigma from the AscLepiape#, which are 
besides more uniform in the structure of the flower and 
fruit. The order is also allied to the Gsnriansa and Ru- 
BIACEX. led ‘ 
