240 NATURAL HISTORY OF PLANTS. 
species, the flowers of which have a short convex receptacle, sup- 
porting five sepals, imbricated in the bud, and five superposed petals, 
scarcely united at their bases, and also arranged in imbricated præ- 
floration. The androceum is formed of an indefinite number of 
hypogynous stamens, the filaments also united for a very short 
Terastremia brevipes, 

Fie. 261. 
Flower, 
Fia. 263. 
Diagram. Long. sect. of flower. 
distance with the base of the corolla and surmounted by basifixed 
anthers, with two adnate cells, almost marginal, dehiscing by two 
longitudinal’ clefts and crowned by a pointed prolongation of the 
connective. The gynæceum is superior, formed of an ovary with two 
cells, tapering above into a conical style, with apex almost undivided, 
and stigmatiferous. In the internal angle of each cell descending 
from the upper part a wide short placenta is inserted, the inferior 
edge supporting from five to eight suspended anatropous’ ovules, 
with micropyle looking upwards and inwards. The fruit is dry, 
accompanied at its base by the persistent calyx, apiculate and inde- 
hiscent ; it encloses a small number of seeds elongated and folded 
upon themselves, like a horseshoe. Under their coats is found a 
fleshy embryo, also hippocrepiform, with cylindrical radicle longer 
than the cotyledons, and surrounded by a fleshy albumen, often thin, 
or even reduced to a simple membrane. 

Guian., 569, t. 227, 225.—Tonabea J., Gen., 252. ? They have two coats. When their anatropy 
— Dupinia Necx., Elem., n. 1042,—Amphania 
Banks, mss. (ex Enpu.).—Cleyera THuNB., F1, 
Jap., 12 (nec DC.).— Reinwardtia Korrn., Verh, 
Nat, Gesch, Bot., 101, t. 12 (nee Bt.).—WAtpP., 
Rep., v. 129 (inel.: Erythrochiton GRIFF., Veele- 
keria Ku. & KARsT.). 
1 The pollen of Ternstræmia is formed of 
ovoid grains analogous to those of the Teas. (H. 
Mout., in Ann. Sc. Nat., ser. 2, iii, 333). It is 
the same in the other genera (Visnea, Caraipa) 
where it has been studied. 
is completed they begin to bend in the inferior 
portion of their raphe, so that the region of the 
chalaza is reflexed upon it outwards and upwards. 
We have noted (in Adansonia, x. 238) that this 
is the first degree of the false campylotropy that 
we have seen so noticeable in the Gyrostemonee 
(page 42, note 6), and which appears alsu in scme 
other groups. 
