CENTRADENIA ROSEA. 
(Rose-flowered Centradenia.) 
Class. 
OCTANDRIA. 
Order. 
MONOGYNIA. 
Natural Order. 
MELASTOMACEiE. 
Generic Character.— Calpx tetragonally-cainpanu- 
late, free ; limb four-parted, lobes broadly triangular, 
acute. Corolla with four petals, inserted in the throat 
of the calyx, and alternate with its lobes, obovate. 
Stamens eight, inserted with the petals, alternate and 
opposite to them, smaller. Anthers elliptical, one-pored. 
Ova7'i/ free, four-celled ; cells many-ovuled. Stple 
short ; stigma subcapitate. Capsule covered by the 
calyx, four-celled ; cells four-valved. Seeds numerous, 
elliptically club-shaped, echinate. 
Spbcific Character. — Plant an evergi'een shrub. 
Leaves ovate-lanceolate, unequal-sided. Racemes sub- 
corj'mbose, terminal, much longer than the leaves. 
Floivers pale pink, twice the length of the ovary. 
Synonyme. — Doncklaeria diver sifolia. 
In the nurseries round London, this attractive little plant is generally known 
by the name of Doncklaeria diversifolia, which is probably a Continental title ; 
though we have no hesitation in identifying it with the Centradenia rosea of Dr, 
Lindley in the Botanical Register. In that work it is described as being intro- 
duced from Mexico by Messrs. Lucombe, Pince, and Co., of Exeter, within the 
last two years. It has, however, been in the Epsom Nursery for four or five years ; 
and Messrs. Young having flowered it there, we had the accompanying figure 
prepared from their plant in March last. 
Speaking of its relations and character, Dr. Lindley mentions that " Mr. 
Bentham has pointed out its near affinity to the Rhexia incequilateris of Schlech- 
tendahl, afterwards called Plagiophyllum by him, and Centradenia by Don. Upon 
turning, however, to an authentic specimen of that plant, we find it different in some 
respects : its leaves are much larger and thinner, its flowers appear to be smaller, 
and are arranged in little terminal racemes, much shorter than the leaves. Neither 
can this be the Plagiophyllum grandifolium of Schlechtendah], which is described 
with leaves as much as six inches long. 
One of the most curious circumstances connected with this plant, is the 
constant abortion, and frequent loss, of one of its two opposite leaves. This ten- 
dency to abortion always alternates along the branches, so that if the first imperfect 
