PuatE 1986. 
CATOSTEMMA FRAGRANS, Benth. 
Matvacez. Subtribe Matisiusz. 
. fragrans, Benth. in Hook. London Jowrn. Bot. ii. (1843) 365, 
1793. 
Dinés Plantarum, Pl. 
Has. British Guiana, banks of ph Schomburgk (No. 280) ; 
Lowe Semeur river, Jenman (No. 4 336). Received in flower and 
finit f from St. Vincent’s, where it still survives in the old Botanic 
Garden, Powell 1891. 
ructus monospermus ellipsoideus, 3-4 poll. a 0k pericarpiam 
crasse coriaceum 3-valve; valvis concavis extus pa omentellis. 
Semen oblongum nubeySekeais v. plus minus epi eum, leve, 
Tu ga albuminosum, 25 poll. longum, i 4-14 poll. diam. ; testa cellu- 
losa cystis mucilaginosis copiosis preedita ; albumine tenui ; ; cotyle- 
ieee crasse plus minus conferruminate, cystigere 
I have little to oi to the ie le description of this remarkable 
tree as given in the works above named. The specimens sent by Mr. 
use in some of the indigenous specimens. Catostemma was 
originally referred by Mr. Bentham to Ternstrcemiaces. In ‘ Genera 
lantarum,’ i. 180, it was rejected from this Order and found pro- 
_Visional Hide, with a few other anomalous genera, at the end of 
to determine its affinity, was led to look into Malwales and 
thao 4 fond: the genus Scleronema, first published by Mr. Bentham in 
a . Journal of the Linnean Society,’ vi. 109, based upon specimens of 
which he rightly regards as se Svea with Daieiewtaae which 
a st stan r. Spruce 
Spruceana, Benth. 1.c.) differs at sight in the elliptical or obovate _ 
eaves with a distinct apiculus and prominent transverse venation. 
senting such marked perigyny of the petals and stamens, I think its 
_ length circumsciss near the base, and the limb splits irregularly into © 
2 to 5 ovate-deltoid, or broader, segments ; which of course are in 
no way imbricate, as Mr. Bentham eae ‘the calyx might be, 
the absence of an "unopened bud.—D. Ot1v 
Fig 2. Petal. 3. Stamen, back and front. 4. hoa section of ovary 
and ix showing perigyny of corolla and stamens. Enlarg 
