CaNTHIUM. RUBIACEJA. 425 
TRIBE VI. COFFEACEE. DC. 
Fruit 2-celled, baccate, containing two 1- (or rarely 2-) seeded bony 
or crustaceous nnclei that are flattish or grooved on the inner side, of- 
ten marked with a furrow on the outer, rarely with only one nucleus 
(and probably so from abortion), containing an erect or depressed or 
laterally attached seed. Albumen horny, sometimes somewhat carti- 
laginous.— Trees or shrubs. Leaves opposite. Stipules interpetiolar, 
two on both sides, and either distinct or combined. 
XXII. CANTHIUM. Lam.; Gertn. fr. 3. t. 196. 
Calyx-tube ovate ; limb short, 4-5-toothed. Corolla with a short tube, 
bearded in the throat ; lobes 4-5, spreading. Anthers 4—5, inserted into the 
throat, scarcely exserted. Style filiform exserted. Stigma thick, ovate-glo- 
bose or mitriform, undivided or bifid at the apex. Drupe globose or com- 
pressed, crowned with the (sometimes inconspicuous) calycine teeth, fleshy, 
2-celled. Seeds solitary in each cell, inserted near the apex, inverted, in- 
curved. Albumen fleshy. Embryo central: radicle long, superior.—Shrubs, 
with branches unarmed or thorny. Leaves opposite, somewhat coriaceous. 
_ Stipules interpetiolar, solitary on both sides. Peduncles axillary, short, se- 
veral-flowered. 
One of the cells in all the following species is occasionally abortive, whence we 
ted little doubt but C. anomocarpum, DC., is a mere state of some more perfect 
plant. 
1301. (1) C. didymum (Geertn. :) shrubby, unarmed : leaves short petioled, 
oval, more or less acuminated, glabrous, coriaceous ; the upper side shining ; 
under with glandular hollows in the axils of the nerves: cymes axillary, pe- 
duncled, much shorter than the leaves: calyx-limb with 5 often inconspicuous 
teeth: tube of the corolla furnished on the inside with rigid scariose hairs 
pointing downwards: stamens 5: stigma much exserted, ovate, mitriform, 
slightly 2-lobed at the apex: drupe (when mature) compressed and some- 
what didymous, broadly obovate and slightly emarginate, wrinkled and tu- 
bereled, marked with a furrow on each side.—Gertn. fr. 3. p. 94. t. 196 ; 
DC. prod. 4. p. 473; Roxb. fl. Ind. 1. p. 535 ; (ed. Wall.) 2. p. 171; Wight! 
eat. n. 1328.—C. cymosum, Pers. syn. 1. p. 200.—Psydrax dicoccos, Gertn. 
fr. 1. p. 125. t. 26. f. 2; DC.I. c. p. 476.—Webera cymosa, Willd. sp. p. 1224. 
—Rondeletia eymosa, Poir. enc. meth. 6. p. 256.— Cupia cymosa, DC. l. c. 
p. 394.—Gardenia parviflora, Poir. suppl. 2. p. 708 (not Kunth nor Smeathm). 
—G. Naum-papata, Roxb. in E. I. C. mus. tab. 15.—# ; young shoots and in- 
florescence softly pubescent.—# ; young shoots and inflorescence glabrous. 
Giertner's figure of the fruit (under Psydrax dicoccos) is much better than 
that given by his son; but this part presents very different appearances ac- 
eording to the stage of growth: at first it is globose without any trace of tu- 
bercles ; it afterwards dilates laterally, though still smooth, and finally the 
nuclei become wrinkled both externally and. internally, marking the seed in 
the same way. We have never observed it so deeply emarginate as in t. 196. 
of Geertner, although we know our plant to be identical with Roxburgh's and 
the Missionaries', preserved in the Banksian herbarium (from which he ob- 
tained his specimen) under the name of Webera cymosa. We — under- 
stand why De Candolle should have kept Web. cymosa (under Cupia), when 
Chamisso and Schlechtendal had already (Linnea, 4. p. 15), from actual ex- 
amination, determined it to be C. didymum. The minute calycine-teeth do. 
