PLEUROSTYLIA. CELASTRINEZ. 157 
492. (1) E. Roxburghii (W. & A.:) leaves elliptical or ovate, crenate- 
serrated, young ones glaucous: cymes lax, dichotomous, divaricated, about 
half the length of the leaves, usually with a solitary pedicellate often cadu- 
cous flower in the forks: drupe 1-celled, obovoid ; nut somewhat crusta- 
ceous and soft.—Wight! cat. n. 485, 487, 488.—E. glaucum, Wall.! L:n. 
4325. a, e, e. —E. ——? Roxb. in E. I. C. mus. tab. 13.—Nerija dichotoma, 
Roxb. fl. Ind. 1. p. 646; (ed. Wall.) 2. p. 444.—Rhamnus Nerija, Spr. syst. 
supp. p. 86. Courtallum, and mountainous parts of the Coromandel coast. 
Roxburgh has erroneously described this as a berry : on cutting aeross the 
fruit, however, the putamen can be distinguished from the thin fleshy coat. 
Wight's n. 488 has more the habit of E. glaucum (in which the drupe is nearl 
spherical, and has a very thick and hard putamen, with occasionally 2 cells) 
than the others, but in the absence of fruit we refer it to E. Roxburghii, as it 
does not appear that the true E. glaucum has been yet found out of Ceylon, 
the plants mistaken for it, from Coromandel, being either referable to the 
present or to the following species. 
493. (2) E. paniculatum (W. & A. :) leaves elliptical, green, bluntly ser- 
rated, on long petioles: panicles dichotomous, without a flower in the fork, 
about as long as the leaves ; ultimate pedicels ternate.— Wight! cat. n. 489.— 
E. mM Wall.! L. n. 4325. f.——Colemala. 
specimens are not in fruit: the whole habit is considerably different, 
but perhaps future observations may prove it to be only a variety of the pre- 
species. 
III. PLEUROSTYLIA. W.& A. 
Calyx 5-parted. Petals 5, with a broad base. Stamens 5: filaments flat, 
shorter than the petals, attached below the margin of the torus: anthers with 
a broad connectivum at the back, opening longitudinally. Torus a thick 
crenulated fleshy disk. Ovary half immersed in the torus, 2-celled, or usually 
from abortion only 1-celled (the other being almost always rudimentary and 
searcely to be traced): ovules 2, collateral, in each perfect cell. Style short 
and thick. Stigma large, somewhat peltate. Fruit indehiscent, 1- (or oc- 
casionally 2-) celled : sarcocarp thin and fleshy, marked on the one side 
about the middle with the remains of the style: endocarp a crustaceous mem- , 
brane, separable from the sarcocarp, with several irregular clefts at the base 
around the hilum of the seed. Seed erect, solitary, covered by the endocarp 
like an arillus. Albumen copious, fleshy. Cotyledons large, orbicular, flat : 
radiele inferior—Shrubs. Leaves opposite, entire and quite entire, whitish, 
shortly petioled. Peduncles axillary or terminal, scarcely longer than the 
petioles, few-flowered. d 
This genus is remarkable for the one cell becoming usually abortive, even in the 
ovary : although, however, it be then perfect, it very rarely indeed comes to maturity. 
The fruit is obviously a modification of a 2-lobed drupe with two l-celled d 
Whereas Eleodendron, when complete, would exhibit a simple drupe wie one 2-cell : 
nut: but independently of that consideration, the peculiar ; putamen, an 
Structure of the s seed, will keep the present genus quite distinct. 
494, (1) P. Wightii (W. & A.:) leaves elliptic-oblong.— Wight ! eat. n. 
481.— Celastrus? V ich qma, wail) L. n. 4323 — Narthamala. Southern 
provinces. 
495. (2) P. Heynei (W. & A.:) leaves narrow oblong-lanceolate.— Wight / 
eat. n. 486.—Celastrus band ei in Rowb. ft. Ind. (ed. Wall.) 2. p. 398 ; 
L. n. 4314 ; Spr. syst. suppl. p. 88.——Southern provinces. 
We have considerable doubts of these two species being distinet. 
