78 : TILIACEJE. GREWIA. 
283. (7) G. Rothii (DC.:) shrubby: leaves 3-nerved, ovate-oblong, acu- 
minated, obtuse or acute, sharply serrulated ; upper side glabrous, under 
hoary with short white tomentum: peduncles 2-3, axillary, slender, 6-8 
times longer than the petioles; pedicels much shorter than the peduncles, 
twice as long as the bracteoles: sepals linear, narrow, twice the length of 
the oblong entire petals: style filiform, a little longer than the stamens; 
stigma dilated: drupe globose, glabrous, with 2 1-2-celled nuts.—DC. prod. 
1. p.509 ; Spr. syst. 2. p.579.—G. bicolor, Roth, nov. sp. p. 240 (not J'uss.)— 
G. leevigata, Heyne (not Vahl.)—G. salvifolia, Roxb. fl. Ind. 2. p. 597 ; in E. I. 
C. mus. tab. 225; Wall.2 L. n. 1090. a.—G. excelsa, Wall.? L. n. 6307. a. 
Our character is principally taken from Roxburgh’s figure. 
284. (8) G. hirsuta (Vahl:) shrubby: leaves 3-nerved, lanceolate, acu- 
minated, obliquely rounded or cordate at the base, acutely and unequally 
serrated ; upper surface slightly wrinkled, hairy or slightly tomentose, of the 
adult leaves nearly glabrous ; under densely tomentose : peduncles 1-3, ax- 
illary, about the length of the petioles, 2-4-flowered : pedicels rather shorter 
than the peduncles: sepals lanceolate, more than twice the length of the 
petals : petals oblong, entire, lower half villous on the back and ciliated on 
the margin : torus scarcely longer than the glands: style filiform, longer than 
the stamens ; stigma large, 2-3-lobed, stellately penicillate: drupes nearly 
globose, hairy, slightly 4-lobed ; nuts 4, each 2-celled.—Vahl! symb. 1. 
p.94; DC.* prod. 1. p. 509; Spr. syst. 2. p. 580; Roxb. fl. Ind. 2. p. 5875 
Wight! cat. n. 256.—G. tomentosa, Roxb. in E. I. C. mus. tab. 226.—G. 
pilosa, Roxb. fl. Ind. 2. p. 588 ; in E. I. C. mus. tab. 1447 (not Lam.)—6G- 
montana, Koen.! in herb. Banks.—G. salvifolia, Wall.! (not Roxb.) L. n. 
1091. a-d.—G. helicterifolia, Wall.! L. n. 1090. c.—G. Roxburghii, G. Don in 
Mill. dict. 1. p. 548 (exclus. syn.)——Jungles near Madras, Wandawasi, 
Nandaradah, &c. 
Roxburgh in his earlier description of this plant as G. hirsuta, and corres- 
ponding drawing as G. tomentosa, has described the fruit as having 1-celled 
nuts, probably from having only cut it transversely : we find them almost con- 
stantly 2-celled : the distinguishing characters, therefore, given at p. 588 of the 
fl. Ind. are insufficient, and we therefore unite the Bengal to the Coromandel 
species; the only difference being that the leaves of G. pilosa are linear-lanceo- 
late, in G. hirsuta approaching to ovate-lanceolate; which differences may arise 
from cultivation, in which state only Roxburgh knew his G. pilosa. Wallich's 
n. 1090. b/, d!, e!, and n. 1091. e/, belong to G. polygama, Roxb.! (fl. Ind. 2. 
P 589, and in E. I. C. mus. tab. 1448), a species differing from G. hirsuta by 
aving male flowers without the least trace of a pistillum on one plant, and 
bisexual flowers on another ; in it also the petals are not hairy on the back, 
and the gland, not the petal, is tomentose round the margin: besides, the 
nuts of the fruit are 1-celled. We do not know Wall. L. n. 1090. a. 
285. (9) G. pilosa (Lam.:) shrubby: leaves 3-nerved, on very short petioles, 
cuncate- or obovate-oblong: lanceolate scarcely acuminated, rounded or slight- 
ly cordate at the base, unequally serrated, rough with stellate hairs ; under 
side with the hairs longer more copious and softer: peduncles 1-3, axillary; 
the length of the petiole, 3-flowered ; pedicels as long as the peduncles and 
the bracteoles; sepals linear, narrow: petals linear, bifid!, as long as the 
stamens and about two-thirds of the length of the calyx : torus scarcely éx- 
ceeding the glands: style filiform, a little longer than the stamens ; stigma 
4-partite, segments filiform: drupes 1—4-lobed, crustaceous, covered with 
stellate hairs; nuts 1-4, 1-3-celled.—Lam. enc. meth. 3. p. 43 (excl. syn. and 
char. of fruit?) ; DC. prod. 1. p. 510; Spr. syst. 2. p. 580 ; Wight ! cat. n. 262. 
—G. carpinifolia, Roxb. (not Juss.) fl. Ind. 2. p. 587; in E. I. C. mus. tab. 
* We are at a loss to know why De Candolle and Mr G. Don have locality of ** Java” to this — 
Vahl received it from Koenig, gathered in the Peninsula. Renters nat ed vas. pop 
allich calls it G. salvifolia ; and, by a singular coincidence, Roxburgh in his Fl. Ind. 2. p. 503, 
both these names to Alangium herapetalum, Linn. 
