174 TEREBINTHACE E. BoswELMA. 
might bring it nearer to the Bytineriacea. But it is by no means certain that th 
calyx is not connate with the arog; and this, from Rheede’s mode of expression, is 
very probable ; in which case an affinity might exist with the Alangice. It has been 
also conjectured that the bractea may be the true calyx, and the apparent calyx a 
tube to the corolla, which would thus have 3 spreading segments. The plant has 
been seen, we believe, by no modern botanist: Rheede found it in “ Parakaroo and 
other provinces of Malabar.” 
XI. BOSWELLIA. Roab. 
Flowers bisexual. Calyx small, 5-toothed, persistent. Petals 5, obovate- 
oblong, very patent, acute at the base, insérted under the margin of the disk : 
sestivation slightly imbricative. Stamens 10, inserted under the disk, alter- 
nately shorter: filaments subulate, persistent: anthers caducous. Torus a 
cup-shaped disk, fleshy, larger than the calyx, crenulated on the margin, 
Ovary oblong, sessile. Style 1, the length of the stamens, caducous. Stigma 
capitate, 3-lobed. Fruit capsular, 3-angled, 3-celled, 3-valved, septicidal : 
valves hard. Seeds solitary in each cell, surrounded by a broad membrana- 
ceous wing. Cotyledons intricately folded, multifid.— Trees produeing bal- 
sam and resin. Leaves deciduous, alternate towards the top of the branches, 
unequally pinnated: leaflets opposite, serrated. Stipules none. Racemes 
terminal or axillary. Flowers on short pedicels, white. 
Although we describe this in conformity to other botanists as a capsule, it is in 
reality a modification of a drupe, the coriaceous sarcocarp splitting at the angles 
inte. valves, which remove with them the endocarp or nuts, each containing a $ 
535. (1) B. glabra (Roxb. :) leaflets broadly-lanceolate, obtuse, serrated, 
glabrous: racemes simple, terminal, fascicled, shorter than the leaves.— 
Roxb. Cor. 3. t. 207 ; fl. Ind. 2. p. 384 ; DC. prod. 2. p. 16.——Mountainous 
districts of Coromandel. 
In the specimen which we have seen (in Hamilton's herbarium) from the 
Calcutta Botanic Garden, marked however “ B. thurifera, Colebr,” the leaflets 
are glabrous, oblong, quite obtuse, and distinctly crenate-serrated : Rox 
says they are sometimes quite entire ; he also quotes Canarium balsamiferum, 
Willd. Rumph. Amb. 2. t.50) which has acuminated entire leaves, but which 
is, according to Rumph’s description, a true Canarium. 
536. (2) B. thurifera (Colebr.:) leaflets oblong, obtuse, serrated, pubes- 
cent: racemes axillary, single, shorter than the leaves.—Colebrooke in As. 
res. 9. p. 377 ; Roxb. fl. Ind. 2. p. 383. Mountainous parts of Coromandel. 
We dare not quote here B. serrata, Stackh. extr. Bruc. p. 19. t. 3, the 
leaves being usually described as ovate-oblong and acuminated: nor, for the 
same reason, can we refer, with Roxburgh, to Rumph Amb. 2. t. 51, 
Canarium hirsutum, Willd., or Bosw. hirsuta, Sm. The different species T°- 
quire however to be re-examined, in order to obtain better specific characters. 
XII. CANARIUM. Linn.; Gertn. Sr. t. 102. 
Flowers polygamous-dicecious. Calyx eampanulate, 3-lobed ; lobes un- 
equal. Petals 3, inserted under the disk, twice as long as the calyx, oblong, 
concave: sestivation imbricative. Stamens 6, inserted under the disk, shorter 
than the petals, unequal. Torus an urceolate disk, at the bottom of the 
calyx. Ovary sessile, ovate-globose, 3-celled. Style very short, or wanting 
Stigma 3-lobed. Fruit a drupe: sarcocarp thin, somewhat fleshy: nut very 
hard, 3-angled, 3-celled (two of the cells often abortive). Seed solitary 1? 
gach perfect cell. Cotyledons intricately folded, deeply 3-cleft.— Trees. 
