placentas in the axis : valvate œstivation of the calyx : monadelphous stamens ; 2-celled anthers : and alternate stipulate leaves. All Indian plants having the above combination of characters, ought to find a place in one or other of the following sections. Tiliaceæ and Elæocarpeæ two very nearly allied orders are distinguished, the former by its distinct stamens, the latter by its lacerated petals.
Tribe I — Helictereae. Flowers bisexual. Calyx irregular. Corolla irregular. Filaments either united into a tube longer than the ovary, or embracing the carpophore (pedicel of the fruit,) and free at the apex; anthers 2-celled, with an obsolete transverse septum. Ovary sessile or stipitate. Fruit with the carpels either distinct, or cohering, one or many seeded. Seeds albuminous. Leaves simple.
- * Eupilectereæ Tube of the stamens elongated, embracing the stipe of the ovary. Filaments free at the point, each bearing a single anther. —To this subsection our Helicteres isora (Isora corylifolia End.) belongs.
Tribe II — Sterculieae, Flowers by abortion unisexual, calyx having the sepals united regularly. Corolla none. Filaments united into a tube adhering to the carpophore anthers imperfectly one or two celled, oblong-curved, situated at the base of the perfect or imperfect ovary. Fruit pod-shaped of several verticelled carpels, opening along the interior sature. Seed sometimes covered with an arillus, albuminous. Trees with simple or digitately compound leaves, with the petiol tumid at the apex.
To this section our genera Sterculia and Heritiera belong. The former, as left by Roxburgh and DeCandolle, is completely broken down and now affords the types of no fewer than nine distinct genera, seven of which appertain to the Indian flora, the characters of these I shall add at the conclusion of this article.
Tribe III— Bombaceœ. Flowers bisexual. Calyx 5-cleft, often irregularly divided, the œstivation then obscure. Corolla regular or none. Filaments united into a tube, covering the ovary. Anthers solitary or several cohering, cells indistinct or confluent, often anfractuose. Ovary sessile. Fruit capsular composed of cohering carpels. Seeds albuminous or exalbuminous, often enveloped in wool or even in pulp. — To this section the genera quoted under the former order belong.
Tribe IV — Byttnerieœ. Flowers bisexual. Calyx 5-partite or 5-sepaled, without an involucel. Petals frequently concave and vaulted at the base, expanded at the apex into a strap-shaped appendage. Stamens 10 — 30, or more ; the five that are opposite to the sepals sterile and strap-shaped ; the others, opposite to the petals, either solitary or pentadelphous, the bundles rarely bearing only one anther. Ovarium sessile, 5 celled ; cells usually 2-ovuled. Seeds sometimes exalbuminose with thick cotyledons ; sometimes albuminose with foliaceous, plane, or convolute cotyledons. Trees, shrubs, or very rarely herbaceous plants. Leaves alternate, entire, or sometimes cut. Stipules twin. Peduncles axillary, opposite to the leaves, and terminal, one or many flowered.
To this section the Cacoa tree, (Theobroma) and the bastard cedar tree (Guazuma) and several other Indian genera belong.
Tribe V — Hermannieœ. Flowers bisexual. Calyx 5-lobed, persistent, either naked or with an involucel. Petals 5, spirally twisted in œstivation. Samens 5, monadelphous, all fertile, opposite to the petals. Carpels united into a single fruit. Albumen between fleshy and mealy. Embryo included: radical inferior, ovate: cotyledons flat, leafy, entire. Shrubs or herbaceous plants. Leaves alternate, simple, or variously cut. Stipules 2, adhering to the petioles. Peduncles axillary, or opposite to the leaves, or terminal, with 1, 3, or many flowers, which are usually in umbels.
To this tribe Riedleia and Waltheria belong. The former as left by DeCandolle a very confused genus, demanding a careful revision, as it certainly includes within itself the types of several. None of those referred to it from India are genuine species. Riedleia truncata I have removed to Malvaceœ, and of our three other species, which I think, should be reduced to one, Dr. Arnott has formed a new genus under the name of Lochennia.
