ILLUSTRATIONS OF INDIAN BOTANY-
115
XCI.— SPHENOGLEACEAE.
This order, if such it be, rests on a single species, the well known Spenoclea Zeylanica, of Willdenow and Roxburgh, very common in moist, cultivated ground in this country. It is one of those out-lying plants that declines associating with any, as yet known, family, but which seems to approach Campanulacece more nearly than any other, wanting, however, several points, deemed essential in that family, which precludes its admission as a genuine member. Dr. Martius, who first indicated it as the type of an order, places it between Rubiaceae and Cam- panulacese. Lindley, in the 2d edition of his Nat. System, adopted it as a sub-order of Campanulacece, in which he was followed by Endlicher. Meisner and De Candolle adopt the order, the latter placing it at some distance from the Campanulaceous orders; and lastly Lindley, in his Vegetable Kingdom, has, for the present, reduced it, placing the genus with two others as "anomalous genera" at the end of his Campanulacea?, waiting the discovery of companions better suited to indicate its true station.
Amidst such diversity of opinion, among the lights of the science, I do not feel myself competent to arbitrate, but at the same time, considering that the readers of this work should be put in possession of all the information I can furnish on such litigated questions, I, for the, present, adopt the views of Martius and De Candolle, as affording an opportunity of furnishing a figure and description of a plant which has provided materials for so much difference of opinion. My own opinion, so far as my imperfect acquaintance with Campanulaceae justifies me in expressing it, is in favour of separation. The deciduous corolla, the insertion of the stamens on its tube, not on the calyx, the short styles, and capitate stigmas, without collecting hairs; the central, pendulous, fungoid placenta?, circumscissile dehiscence, and, above all, the very sparingly albumenous or almost exalbumenous seed; form such a combination of characters, as leave no room to doubt the propriety of keeping this genus altogether distinct, rather than placing it at the tail of another order, waiting the discovery of companions, as, in that situation, its peculiarities may not be so clearly brought to light. De Candolle thus defines the order, with the exception of one or two slight modifications I have introduced.
Character of the Order. Tube of the calyx adnate to the ovary, limb 5-parted, lobes round on the margin, inflexed, persistent, finally connivent over the ovary. Corolla deciduous, 5-parted, lobes inflexed. Stamens 5, sub-sessile on the sinuses of the corolla; anthers roundish, 2-celled, dehiscing longitudinally. Ovary 2-celled, many-ovuled. Styles very short, stigma, capitate, bilobate, glabrous. Capsule membranaceous, 2-celled, cunif'orm at the base, many- seeded, circumscissile. Placenta? fungose, pendulous from the apex of the septum. Seed tuberculate, minute, terete, sparingly furnished with fleshy albumen. Embryo straight, terete, radicle about twice as long as the cotyledons. — Herbaceous, erect, ramous, glabrous, annuals. Leaves alternate, lanceolate, exstipulate, entire. Spikes terminal or leaf opposed, peduncled, cylindrical. Bracts 3 or 3-partite under the flower. Flowers small, white.
Affinities. On this head I have nothing to offer. That this family, if such it be, is nearly allied to Campanulacece, seems certain, but that it does not belong to it is admitted on all hands. The placenta? are curious, two spongy, egg-shaped bodies, pendulous from the apex of the septum. My dissections of the seed tend to confirm Ga?rtner 1 s description of its structure, as I find the embryo enclosed in a thin, fleshy albumen and very delicate, translucent, tuber- culed testa.
Geographical Distribution. So far as yet known, this order rests on a solitary species, found growing in wet or marshy soils, in India, Ceylon, Eastern Islands, Egypt, Mexico and the West Indies ; thus, like other aquatic and sub-aquatic plants, enjoying a very extended geographi- cal distribution.
Properties. Unknown.
