Page:Illustrations of Indian Botany, Vol. 1.djvu/439

ILLUSTRATIONS OF INDIAN BOTANY.

217 As stated above, America may justly be considered the head quarters of this tribe, upwards of 600 species having been obtained from that quarter out of about 800 described by DeCandollein his Prodromus. Since the publication of that work many genera and species have been added, whence I believe I under estimated the number of species now known at one thousand. Of this number probably about 100 are natives of continental India and Ceylon ; of these Roxburgh seems only to have known 14, as 7 out of 21 described by him are from Pulo Penang and the Moluccas. Blume has added largely to the list from Java, having probably nearly doubled the Indian list, that is, supposing that the other genera are somewhat in proportion to Medinilla, of which that Island produces 21 species. In Ceylon Mrlastomaceae are also numerous, and several new and very curious species have recently, through the researches of Colonel Walker, been brought to light; among these are several species of Osbeckia, five or six of Sonerila and three or four of Medinilla. One species of this last genus, but differing somewhat from the generic character especially in the anthers being spurred only, not auricled and spurred, at the base, is among the handsomest plants I have ever seen. It is an enormous creeper, adheres firmly to the trees on which it grows, climbs to the tops of the highest trees of the forest, and covers them with a profusion of large crimson flowers. This species T first saw in company with Colonel Walker (in a. dense forest above Ramboddy) to whom I have dedicated this noble plant Medinilla f Walkerii. On the alpine ranges of the southern provinces and in Malabar, I collected many specimens, several of which, I then supposed were new species, but which, on a more careful examination and comparison with others, must I fear be considered as varieties only, at least I have not yet been able to find satisfactory distinctive marks by which to raise them to the rank of species. Figures of some of these shall shortly appear in the Icones. In addition to these from the south, a few species spread northwards and even extend to the foot of the Himalayas, but these are so few in number as scarcely to form an exception to the general rule that this is truly a pre-eminently tropical order.

Astringency is said to be the predominating quality of the order, but little seems yet known regarding them. Lindley remarks that the order " though one of the most extensive known is entirely destitute of any unwholesome species. The succulent fruit of many is eatable that of some dyes the mouth black, whence the name Melastoma. It may be here remarked as somewhat curious that the genus established by Burman under this name, for which he is quoted as the authority and of which he figured two species, is not that now called Melastoma, but Osbeckia, the original name being limited to a genus, of which it does not appear he ever saw a single plant.

The Indian genera are few and easily distinguished, I do not therefore think it necessary to devote much space to this division. Osbeckia has either a quaternary or quinary order but the stamens are all alike. Melastoma, which in general appearance it resembles is easily distinguished by the stamens, one-half having the anthers sessile on the apex of the filament, the other having them supported on a long pedicel or stipe, apparently formed from a prolongation downwards of the connectivum : by this mark alone these allied genera can always be distinguished at a glance. Sonerilla is separated from both by its ternary order of parts — Medinilla has a quaternary series, the anthers with auricles and a spur at the base, in M. Walkerii the auricles are wanting, but the spur is considerable, on which account I have marked the genus with a doubt. Among my collection of Mergui plants communicated by Mr. Griffith I find a second species presenting the same peculiarity, agreeing in that respect with Pternandra, but distinguished by the anthers of the latter opening by a longitudinal slit not by a pore as in the other. The genus Triplectrum abounds about Courtallum but seems rarely to flower as I, in the course of several visits to that place, could not find a single plant in flower though I walked over acres of ground covered with the plant. This disappointment I the more regretted, as the genus is founded on a single specimen and that not a very good one, whence I fear it may only be a modified state of Medinilla.