Page:Illustrations of Indian Botany, Vol. 1.djvu/53

Rh Fresh gathered from the tree, they are agreeable to the palate, and grateful to the stomach.

The bark is also richly impregnated with this juice, as well as the wood, and both yield a very agreeable bitter in the mouth while fresh ; but. that diminishes greatly after they are dried. The bitter quality of this tree is communicated with great facility. A handful of the shavings immersed in water, and instantly taken out again, will render it of a very bitter taste. Sugar sent over in hogsheads made of this wood, was so bitter that no person would purchase it. Bedsteads, and presses made of it are proof against cockroaches and other insects."* A decoction of it is said to create appetite, and possess the usual tonic properties of bitters. These examples are I presume, sufficient to establish the value of the properties inherent in members of this family, and go far to prove, that among the numerous Indian species belonging to it, we have good reason to expect, some, when properly investigated, will be found not less valuable,, than the better known American and African ones.

, &c. Of this order Dr. Roxburgh, in his Flora Indica, gives characters of 27 species, classed under three genera, namely, Anona, Uvaria, and Unona ; Guatteria, was afterwards added to the catalogue, of Indian genera, by referring several of Roxburgh's Uvaria's to that genus. These genera were distinguished by characters almost entirely taken from the fruit, without reference to the ovary, or indeed to any other part : Anona, being distinguished by having a number of carpels, united into a single compound fruit: Uvaria, by having its carpels distinct, stipitate, fleshy, many-celled, and many seeded ; with the seeds ranged in two rows : Unona, by having distinct, but dry, many-seeded, carpels ; the seeds ranged in a single row, and often resembling, by the contraction of the carpel between them, a necklace : Guatteria, by having dry, globose, stipitate, one-seeded, carpels. These simple, and, at first sight, apparently all-sufficient characters, were found on more careful examination to be exceedingly incorrect. Uvaria, for example, to which was attributed a many-celled fruit, and two rows of seed, was found to have a one-celled ovary, and the rows of ovules, if two, so close as scarcely to be distinguishable, and in truth forming a single line attached to the inner angle ; hence the many cells, and two rows of seed observed in the mature carpel, must be produced by mere condensation of the surrounding pulp, and the divergence of the free extremities, of the seed, since they are all attached along the same angle of the seed vessel. In Unona, the structure of the ovary is the same, with probably fewer ovules, hence it follows, that the abortion of a few ovules (by leaving more room for the regular development of the remainder) may convert a Uvaria,'mto&Unona; and vice versa, an unusual number of ovules, or any hindrance to the usual course of development of the seed vessel, might equally change Unona, into Uvaria ; the differences between the two genera, thus rest on adventitious, not structural, differences. In support of the justice of this position, it is only necessary to state, that a large proportion of the species of the latter genus, have, since the publication of DeCandolle's sy sterna, been removed to the former. The propriety therefore, nay, the necessity, of uniting the species of both, and of two American genera, having similar ovaries and fruit, (Asimia and Porcilia) into one genus, as ably advocated by M. Richard, in some remarks on the subject in the Flore Senegambie, becomes evident. Blume, however, as appears from DeCandolle's memoir, has revised the character of Uvaria, and still keeps them distinct ; but as I have not his work to consult, I am unable to state with what propriety. Guatteria, which in like manner is characterized from the mature fruit, without reference to the ovary, may be simulated by species of Uvaria, or Unona, through the abortion of all the ovules but one, a modification of which my collection presents specimens.

Swayed by these facts, M. Richard proposes an amended character for Uvaria, in which the one- celled, many ovuled ovaries, with the ovules attached along the inner angle, forms the essential distinction; a modification which admits of the association of all the species now referred, to the four genera above named. The character of Guatteria, might be similarly modified with advantage, and would then, perhaps, be found to separate the American, from the Indian, division of the genus. All the Indian ones I have yet examined have a single, erect, ovule, attached to the bottom of the ovary. Whether or not the American species referred to this genus, possess this structure, I am unable to say, but in the following Indian ones, I have