is to have quaternary sepals and petals, 8 stamens, and a 3- celled ovary, with 2 superposed ovules in each. To admit my plant it would require to stand thus— Calyx 4-5 lobed, petals 4-5. Stamens 8-10, free or united at the base Ovary 3 or 5-celled, with 2 superposed ovules in each. Scandent armed shrubs, with ternate, or simple leaves, and axillary solitary, or racemose, " pretty large, white, fragrant flowers" (Roxb.) Such a union is however very objectionable as combining two plants, one with the number of petals and cells of the ovary equal, and the other with them unequal. I therefore pursue what I consider the more judicious course, that of forming a genus for the reception of this plant.
Luvunga. — Hamilton. — The following is Roxburgh's description of the fructification of his Limonia scandens, now Luvunga, Flora Indica. 2 page 380. " Calyx 1-leaved, cylindric with the mouth cut into four short, truncate divisions. Petals 4, linear oblong, fleshy, recurved. Filaments 8 ; the lower half united into a firm, fleshy tube. Anthers linear, incumbent. Germ conical, elevated on a fleshy receptacle, 3-celled with 2 vertical ovula in each, attached to the axis. Style cylindric. Stigma entire, roundish. Berry oblong, somewhat 3-lobed, size of a pigeon's egg, pretty smooth, pulp of a resinous nature, and odoriferous, 3-celled. Seed solitary, oval, somewhat pointed at the apex, covered with a single greenish-veined integument. Perisperm none. Embryo conform to the seed, inverse. Cotyledons oblong, green, fleshy. Plumula 2-lobed. Radicle ovate, superior."
Two species of this genus are named in Wallich's list, namely, L. scandens and L. Tavoyana— Roxburgh thus characterizes the former, his Limonia scandens " Shrubby, scandent, armed, leaves ternate, leaflets lanceolar, entire : [flowers racemose,] : berries 3-seeded" Roxb.
The following character will I think distinguish my new genus from all the others of the order.
Paramignya R. W. —Calyx urceolate, obtusely 5-lobed at the apex. Petals 5, lanceolate. Stamens 10, free : filaments compressed, subulate at the point : anthers oblong, obtuse, attached by the back. Torus fleshy, cup-shaped, lobed at the apex ; embracing the base of the ovary. Ovary obtuse. 5 celled, partitions fleshy, not twisted. Ovules 2, superposed, in each cell, attached near the middle of the axis. Style thick, length of the stamens. Stigma capitate, obtuse, somewhat lobed at the apex. Fruit pomacious ! endorcarp 5-angled. between coriaceous and fleshy, surrounded by cellular pulp and rind, one (always ?) celled by the rupture and absorption of the partitions. Seed.— Scandent, armed shrubs ; spines axillary reflexed. Leaves simple, ovate, lanceolate, acuminated, glabrous. Flowers axillary, solitary, of rarely 3 or 4 together, rather large, white, very fragrant. The fruit ovate, obtusely 5-angled and furrowed between, clothed with short matted pubescence, the seed I have not seen.
The scandent habit, associates this genus with Luvunga, from which it is removed by the symmetrical petals and cells of the ovary, the predominance of the quinary not quaternary number of parts, and the free not united filaments, but above all, by the distinct endocarp of the fruit. I first found it in the jungles about Courtallum its long decumbent branches, spreading in all directions among the surrounding bushes, forming a nearly impenetrable brush-wood ; I afterwards received fine specimens through the kindness of Coionel Walker from Ceylon, and very recently one in fruit, but in a bad state for examination, from Mr. Nimmo, of Bombay. This last is the only fruit I have seen.
Only one species is yet known, the one here figured — It varies however in having larger, and more decidedly ovate leaves, and smaller thorns, differences depending I apprehend on luxuriance only.
The species of Micromelum the examination of which enabled me to determine, that the above was really a distinct genus, is a native of Ceylon, and does not seem to differ from Blume's M. pubescens.
ATALANTIA.
The following brief characters will sufficiently distinguish my two species of Atalantia.
A. monophylla DeC. Style as long as the stamens, stigma subcapitate, lobed.
A. platystigma. R. W. Style wanting, stigma large, flat, appressed to the apex of the ovary.
The habit and general appearance of both plants are the same.
GLYCOSMIS.
I mentioned above, that I had met with what I consider a new species of this genus. The following character may perhaps suffice to distinguish it from the preceding species.
