Of these 10, four are certainly natives of the Indian Peninsula, namely, Tetracera, Dillenia, Colbertia, and Acrotrema : Colbertia however, heing only separated from Dillenia, by having a few of the stamens sterile and longer than the rest, and fewer pistils. This last character is now of no value, owing to some species recently referred to the genus, on account of their sterile stamens, having as many as 12 styles, the remaining distinction, sterile stamens, does not seem sufficient to authorize its removal as a separate genus, on which account, Dr. Arnott and myself, following Roxburgh, reunited it to the older genus, thus leaving only three for the Peninsula. Delima Sarmentosa quoted by Roxburgh as a Peninsular plant on the authority of Rheede's figure (Hort.Mal. 7 tab. 53) has not yet been found on the continent of India, Kheede's figure representing a plant not even referable to the order. The genera Acrotrema and Schumacheria, the ones here figured, require a somewhat more detailed notice. The former was established by the late Dr. Jack in the Malayan miscellanies for a Malay plant first discovered by him : since then two species have been discovered in Malabar, and some others in Ceylon. Some of the Ceylon ones I have seen, and think quite distinct from the Malabar plants, but owing to my not having specimens I do not attempt to characterize them. The genus was thus briefly defined by its discoverer, "Calyx pentaphyllus. Corolla 5-petala, patens. Stamina 15, erecta, filamentis brevibus, antheris longis, linearibus apice biporis. . Ovaria 3, distincta, 2-spora, ovulis angulo interiori affixis. Stili 3. Stigmata simplicia. Capsulae uniloculars — Herba acaulis, pilosa, pedunculis racemosa multifloris." With this character both A. costatvm and Wightianum sufficiently agree, but the species now added calls for some modifications to admit of its being received into the genus. Thus in A. Arnottianum, in place of 15 there are nearly 30 stamens, in place of 2 ovules they are very numerous, and in place of one or at most two seeds I have observed nearly 20 in one carpel. The inflorescence also differs ; for in place of a short raceme, on the extremity of an erect naked scape they are borne on a procumbent sucker-like branch, covered throughout with appressed scale-like bracts, from the axils of which, the long filiform pedicels spring.
The species equally require revision since the addition of A. Arnottianum, but that I feel averse to attempt until I receive specimens of the Ceylon species, I shall therefore content myself for the present by stating that A. costatum and A. Wightianum seem, from description, very nearly allied, if indeed distinct, (our former character is referable, partly, to two, very distinct, species, owing to our specimens of both, being so very imperfect as to prevent their being recognized as distinct) while A. Arnottianum is so widely removed from both in every thing but habit, that it may almost be considered a distinct genus. In the former, the flowers are borne on an erect scape, and the carpels 1-2 seeded; in the latter, they spring from, apparently, an abortive surculus (sucker) and the carpels are many seeded.
The genus Schumacheria was originally established by Vahl, and published in a German periodical, apparently little known, and seems to have been so imperfectly described, that DeCandolle with all his care and research, seems either to have overlooked it, or was unable to determine its affinities and place in the system of plants, as it is not taken up in his Systema Naturale. Springel has equally overlooked it, and has even published another, and very different genus, under the same name : hence we may conclude the genus was virtually lost until resuscitated by Dr. Arnott, who published a revised character in Jameson's New Philosophical Journal for April 1834. His character, though detailed and accurate, seems still to have left some point doubtful, as Professor Lindley in the second edition of his excellent "Natural system of Botany," has placed it at the conclusion of his arranged list of genera of the order, as if uncertain, either of its proper place or whether it actually belonged to the order. A genus so little known, and which may be found to merit a different fate, I have thought a suitable subject for this work. In the specific name I have followed Dr. Arnott, this being undoubtedly his plant, though it is possible, Vahl's may be a different one, as I am informed by Colonel Walker, that he has recently met with several other species, all quite distinct from the one here represented, descriptions of which I hope by and bye to have an opportunity of introducing.
Acrotrema.
Generic Character. Stamens 15-30, erect : filaments short: anthers adnate, long, linear, opening by terminal pores. Ovaries^, adherent at the base only, each terminated with a style and simple stigma: ovules few or many in each. Carpels 3, capsular: seeds 1-20, furnished with a membranous aril.
Herbaceous low plants. Leaves all radical. Petiols short, their margins dilated into membranacious, somewhat sheathing auricles. Peduncles either erect scapes, bearing a short terminal umbel-like raceme ; or sucker-like, clothed ivith dry scale-like bracts, from the axils of which the long, slender, one-flowered peduncles, spring.
