bursting elastically. Albumen between cartilaginous and fleshy. Embryo straight, as long as the albumen: radicle long; next the hilum : cotyledons foliaceous.—Leaves compound (or by abortion simple), alternate, seldom opposite or whorled."
Affinities. The affinity of this with the three preceding orders, has been already adverted to. Formerly it was united with Geraniaceae, and is still considered by some not sufficiently distinct. DeCandolle however thinks it more nearly allied to Zygophylleae, though quite distinct. Its compound leaves and albuminous seed in both of which respects it differs from Geraniaceae but associates with Zygophylleae, seem to confirm this view.
Geographical Distribution. The genus Oxalis is principally confined to the Cape of Good Hope and America, but is also though sparingly met with in Europe, Asia, and New Holland. The genus Averhoa is confined to India and the adjoining islands : of Biophytum one species is found in the West Indies and the rest in India.
Properties and Uses. Acidity is the predominating quality of this order. The Oxalis acetosella or wood sorrel, is well known in Europe, and esteemed on account of its cooling and refreshing properties, the expressed juice of which furnishes, when crystalized, the so-called essential salt of Lemons, or binoxalate of potash. Both species of Averhoa, namely, A. Carambola and A. Bilimbi, are equally well known on account of their intense acidity. The kind of acid which imparts the taste I have never seen mentioned, but presume it is oxalic, the one which predominates in the family. The fruit of the former is considered cooling and aperient.
Remarks on Genera and Species. Originally two genera only were referred to this family, Oxalis and Averhoa. The former of these DeCandolle split into two, more perhaps, on account of the marked difference of habit than from any well marked differences of structure. The principal points of difference consist in the filaments, being free to the base in the one (Biophytum) while they are united into a tube nearly half their length in the other (Oxalis). They also differ in the form of their fruit, the former having its capsules approaching to globose, while in the latter they are cylindrical. These distinctions, added to the very marked difference of habit have induced me to revert to DeCandolle's division, from which we departed in the Prodromus The species of the genus Biophytum are of difficult discrimination, and it has been doubted whether there are more than one in India—on this point I now feel quite satisfied, and think we may certainly acknowledge two, I think several species. Supposing we fix upon the formei number there can be no difficulty in distinguishing them, the one being marked by having a single terminal tuft of leaves on a more or less elongated simple unbranched palm-like stem—the other, by having a diffuse ramous proliferous stem, each branch terminating in a tuft of leaves. The various forms might then be ranged under these, as so many varieties. The difference of habit and station which some of these forms affect seem to stand in opposition to this extent of simplification, for example—Biophytum sensitivum is only found on the plains usually in very open ground exposed to the full blaze of the sun's light, whereas, the form figured in plate 62 is only found in cool alpine situations under the shade of thick jungles, both here and in Ceylon. These circumstances may induce the belief that the plant is the same, only altered by local circumstances, an opinion which, if urged, 1 confess I have not the means of controverting, since it can only be set aside by making the two plants change places, and in that way determining whether or not their forms would alter also. The B. sensitivum is further distinguished by being glabrous, while the other is very generally clothed with hairs, sometimes, especially on the rachis of the leaves and peduncles, very densely. This character, however, is not sufficiently constant in either form to admit of much weight being attached to it. Of the palm-like forms there are again two varieties distinguishable by the form of their leaflets—the form fig. 8 in plate 62 represents the one, that of fig. 10 nearly corresponds with the other, though not taken from it— fig. 8 is a slightly magnified leaflet of B. Candolianum—fig. 10 is that of a species to be afterwards mentioned, but, with the exception of its being a little more tapering and less distinctly mucronate at the point, gives a pretty good idea of the form of Dr. Arnott's B. (Oxalis) nudum, and will assist in distinguishing these two perhaps too nearly allied species.
