"From Aceraceae these scarcely differ, except in their alternate leaves and petals, which have almost always an appendage on their surface. In some respects near Meliaceae, which agree in habit and in their pinnated leaves, but which are known by their monadelphous stamens and symmetrical flowers. To Polygalaceae they are no doubt akin in the singular combination of 8 stamens with 5 unequal sepals, and an uncertain number of petals; and also in their aril, which may be compared to the caruncula of Polygalaceae, although somewhat different in its origin. The dried leaves resemble, as DeCandolle remarks, those of Connaraceae. Their climbing habit and tendency to produce tendrils indicate a relation to Vitaceae, which, however, is not very near. Brown remarks, that although in the far greater part of this family the ovule is erect, and the radicle of the embryo inferior, yet it includes more than one genus, in which both the seeds and embryo are inverted."
To me it is matter of surprise that there is no allusion in any work I have had an opportunity of consulting to any affinity existing between this order and Euphorbiaceae, though some of the genera of each order seem so closely allied that it is difficult to say to which order they ought to be referred.
Geographical Distribution. This large order is nearly confined to the tropics, or extends but a short way beyond, being still limited to the warmer latitudes on either side. The greater number of species are natives of Equinoctial America, and India; Africa also has many of them, but they are unknown as natives in Europe and the United States of America, while the genus Dodonaea alone represents them in New Holland. We have not the same data for estimating the number of Indian species as in most of the other orders yet gone through, as they are not included in Wallich's list of Indian plants. The number of Peninsular ones known to us amounted however to only 14, and these have not since been augmented. This I should suppose falls greatly short of the actual number as Blume in his Bijdr : Florae Java?, has no fewer than 23 species, and in Ceylon there are several that have not yet been met with on the continent, though it seems to me they are such as might be expected. Those met with on the continent occupy very different stations. Thus Cardiospermum Halicacabum is commonly met with in hedges and corn fields, while C. canescens is rarely met with except in dry jungles, where it appears as a very extensive climber, and when in full flower, which it is the greater part of the year, is really a pretty plant. The Sapindus emarginatus is generally met with as a cultivated plant, but is not rare in subalpine jungles in a wild state. The same remark applies to Schleichera trijuga. Nephelium Longanum on the other hand, a species very closely associated with Litchi, I have only seen in jungles, and usually at a considerable elevation. The Schmidelias which are shrubby very ramous plants, sometimes becoming small trees, and bear a small red berry, are always so far as I have seen, jungle plants. Our only species of Cupania, C. canescens, is never I believe met with in cultivation, but is a common plant in subalpine jungles and has a wide distribution over India.
Properties and Uses. This family is remarkable on account of the leaves and bark, and even the fruit of some of its species being possessed of active medicinal or even decidedly poisonous properties, while the fruit of others, is highly esteemed for the dessert. Among these last are the Litchi, the Rambutan, the Longan, and a variety of others. The root of Cardiospermvm Halicacabum is aperient. The succulent capsule of Sapendus emarginatus, the common Soapnut, is considered by the native practitioners an excellent expectorant, an opinion which Dr. Ainslie thinks correct, it is also, as the English name implies, a useful detergent, and much used as such. The root Schmidelia serrata (Ornithrophe Roxb.) is according to Roxburgh a mild astringent, and prescribed by the natives in cases of diarrhasa, while the small red subacid berries are eaten by the natives. So also is the subacid aril of the seed of Schleichera trijuga, a large and handsome tree, not uncommon in our jungles; and from the seed themselves a lamp oil is expressed in Malabar.
Remarks on Genera and Species. The discrimination of the genera of this order is often most difficult, unless the specimens under examination are very complete. When furnished with both flowers and fruit the characters are more easily made out, but without fruit the reverse is often the case, as for example, between some of the species of Cupania and Sapindus. With these exceptions the Sapindaceae of the Indian Peninsula are for the most
