"' Stigma downy. Without the cha racter of the foregoing section. Phase olus, Dolichus, Orobus, Pisum, Lathyrus, Vicia, to which Dr. Smith has added Eroum, after separating from the latter some species erroneously referred to it. See Flora Britanica, 776.
*** Legume imperfectly divided into two cells. Always without the charac. ter of the preceding sections. Biserrula, Phaca, Astragalus, the last a very exten sive and intricate genus.
**** Legume with scarcely more than one seed. Psoralia, Trifolium, the latter a very irregular genus in character, though distinct in habit, sufficiently known for its agricultural uses.
**•** Legume composed of single valved joints, which are rarely solitary. Hedysarum, Hippocrepis, Coronilla, Smi thia, the latter furnished with irritable leaves like the true Sensitive plant or Mimosa.
****** Legume of one cell with seve ral seeds. Many species of Trifolium pro perly belong here, and have been sepa rated by some authors under the name of Melilotus: also the valuable Indigofera, with Cytisus, Robinia, Lotus, and Medi cago.
XVIII. Polyadelphia. Stamens united Xviii. Polyadelphia. Stamens united by their filaments into more than two parcels. Orders characterised by the number or insertion of their stamens. In this class Dr. Smith has made many cor rections, and the orders in his Introduc tion to Botany stand as folloWs.
1. Dude candri a. Stamens, or rather an thers, from twelve to twenty or twenty live. Their filaments unconnected with the calyx. Theobroma, the chocolate tree, Bubroma, Abroma, Monsonia, and Citrus.
2. Icosandria. Stamens numerous, their filaments inserted into the calyx, in seve ral parcels of course, as Melaleuca, a fine aromatic genus, principally from New Holland.
3 Polyandria. Stamens very numerous, unconnected with the calyx. Ifypericum is the principal gems here.
XIX. AS'ungenesia. Anthers united into a tube. Flowers compound. This is en tirely a natural class, and its orders like wise are founded on natural characters.
1. Polygamia i-Equalis. Every one of the florets which constitute the compound flowers is, in this order, perfect within itself, having perfect stamens, and pistil with one seed. The florets are either ligulate, as in the dandelion ; tubular, forming a globose head, as in the thistle ; or tubular, and level at the top , or dis coid, as in lavender cotton, eantolina.
2. Polygamia Superflua. Florets of the disk like the discoid, flowers of the last or der, and, like them, perfect within them selves. Those of the margin furnished with pistils only, but all the florets pro duce perfect seed. In this order the mar
ginal florets are sometimes minute and inconspicuous, but they are for the most part ligulate, and form diverging rays, as in the daisy, aster, chrysanthemum, &c.
3. Polygamia Frustranea. Differs from the last order only in having the florets of the margin abortive or neuter ; in the for mer case there are no rudiments of a pis til in these florets, as in centaurea ; or there are abortive pistils, as in the sun flower. This order is considered by Dr. Smith as not essentially different from the last.
4. Polygamia necessaria. Florets of the disk furnished with perfect stamens only, those of the margin with perfect pistils only, as in the garden marigold, calendula.
5. Polygamia Segregata. Several flow ers, either simple or compound, with unit ed anthers, and a partial calyx, all in eluded in one general calyx, as die globe thistle, &c.
Another order follows in Lintueus, Monogamia, consisting of simple flow ers with united anthers; but this order is now generally abolished. The cireum.
stance of the union oldie anthers in sim ple flowers being extremely various and uncertain, though in compound ones scarcely liable to any exception.
XX. Gynandria. Stamens inserted either upon the style or germen. Such is the true idea of this class, and its charac ter, thus understood, is as much founded in nature and reality as that of any other ; by which we do not mean, that the class is a natural one, like the 19th, as it, in fact, comprises several natural families, whose allies may happen to be in other classes. Linnaeus, in his idea of this class, has understood as belonging to it, many plants, whose stamens did not really grow out of the germen, as the passion-flower, the sisyrinchium, &c. Ilence Thumberg, and some other bota nists have judged the class altogether un tenable. In the orders, some alterations have recently been made by Dr. Smith, the reasons for which are more fully par ticularized in his Introduction to Botany, than we have room here to explain. These orders are distinguished by the number of stamens. Monandria, the first of them, contains almost all the Orchis tribe. To the fifth, Pentandria, Dr. Smith refers many of the natural family of Contortx, as Pergularia, Cynanchum, and Asclepias, a curious tribe, the structure of whose organs of impregnation is extremely puz zling even to the botanical adept. They have hitherto been placed in the fifth class, and some have thought they should be referred to the tenth. In the sixth or der of this class, Hexandria, we find the aristolochia, or birth-wort.