6. GYNOBASEOSE. Carpels not exceeding five, diverging at the base, arranged in a single row around an elevated axis, or gynobase (337. 6). Stamens usually separate from the calyx (hypogynous* of authors).

7. APOCARPOSE. None of the characters of the other groups, but with the carpels distinct; or separable by their faces; or solitary.

N. B. In the succeeding pages the first column contains a brief character of the Natural Order; the second the name of the Order; the
third its sensible properties, with some officinal example in italics within brackets, when any is to be found. When the third column is blank
nothing is known of the sensible property.

The orders marked (eighty-nine in number) are the most important, and those with which all students should be made acquainted.

Floral envelopes in threes or fives. Sap transparent

Floral envelopes in twos or fours. Sap usually milky

Embryo enclosed in a vitellus. Floaters

Ovaries concealed in a fleshy receptacle. Floaters Stamens perigynous

Polyspermous, with basal many-seeded placenta.

? Dionæa.

* It is to be remarked, that in this and the following characters, the term hypogynous is applied to all plants with a superior ovary, and the term epigynous to those with an inferior ovary. Therefore the term hypogynous is here equal to perigynous and hypogynous of other writers (313.). † Apocarpous always means single carpels, as well as a collection of separate or separable ones.

Alliance 2. Anonales. Apocarpous woody plants.

Flowers unisexual, three-lobed. Stamens columnar 6. Myristicacea

Leaves with stipules, without dots

Leaves with stipules and transparent dots

Flowers hermaphrodite, three-parted.

7. Magnoliacea

Stamens

usually distinct

Alliance 3. Umbellales. hollow. Carpels two

Flowers epigynous (314.), arranged in umbels.

in umbels. Stems usually

Carpels more than two

12. Araliacea

Alliance 5. Berberales. Apocarpous, with the valves of the anthers curved backwards.

16. Berberaceæ

. Acid, astringent (Berberry).' Dye yellow.

Magnolia;

A group in appearance truly natural, and agreeing with its technical character in all respects, with the exception of Nelumbium, which has no albumen; and the genus Berberis, in which the embryo is much larger in proportion to the albumen than in any other instance. Dionæa is a paradoxical plant, whose true affinities are doubtful. It appears most nearly allied to Ranunculaces, or Cephalotaceæ, and is perhaps analogous to the genus Adrastaa among

Anonaceæ.

Some connecting links are obviously wanting in this group; and, until it is well considered, it will appear less natural than it really is, especially if we compare such plants as the Vine with the Crowfoot, or either with Sarracennia. Nevertheless, it is to be observed, that it very nearly agrees with De Candolle's Thalamiflorous subclass, and that the mutual affinities of the alliances may be demonstrated. Take Anonales and Ranales for the centre of the circumference of a circle composed of the six foregoing alliances:

and the relative position of the alliances will be thus:

Umbellales Berberales.

There is no difficulty whatever in pointing out the various gradations that connect the genera belonging to the orders comprehended in the Albuminous group; but this is not the place for the discussion. The most paradoxical part of the combination is the union of baccatefruited with dry-fruited plants: but even Vitaceae pass into Umbellacea through Leea; and the petals of the genus Vitis itself are inflected at the points, in the way of Umbellaceæ.

These plants seem to be all connected by a general natural relationship; and yet it is extremely difficult to fix the limits of their alliances. They appear to be connected with the syncarpous group through Melastoma and Lythraceæ, and with the albuminous group by the genus Eupomatia, and even by Cactaceae, which evidently touch upon Gros

sulaceæ. They also pass into Monopetala by Melastomacea, which join them with Gentianaceæ. I entertain no doubt about this being the true position of Begoniaceæ, among which we may expect to find some climbing species.