Morison, in a second edition of the Hortus Regius Blesensis, gave, in 1669, the rudiments of a method founded on the fruit; but the remembrance of this proposal has been obliterated by the splendour of Ray, who published his Methodus Plantarum Nova, in tables. He divided the vegetable kingdom into trees, shrubs, and herbaceous plants; classing with the latter the suf-frutices, the lesser shrubs. His great object was to connect the natural families, or the plants united by a similarity of fructification 'and general habit; but his genera were more loosely and imperfectly grouped; and in this first imperfect attempt we wanted both descriptions and synonyms. His system appeared in a more accurate and complete state in his Synopsis of British Plants, and his History of Plants. Two editions of the Synopsis were, we believe, published during his life. The most complete is the third, published in two vols. 8vo, 1724, near twenty years after his death.

It belongs to the general history of botany to pursue the various modes of classification proposed and adopted by different authors. It is our object in this sketch, chiefly to point out the progress of Medical Botany, and of the arrangement of plants in natural families, the means as we think of improving it. Though we greatly admire the ingenuity of the system of Tournefort, and, with all their imperfections, find considerable merit in those of Sauvages and Haller, we must chiefly confine ourselves to Linnaeus. From the year 1735 to 1737, the different publications which announced his system appeared; and it was gradually improved to the last period of his life. It is styled the sexual system, because he supposed plants distinguished by sexes, the antherae being in his opinion the male, and the pistils the female organs. From the number, situation, and connection of the former, his classes are chiefly denominated and distinguished; from the number of the pistils in general, the orders. It is confessedly an artificial system, but many of the families of plants are preserved undisturbed. Whatever are its imperfections must be now overlooked, since in the present vast extent of botanical science, this system only is commensurate with it. Various attempts have been made to correct and improve it. Thunberg attempted to render it more simple by diminishing the number of classes; and Gmelin has added to them. Botanists, however, have received each offered improvement with caution; and, as Dr. Smith has almost scrupulously adhered to the arrangement of the Swede, and Wildenow in his new edition of the species followed it without alteration, there is little probability of its being now disturb.

A natural method, however, appeared always from the confession of Linnaeus himself, the first and last object of the botanist; and in his Philosophia Botanica, a work published in 1751, often highly praised, but scarcely ever in proportion to its merits, he introduces what he styles ' Fragmenta Methodi Naturalis,' viz. fifty-seven natural orders or families of plants. We mean not to say that this was the first example of a natural method. Morison's and Ray's were certainly such; and previous to these, in 1626, Laurenberg published at Rostock his Botanotheca, in which he distinguished ten natural families of plants with tolerable accuracy. Ray's four classes, taken from the cotyledons, branched out into twenty-nine families, the greater number of which were natural groups. The cryptogamiae and capillary plants we of course exclude. Besides our countryman, Reyes of Leyden, in 1740, gave an elegant natural arrangement; Haller, a more laboured and a less useful one; and Wachendorff of Utrecht, a system not inelegant, but deformed by titles peculiarly complicated and compounded. At a later period, viz. in 1766,'crantz of Vienna published his Institutiones Rei Herbariae; in which, by an arrangement neither neat nor convenient, he has, however, retained many of the more natural groups or families-of plants. About the same time, Adanson formed his natural classes; but these are so strictly natural, that his definitions become descriptions, tedious and useless from their extent, since he includes a similarity in every part. Gaertner, in his arrangement from the seeds, without pretensions to natural orders, has formed numerous natural groups which well merit the attention of the botanist. The most successful modern attempts are those of Jussieu and Murray; the former more strictly botanical, the other subservient to his excellent work on the Materia Medica (Apparatus Medicaminum). Anthony de Jussieu is the nephew of Bernard, demonstrator in the royal gardens at Paris, to whom the merit of the first attempt is due. It was from Bernard Jussieu that we received the first systematic work of this kind in 1789, viz. Genera Plantarum secundum Ordinesnaturalesdis-posita; and from him Adanson, who was his pupil, probably derived his ideas. This was the foundation of the last and best natural botanical system that has appeared, viz. Tableau du Regne Vegetale selon la Methode de Jussieu, par E. P. Ventenat. The system of Murray we shall soon notice.

We have enlarged on this part of the science of botany, because we perceive a very intimate connection between-the natural orders and the medical properties of plants; and because this part of the subject has not yet received the attention which it so justly merits in this point of view. The Linnaean system we have said is artificial; but it is not to be rejected on this account. With its aid we can at any time distinguish a plant, convey an accurate idea of its general form, and arrange it so as that future observations may be directed to their proper object; in short, for the purpose of discrimination, and the reference of synonyms, it is highly valuable. As an artificial method also it is always complete, so far as our knowledge extends. There are no lacunae to be filled up, there are no links to be supplied. A natural method cannot possess such advantages. In reality, the further our knowledge extends, the more imperfect it becomes, for 'nature makes no leaps;' and the lacunae, which, in the present state of science, give limits to the orders, will, when supplied, no longer enable us to discriminate them. In a general view, therefore, the clas-m m 2 sification and nomenclature of plants may be advan-geously supplied by the Linnaean system; the union of properties and virtues by the natural families.