steps of classificatory knowledge and skill, as he became acquainted with a greater number of facts.

The acknowledged superiority of Linnæus in the knowledge of the matter of his science, induced other persons to defer to him in what concerned its form; especially when his precepts were, for the most part, recommended strongly both by convenience and elegance. The trivial names of the Species Plantarum were generally received; and though some of the details may have been altered, the immense advantage of the scheme ensures its permanence.

Sect. 4.-Linnæus's Artificial System.

WE have already seen, that, from the time of Casalpinus, botanists had been endeavouring to frame a systematic arrangement of plants. All such arrangements were necessarily both artificial and natural: they were artificial, inasmuch as they depended upon assumed principles, the number, form, and position of certain parts, by the application of which the whole vegetable kingdom was imperatively subdivided; they were natural, inasmuch as the justification of this division was, that it brought together those plants which were naturally related. No system of arrangement, for instance, would have been tolerated which, in a great proportion of cases, separated into distant parts of the plan the different species of the same genus. As far as the main body of the genera, at least, all systems are natural.

But beginning from this line, we may construct our systems with two opposite purposes, according as we endeavour to carry our assumed principle of division rigorously and consistently through the system, or as we wish to associate natural families of a wider kind than genera. The former propensity leads to an artificial, the latter to a natural method. Each is a System of Plants; but in the first, the emphasis is thrown on the former word of the title, in the other, on the latter.

The strongest recommendation of an artificial sys

tem, (besides its approaching to a natural method,) is, that it shall be capable of easy use; for which purpose, the facts on which it depends must be apparent in their relations, and universal in their occurrence. The system of Linnæus, founded upon the number, position, and other circumstances of the stamina and pistils, the reproductive organs of the plants, possessed this merit in an eminent degree, as far as these characters are concerned; that is, as far as the classes and orders. In its further subdivision into genera, its superiority was mainly due to the exact observation and description, which we have already had to notice as talents which Linnæus peculiarly possessed.

The Linnæan system of plants was more definite than that of Tournefort, which was governed by the corolla; for number is more definite than irregular form. It was more readily employed than any of those which depend on the fruit, for the flower is a more obvious object, and more easily examined. Still, it can hardly be doubted, that the circumstance which gave the main currency to the system of Linnæus was, its physiological signification: it was the Sexual System. The relation of the parts to which it directed the attention, interested both the philosophical faculty and the imagination. And when, soon after the system had become familiar in our own country, the poet of The Botanic Garden peopled the bell of every flower with Nymphs' and 'Swains,' his imagery was felt to be by no means forced and far-fetched.

The history of the doctrine of the sexes of plants, as a point of physiology, does not belong to this place; and the Linnæan system of classification need not be longer dwelt upon for our present purpose. I will only explain a little further what has been said, that it is, up to a certain point, a natural system. Several of Linnæus's classes are, in a great measure, natural associations, kept together in violation of his own artificial rules. Thus the class Diadelphia, in which, by the system, the filaments of the stamina should be bound together in two parcels, does, in fact, contain many genera which are monadelphous, the filaments of the

stamina all cohering so as to form one bundle only; as in Genista, Spartium, Anthyllis, Lupinus, &c. And why is this violation of rule? Precisely because these genera all belong to the natural tribe of Papilionaceous plants, which the author of the system could not prevail upon himself to tear asunder. Yet in other cases Linnæus was true to his system, to the injury of natural alliances, as he was, for instance, in another portion of this very tribe of Papilionacea; for there are plants which undoubtedly belong to the tribe, but which have ten separate stamens; and these he placed in the order Decandria. Upon the whole, however, he inclines rather to admit transgression of art than of nature.

The reason of this inclination was, that he rightly considered an artificial method as instrumental to the investigation of a natural one; and to this part of his views we now proceed.

Sect. 5.-Linnæus's Views on a Natural Method.

THE admirers of Linnæus, the English especially, were for some time in the habit of putting his Sexual System in opposition to the Natural Method, which about the same time was attempted in France. And as they often appear to have imagined that the ultimate object of botanical methods was to know the names of plants, they naturally preferred the Swedish method, which is excellent as a finder. No person, however, who wishes to know botany as a science, that is, as a body of general truths, can be content with making names his ultimate object. Such a person will be constantly and irresistibly led on to attempt to catch sight of the natural arrangement of plants, even before he discovers, as he will discover by pursuing such a course of study, that the knowledge of the natural arrangement is the knowledge of the essential construction and vital mechanism of plants. He will consider an artificial method as a means of arriving at a natural method. Accordingly, however much some of his followers may have overlooked this, it is what Linnæus himself

always held and taught. And though what he executed with regard to this object was but little,12 the distinct manner in which he presented the relations of an artificial and natural method, may justly be looked upon as one of the great improvements which he introduced into the study of his science.

Thus in the Classes Plantarum, (1747,) he speaks of the difficulty of the task of discovering the natural orders, and of the attempts made by others. 'Yet,' he adds, ‘I too have laboured at this, have done something, have much still to do, and shall labour at the object as long as I live.' He afterwards proposed sixty-seven orders, as the fragments of a natural method, always professing their imperfection. 13 And in others of his works 14 he lays down some antitheses on the subject after his manner. 'The natural orders teach us the nature of plants; the artificial orders enable us to recognize plants. The natural orders, without a key, do not constitute a Method; the Method ought to be available without a master.'

That extreme difficulty must attend the formation of a Natural Method, may be seen from the very indefinite nature of the Aphorisms upon this subject which Linnæus has delivered, and which the best botanists of succeeding times have assented to. Such are these; the Natural Orders must be formed by attention, not to one or two, but to all the parts of plants; the same organs are of great importance in regulating the divisions of one part of the system, and of small importance in another part;15—the Character does not constitute the Genus, but the Genus the Character; the Character is necessary, not to make the Genus, but to recognize it. The vagueness of these maxims is easily seen; the rule of attending to all the parts, implies, that we are to estimate their relative importance, either by physiological considerations,

12 The natural orders which he proposed are a bare enumeration of genera, and have not been generally followed.

13 Phil. Bot. p. 80.

14 Genera Plantarum, 1764. See Prælect. in Ord. Nat. p. xlviii. 15 Phil. Bot. p. 172.

(and these again lead to arbitrary rules, as, for instance, the superiority of the function of nutrition to that of reproduction,) or by a sort of latent naturalist instinct, which Linnæus in some passages seems to recognize. "The Habit of a plant,' he says,16 'must be secretly consulted. A practised botanist will distinguish, at the first glance, the plants of different quarters of the globe, and yet will be at a loss to tell by what mark he detects them. There is, I know not what look,-sinister, dry, obscure in African plants; superb and elevated, in the Asiatic; smooth and cheerful, in the American; stunted and indurated, in the Alpine.'

Again, the rule that the same parts are of very different value in different Orders, not only leaves us in want of rules or reasons which may enable us to compare the marks of different Orders, but destroys the systematic completeness of the natural arrangement. If some of the Orders be regulated by the flower and others by the fruit, we may have plants, of which the flower would place them in one Order, and the fruit in another. The answer to this difficulty is the maxim already stated;-that no Character makes the Order; and that if a Character do not enable us to recognize the Order, it does not answer its purpose, and ought to be changed for another.

This doctrine, that the Character is to be employed as a servant and not as a master, was a stumblingblock in the way of those disciples who looked only for dogmatical and universal rules. One of Linnæus's pupils, Paul Dietrich Giseke, has given us a very lively account of his own perplexity on having this view propounded to him, and of the way in which he struggled with it. He had complained of the want of intelligible grounds, in the collection of natural orders given by Linnæus. Linnæus 17 wrote in answer, 'You ask me for the characters of the Natural Orders: I confess I cannot give them.' Such a reply naturally increased Giseke's difficulties. But afterwards, in

16 Phil. Bot. p. 171.

17 Linnæi Prælectiones, Pref. p. xv.