not say prodigious, where faithful and probable accounts can be obtained, but superstitious), together with the experiments of natural magic. For we would not accustom philosophy in her infancy, whose very nurse is natural history, to old wives' tales. A time may come (after a deeper investigation of nature) when such matters may be lightly touched upon, so as to extract and lay up for use such natural knowledge as may lurk in their dregs, but till then they are to be put aside. In like manner the experiments of natural magic are to be diligently and rigidly sifted before their adoption, especially those which are wont to be derived from vulgar sympathies and antipathies, owing to the indolence and credulity of both believers and in

ventors.

It is no slight matter to have thus relieved natural history of these three vanities, which might otherwise have hereafter filled volumes.. Nor is this all; for it is as essential to a great work, that that which is admitted be briefly described, as that the superfluous should be rejected, although it must be obvious that this chastened and precise style must afford less pleasure both to the reader and to the author. But it is ever to be repeated, that the object is to prepare a mere granary and warehouse, in which no one is to loiter or dwell for amusement, but only to visit as occasion may require, when any thing is wanted for the work of the interpreter, which follows next in order.

Iv. One thing, above all others, is requisite for the history we design; namely, that it be most extensive, and adapted to the extent of the universe. For the world is not to be narrowed down to the measure of the understanding (as has hitherto been done), but the understanding is to be expanded, and opened for the admission of the actual representation of the world as it is. The maxim of examining little and pronouncing on that little has ruined every thing. Resuming then our late partition of natural history, into that of generation, præter-generation, and the arts, we divide the first into five parts: 1. The history of the sky and heavenly bodies. 2. Of meteors and the regions (as they are termed) of the air, that is to say, its division from the moon to the earth's surface, to which division we assign every kind of comet either superior or inferior (however the actual fact may be), for the sake of method. 3. The history of the earth and sea. 4. Of the elements, as they are called, flame or fire, air, water, and earth; considering them, however, under that name, not as the

first principles of things, but as forming the larger masses of natural bodies. For natural objects are so distributed, that the quantity or mass of certain bodies throughout the universe is very great, owing to the easy and obvious material texture required for their conformation, whilst the quantity of others is but small and sparingly supplied, the material, being of a diversified and subtile nature, having many specific qualities, and being of an organised construction, such as the different species of natural objects, namely, metals, plants, and animals. We are wont, therefore, to call the former greater colleges, and the latter lesser colleges. The fourth part of our history then is of the former, under the name of elements. Nor is there any confusion between this and the second or third parts, although we have spoken of air, water, and earth in each. For in the second and third they are spoken of as integral parts of the world, and in relation to the creation and configuration of the universe; but in the fourth is contained the history of their own substance and nature, as displayed in the homogeneous parts of each, and not referred to the whole. Lastly, the fifth part of natural history contains the lesser colleges or species, upon which alone natural history has hitherto been chiefly occupied.

As to the history of præter-generation, we have already observed that it may, with the greatest convenience, be combined with that of generation, including that which is prodigious only, not natural. For we reserve the superstitious history of miracles (such as it may be) for a separate treatise, nor is it to be undertaken immediately, but rather later, when more way shall have been made in the investigation of nature.

We divide the history of the arts, and of nature's course diverted and changed by man, or experimental history, into three parts. For it is derived either: 1. from the mechanical arts; or, 2. from the practical part of the liberal sciences; or, 3. from various practical applications and experiments, which have not yet been classed as a peculiar art, nay, sometimes occur in every day's experience and require no such art. If, then, a history be completed of all these which we have mentioned, namely, generation, præter-generation, the arts and experiments, nothing appears omitted for preparing the senses to inform the understanding, and we shall no longer dance, as it were, within the narrow circles of the enchanter, but extend our march round the confines of the world itself.

v. Of those parts into which we have divided natural history, that of the arts is the most useful, since it exhibits bodies in motion, and leads more directly to practice. Besides this, it lifts the mask and veil, as it were, from natural objects, which are generally concealed or obscured under a diversity of forms and external appearance. Again, the attacks of art are assuredly the very fetters and miracles of Proteus, which betray the last struggle and efforts of nature. For bodies resist destruction or annihilation, and rather transform themselves into various shapes. The greatest diligence, therefore, is to be bestowed upon this history, however mechanical and illiberal it may appear, laying aside all fastidious arrogance.

Again, amongst the arts those are preferable which control, alter, and prepare natural bodies, and the materials of objects, such as agriculture, cookery, chymistry, dyeing, manufactures of glass, enamel, sugar, gunpowder, fireworks, paper, and the like. There is less use to be derived from those which chiefly consist in a delicate motion of the hands, or of tools, such as weaving, carpentry, architecture, mill and clock work, and the like; although the latter are by no means to be neglected, both on account of their frequently presenting circumstances tending to the alteration of natural bodies, and also on account of the accurate information they afford of translatitious motion, a point of the greatest importance in many inquiries.

One thing, however, is to be observed and well remembered in this whole collection of arts, namely, to admit not only those experiments which conduce to the direct object of the art, but also those which indirectly occur. For instance, the changing of the lobster or a crab when cooked from a dark to a red colour has nothing to do with cookery, yet this instance is not a bad one in investigating the nature of redness, since the same thing occurs in baked bricks. So, again, the circumstance of meat requiring less time for salting in winter than in summer is not only useful information to the cook for preparing his meat, but is also a good instance to point out the nature and effect of cold. He therefore will be wonderfully mistaken, who shall think that he has satisfied our object when he has collected these experiments of the arts for the sole purpose of improving each art in particular. For although we do not by any means despise even this, yet our firm intention is to cause the streams of every species of mechanical experiment to flow from all quarters into the ocean of philosophy. The

choice of the most important instances in each (such as should be most abundantly and diligently searched and, as it were, hunted out) must be governed by the prerogative instances.

VI. We must here allude to that which we have treated more at length in the ninety-ninth, one hundred and nineteenth, and one hundred and twentieth aphorisms of the first book, and need now only briefly urge as a precept, namely, that there be admitted into this history, 1. the most common matters, such as one might think it superfluous to insert from their being so well known; 2. base, illiberal, and filthy matters (for to the pure every thing is pure, and if money derived from urine be of good odour, much more so is knowledge and information from any quarter), and also those which are trifling and puerile; lastly, such matters as appear too minute, as being of themselves of no use. For (as has been observed) the subjects to be treated of in this history are not compiled on their own account, nor ought their worth, therefore, to be measured by their intrinsic value, but by their application to other points, and their influence on philosophy.

VII. We moreover recommend that all natural bodies and qualities be, as far as is possible, reduced to number, weight, measure, and precise definition; for we are planning actual results and not mere theory; and it is a proper combination of physics and mathematics that generates practice. The exact return and distances of the planets, therefore, in the history of the heavens, the circumference of the earth, and the extent of its surface compared with that of water, in the history of the earth and sea, the quantity of compression which the air will suffer without any powerful resistance in the history of air, the quantity by which one metal exceeds another in weight, in that of metals, and a number of like points are to be thoroughly investigated and detailed. When, however, the exact proportions cannot be obtained, recourse must be had to those which are estimated or comparative. Thus, if we distrust the calculations of astronomers as to distances, it may be stated that the moon is within the shadow of the earth, and Mercury above the moon, &c. If mean proportions cannot be had, let extremes be taken, as that the feeblest magnet can raise iron of such a weight compared with its own, and the most powerful sixty times as much as its own weight, which I have myself observed in a very small armed magnet. For we know very well that determinate instances

do not readily or often occur, but must be sought after as auxiliary, when chiefly wanted, in the very course of interpretation. If, however, they casually occur, they should be inserted in natural history, provided they do not too much retard its progress.

VIII. With regard to the credit due to the matters admitted into our history, they must either be certain, doubtful, or absolutely false. The first are to be simply stated, the second to be noted with a "report states," or "they say," or "I have heard from a person worthy of credit," and the like. For it would be too laborious to enter into the arguments on both sides, and would too much retard the author, nor is it of much consequence towards our present object, since (as we have observed in the hundred and eighteenth aphorism of the first book) the correctness of the axioms will soon discover the errors of experiment, unless they be very general. If, however, there be any instance of greater importance than the rest, either from its use, or the consequences dependent upon it, then the author should certainly be named, and not barely named, but some notice should be taken as to whether he merely heard or copied it (as is generally the case with Pliny), or rather affirmed it of his own knowledge, and also whether it were a matter within his own time or before it, or whether such as if true must necessarily have been witnessed by many; or, lastly, whether the author were vain and trifling, or steady and accurate, and the like points, which give weight to testimony. Lastly, those matters which are false, and yet have been much repeated and discussed, such as have gained ground by the lapse of ages, partly owing to neglect, partly to their being used as poetical comparisons, for instance, that the diamond overpowers the magnet, that garlick enervates, that amber attracts every thing but the herb basil, &c. &c. all these ought not to be silently rejected, but expressly proscribed, that they may never trouble science

more.

It will not, however, be improper to notice the origin of any fable or absurdity, if it should be traced in the course of inquiry, such as the venereal qualities attributed to the herb satyrium from its roots bearing some resemblance to the testicles. The real cause of this formation being the growth of a fresh bulbous root every year, which adheres to that of the preceding year, and produces the twin roots, as is proved by the firm juicy appearance which the new root always presents, whilst the old one is withered and