amazement and delight, are so clear and interesting, that we shall present them to the reader.

But

'It is evident that sight and touch are senses by which the relations of space are perceived, principally or entirely. It does not appear that an odour, or a feeling of warmth or cold, would, independently of experience, suggest to us the conception of a space surrounding us. when we see objects, we see that they are extended and occupy space : and when we touch them, we feel that they are in a space in which we also are. We have before our eyes any object, for instance, a board covered with geometrical diagrams; and we distinctly perceive, by vision, those lines of which the relations are the subjects of our mathematical reasoning. Again, we see before us a solid object; a cubical box for instance; we see that it is within reach; we stretch out the hand and perceive by the touch that it has sides, edges, corners, which we had already perceived by vision.

Probably most persons do not generally apprehend that there is any material difference in these two cases; that there are any different acts of mind concerned in perceiving by sight a mathematical diagram upon paper, and a solid cube lying on a table. Yet it is not difficult to show that, in the latter case at least, the perception of the shape of the object is not immediate. A very little attention teaches us that there is an act of judgment as well as a mere impression of sense requisite, in order that we may see any solid object. For there is no visible appearance which is inseparably connected with solidity. If a picture of a cube be rightly drawn in perspective, and skilfully shaded, the impression upon the sense is the same as if it were a real cube. The picture may be mistaken for a solid object. But it is clear that in this case the solidity is given to the object by an act of mental judgment. All that is seen is outline and shade, figures and colors on a flat board. The solid angles and edges, the relation of the faces of the figure by which they form a cube is a matter of inference. This, which is evident in the case of the pictured cube, is true in all vision whatever. We see a scene before us on which are various figures and colors, but the eye cannot see more. It sees length and breadth, but no third dimension. In order to know that there are solids, we must infer as well as see. And this we do readily and constantly; so familiarly, indeed, that we do not perceive the operation. Yet we may detect this latent process in many ways; for instance, by attending to cases in which the habit of drawing such inferences misleads us. Most persons have experienced this delusion in looking at a scene in a theatre, and especially that kind of scene which is called a diorama, when the interior of a building is represented. In these cases, the perspective representations of the various members of the architecture and decoration impress us almost irresistibly with the conviction that we have before us a space of great extent and complex form, instead of a flat painted canvass. Here, at least, the space is our own creation; but it is manifestly created by the same act of thought as if we were really in the palace or the cathedral of which the halls and aisles thus seem to enclose us. And the act by which we thus create space of

three dimensions out of visible extent of length and breadth, is constantly and imperceptibly going on. We are perpetually interpreting in this manner the language of the visible world. From the appearances of things which we directly see, we are constantly inferring that which we cannot directly see, their distance from us, and the position of their parts.

The characters which we thus interpret are various. They are, for instance, the visible forms, colors, and shades of their parts, understood according to the maxims of perspective (for of perspective every one has a practical knowledge as every one has of grammar); the effort by which we fix both our eyes on the same object, and adjust each eye to distinct vision; and the like. The right interpretation of the information which such circumstances give us respecting the true forms and distances of things is gradually learned; the lesson being begun in our earliest infancy, and inculcated upon us every hour during which we use our eyes. The completeness with which the lesson is mastered is truly admirable; for we forget that our conclusion is obtained indirectly, and mistake a judgment on evidence for an intuitive perception. We see the breadth of the street as clearly and readily as we see the house on the other side of it; and we see the house to be square, however obliquely it be presented to us. This, however, by no means throws any doubt or difficulty on the doctrine that in all these cases we do interpret and infer. The rapidity of the process, and the unconsciousness of the effort, are not more remarkable in this case than they are when we understand the meaning of the speech which we hear or of the book which we read. In these latter cases we merely hear noises or see black marks, but we make, out of these elements, thought and feeling, without being aware of the act by which we do so. And by an exactly similar process we see a variouslycolored expanse, and collect from it a space occupied by solid objects. In both cases the act of interpretation is become so habitual that we can hardly stop short at the mere impression of sense.

But yet there are various ways in which we may satisfy ourselves that these two parts of the process of seeing objects are distinct. To separate these operations is precisely the task which the artist has to execute in making a drawing of what he sees. He has to recover the consciousness of his real and genuine sensations, and to discern the lines of objects as they appear. This at first he finds difficult; for he is tempted to draw what he knows of the forms of visible objects, and not what he sees: but as he improves in his art, he learns to put on paper what he sees only, separate from what he infers, in order that thus the inference, and with it a conception like that of the reality, may be left to the spectator. And thus the natural process of vision is the habit of seeing that which cannot be seen; and the difficulty of the art of drawing consists in not seeing more than is visible.' -Ib. pp. 108-111.

Nor are other portions of the work less deserving of commendation. The chapter on the 'Successive Attempts at the Scien'tific Application of the Idea of a Medium,' and that on the

'Measure of Secondary Qualities' (though rather too historical, as already remarked, for the present work), are full of interesting matter. One of the best portions of the work, in our opinion, are the first and second chapters on the Philosophy of Chemistry,' entitled 'Attempts to Conceive Elementary 'Composition,' and the Establishment and Development of 'the Idea of Chemical Affinity.' The manner in which the Conception of Chemical Affinity evolved itself, after many crude and fanciful hypotheses, is very clearly and succinctly described. From the former of these two chapters we extract the following paragraphs. There are few things more amusing than exploded hypotheses, once gravely propounded and zealously maintained.

The mode in which elements form the compound bodies and determine their properties was at first, as might be expected, vaguely and variously conceived. It will, I trust, hereafter be made clear to the reader that the relation of the elements to the compound involves a peculiar and appropriate Fundamental Idea, not susceptible of being correctly represented by any comparison or combination of other ideas, and guiding us to clear and definite results only when it is illustrated and nourished by an abundant supply of experimental facts. But at first the peculiar and special notion which is required in a just conception of the constitution of bodies was neither discerned nor suspected; and up to a very late period in the history of chemistry, men went on attempting to apprehend the constitution of bodies more clearly by substituting for this obscure and recondite idea of elementary composition, some other idea more obvious, more luminous, and more familiar, such as the ideas of resemblance, position, and mechanical force. We shall briefly speak of some of these attempts, and of the errors which were thus introduced into speculations on the relations of elements and compounds.

Compounds assumed to resemble their Elements.

The first notion was that compounds derive their qualities from their elements by resemblance :—they are hot in virtue of a hot element, heavy in virtue of a heavy element, and so on. In this way the doctrine of the four elements was framed; for every body is either hot or cold, moist or dry; and by combining these qualities in all possible ways, men devised four elementary substances, as has been stated in the history.

This assumption of the derivation of the qualities of bodies from similar qualities in the elements was, as we shall see, altogether baseless and unphilosophical, yet it prevailed long and universally. It was the foundation of medicine for a long period both in Europe and Asia; disorders being divided into hot, cold, and the like; and remedies being arranged according to similar distinctions. Many readers will recollect, perhaps, the story of the indignation which the Persian physicians felt towards the European, when he undertook to cure the ill effects of cucumber upon the patient by means of mercurial medicines, for cucumber, which is cold, could not be counteracted, they

maintained, by mercury, which in their classification is cold also. Similar views of the operation of medicines might easily be traced in our own country, A moment's reflection may convince us that when drugs of any kind are subjected to the chemistry of the human stomach, and thus made to operate on the human frame, it is utterly impossible to form the most remote conjecture what the result will be from any such vague notions of their qualities as the common use of our senses can give. And in like manner the common operations of chemistry give rise in almost every instance to products which bear no resemblance to the materials employed. The results of the furnace, the alembic, the mixture, frequently bear no visible resemblance to the ingredients operated upon. Iron becomes steel by the addition of a little charcoal; but what visible trace of the charcoal is presented by the metal thus modified? The most beautiful colors are given to glass and earthenware by minute portions of the ores of black or dingy metals, as iron and manganese. The worker in metal, the painter, the dyer, the vintner, the brewer, all the artisans in short who deal with practical chemistry, are able to teach the speculative chemist that nothing can be so false as to expect that the qualities of the elements shall be still discoverable, in an unaltered form, in the compound. This first rude notion of an element, that it determines the properties of bodies by resemblance, must be utterly rejected and abandoned before we can make any advance towards a true apprehension of the constitution of bodies.'-Ib. pp. 362-364.

Compounds assumed to be determined by the figure of Elements.

I pass over the fanciful modes of representing chemical changes which were employed by the alchemists; for these strange inventions did little in leading men towards a juster view of the relations of elements to compounds. I proceed for an instant to the attempt to substitute another obvious conception for the still obscure notion of elementary composition. It was imagined that all the properties of bodies and their mutual operations might be accounted for by supposing them constituted of particles of various forms, round or angular, pointed or hooked, straight or spiral. This is a very ancient hypothe sis, and a favorite one with many casual spectators in all ages. Thus Lucretius undertakes to explain why wine passes rapidly through a sieve and oil slowly, by telling us that the latter substance has its parcles either larger than those of the other, or more hooked and interwoven together. And he accounts for the difference of sweet and bitter by supposing the particles in the former case to be round and smooth, in the latter sharp and jagged. Similar assumptions prevailed in modern times on the revival of the mechanical philosophy, and constitute a large part of the physical schemes of Descartes and Gassendi. They were also adopted to a considerable extent by the chemists. Acids were without hesitation assumed to consist of sharp, pointed parti. cles; which I hope,' Lemmery says, 'no one will dispute, seeing every one's experience does demonstrate it: he needs but taste an acid to be satisfied of it, for it pricks the tongue like anything keen and finely Such an assumption is not only altogether gratuitous and useless, appears to be founded in some degree upon a confusion in the

cut.

but

metaphorical and literal use of such words as keen and sharp. The assumption once made, it was easy to accommodate it, in a manner equally arbitrary, to other facts. A demonstrative and convincing proof that an acid does consist of pointed parts is, that not only all acid salts do crystallize into edges, but all dissolutions of different things, caused by acid liquors, do assume this figure in their crystallization. These crystals consist of points differing both in length and bigness one from another, and this diversity must be attributed to the keener or blunter edge of the different sorts of acids: and so likewise this difference of the points in subtlety is the cause that one acid can penetrate and dissolve with one sort of mixt, that another can't rarify at all. Thus vinegar dissolves lead, which aqua fortis can't: aqua fortis dissolves quicksilver, which vinegar will not touch; aqua regalis dissolves gold, whereas aqua fortis cannot meddle with it; on the contrary, aqua fortis dissolves silver, but can do nothing with gold, and so

of the rest.'

The leading fact of the vehement combination and complete union of acid and alkali readily suggested a fit form for the particles of the latter class of substances. This effect,' Lemery adds,' may make us reasonably conjecture that an alkali is a terrestrious and solid matter whose forms are figured after such a manner that the acid points entering in do strike and divide whatever opposes their motion.' And in a like spirit are the speculations in Dr. Mead's Mechanical Account of Poisons (1745). Thus he explains the poisonous effect of corrosive sublimate of mercury, by saying that the particles of the salt are a kind of lamellæ or blades to which the mercury gives an additional weight. If resublimed with three-fourths the quantity of mercury, it loses its corrosiveness (becoming calomel), which arises from this, that in sublimation the crystalline blades are divided every time more and more by the force of the fire;' and 'the broken pieces of the crystals uniting into little masses of different figures from their former make, those cutting points are now so much smaller that they cannot make wounds deep enough to be equally mischievous and deadly and therefore do only vellicate and twitch the sensible membranes of the stomach.'

Among all this very fanciful and gratuitous assumption we may notice one true principle clearly introduced, namely, that the suppositions which we make respecting the forms of the elementary particles of bodies and their mode of combination must be such as to explain the facts of crystallization, as well as of mere chemical change. This principle we shall hereafter have occasion to insist upon further.

I now proceed to consider a more refined form of assumption respecting the constitution of bodies, yet still one in which a vain attempt is made to substitute for the peculiar idea of chemical composition a more familiar mechanical conception.

Compounds assumed to be determined by the Mechanical Attraction of the Elements. When, in consequence of the investigations and discoveries of Newton and his predecessors, the conception of mechanical force had become clear and familiar, so far as the action of external forces upon a body was concerned, it was very natural that the mathematicians who had pursued this train of speculation should attempt to