CAUSERIES.

EVERY one has heard of, and marvelled at, the discovery of metals, in the atmosphere of the sun and stars, identical with those common to our earth-a discovery established by means of what is called spectrum analysis: a beam of light decomposed into its constituent colours is made to betray by certain markings the nature of the source whence it proceeds, and of the bodies existing in the atmosphere through which it passes. A recent discovery by M. Janssen has established that not only metals in a gaseous state produce the spectral markings, but that the vapour of water, especially when sufficiently dense, has a similar effect. Three and thirty years ago Sir David Brewster noticed the effect of our atmosphere on the solar spectrum-noticed that when the sun was near the horizon the spectrum was marked with new dark bands. The splendid discovery of Kirchoff and Bunsen, by proving that markings on the spectrum were due to the sun's atmosphere, carried speculation away from the facts observed by Brewster. But M. Janssen saw that between the attempt of Brewster to explain the spectral markings by the action of our atmosphere, and the attempt of Kirchoff to explain them by the sun's atmosphere, there was no irreconcilable discrepancy; both might be true. Both are true. The action of the sun's atmosphere is indubitable. The action of our atmosphere, especially of its vapour, M. Janssen undertakes to prove. He finds certain definite lines and bands on the spectrum, varying in intensity according to the variable position of the sun, that is, according to the greater or less quantity of vapour in the atmosphere its beams have to traverse. He found that as he ascended the Faulhorn (that is in proportion as the solar beams had to traverse a thinner atmosphere), the markings became feebler. I must refer the curious to his note in the Comptes Rendus (Aug. 13, 1866) for a detail of the experiments. It is enough here to say that he finds our atmosphere to have a decided action on the spectrum, but its action is different from that of the sun's atmosphere. The first produces in the red, orange, and yellow of the spectrum ten times as many lines as the second. On the other hand, in the green, blue, and violet the solar lines predominate. Thus our atmosphere acts mainly on the rays which have a long wave, the solar atmosphere on rays which have a short wave. He concludes, moreover, that the effect of the vapour of water is to cause the red rays to be transmitted; which will account for the redness of the rising and the setting sun, since the greater the thickness of the vapour the deeper the redness of the transmitted rays.

M. Janssen's restitution of Brewster's neglected observations prompts me to communicate for the reader's meditation an idea which has latterly been taking shape in my mind connected with the Logic of Science.

Science is distinguished from common knowledge not only by its wider reach and more systematic structure, but also by its employing consciously the artifices which the necessary infirmity of our faculties render indispensable, and which the unscientific mind employs unconsciously. Science not only employs methods, but reflects upon them, and systematises them. One of the great artifices of research is Abstraction. The man of science is conscious of what he is doing when he abstracts certain phenomena from the mass of phenomena presented to observation, and proceeds to deal with those abstracts

as if they were the whole of the reality. The ordinary man does the same thing, but is not conscious of doing it. Why must both make this preliminary departure from the actual facts, in order eventually to understand the facts? Why must the search of truth begin with a falsehood?

The answer is simple. Unless some such beginning is made the search is hopeless. The parrots of Bacon chatter about Observation; but Observation of cases, however patient and prolonged, cannot detect the laws which are enveloped in the cases; because Laws are the constants of phenomena, and these can only be separated from perturbations (due to other Laws) by a process of abstracting all the accidents and individual peculiarities that determine special cases. Observation necessarily includes both ore and dross together, i.e., both the constants, which will be found in every case, and the accidents, which are found only in particular cases. The mineralogist has to separate the ore from the dross; but he must know the one from the other before he can separate them. How does the philosopher know the law? Simple observation cannot discriminate between the constant and the accident, but it can and does furnish, through comparison, the data for such discrimination. Even the laws of Motion and Gravitation, universal as they are, could never have been discovered by observation of cases of motion and gravity; but by a process of abstraction, which eliminated all consideration of the variable resistances. The laws of chemical affinity were still more dependent on a process of abstraction, each element having to be forcibly torn away from every other, and studied in itself. If Kepler and Newton had not boldly set aside all consideration of planetary perturbations, they could never have established their laws. But this was a preliminary falsifaction; and it was only rectified by their successors, who deduced the perturbations from secondary gravitations. Again, in establishing mechanical laws philosophers always falsify the facts to the extent of assuming that the lines of direction are undisturbed, and that materials are uniform and perfect; but the practical mechanic has to rectify this statement of ideal facts by reintroduction of the discarded facts: he has, at peril of ignoble failure, to ascertain what are the actual lines of direction as produced by the law combined with the resistances; he has to ascertain to what extent the materials are uniform and perfect.

Now inasmuch as Science consciously employs the spontaneous artifices of ordinary search, one great principle of scientific teaching should be the clear recognition of such artifices. Hence I would propose a new logical canon, namely, that every theory should be pronounced incomplete until the preliminary abstraction has been rectified by a secondary restitution.

Two illustrations will suffice to exhibit the importance of this canon. The undulatory theory of light and heat is justly regarded as one of the triumphs of modern science. But what does it start from? The assumption of oscillating atoms having no dimensions-mere points without form or size! This is a sufficiently bold disregard of concrete observation; it is an abstract so entirely removed from reality as to be unimaginable objectively. Nevertheless mathematical analysis, occupied solely with the oscillations and wholly disregarding the atoms, has furnished vibratory laws which explain many of the most remarkable phenomena of light and heat, such as refraction, polarisation, and interference of rays. So far the abstraction has justified itself. But the incompleteness of the theory is equally evident in its failure to account for other important phenomena. Here then we are recalled to the necessity of

reintroducing some of the discarded elements of the problem, and perhaps Restitution will furnish a solution of the outlying cases. The theory has excluded all consideration of size and form; but if the atoms exist at all, it is eminently probable that they have both size and form; and a new line of inquiry is thus opened, namely, what are the different varieties of movement which these ponderable atoms are susceptible of receiving from the influence of external and internal impulses ?

I do not pretend to say what mathematical analysis may or may not discover in this direction; I only affirm that the form and volume of an atom must influence its movement; and that the present condition of the undulatory theory is incomplete, because the laws of oscillation only, and not the laws of atomic movement, constitute its object.

The second illustration shall be the popular question of Species. Are species variable or invariable? This is very much like the case of planetary perturbations. The abstract law of reproduction (like produces like) points to fixity of species, as a fundamental biological truth. But this abstraction ignores the reaction of the medium on the organism-such as daily exhibits itself in a hundred different ways-it ignores the concrete facts of the struggle for existence-all of which act as perturbations of the biological law. The question then arises: what is the sweep of the perturbation? Can these perturbations, by accumulation, change the primary law? To answer this, the research must be guided by the canon just laid down. I do not assert that in this case there is any danger of men neglecting to correct the preliminary abstraction by a subsequent reintroduction of the discarded elements, such as were noted with respect to the undulatory theory; the naturalist is, like the practical mechanic, more disposed to concrete observation than to abstraction; but I will suggest that the final settlement of the controversy on Species can only issue from the establishment of the theory of the organism, and the subsequent establishment of the theory of the medium in its modifying influences. At present we have two groups of indisputable facts, the group which proves the constancy of forms and the group which proves the variability of forms; a complete scientific theory must include and reconcile both groups. And this theory will best be reached by a preliminary abstraction of the biological laws, and a subsequent restitution of the perturbations due to the reaction of the medium on the organism. The separation is but an artifice to aid our infirmity, but if systematically adopted, it will be found of eminent utility.

From Science to Criticism is a long step, yet in many quarters the question is being raised, Why have we no science of Criticism? Nothing can be more patent than the fact that such a science is absent, but I am very far from thinking that such a science is desirable. A writer in the Pall Mall Gazette, pointing out the diametrically opposite dicta of two critical journals on the same work, asks, Are there no canons of criticism? apparently wishing that such canons were in force as would prevent contradictory judgments being delivered. A writer in the last Revue des Deux Mondes also urges the necessity of a doctrine in Literature.

My first objection to anything like a doctrine in Literature is, that it must necessarily be so incomplete as to be tyrannically oppressive; because at the best it could only exhibit the laws which great artists had followed, it could not embrace the laws which great artists to come would follow. Thus, sup

posing the science perfect in its construction (a large supposition!), it could only explain the works and processes of an art that had developed itself up to a certain point; it could not explain, it could not even divine what would be the new evolutions of the art under the new conditions of advancing civilisation. Let us take the case of Music, and suppose that an Aristotle had constructed a perfect science of musical criticism, out of the musical productions then known in Greece. Had there been a musical doctrine, with canons which all critics would enforce, the consequence would have been that progress would have become impossible. It is now known, and has been demonstrated, that the Greek music, from the very nature of its gamut, could have no Harmony. All the magnificent developments of modern music which spring from its enlarged gamut would, therefore, not only have been unsuspected by the critics, but would have been arrested in the early stages because "contrary to rule." The innovator would have been repressed. In like manner the Greek Drama is constructed on principles so narrow compared with those of the modern drama, that an application of the canons of the one to the productions of the other can only be an oppressive mistake.

I have been arguing on the supposition that the Science would be a true exposition of the laws of art. If even on that supposition the effect of canons would be disastrous, what would be the effect of canons that were false? We have had one striking lesson. Europe once had a literary doctrine, which it accepted from France; and the effect of that doctrine in repressing all originality and all progress is familiarly known to the most superficial student of literary history. A somewhat similar oppression is exercised in Germany by the so-called philosophic criticism, which views a work of art in relation to certain philosophic ideas, not in relation to the effect on the emotions of the audience.

Because the Laws of Nature are more or less discoverable and reducible to a system, it is supposed that the Laws of Art must be equally discoverable. There is, however, this difference: Art is in a state of perpetual evolution, new forms arise under new conditions, and new inventions introduce new laws. Now it is certain that if men of science had the power, they would suppress all the facts they were unable to explain; whatever disturbed the symmetry of their doctrine would be set aside as chaotic and unworthy of a place in orderly creation. They have not this power, and so are forced reluctantly to accommodate their doctrine to the facts, to enlarge their doctrine with enlarging knowledge. But critics would have the power of suppressing originality; and would brand as "chaotic," "unworthy a place in orderly Art," whatever disturbed the symmetry of their system, whatever was not amenable to their canons.

Hence I maintain that the present state of anarchy in Criticism is preferable to a state of dogmatic authority. Criticism may suffer; but Art is freer. This is by no means asserting that one critic's judgment is as good as another's, or that every man may set up his individual judgment as a standard. One judgment is not so good as another, because it will not be founded on equal insight, equal knowledge; nor can every man make his judgment a standard for others, but only for himself and for those who think and feel like him. In every work there are certain general principles involved, and certain technical principles; the best critic is he who best understands both principles, and whose sagacity enables him to appreciate their application. The technical principles which are

involved in the drama are not the same as those involved in the novel, and therefore an effect in the one may be a defect in the other. But there are certain general principles common to both, and these the public at large can judge as well as the best critic.

Inasmuch as every Art has its rules, general and technical, there might be a codification of the various rules which would be of service, and might stand for a Science of Criticism; but were this codification effected, we should still have to remember that Criticism is itself an art and not a science, and that nothing could be more disastrous than the establishment of a Doctrine of Criticism, with its rigorous canons, which would suppress originality merely because originality was a violation of some canon.

Shakespeare may be said to be the opprobrium of Criticism. He has from first to last been the subject of more criticism and more critical nonsense than any other writer. He is a puzzle to all critics, eluding all their canons, either by the glaring discrepancy between his effects and their rules, or else by mashing all their rules into a general mush of admiration. But if Shakesperian critics are unprofitable labourers, Shakespearian editors are men who claim our gratitude. There has been no lack of them, thanks to the perennial interest in the subject, and the hopeless difficulty of a perfect text. Of the two latest editions, having high pretensions, that by Messrs. W. G. Clark and W. Aldis Wright, known as the Cambridge Shakespeare (Macmillan and Co.), is now completed by the publication of the ninth volume; and that by Mr. Alex. Dyce, also in nine volumes (Chapman and Hall), only awaits the last volume, which is to be devoted to an extensive glossary. These two admirable editions, remarkable for the scholarly scrupulosity with which the text is collated and established, and for their sumptuousness of paper and print, demand a mature and minute examination, which the REVIEW will endeavour to offer ere long. Meanwhile it is enough to announce their completion.

Students of History, especially those who are more immediately seeking to understand European development during the Seventeenth Century, are recommended to take in hand the small volume by Dr. Bridges, entitled "France under Richelieu and Colbert" (Edmonstone and Douglas). It is a republication of lectures delivered at the Philosophical Institution of Edinburgh. Dr. Bridges is known as a disciple of Auguste Comte; and as the author of a very remarkable survey of our relations with China, in the "Essays on International Subjects" (Chapman and Hall). He has a mind happily endowed for exposition of great subjects, being at once philosophical and sympathetic, always demanding the breadth of view which can only be given by general principles, yet having also a sufficiently vivid interest in concrete facts to supply him with ready illustrations of his principles. This combination is The oxy-hydrogen light, brilliant as it is, if it fell on no objects, would be a more fruitless illumination than a dark-lantern which showed us our stepping-stones; and to minds of the ordinary class, great thinkers, like Comte, who have much to say in brief space, are often dark from the excess of theoretic light. All the more precious, therefore, are the successors who can expound and illustrate, and find varied demonstrations for the same formula; who come from supping with the Gods to join fellowmen at the shilling ordinary, and have a sufficiently keen fellowship to feel no sort of stupidity altogether foreign

rare.