III
VIBRATION
The most splendid achievement of the nineteenth century in painting, and its best legacy to the future, was the discovery of the technical means by which the scintillating effect of living light could be transferred to the dead and rigid surface of a canvas. Of this the old masters had absolutely no conception. The discovery belongs to our generation, and is a distinction of which any age might well be proud for it is the only important step in advance made since the great Renaissance of the fifteenth century. Without it landscape art had hardly been possible landscape art, that is, in the modern sense in which we know it. There were indeed many landscape painters among the older masters—Ruysdael and Cuyp, Hobbema, Salvator Rosa, Claude, and even Rembrandt on occasion. But, owing to a curious psychological phenomenon, none of these men were able to see straight out of their eyes once they were in the open air. They painted landscape, but landscape in which the fields and the hills and the trees bore no relation to the skies that overhung them, in which the shadows were warmer in color than the lights, in which browns took the place of violets, and in which (owing to ignorance of the laws of vibration) the surface of the canvas never entirely disappeared from view.
As I have previously stated, the dawn of the new movement was seen in England, when Constable and his confreres carried their easels into the open, and brought back studies wherein the pearly tones of out-of-door nature were for the first time accurately seen and noted.
A few of these pictures finding their way to France, were eagerly studied by a group of young Frenchmen, who, tired of the hide-bound conventions of David and Delaroche, were quick to recognize and absorb the new light. Armed with this fresh knowledge, these men in their turn went out into the fields, and looked and studied and painted; and thus grew up the great school of Barbizon.
A little later the artistic world was startled by the appearance of the French impressionists or luminarists. According to them, nature had spread her palette upon the heavens in the form of the rainbow, where all who looked might see and understand it. And everywhere and always, on hill and dale, on rock and tree, so long as light endured there must also be the rainbow attenuated and diminished in power, it is true, but with its three primary and prismatic colors, locking and interlocking, shifting and shimmering and playing across one another in an iridescent dance of color that was, or should be, always clearly visible to the eye of the trained artist. And as they saw nature so these men painted their pictures, laying the pure pigments side by side upon the canvas in strokes and dots or dashes of red and yellow and blue which, seen at the proper distance, were supposed to fuse into the desired tones and masses, while at the same time retaining a luminous quality of their own never before seen upon canvas.
I can remember the first exhibition which these men gave in Paris in the 
From a photograph by Durand-Ruel & Fils
Claude Monet—"The Bridge at Argenteuil"
little rotunda behind the Palais de l'Industrie; and the bewilderment and scorn with which it was received by the critics and the older painters. I can remember also the heroic struggle which they made against apparently hopeless odds; and we all know how they finally won the long fight, proving their point so conclusively that no one to-day thinks of questioning it.
But while all painters now admit that the prismatic theory of light as applied to the art of painting is both scientifically correct and artistically admirable—that it is practically impossible to secure luminosity in a picture without some sacrifice to the principle, it is nevertheless open to question if the crude and primitive method invented by the French Impressionists is necessarily the last word on the technical side of the matter. We must have "vibration" in a picture, it is true, because without vibration there can be no light, but may it not be possible to secure the necessary vibration without loss of "quality," that charm of surface with which we would not willingly part?
There are many, many paths by which the problem may be approached. Indeed, one of the chief delights of the art of painting lies in the fact that each artist does, and of necessity must, invent his own technique; for his personal technique is an inalienable part of the personal vision which makes his art his own. Nevertheless there are in a broad sense only four general methods of painting with oil colors, from which (used either in their direct and simple expression or infinitely varied and compounded) all of our personal technical methods must be drawn. First we may mention the method used by so many of the old masters, which consisted in a solid underpainting in black and white with a slight admixture of red. In this method the whole scheme of the picture was built up with these three pigments, and all of the drawing and modelling was accomplished without any attempt at color. Then, after a very thorough drying, the work was completed and the color obtained by a series of very thin glazes drawn over the dried and hardened surface. This method, although wonderfully sound in itself and lasting in its results, must of course be discarded by the modern painter for the reason that it precludes all possibility of vibration.
Of the three remaining systems one other is entirely bad for the same reason—it does away with vibration. This system consists in mixing the tones evenly and applying them to the canvas in smooth flat masses in much the same manner as a house painter paints his door or cornice. There remain then practically but two systems from which the modern painter is at liberty to choose. The first of these is the spot and dash method used by the Impressionists and their school. It must be clear to any one that this system, while giving beautiful results in the way of luminosity, does not logically follow the forms of nature, or reproduce her surfaces, and it must therefore be regarded as an imperfect and a temporary manner which is destined to be superseded in time by some more supple and expressive technique.
The last of the four systems mentioned and one which has gradually come to be adopted by the vast majority of our best landscape painters is one in which vibration is obtained by means of a cool overtone painted freshly into a warm undertone, care being taken not to mix or blend the two coats and not to cover up completely the undertone, rather letting it show through brokenly all over the canvas; the vibration being secured, naturally, by the separate play of the warm and the cold notes. Neither alone would accomplish this purpose, nor would the neutral gray that would result from a too thorough mixing of the tones in the final brush-work.
This method has first of all the great advantage of being thoroughly logical; for in nature herself the undertones are represented by the local color of the various units—leaves, grass, rocks, and good rich earth; and these are always warmer and more vivid in color than the lights dropped upon their surfaces by the over-arching sky. But the method has the still greater advantage of being wonderfully supple and responsive—lending itself not only to the infinite variations of technique demanded by differing temperament in the artist, but allowing endless latitude for any and all desired changes in composition or mass after the picture is placed on the canvas; for all of these changes can be made in the undertone itself before the overtone is applied, and therefore before any attempt to secure vibration has been made. Indeed the whole picture in all its exact values can and should be built up in this preliminary covering of the canvas, for the value of the overtone must in every case exactly match the value of the undertone. While we wish to secure broken color, we must avoid broken values, for they utterly destroy atmosphere. Any one who wishes to prove this to his own satisfaction can readily do so by making the following experiment. Paint a sunny sky in two simple tones, using, say, delicate gray pink for the underlay and blue green or green blue for the overlay, varying the color from the horizon up as it occurs in nature. In the first experiment mix the overlay with extreme care until its value exactly matches that of the underlay. Then mix another lot to the green blue either slightly darker or slightly lighter than the underlay. Apply these tones each to one-half of the prepared sky, and you will find that the sky painted with the perfectly matched tone will fly away infinitely, will be bathed in a perfect atmosphere, while the other half of the canvas will remain merely paint and canvas, and will have no atmospheric quality whatever. The explanation of this is very simple—nature deals in broken color everywhere, but she never deals in broken values. The color dances, but the values "stay put."
As to the general tint of color of the undertone no rule can be given, for it can never in any two pictures be alike. It will vary infinitely, according to the effect to be painted, and also according to the temperament of the artist. There would seem to be only two rules that cannot be broken: first the undertone must be warmer than the overtone, and second it must never be brown; and this for the excellent reason that out-of-door nature abhors brown, and never uses it. Even the house-painter's most venomous effort in this direction is generally met by kindly and all-forgiving mother nature with some gray reflection from the sky to mitigate its worst virulence. The one weak spot in the technical armor of the Barbizon painters was their tenacity in clinging to the traditional recipe of the brown rub-in. And although this was allowed to dry thoroughly and was then completely painted over with pearly tones that were true to nature, the browns are now beginning to strike through to the surface—to the serious detriment of some of the finest pictures on earth.
Now when the fullest acknowledgment has been made of our stupendous indebtedness to the discoverers of prismatic painting, it will be wise for us to recognize the limitations of the system; to admit that there are very many effects in which it must be used with extreme caution, and others in which it had best not be employed at all. If we frankly envisage the fact that its chief function is to endow our dead pigments with life, with the power to convey in a picture the joyous impression of dancing light, we shall understand where these limitations begin. As the system gives its best results in the translation of brilliant sunlight, so, as the light decreases its value decreases, until in a low-toned moonlight it may become positively detrimental. It can easily be seen that in this subdued light the sibilant vibration of powerful color-tones would be fatally out of place and their use detract seriously from the brooding sense of mystery which gives to night its most poignant charm.
We must not forget, moreover, that another weakness inherent to the system lies in the physical impossibility of securing with pigments and brushes any approximation to the infinitely fine and delicate color vibration of nature—where no spot or dash or stroke of pure color is anywhere visible; and that our best efforts in this direction are 
Winslow Homer—"The Fog Warning"
therefore only a compromise—that owing to this compromise our best technique of vibration remains at the present time more or less obtrusive, and that any technique which obtrudes itself is to that extent bad technique; for technique, as Millet so truly said, "should always hide itself modestly behind the thing to be expressed."
Finally let us frankly admit the fact that vibration has little to do with atmosphere in a picture (in spite of much wordy argument to the contrary). A Whistler nocturne, for instance, which is painted without the slightest vibration, or any attempt at broken color, may swoon in the most exquisite bath of atmosphere, while a vibrant Monet, with a few hard edges, may lack all atmospheric quality.
Atmosphere in a painting is only secured by the use (conscious or unconscious) of the laws of "refraction," a much more subtle and elusive visual phenomenon of which I will say a word in the following chapter.