them in the dark in a portfolio or betwixt the leaves of a book, and frame the other half and hang it up to the light in a window casement as before. At the expiration of the same period-though much less than a year would really suffice-let the two halves be brought together again and compared as in the former illustration. Whilst in some instances the tints in the respective halves will be found to be unchanged and exactly alike, others in the exposed portion will have absolutely vanished off the paper, or at best left but a faint and dirty stain, generally due to some extraneous impurity in the pigment, more permanent in its nature than the substance which it adulterated.

Nearly all the perennialists, if I may so term them, admit that exposure to direct sunlight is detrimental in some unexplained way to water-colour drawings; but direct sunlight is only the most intense and searching measure of daylight. It must be obvious, inasmuch as the sun's light is the same, whether direct or infinitely lowered or mitigated, that it must in the exact measure of its intensity possess precisely the same properties.

I have some misgivings as to whether any pigment, if exposed for very long periods to the fullest blaze of the sun's rays, would be found to be entirely unaffected; the experiment at any rate could scarcely be tried in this comparatively sunless country. For all practical purposes, however, the ascertained constancy of certain colouring substances is great enough to entitle them to be considered quite permanent.

As regards the relative durability of fugacious pigments, the action of direct sunlight is very rapid and notable. The exposure of any water-colour drawing of Turner, for instance, for a few days even to direct sunlight would be sufficient to effect visible deterioration. Probably, indeed, more actual mischief may be caused by accidental exposure to the direct rays of the sun during so short a time even than would be the case in a series of years in the low half-light of an ordinary apartment.

From the infinitely varied, mutable, and uncertain conditions under which drawings have been preserved, and of which in the nature of things, in scarcely any instance, accurate record can have been kept, it is clear that appeals to the evidence of particular specimens in support of this or that theory must always be of a dubious. or fallacious nature.

Considering, then, the striking differences in regard to conservation which drawings display, it is no wonder that unobservant and unreflecting persons should still be found to deny the fact that watercolour drawings are liable to fade, as well as obstinate partisans who when virtually convinced of it, but against their will, remain of the same opinion still. When these considerations are borne in mind, it will, I think, be admitted to be unnecessary for me

to notice many objections, supported by inconclusive or irrelevant instances, which the recent controversy has elicited.

If absolute immunity from the bleaching effects of light could be secured for the pigments made use of, so many and weighty are the recommendations and advantages of the water-colour process that I incline to think it would on the whole be even preferable as a means of art to oil painting.

The stumbling blocks and disadvantages which beset the latter method, indeed, meet both the artist and the picture collector at every turn.

Oil painting is a vastly more complex, cumbersome, and tedious process, yet not necessarily a more durable one, and it may be questioned whether it has any inherent charms or physical capabilities which the water-colour process, as carried out in these days, does not possess in equal measure. One thing, however, is certain, and it is desirable to place it in a clear light: whatever may be the especial drawbacks of the art, oil pictures are much less liable to deterioration from the influence of light than water-colour drawings as they have been heretofore executed. There are, as a matter of fact, in oil painting counteracting agencies at work, tending to neutralise more or less, according to circumstances, the fading influence of light on the colours employed. One of these causes of greater stability in the pigments is probably the fact that they are employed in much greater volume than in watercolour painting. In the latter process the tints are for the most part mere washes or stains of impalpable tenuity; in oil paintings, on the contrary, the colours are often piled up,' or loaded, in quite measurable thickness; there is therefore a much greater substance of colouring matter for the light to act upon. More efficacious still, in the sense of preservation, is the fact that the colouring substances in oil painting are effectually enveloped or locked up' with the oils and varnishes with which they are mixed, the separate particles being each surrounded with a protecting medium in the oil or varnish employed, vastly more solid and efficacious than the gum or size which binds together the colouring atoms in watercolour painting. It is true that the oleo-resinous vehicles of oil painting are in other ways themselves the causes and media of decay and alteration often as fatal to the artist's work as light is to water-colour pictures, but to go into this matter would involve a separate treatise.

6

To the fading of pigments, however, these oleo-resinous enveloping vehicles oppose a direct resistance in another way. Whilst it is untrue— physically impossible, in fact that simple water-colour pictures can deepen in tone by age, it is, on the other hand, just as certain that oil pictures do become darker in aspect as time goes on; but this increasing depth of tone is caused in most instances mainly, and in every instance to a great extent, by the alteration in colour of the

oleo-resinous vehicles, not by any mutation of the pigments locked up in them, which latter, if they undergo any alteration at all, change in the sense of diminution, not increase of intensity of tint. Oils and varnishes, though often quite colourless at first, undergo inevitable chemical changes, accelerated or retarded, superinduced, and even remedied again by modifying external influences. They absorb oxygen from the air, and a process analogous to the slow combustion of their particles takes place, the result being that the originally colourless vehicles gradually acquire a yellow or brown tint of varying degrees of intensity, and tend to become more or less turgid and opaque. This darkening process goes on most rapidly in the dark, and it is retarded or remedied again by re-exposure to light. A bleaching action, in fact, is then established. As I have intimated I am no authority on the actinic properties of light. The processes of photography have, however, made certain phenomena familiar to us. Everybody knows that it is the rays at the blue end of the spectrum which bleach and disintegrate, whilst the yellow and red rays are inert or protective.

The particles of colouring matter, then, in oil painting are 'locked up' and protected by a semitransparent medium of the precise tint, which tends to neutralise the fading effects of the sun's light on the pigments, and a virtual equilibrium of forces seems thereby to be established. From this it results that whilst it is good for watercolour drawings that they should be kept in the dark as much as possible, in portfolios rather than in glazed frames exposed to the light, it is, on the contrary, bad for oil pictures to hang them in dark corners or for periods of time in shut-up rooms, the privation of light in those cases bringing about increasing darkness, dulness, and opacity.

The action of the electric light on water-colour pigments is a question of much interest, and it is greatly to be desired that some competent scientific authority should practically investigate it. The feasibility of the exhibition of water-colour drawings in our public collections literally depends upon a favourable solution of this question, for nothing is more certain than that the continuous exhibition of water-colour drawings of past epochs by daylight inevitably entails their rapid, irremediable, and more or less complete and final deterioration. It is not unlikely that the white or bluish are' light would, though in a far less degree, be as destructive as the sun's light to mutable pigments; but the yellow incandescent' light, on the contrary, would probably be found to exercise no perceptible bleaching influence.

By way of fencing with obvious and glaring facts which cannot be ignored it is not unusual to find people, especially professional artists, assert in respect to faded drawings that they have 'gone' as far as they ever will, and that they will fade no further. The

absurdity of such a doctrine, as it is enunciated and understood by such persons, is self-evident; nevertheless it may be literally true in certain circumstances. When, for instance, a drawing executed partly with permanent and partly with fugacious pigments is exposed to the light for very long periods, it may well happen that the time will come when it can fade no further, for the simple reason that the inherently fugacious tints will have all finally vanished and fled away, leaving nevertheless the others unaltered as at first. Drawings in this predicament have indeed not unfrequently come under my notice; they are, of course, reduced to the condition of mere disjointed wrecks.

It must be remembered that in the case of water-colour drawings such works are still in their first youth, so to speak, for the art itself is new; but we have remaining to us innumerable drawings and studies, of much simpler technique it is true-works of the 'old masters—over some of which four or five centuries even have passed. Some of these are still fresh and brilliant, while the others are but 'pale ghosts of their former selves.' How is this remarkable diversity of condition to be accounted for? For, be it observed, the same simple pigments employed in those former times are those still in use for similar purposes. A few remarks about 'ancient drawings' will incidentally answer this question, whilst at the same time relevant to our subject in general and instructive in various ways. By common consent of directors and curators of public collections throughout Europe of late years, certain classes of such drawings have been withheld from exhibition, for it has been universally recognised that the continuous exposure to daylight is fatal to them. Ancient drawings' are executed mainly in two distinct technical methods, both extremely simple; for the most part either they are drawings in black or red chalk or pen drawings in bistre or sepia, sometimes washed or shaded with the same or with other colours. Now black and red chalks are quite permanent substances, and the pale, vaporous aspect of some ancient chalk drawings has been brought about by simple wear and tear: friction and abrasion in portfolios, &c. They have, in short, been partially rubbed out. Pen and washed drawings, on the other hand, have suffered comparatively little from abrasion; but, inasmuch as both bistre and sepia are fugacious pigments, fading from exposure to light in the irregular and intermittent intervals when they have been exposed during the course of centuries has been the real cause of the weakening of the tints.

The comparatively fresh and uninjured pen drawings of sixteenth and even fifteenth century origin, which undoubtedly still remain to us, owe their preservation unquestionably to the fact that it was formerly almost the universal practice to preserve such works in the dark, either mounted in books or in portfolios, framing and glazing and the consequent continuous exposure to light being comparatively unknown in

ancient times, from the simple reason that glass of flat, even surface and sufficient size, was difficult to obtain and very expensive; whilst to hang up valuable drawings without any such protection was obviously to expose them to certain destruction within very short periods. Generally speaking, therefore, the deterioration of ancient drawings from fading has mainly happened during the course of the last hundred years or so, during which time the practice of framing and glazing drawings and prints has grown up.

I observe that a recent writer on artists' pigments states that bistre and sepia are stable and reliable colours; but in this I differ from him entirely, and I apprehend that his observations and experiments have not extended over a sufficient range and length of time to have enabled him to ascertain the real truth of the matter. A very pertinent, and indeed amusing, instance in illustration of the fugacious nature of bistre and sepia (it is not easy, by the way, to decide in all cases which of these colours has been employed) occurs to me on the spur of the moment. In the Oxford University collection of drawings by the ancient masters is a large bistre pen drawing, an elaborate copy of Michel Angelo's 'Last Judgment' by a contemporary sixteenth-century artist. This drawing has been exposed to the light for a long period under glass, both at Oxford and before it came there; consequently it has waned and dwindled to a very pale and shadowy status. One figure in the composition, and one only, nevertheless retains its pristine force of tint, standing out like a dark rock against a vaporous sky. This is the figure of Charon, on the lower part of the composition, ferrying over condemned souls in his boat and striking them with his oar. The fact is, the simple-minded artist, anxious to invest the evil one with superabundant terror, drew the grisly fiend with black pigment, doubtless Indian ink or lampblack, and this colour, being in its nature quite unchangeable, has stood its ground perfectly whilst all the rest of the work executed in bistre has almost faded off the paper. Thus the devil in this instance at all events remains just as black as he was painted!

Sir James Linton and others have alluded to the behaviour of two pigments very frequently used in combination by the English watercolourists of the earlier part of the present century. These are Indian red and indigo, chiefly employed to form compound tints, in skies, distances, &c., of varying degrees of cool grey and purple. The strangest possible changes have taken place in many of the drawings in which these colours have been made use of. Cold grey skies with dark rolling clouds and distant purple mountains, for instance, have been utterly transformed and metamorphosed; they have often become bright 'foxy' red: in short, the effect of brilliant, glowing sunsets has sometimes replaced that of the lowering skies originally depicted. It was impossible for the President of the Water-Colour Institute and his supporters to ignore these striking and significant phenomena, and VOL. XIX.-No. 112.

3 M