Page:Scottishartrevie01unse.djvu/183

Rh investigation necessarily requires much time ; many of the changes, more especially the chemical ones, can only be brought about by brilliant sunlight, acting for a considerable length of time. The report is therefore divided into two sections, the first of which is now presented, dealing especially with the nature of the optical changes. As this is a question of very great interest to water-colour painters, the following resume of the report may not prove uninteresting. It was expedient in conducting the experiments to choose the light which would most readily adapt itself to giving a clue as to which colours were affected by exposure in a time which would be measured by months instead of years. After deli- berate consideration, the experimenters decided to avail themselves of as much sunlight as could be secured, together with the diffused and sky light when sunshine was absent. Some writers have declined to accept deductions as to the fading of pigments when exposed to the bright liglit of the sun, but they have never given any serious reasons for their doing so. Their arguments have usually been based upon their own convictions rather than on experimental proof of any kind, or if experi- mental proof has been quoted from other writers, half the truth or more is most frequently and pro- bably unwittingly concealed. A popular fallacy is that if the light be very feeble, a bleachable colour, no matter how long it may be exposed, will not fade. It has been fully proved by experiment that, if a certain tint be exposed to an intensity of light, say of 100, and is bleached by it in, say one hour, a similar tint, exposed to an intensity of 1, would, with 100 hours exposure, undergo the same bleaching. As to the liglit to which pictures are ordinarily exjDosed in a room, there is no doubt that they are, as a rule, carefully protected from direct sunlight, but it is nevertheless true that the greater portion of the light they receive is reflected sunlight. On a bright day clouds reflect sunlight, and on a dull day the principal part of the diffused light is also sunlight, which is reflected according to the laws of geometrical optics from particle to particle, a certain percentage eventually reaching the earth througji the clouds. There is of course also a fair proportion of the light due to the upper sky, and this light is bluer than reflected or diffused and weakened sunlight. In cases where the windows of a gallery are in the vertical walls, and have an uninterrupted view of the horizon, the blue light reflected is com- paratively small, the light near the horizon being distinctly more like sunlight than is that nearer the zenith. In galleries lighted like those at South Kensington by skylights, the light to which pictures are subjected is on the whole bluer. The artificial lights to which water-colours are exposed are gas-light, the arc and incandescence electric liglits, and, as will be seen by the results, the first and last of these are very deficient in blue rays. Doubtless, to the eye, the hue of the lights men- tioned above differ considerably ; but unless the cause of the difference had been tracked out experi- mentally, and with scientific exactness, it would have been unwise to have selected any one of them with which to conduct experiments, since the results obtained with it might not be applicable to any other. Happily, however, for such work, the spec- troscopic analysis of light furnishes irrefutable evidence that, from the results obtained from ex- posure to one light, correct deductions may be made as to what would happen* were the exposure made to another. If, by a prism, we analyse all the different kinds of light mentioned above, we find that in the visible spectra so obtained no colour is absent ; but, if we compare the intensity of the same colours in the different spectra, we find that there is a variation. Since, then, all these sources of light emit the same rays, but of different intensities, which can be measured, it follows that if we know which rays are cliemically active, and the amount of work which, wlien of a certain intensity, they perform, we can, from the work done by the light from one source, deduce the work that would be done by another. The most perfect manner of noting the action of light would be to expose the pigments for a given time to the action of the spectrum formed by an unvarying source of light, and to measure the amount of chemical action (fading of the colour in most cases) which had taken place under every part of the spectrum. When the relative intensities of the different parts of the spectra from other sources of light compared with this standard spectrum were known, then the length of time during which it would be necessary to expose the colour to any one of them to produce that same total effect could be calculated. Months being often required, however, even in full sunlight, to effect a visible chemical action on some of the pigments, resort was therefore had to the use of coloured glasses, in order to hasten the investigations, to ascertain the part of the spectrum which was most active in producing the fading action. The glasses used were red, green, and blue. In every case where any fading took place, it was always found beneath the blue glass, much less often, and to a far less degree, under the green, and only twice under the red glass, and was then barely perceptible.