Thus, although Du Hauron actively continued his experiments, he failed to produce any satisfactory results, and for nearly twenty years from the publication of his treatise the practicability of his methods remained unproved. The credit of first practical success belongs to Mr. Frederic Eugene Ives, of Philadelphia, who, after many years of patient research and experiment, not only produced light-filters of the necessary degree of accuracy, but devised a most ingenious camera, by means of which he simultaneously exposed three specially prepared colour-sensitive plates, thus obtaining triple colour-sensation negatives by what was practically one exposure. These negatives, of course, were entirely destitute of colour in themselves, as also were the positives made from them; but when the latter were viewed by light passed through suitably coloured light-filters in Ives's cleverly devised "Kromskop," or projected on the lantern screen by his equally ingenious triple-projection lantern, the effect was a startlingly lifelike representation of the objects photographed, in all their natural colours. In the stereoscopic Kromskop the illusion of reality is particularly striking. The writer recollects inviting a lady to look into such an instrument at a photograph of a vase of flowers. While she did so, he endeavoured to explain how the colour effect was produced. "But," said she, looking up from the instrument when the explanation was concluded, "that's not a photograph; it's a real vase of flowers."
But beautiful as were the effects produced by Ives's apparatus, it was too complicated and too expensive to become popular. What was wanted was such a colour-photograph as could be viewed in the hand, without apparatus, or used as an ordinary lantern slide in any single projection lantern. The Ives process having proved the practicability of the triple-negative method, other experimenters attacked the problem assiduously, and before the end of the nineteenth century Mr. Sanger Shepherd in England, and Messrs. Lumiere in France, had brought out, on commercial lines, processes of transparency-making in colours based on the Ives method. These processes are still in active operation, and for many purposes remain superior to any of the more recent methods.
In the Sanger Shepherd process, three negatives of the subject are obtained on colour-sensitive plates by exposure in the camera through red, green and blue-violet light-filters respectively. A positive, on an ordinary lantern-slide or transparency plate, is then made from the negative taken through the red filter, and toned to a greenish-blue. Transparencies from the negatives taken through the green and blue-violet light-filters are then made on thin celluloid films coated with bichromated gelatine, the picture being reversed by printing through the celluloid film. These are developed by dissolving out the unhardened gelatine with warm water, after the manner of carbon prints. The gelatine image obtained from the green-filter negative is then stained magenta-pink, and that obtained from the blue-filter negative, yellow. The two stained film positives are carefully superimposed upon the blue-toned glass positive, and the whole bound up with a cover glass in the ordinary way.
It will be noted that the positives are not stained in the colours of the light-filters through which the negatives were taken, but in colours complementary to them. The necessity for this will be made clear if we consider the effect of photographing a single colour - a disc of green, for instance - on a white ground. In the negatives taken through the red and blue-violet filters, it will appear as a transparent patch on an opaque ground; but the negative obtained through the green filter will be opaque throughout, because the green rays from the disc and those from the white ground have alike been passed by the filter. A bichromated gelatine film exposed under such a negative would, of course, be entirely protected from light-action, with the result that the gelatine, on development, would be entirely dissolved away, leaving nothing to take up the dye. In the positives from the red-filter and blue-filter negatives, however, the green disc would be represented by undissolved gelatine. Supposing these were dyed with the colours of the filters through which the respective negatives were taken, and then superimposed, we should have a red disc upon a blue one. The effect would be this - the red would cut off the blue and green rays, and the blue would cut off the red rays; so, instead of the disc appearing green, it would appear black, all the light being cut off by the two superimposed colours. If, on the other hand, we dye the red-filter positive greenish-blue, and the blue-filter positive yellow, the correct result is obtained, thus - the greenish-blue cuts off the red rays, and the yellow cuts off the blue rays; the green rays are therefore the only ones to reach the eye, and the true colour-sensation is produced.
The production of colour photographs by the superposition of dyed positives is known as the subtractive method, because the final colours are obtained by the action of the dyes in each subtracting, from the white light falling upon the photograph, the rays that are not required. The Sanger Shepherd process may be taken as typical of the many processes for the production of transparencies by the subtractive method, differing from each other only in detail.