The subtractive method is also applicable to the production of colour photographs on paper, and many processes have been devised by which this can be successfully accomplished. Ducos du Hauron suggested in 1869 the use of three carbon tissues, pigmented respectively in red, yellow and blue. This suggestion has been successfully put into practice within the last few years, and very beautiful results have been obtained. To facilitate exact superimposition of the prints, the carbon tissues are supplied on temporary supports of transparent celluloid film. The process has been introduced commercially in England by the Rotary Photographic Co. and the Autotype Co., and all the necessary materials are readily obtainable.

In the Lumiere process, introduced by Messrs. Lumiere, of Lyons, a print on glossy bromide paper is made from the red-filter negative, and toned blue. Prints from the other two negatives are then made in bichromated gelatine on a transparent celluloid support, and thes, after being stained yellow and pink respectively as described above, are superimposed on the blue print, each being stripped from its celluloid support after being brought into exact register with the blue print.

Messrs. Sanger Shepherd & Co.'s very successful Imbibition Process is based on the discovery that a thin film of soft gelatine on damped paper, placed in contact with a dyed gelatine relief, will quickly absorb all the dye from the latter. The process consists in printing from the triple negative in bichromated gelatine on a transparent celluloid support. The developed prints are stained in the usual colours, and successively brought into close contact with a sheet of damped gelatine-coated paper, by which the stains are absorbed, the gradations of each positive being exactly represented by the depth of the staining. A complete colour print is thus produced on the paper, and the celluloid positives, entirely cleared of their dyes, are ready for use in making another paper print, and may be used again and again, there being no necessity, as in the processes before described, to make a fresh set of positives for every print. Another advantage of this process is that the three dyed positives may be viewed in superposition over a sheet of white paper, as a test of the colours, before bringing any of them into contact with the printing paper, and any errors in depth of staining may thus be detected and rectified in time to prevent spoiling a print.

"Pinatype," introduced by Dr. Konig in 1905, is another process in which colour prints are produced by transference of dyes to gelatine-coated paper from bichromated gelatine "printing plates." The peculiarity of this process is in the dyeing of the printing plates. In the Sanger Shepherd process it is the hardened gelatine image that is dyed; the Pinatype dyes, on the contrary, are taken up only by the unhardened gelatine. The printing plates are therefore made by exposure under positives, instead of direct from the negatives, and do not require development, as the unhardened gelatine must not be removed. As in the Sanger Shepherd process, the printing plates may be used over and over again.