The first commercial collotypes were produced in 1868 in Germany by Josef Albert and in 1869 in England by Ernst Edwards. Until the development of the half-tone screen it was the only photomechanical process, besides hand photogravure, capable of reproducing tone.
Collotype is the most accurate and beautiful method of photomechanical reproduction yet invented. It has the advantage that it can render continuous graduations of tone without the intervention of a screen. But the making and printing of collotype plates is skilled and expensive work and can easily go wrong. Variations in humidity are likely to upset the balance of moisture in the gelatin causing it to swell or shrink. The surface is too delicate to produce more than two thousand impressions. For these reasons, collotype has been used for luxury publications and, since World War II, has been largely abandoned for other commercial purposes. Collotype is a photographic process in which a film of gelatin provides the printing surface. The technique is dependent on the fact that light sensitized gelatin hardens in proportion to the amount of light to which it is exposed.
A solution of gelatin is poured over a sheet of plate glass. When dry the plate is placed in contact with a reverse negative and exposed to light. In proportion to the strength of the transmitted light the gelatin dries and hardens, while the unexposed areas remain capable of absorbing moisture. The surface can then be printed as a lithograph. The gelatin accepts the ink in an inverse proportion to the amount of moisture the surface retains, the driest areas accepting the most and therefore printing darkest.
Before starting the process of facsimile reproductions, two vital decisions needed to be made. First, how many base colors were to be printed in collotype to supply the framework for the facsimile. Second, how many stenciled colors would be required to reproduce an accurate watercolor copy of the original.
Then the original plate was photographed and a reverse negative was made for each base collotype color. Each negative was reworked by hand so that all the colors other than the base color for that negative were eliminated and the values of that base color were attenuated. Each negative required work, up to sixteen hours involving retouching and masking. The negative was then developed onto a previously prepared gelatin plate, which was baked in order to harden it sufficiently to take up to five hundred good impressions.
The success of the process depended on the Colorist analysis of colors for the base print, the subsequent layering of colors by the pochoir process (hand stenciling process) and on continuous comparison with the original plate. The layering of the base colors provided the skeleton of the work to be reproduced. Often up to three weeks' work by a skilled craftsman was required to prepare and make trials with the negatives before any actual printing was started. It was important for the stencil work to follow that these collotype printings be as perfectly 'modeled' as possible, and that the same craftsman who would later cut the stencils be in charge of preparing and printing the collotype negatives.
After the plate was put on the press, at least half a day was needed to 'work up' the color values. The printer could modify values by the application of chemicals onto the plate. If there were no accidents from the action of the chemicals that were used in the course of 'working up' the plate or from other hazards in the shop, and the correct temperature and humidity were maintained, the printer could expect to print about 350 sheets per day.
The hand stenciling process began after the base collotype was printed. In determining the base colors for the collotype printing, the Colorist also chose the amount and order for the application of colors by the pochoir process. The zinc stencils were cut by hand for each shade of color in the original, using a sheet of the collotype printing as a guide. A watercolor wash was then applied by hand for each color, using large beaver stencil brushes. A certain amount of direct hand work was needed to soften the edges. For this the stencil was lifted and a smaller brush was used.
Pochoir proofs were obtained by the superimposition of all the stenciled colors required for each plate over the base collotype. These proofs were corrected from the original and alterations in color and improvements in color order were made. The time required for obtaining a satisfactory first proof was about a month for a highly skilled craftsman. It then took approximately 7 - 9 weeks to apply the colors by hand through stencils for an edition of 400 copies, one plate only. Finally when the proofs were considered satisfactory, they served as models for the coloring of the collotype plates.
A Zinc stencil
A Guide sheet with stenciling information
The stencil was a zinc sheet cut in conjunction with the guide sheet, one for each layer of color to be washed onto the base print. Notice the corrections made to the stencils, areas covered up or more cut out than on the original guide sheet. These corrections were made after the proofs were colored and compared with the original.
Guide sheets were made from the base collotype print. One guide sheet was cut for each layer of color used in the stenciling process. The zinc stencils were placed under the base collotype print and were cut at the same time the guide sheet was cut.
Stenciler information sheets were the proof sheets used to correct colors and the order in which they were laid down. Stencil corrections were made from these sheets. These sheets were compared continuously with the original print and corrected until the colors matched. They were then used in lieu of the original as comparisons for the coloring of the rest of the edition.
Color decompositions were a record of the color applications . These sheets show step by step the subtle layering of the watercolors in the pochoir process. Each sheet is one color layer different from the one before. The final result is a finished print with great depth of color.