It is advisable to make up stock solutions of the dyes and add these to plain gelatine solution. The stock solution for the blue filter is made from:

Crystal violet 3 g.

Methylene blue 1 g.

Glacial acetic acid 5 to 6 drops.

Warm water 100 ccm.

A 6 per cent solution of plain gelatine will also be required. The actual dyed gelatine should be:

Gelatine solution 747 ccm.

Dye solution 23 ccm.

This is sufficient for 1 square meter of filter surface.

For the green filter the following stock solution should be prepared:

Tartrazin 6 g.

Patent blue 1 g.

Naphthol green 2 g.

Warm water 180 ccm.

The green filter is the one that gives the most trouble and requires the most adjustment so as to obtain white light, for the three filters must give white light at the eyehole 0. It is as well, therefore, to make about five filters of different depths of coloring, which can be done by adding varying quantities of the dye solution to the plain gelatine:

Gelatine solution 690 ccm.

Dye solution 8 to 16 ccm.

It is as well to start with 8 ccm of the dye solution, and then make other mixtures with 10, 12, 14 and 16 ccm of dye solution, so that five different depths of filters will be obtained. The above quantity is sufficient for 1 square meter of filter surface.

The red filter should be made with the following stock solution:

Tartrazin 4 g.

Rose Bengal 3.5 g.

Water 150 ccm.

The actual dyed gelatine will be:

Gelatine solution 432 ccm.

Dye solution 38 ccm.

This is sufficient for 1 square meter of surface.

As the mirrors M and M' must not only reflect the pictures at L and L', but also transmit that at L", ordinary mirrors are not suitable. Half silvered or platinized mirrors may be used, but these are costly and easily tarnish, so that plate glass should be used, and as they should not reflect the pictures from the back surfaces they ought to be colored. The easiest way to make these is to coat them with dye solutions, and they should be cemented to plain glasses with Canada balsam.

The dyed gelatine for the green mirror should be made from the following stock solution:

Tartrazin 0.5 g.

Naphthol green 0.8 g.

Patent blue 1.2 g.

Water 250 ccm.

To every 100 ccm of the plain gelatine solution should be added 8 ccm of the dye solution, and the mixture filtered. For a mirror 135 x 135 mm, 13 ccm of the dyed gelatine should be used.

For the blue mirror use the following:

Patent blue 1 g.

Warm water 100 ccm.

To every 100 ccm of gelatine solution allow 4 ccm of the dye solution and coat the same quantity as above, namely 13 ccm, on the same size mirror.

These colored mirrors should now be fastened in the box at an angle of 45 degrees with the base. The blue mirror should be placed under the red filter, and the green mirror under the blue filter.

The instrument should be inclined towards the sky and the inclination of the mirrors altered until the images of the filter apertures are seen as far as possible in coincidence. It is advisable to cover one of the apertures and adjust one mirror first, and then, having obtained satisfactory agreement (narrow colored fringes at the margins can be ignored), cover up the aperture just brought into coincidence and use the third aperture in the same way.

If the filters are correct, the field of view should be whitish and very bright; but care must be taken to use the reflector, as if this is not done one or other of the colors will predominate. If the field is too red the green filter is too dark; with a greenish field the green filter is too bright. And it is here that the use of the extra filters comes into play, as one can substitute one for the other till a good white is obtained. If the field is yellowish, the blue filter is too dark; if it is blue the red filter is too dark or the blue filter is too bright. As a rule if the filters are prepared as advised above it is only the green filter that will require adjustment. An absolute white is extremely difficult to obtain, but a great preponderance of one or the other color can be avoided.

As soon as the correct filters have been found they should be permanently fixed in the apertures. To hold the transparency at L", two wooden or metal grooves should be provided. At L and V spring metal clips can be used.

The green filter should be placed at L", the red may be either at L or L', and the transparency from the green-filter negative should be placed here. The red-filter transparency should be placed on the red filter and the blue on the blue filter. The latter should be temporarily covered with an opaque card, and the red transparency shifted till the outlines of the images coincide, and then the card removed from the blue picture and this adjusted in register.

If the measurements of the instrument are not exactly correct, the three images may coincide but may not lie in the same plane; for instance, the red image may be in front or behind the green.

If the red image is in front of the green transparency, the distance LM is too small; if it lies behind, LM is too great. This may be remedied by shifting the mirror M backwards, or by altering the distance LM or GH. It is as well to bear this in mind when making the apparatus and arrange so that the heights GH and CJ may be slightly altered by inserting a wooden frame or planing off a little. This regulation requires to be done only once, while the focusing of the images must be done each time. To avoid adjusting it each time, the following plan may be adopted: the red and blue transparencies should be provided with narrow card margins, which should be stuck on. On the edges GK and JE should be fastened a T-square of wood or metal, only a few millimeters long, against which the picture edge should be laid. It will now be easy by careful and gradual paring away of the card to so arrange that the transparencies at once coincide, and they will then only have to be pressed up against the T-square to fall in register.

Frequently great improvement may be effected by placing over the green filter other green glasses of faint tint, which can be prepared by coating glasses with varying amounts of the green-dyed gelatine. These compensate for the varying composition of daylight. It is not possible to use artificial lights with this instrument, unless the composition of the light is brought to approximately the same as daylight. The instructions given for the preparation of the viewing filters for screen-plates on page 165 may be followed.

With such an instrument one may perform many interesting experiments, which, although not color photography, are well worth trying. For instance, by placing the transparencies in the wrong apertures all sorts of colored results may be obtained. Negatives may be used and then images in the complementary colors will be seen. And obviously by covering up any aperture the complementary color is at once seen from the admixture of the other two. By using cut out patterns, too, some wonderful color effects can be obtained.