UNFORTUNATELY there is no reasonably priced three-color camera on the market. So, unless one is prepared to make a camera, the only alternative is a sliding back, and this limits the work in the main to subjects in which there is no movement, as obviously three successive and not simultaneous exposures must be given.

Many complicated forms of cameras have been devised, with costly optical arrangements for obtaining the three constituent negatives with one exposure, but these we can ignore. Possibly the simplest form of one-exposure camera is the two-step, of which the best constructive details were given by C. E. K. Mees, and his instructions are here followed. The camera was originally devised by E. T. Butler, on the lines of the photochromoscope suggested by Cros in 1871. The camera takes a seven-inch lens, but can be used with any lens that has not a shorter focus than this. The size of the plate determines the extension of the camera, and also decides as to whether the longer or shorter axis of the plate is vertical or horizontal. The lens should be one of the newer anas-tigmats and work at as large an aperture as possible, for the filters prolong the time of exposure. A reflector is placed behind the lens at an angle of 45 degrees to the axis of the lens. Part of the light is reflected and passes through a filter to the plate, while the remainder which passes through the first reflector meets a second reflector placed parallel with the first. Here the light is again divided; the reflected part passes to the second plate through another filter, while the rest of the beam goes onward to the third plate behind the second reflector. It is necessary to use the three filters, and the images must be of identical size; in addition, the reflectors must not give two images, so that they must be so constructed that the second image, due to reflection from the back of the glass, is rendered harmless.

All the images have to be in focus at the same time, and if a black and white object be taken it must be rendered of the same value in all three negatives Only one kind of plate must be used, that is, one cannot use a panchromatic plate for the red exposure, an orthochro-matic for the green and an ordinary for the blue. As regards the nullification of the double image this can be attained by using colored reflectors, and obviously this can be most satisfactorily done by coating the back of the glass with colored gelatine. This coloration must be the minus color of the taking filter; for instance, in the diagram (Fig. 10) it will be seen that the first reflector, which gives the red record, is designated as a "minus red reflector," and a minus red must be blue-green. In like manner a minus blue must be yellow. It will be found that the arrangement suggested in the diagram will be the most satisfactory, though the red and green sensitive plates may change places. Then, of course, the colors of the back coatings of the reflectors must also be changed, and that for the first reflector would be a minus green, or crimson. Identity of size of the images can be secured by altering the angle of the reflectors, as if they are raised the central beam of light is shortened. The reflectors must be separately adjusted so that the focus of the reflected images is the same in each case, which means that the length of the optical path in glass must be equal for each plate. For this purpose, the thickness of the filters must be adjusted so that the length of glass through which each beam travels is equal. The length of the glass through which the direct beam travels is equal to the thickness of the two reflectors taken at an angle of 45 degrees to the axis of the lens. That is to say, it is 1.41 times the actual thickness of the filters. The thickness of the red filter is, therefore, equal to 1.41 times the thickness of the two reflectors, while the thickness of the blue filter will be 1.41 times the thickness of the yellow reflector.

Fig. 10.

In order to adjust the color of the filters so that the correct ratio may be obtained, pieces of film should be placed between the glasses and a black and white chart photographed till equal density results are obtained on all three plates. It will be seen that the adjustments are rather delicate and involve a considerable amount of work by trial and error, but this should not deter anyone from attempting to make such an instrument, as it is the cheapest and most convenient camera for a single exposure that can be devised. It is quite possible also that a camera constructed on the line of the chromo-scope, described on p. 125, would be simpler, though not quite so compact. The dimensions are given there, so one can adapt it for negative work, but as the depth of the instrument is 10 2/5 inches, it obviously entails the use of a lens of focal length not less than this, though this depth can obviously be cut down.

As an alternative to the above described camera, the following may be adopted. This has some advantages, particularly as regards the evenness of illumination of the reflected images, which is frequently faulty with parallel reflectors. This type was patented in 1896, and some changes in the disposition of the filters have been recently suggested by H. E. Rendall. The internal construction is shown in the accompanying diagram (Fig. 11), in which L is the lens in a focusing jacket, or obviously a short length of bellows could be placed here. A, the first filter, of a mauve color, that is, one transmitting red and violet, is placed at an angle of 45 degrees to the optical axis of the lens, and from it some of the light is reflected to a plate at P1, in front of which is a green filter G. The remainder of the light passes through A and some is reflected from B, a blue filter at right angles to A, to the second plate P2, in front of which is an orange filter O. The remainder of the light passes through B to P3, in front of which there is no filter, the reflector B acting as filter.