December i i, 1902] 



NA TURE 



129 



is for the production of printed illustrations, but for pur- 

 poses of demonstration its application to the production 

 of pictures for exhibition by the lantern is much more 

 convenient. By the use of a triple lantern the light from 

 a single source can be divided up into three beams. If 

 in the path of the beams we place screens of coloured 

 glass of colours corresponding with the three primary 

 colour sensations — red, green and blue — we have, of 

 course, a disc of each colour projected on the lantern 

 screen. If by moving the lantern lenses the three discs 

 are caused to overlap, the colours will be mixed and 

 combined. Where all three colours overlap there will 

 be a white patch. Where only two overlap there will be 

 a patch caused by the combination of those two colours, 

 and this of necessity will be complementary to the third. 

 We have therefore on the screen a coloured pattern 

 showing white, the three primaries and their three com- 

 plementary colours. 



If in front of each lens of the lantern we introduce a 

 simple pattern cut out of black paper, we shall, when 

 the three images are separated on the lantern sheet, get 

 three coloured reproductions of the three patterns. If 

 they are of a suitable shape and suitably arranged we 

 can combine these into a variegated pattern on the screen. 

 We may take, for instance, such a simple pattern as a 

 half-circle ; then if we arrange the three half-circles in 

 such a way that they do not coincide when projected 

 together on the lantern sheet, but combine and overlap 

 so as to form one complete circle, this circle will be 

 divided into six sectors, three of which will show the 

 primary colours and the other three their complemen- 

 taiies. 



This simple experiment shows that it is possible to 

 get a coloured picture by means of a black and white 

 pattern and the three coloured glasses. In it, however, 

 only the complementary colours are shown, because 

 equal amounts of the primaries are combined. To get 

 other tints, varying amounts of one or more of the 

 component colours have to be used. Experimentally 

 this is easily done by introducing in front of one of the 

 lenses of the lantern an optical wedge — a sheet of glass 

 coated with a neutral-tinted film, graduated from trans- 

 parency at one end to opacity at the other. By cutting 

 out, say, more or less of the red, we get a series of 

 browns, greyish blues, &c. ; by diminishing the green 

 we get salmon colour, yellow ochre, &c. By this means 

 it is evident that any desired tint which the human eye 

 can appreciate can readily be produced. 



Now a picture is only a complicated coloured pattern, 

 and if we can analyse a picture and resolve its colours 

 into the three components, arranged in their proper 

 shapes, the combination of these three components will 

 reproduce the picture as regards both shape and colour. 

 Such analysis is possible by photography. A photo- 

 graph taken through a red screen gives us the red com- 

 ponent, and by using blue and green screens the blue 

 and green components can be obtained. It is to be 

 remembered that these photographs are merely mono- 

 chrome photographs. They are simply ordinary photo- 

 graphs taken by a portion of the light of the spectrum, 

 instead of by the whole of it. Making positive prints 

 from negatives thus produced and projecting them on 

 the screen, they show like ordinary lantern slides, except 

 that each picture looks rather incomplete. In the red- 

 light picture blue objects are but faintly reproduced. In 

 the blue-light picture the red objects appear but feebly. 

 When the coloured glass screens are interposed in front 

 of the monochrome positives we get three pictures 

 coloured red, green and blue respectively, and a com- 

 bination of these on the sheet shows the original object 

 in all its varied colours. 1 



_ 1 The ingenious photochrotnoscope of Mr. Ives works or. precisely 

 similar principles, except thai the three coloured pictures are combined in 

 the eye of the observer, instead of on the lantern screen. 



NO. 172S, VOL. 67] 



The use of the triple lantern, however, is not very con- 

 venient, and there are certain drawbacks to its employ- 

 ment, though it suggests a possible means for the produc- 

 tion of kinematograph pictures in colour. This is not yet 

 possible, but it is conceivable that photographic films 

 might be made capable of taking instantaneous pictures 

 through the coloured screens, and that mechanism of suffi- 

 cient accuracy could be constructed to register a series 

 of three such pictures on a screen, so that they might be 

 shown in the way animated photographs are now shown. 



For practical purposes it is more convenient if we can 

 have our coloured pictures in the form of an ordinary 

 slide, which can be shown in the ordinary single lantern. 

 Now it is quite obvious that with a single lantern we 

 cannot use three coloured screens, one in front of the 

 other. In the triple lantern we are mixing coloured 

 lights, adding colour to colour. The superposition of 

 one screen upon another in a single lantern merely means 

 that only those rays will pass which can get through both 

 screens, and the three screens together in the lantern 

 would, of course, obstruct all the light, and the result 

 would be nothing but darkness, With the triple lantern 

 we are using a method of addition ; with a single lantern 

 we must use a method of subtraction or absorption. 



The end can, however, be attained by the use of a 

 film of bichromated gelatin, coated on a celluloid support. 

 The film is printed and washed in the usual manner 

 of carbon printing. The resulting relief in colourless 

 gelatin is then stained the complementary colour to that 

 by which the negative was taken. The need for em- 

 ploying the complementary colour is not difficult to 

 understand. The bright parts of the red-screen 

 positive represent bright red light. The dark parts 

 represent the absence of red light, red shadows. 

 When the film is stained, the transparent parts take 

 little or no stain, the denser and thicker parts take the 

 stain in proportion to their thickness. They should 

 therefore be stained the opposite to red, the comple- 

 mentary to red (it is convenient to think of it as "minus 

 red''), or blue-green. So the green-screen print must be 

 stained " minus green," or pink, and the blue-screen 

 print must be stained "minus blue,'" or yellow. 



If we now take the three films and put the blue film in 

 the lantern, we get a blue picture on the sheet. Putting 

 in front of this the yellow film, our picture becomes 

 partly blue, partly yellow and partly green, and we have 

 some accession of detail. Adding again to this the pink 

 film, we get at once all the different colours of the 

 original object, and the picture is recognised as a practi- 

 cally correct reproduction of the original. 



If the three films, instead of being mounted in such a 

 way that they can be shown in the lantern, are stripped 

 from their supports and superposed one above the other 

 on a sheet of white paper, we get a coloured picture 

 suitable for use as a book illustration. This process is 

 quite practical, but it is by no means easy, and, of course, 

 it is useless for the production of large numbers. For 

 commercial purposes no process can be of much service 

 which is not applicable to the printing-press. Now it 

 must be familiar to most people that a printing-block 

 can be produced from any photographic negative. The 

 methods by which this is effected are well known, and 

 they are in constant use, the great bulk of the black and 

 white illustrations in magazines and newspapers being 

 now produced by them. It is, therefore, not difficult to 

 see that if from each of our negatives we make a 

 printing-block and use the three blocks to print— the 

 blue-screen block in yellow ink, the green-screen block 

 in red ink and the red-screen block in blue ink — we are 

 merely varying the process by substituting films of 

 printing-ink for films of stained gelatin. This is, 

 indeed, in barest outline the method by which the very 

 numerous coloured illustrations made by the three-colour 

 process are all produced. 



