THE PHOTOGRAPHY OF COLOUR 365 



red lines, leaving it exactly filling the spaces between them. 

 The plate is put into a solution of a blue dye to stain the new 

 lines. By this method the lines can be made more than twice as 

 fine as by ruling (more than 600 to the inch), and the troubles of 

 overlapping and of uncoloured gaps are entirely obviated. The 

 method of using these plates, when they are issued to the public, 

 w^ill doubtless be similar to that of the Lumiere " Autochrome " 

 plates, described in the next paragraph. Their special 

 advantages will be referred to later. 



It may be here mentioned in passing that during the last 

 twelve years it has been shown by many workers to be possible 

 to get three colours that will serve more or less w^ell by the use 

 of prisms or diffraction gratings instead of dyes or pigments. 

 But these methods do not appear to lend themselves kindly to 

 practical work, even if no exception is taken to the principles 

 upon which they are based. 



Instead of arranging the three colours in lines, they may be 

 disposed in small hexagon or square patches. These methods 

 have been patented, but nothing practical has 3^et come of the 

 idea. The third alternative is a random grain, and a method of 

 doing this by means of starch granules, which was described by 

 Messrs. Lumiere, of Lyons, rather more than three years ago, 

 has just been commercially perfected. Quantities of starch 

 granules of approximately uniform size are stained respec- 

 tively red, green, and blue, mixed as thoroughly as can be in 

 such proportions as present a neutral grey to the eye, and 

 dusted over a prepared plate so that they adhere in a single 

 layer, which is then pressed or dusted with a black powder (or 

 both) to fill up the small spaces between the granules. A pro- 

 tective waterproof varnish is applied, and on this is spread an 

 orthochromatic emulsion, so that the plate is complete in itself — 

 photographic plate and colour screen in one. It is exposed on 

 the camera exactly as usual, except that the glass side of the 

 plate is put towards the lens, so that the light has to pass 

 through the laj^er of coloured starch granules before it affects 

 the sensitive film. The plate is developed, and the resulting 

 negative, instead of being fixed, has the metallic silver that 

 constitutes the image dissolved away, and the remaining silver 

 salt reduced to the metallic state, thus transforming the negative 

 into a positive. When viewed as a transparency, the colours as 

 well as the form of the original are seen. The preparation of 



