1896.] on Colour FJiotograjphy. 155 



«very other position we see nothing but a colourless negative. Now, 

 as you are aware, the colours of pigments are seen in any direction. 

 By projecting again a photograph of the spectrum, and turning it to 

 and fro, I can show you that the colours are visible only in one 

 direction. 



Thirdly, if we change the incidence of the illuminating rays, that 

 is, if we look at the plate first in a normal direction, then more and 

 more slantingly, we find that the colours change with the incidence 

 exactly as they do in the case of soap bubbles, or of Newton's rings ; 

 they change according to the same law and for the same reasons. 

 The red end of the spectrum turns successively to orange, yellow, 

 green, blue and violet. The whole system of colours, the image of 

 the spectrum, is seen to move down into the part impressed by the 

 infra-red. This is what we expect to happen with interference 

 colours, and what again we cannot obtain with pigments. 



Fourthly, if while looking at the film normally, we sufier it to 

 absorb moisture — this can be done by breathing repeatedly on its 

 surface — we see that the colours again change, but in an order oppo- 

 site to that above described. Here the blue end of the spectrum is 

 seen to turn gradually green, yellow, orange, red, and finally infra- 

 red, that is, invisible. The spectrum this time seems to move up into 

 the ultra-violet part of the improved film. By suffering the water to 

 evaporate, the whole image moves back into its proper place ; this 

 experiment may be repeated any number of times. 



The same phenomenon may be obtained with Newton's apparatus, 

 by slowly lifting the lens out of contact with the plane surface. The 

 explanation is the same in both cases. The gelatine swells up when 

 imbibing moisture. If we consider, for instance, the violet of the 

 spectrum, the small intervals between the strata corresponding to 

 violet rays, gradually swell up to the values proper for green, and for 

 red, and for infra-red ; green, then red, then infra-red are therefore 

 successively reflected. 



We will wet this photograph of the spectrum with^water, project 

 it on the screen, and watch the colours coming back in the order pre- 

 scribed by theory. 



It is necessary to use a transparent film, since an opaque one, 

 such as is commonly in use, would hide the mirror from view ; the 

 sensitive substance must be grainless, or at least the grains must be 

 much finer than the dimensions of the strata they are intended to 

 form, and therefore wholly invisible. The preparation of transparent 

 layers gave me at first much trouble ; I despaired for years to find a 

 proper method for making them. The method, however, is simply 

 thus : if the sensitive substance (the silver bromide, for instance) be 

 formed in presence of a sufficient quantity of organic matter, such as 

 albumen, gelatine or collodion, it does not appear as a precipitate ; it 

 remains invisible ; it is formed, but seems to remain dissolved in the 

 organic substratum. If, for instance, we prepare a film of albumeno- 

 iodide in the usual way, only taking care to lessen the proportions of 



