to Three-colour Work, 



:u 



so formed can be increased in any part of the spectrum by printing 

 all three inks less fully, and the best result will be obtained when the 

 added white is everywhere proportional through the spectrum to the 

 luminosity of the colour. , 



The three inks we have been discussing are supposed to absorb the 

 spectrum thus : 



Y Or B 



BUie Ink. 



Y fir B 



Pink Ink. 



"f Or B 



YeLUiw Ink 



The curves to be used in printing these must be adjusted so that 

 the spectrum may be matched in luminosity as well as hue, the added 

 white being itself (roughly) proportional to the luminosity. To adjust 

 the luminosity of the colours, it will be observed that the colours 

 which are relatively too bright can be reduced by printing more fully 

 with all three inks. The curves are roughly 



R ^ ^ ^ WE 



BitjelnH FiUer Curve. Pink Ink Fitt-er Curve. YeU^ow Ink Filter Curve. 



I have calculated the exact curves for such inks to enable them to 

 match exactly the colour sensations as determined by Abney, and the 

 results are given in the second part of this paper. 



4. The Effect of the Addition of White. 



There are several ways in which this can be shown not to be im- 

 portant when the amount of white is not very great. 



In the first place, we are accustomed to use colours which are diluted 

 with white. According to Abney the light reflected from vermilion 

 contains 2 '5 per cent, of white, from cobalt 55 per cent., from French 

 ultramarine blue 61 per cent., and from chrome yellow 26 per cent, of 

 white (p. 166, ' Colour Measurement ') ; whilst in the same way the 

 light transmitted by a ruby glass contains 2 per cent, of white, by a 

 canary glass 26 per cent, of white, by a green glass 31 to 61 percent., 

 by a cobalt glass 42 per cent, of white. That is to say, the trans- 

 mitted light from a cobalt glass can be matched by 58 per cent, of a 

 spectrum colour with 42 per cent, of white. 



Then again about 3 per cent, of white can be added to orange, 

 rather less to green, and so less along the spectrum till 0*8 per cent, in 



