470 JAMES CLERK MAXWELL. 



disc could be made visible from above. The top consisted 

 of a disc of metal, covered with white paper and mounted on 

 a suitable spindle, the circumference of which was divided 

 into 100 equal parts. Two sizes of discs were employed ; 

 the larger discs having been adjusted as required, were 

 placed on the graduated plate, and the smaller discs above 

 them. The plate then presented an appearance similar to 

 that sketched in the adjoining figure. The employment 

 of the apparatus depends on the fact, observed by Hartley 

 the psychologist, that visual impressions remain for some 

 time on the retina even after the source of light has been 

 removed. Thus, suppose that the larger discs are respect- 

 ively red, green, and blue, and suppose that we look at the 

 middle of the red sector and then spin the top. If the 

 spinning is sufficiently rapid the impression produced by the 

 red light will not have diminished sensibly before the green 

 paper takes the place of the red, and the green light pro- 

 duces its effect on the same portion of the retina, and this 

 again is followed by blue before either the red or the green 

 sensations have sensibly diminished, and then the whole 

 process is repeated many times in a second. The impressions 

 due to the three coloured lights are therefore blended on the 

 retina, and it can be shown experimentally that the tint 

 observed is the same as when the three kinds of light in 

 the same relative proportions are allowed to fall simultan- 

 eously upon the eye. The colour top, therefore, enables us 

 to mix together, in any definite proportions, which can be 

 changed at will, the lights from differently coloured papers, 

 and to observe the effect produced by the mixing. 



Suppose that the colours of the larger discs are red, 

 green, and blue, and that these are arranged so that the 

 coloured lights may be mixed in the proportions necessary 

 to form white light. Then the colour of the ring exhibited 

 by the larger discs when the top revolves can differ from 

 that of white paper only in respect of illumination. If, 

 therefore, we can diminish the apparent brightness of the 

 central circle at will, we may produce a complete match. 

 This may be effected by combining a black disc, which 



