300 



•37 Vermilion, -( J .^g White 

 + •27 Ultramarine, V = < .^^ j., , 

 + -36 Emerald green, J I + ^"^ ^^^^^ 



By experiments on various individuals, it was found (1.) that a 

 good eye could be depended upon within two of these divisions, or 

 hundredths, at most ; and that by repetition of experiments the 

 average result might be made much more accurate. 



(2.) That the difference of the results of experiments on different 

 individuals was insensible, provided the light used remained the 

 same. 



(3.) That when different kinds of light were used, or when the 

 resultant tints were examined with coloured glasses, the results were 

 totally changed. 



It follows from this that the cause of the equality of the result- 

 ing tints is not a true optical identity of the light received by the 

 eye, but must be sought for in the constitution of the sense of sight. 

 The materials for this inquiry are to be found in the equations of 

 colour of which the above is an example, and these are to be viewed 

 in the light of Young's theory of a threefold sensation of colour. 



The first consequence of this theory is, that between any four 

 colours an equation can be found, and this is confirmed by experi- 

 ment. 



The second is, that from two equations containing different colours 

 a third may be obtained by the ordinary rules, and that this also 

 will agree with experiment. This also was found to be true by ex- 

 periments at Cambridge which include every combination of five 

 colours. 



A graphical method was then described, by which, after fixing 

 arbitrarily the positions of three standard colours, that of any 

 other colour could be obtained by experiments in which it was made 

 to form a neutral gray along with two of the standard colours. In 

 the diagram so formed, the position of any compound tint is the 

 centre of gravity of the colours of which it is composed, their masses 

 being determined from the equation, and the resultant mass of 

 colour being the sum of the component masses. The colour-equa- 

 tions represent the fact that the same tint may be produced by two 

 different combinations. This diagram is similar to those which 

 have been given by Meyer, Hay, and Professor J. D. Forbes, as the 

 results of mixing colours. It is identical with that proposed by 



