1835.] Composition of White Light. 109 



The observations and experiments which have since con- 

 firnied these views are so numerous, that there is some 

 danger of my being tedious, though it is my intention to 

 select only what I consider necessary to render the proof 

 satisfactory. In order to avoid this I shall give a preference 

 to subjects which appear to be capable of deriving improve- 

 ment from further investigation ; and thus endeavour to 

 give an interest to the inquiry, where it might otherwise 

 be deficient of it. 



It is evident that if we could ascertain the complemen- 

 tary colours of the different colours which we suppose to 

 be primary, it would set the question at rest ; but, simple 

 as the means are which have been proposed for this pur- 

 pose, there is a discrepancy in the results hitherto obtained, 

 which is calculated to weaken our confidence, and to con- 

 fuse rather than elucidate the point at issue. 



The la\Y of accidental colours, is thus stated by Sir 

 David Brewster, in his Treatise on Optics: — " The acci- 

 dental colour of any colour in a prismatic spectrum, is that 

 colour which in the same spectrum is distant from the first 

 colour, half the length of the spectrum." 



In another part of the same work, the following is given 

 as the results obtained by Fraunhofer, with flint glass ; and 

 by Newton, with a prism, the composition of which is not 

 stated. 



360 360 



Now, if we take the centre of Newton's red at 23, the 

 orange 27, the yellow 40, and the green 60, we require, 

 according to this law, 30 of the blue to make half the 

 length of the spectrum ; so that if blue be the complemen- 

 tary colour of red, the law in this instance, which brings 

 the centres of these two colours to correspond, is correct. 



