296 Ml'. Charles Tomlinson on 



colours, and if 1 include the theories which seek to explain 

 coloured shadows, Sec, the number will be nearly doubled. 

 It unfortunately happens, too, that each class of facts has 

 its own theory or theories attached to it, and the great 

 fault of all, including M. Plateau's, is the want both of 

 simplicity and comprehensiveness. If the theory that I 

 propose has any merit it is that of being very simple, and 

 at the same time it comprehends all existing facts, as far 

 as I am aware, (except those facts which belong to and are 

 explained by polarization,) for the more these various 

 phenomena are studied, the stronger, I am sure, will be 

 the conviction that they are all due to one cause, or 

 analogous causes, and must and ought to be explained by 

 one theory. 



28. In these two papers I have assumed that white light 

 is composed of red, yellow and blue, one or more of which 

 colours, by itself, or by combination, in various definite 

 proportions, produce all the ordinary effects of colour. 



I have endeavoured to establish the following propo- 

 sitions : 



1st. That two colours are complementary to each other, 

 when, from a combination by direct impression, white is 

 produced ; and by superposition, black. 



Thus, if a disk be divided into two parts, one painted red 

 and the other green, and the disk be rapidly rotated, the 

 colours will combine and produce white. This is an 

 instance of combination from direct impression. 



If we view red lead through a disk of green glass the red 

 powder will appear as black as lamp black. This is an 

 instance of combination by superposition. 



2nd. That if the fundamental colour be simple, the com- 

 plementary colour is compound ; and if the complementary 

 colour is simple, then the primary or fundamental colour 

 is compound ; but in certain cases both the fundamental 

 and complementary colours are compound, but never simple, 

 (See table 6.) 



3rd. That white light in passing through coloured trans- 

 parent media is decomposed, the second surface of such 

 media reflecting rays of the same colour as the medium 

 itself, and the first surface reflecting the other rays, which. 



