KEFLEXION AND REFRACTION OF LIGHT. 49 



David Brewster have both remarked that the flame of oil, urged 

 by the blowpipe, consists chiefly or wholly of yellow rays. The 

 same fact was long since observed by Mr. Melville with respect to 

 the flame of alcohol into which nitre, muriate of soda, and other 

 salts had been introduced ;* and Sir David Brewster has found 

 that the quantity of yellow light given out by burning bodies 

 increases with their humidity, the flame of alcohol diluted with 

 water being nearly a homogeneous yellow, f It is more important 

 to remark, however, in illustration of the undulatory view of the 

 phenomenon of emission, that the colour of flames is often found 

 to depend on the presence of something which is itself unaltered 

 in the process of combustion. Thus Mr. Talbot has remarked, 

 that when a small quantity of muriate of lime was placed on the 

 wick of a spirit lamp, it gave out red and green rays during an 

 entire evening, though the salt was not sensibly diminished. J 

 The absence of definite rays in certain lights, and the fixed lines of 

 the solar spectrum, have been referred by Sir John Herschel to 

 the same principle by which he has explained the absorption of 

 specific rays. 



In what has preceded we have assumed the truth of the re- 

 ceived theory with respect to the composition of solar light, and 

 the connexion between the colour of a ray and its refrangibility. 

 This theory, however, has been recently opposed by Sir David 

 Brewster. According to this philosopher, white light consists of 

 but three simple colours red, yellow, and blue ; and the solar 

 spectrum is composed of three overlapping spectra of these colours, 

 the intensity of each of which is greatest at the point where that 

 colour is strongest in the compound spectrum. According to this 

 view, then, all the colours in the solar spectrum are compound, and 

 consist of red, yellow, and blue light, in different proportions. 

 These compound colours cannot be analyzed by the prism, inas- 

 much as the rays of Which they consist at any point of the spec- 

 trum have the same refrangibility ; and it is only by the different 

 action of absorbing media on their constituent elements that their 

 compound nature can be detected. Each of them may be con- 



* Edinb. Estays. 



t On a Monochromatic Lamp, Ibid. 



I Edinb. Journ. of Science, v. 77. 



Phil. Mag., Third Series, vol. iii. p. 407. 



