202 Dr Brewster on a Nezv Analysis of Solar Light. 



being indecomposable by the prism, and of being pre-eminent- 

 ly adapted for the nicest purposes of vision. The existence of 

 such light has never even been conjectured, and its insulation 

 at any point of the spectrum becomes a proof of the existence, 

 at that point, of red, yellow, and blue rays of equal refrangi- 

 bility. 



Having thus given a general view of the structure which I 

 have found in the spectrum, I shall now proceed to state the 

 experimental evidence from which it has been deduced. 



From the simple inspection of the coloured spaces, it is ob- 

 vious that red light exists in the red, orange, and violet divi- 

 sions of the spectrum ; but, according to Fraunhofer's mea- 

 surements, these three spaces occupy 190 parts when the whole 

 length of the spectrum is 360; hence, red rays are observed 

 in more than one-half of the whole spectrum. If we examine 

 the blue and indigo spaces through certain yellow fluids, such 

 as oil of' olives, they acquire a distinct violet tint, so that these 

 fluids must have absorbed some rays which had neutralized or 

 masked the red. Red light, therefore, exists in the blue and 

 indigo spaces ; and, as I shall afterwards show that white light, 

 which necessarily includes red, may be insulated both in the 

 green and yellow spaces, it follows that red light exists in all 

 the seven coloured spaces into which the spectrum is divided. 



Yelloxc light is distinctly recognized by the eye in the orange, 

 yellow, and green spaces which occupy 77 parts of the spec- 

 trum whose length is 360. When the spectrum is examined 

 with a deep blue glass, the green light is distinctly seen at F, 

 one of Fraunhofer's lines ; and as a green transparent wafer 

 of gelatine produces a whitish band beyond F, and in the blue 

 space, it is clear that a certain portion of yellow light must 

 exist there. We have already seen that the action of oil of olives 

 on the blue and indigo spaces absorbs certain rays and leaves 

 a violet tinge. These rays cannot be red, and they are not 

 blue, because blue taken from blue would not leave violet. 

 They must, therefore, be a small portion of yellow rays, which, 

 forming white with the red, and a portion of the blue, had the 

 effect of diluting the predominant blue light. The existence 

 of both yellow and red rays in the blue and indigo spaces, 

 may be inferred from another experiment. When we trans- 



