136 NEWTON'S THEORY OF COLOURS. 



of the colours over the spectrum, and regard white light as 

 consisting only of three rays, which in the prismatic images 

 overlap each other ; and from these red, yellow, and 

 blue all the others can be formed by combination in 

 varying proportions. The truth will probably be found 

 to be, that the ordinary prismatic spectrum is a com- 

 pound of two spectra : that is, as we have the ordinary 

 rainbow, and a supplementary bow, the colcurs of which 

 are inverted, so the extraordinary may be somewhat 

 masked by the intense light of the ordinary spectrum ; 

 and yet by overlapping produce the variations of colour 

 in the rays. We have already examined the heating 

 power found in these coloured bands, which, although 

 shown to be in a remarkable manner in constant 

 agreement with the colour of a particular ray, is not 

 directly connected with it ; that is, not as the effect 

 of a cause, or the contrary. The chemical action of the 

 solar rays, to which from its important bearings we shall 

 devote a separate chapter, has, in like manner with 

 heat, been confounded with the sun's luminous power ; 

 but although associated with light and heat, and modi- 

 fied by their presence, it must be distinguished from 

 them. 



We find the maximum of heat at one end of the 

 spectrum, and that of chemical excitation at the other 

 luminous power observing a mean point between them. 

 Without doubt we have these powers acting recipro- 

 cally, modifying all the phenomena of each other, and 

 thus giving rise to the difficulties which beset the in- 

 quirer on every side. 



We have beautiful natural illustrations of luminous 

 refraction in the rainbow and in the halo : in both cases 

 the rays of light being separated by the refractive power 

 of the falling rain drop, or the vesicles which form the 

 moisture constituting a fog. In the simple toy of the 

 child the soap-bubble floating upon the air the phi- 



