424 SIR DAVID Brewster's description of the lithoscope. 



] St, By dividing the reflecting surface of the prism into two, three, or more 

 parts, by means of grooves, so that two, three, or more oils may be used at 

 once; and, 



2d, By using a prism composed of two, three, or more portions of glass 

 with different refracting powers. 



By either of these means, or by the two in combination, the experimental 

 results may be more readily and accurately compared. 



In the preceding experiments our attention is called to two different pheno- 

 mena — the intensity and the colour of the reflected pencil. When the pencil is 

 colourless, which is generally the case when one of the surfaces is that of glass, 

 its intensity depends on the index of refraction of the surfaces in contact, which 

 is always equal to the quotient of the greater index of refraction divided by the 

 lesser. 



The phenomena of colour produced by feebly reflecting surfaces, and first 

 described in my paper of 1819, already referred to, arise from two different 

 causes — 



1st, In doubly refracting crystals, from the influence of the doubly refracting 

 force in turning through different angles the planes of the polarisation of the 

 differently coloured rays ; and, 



2d, In the same class of bodies, and in those which have no double refraction, 

 from the different dispersive powers of the surfaces in contact, and also from the 

 irrationality of dispersion, in consequence of Avhich the coloured spaces in different 

 spectra have not the same proportion to each other. 



To the subject of crystalline reflexion in which the phenomena of colour are 

 exhibited, I shall have occasion to direct the attention of the Society in another 

 paper. When the colour is produced by difference of dispersive power, or by 

 irrationality in the spectra, it is easily explained. In a combination, for example, 

 of flint glass and oil of cassia, the index of refraction for red light is the same in 

 both bodies, but different for all the other colours. The i^ed light, therefore, in a 

 perfectly colourless beam, will pass through the surface of the oil and the glass, 

 without suffering reflexion ; and, consequently, the light reflected must be bluish, 

 while the transmitted light will be yellowish. As there is no reason for believing 

 that in any two bodies, whether gaseous, fluid, or solid, the dispersive powers are 

 exactly the same, or the coloured spaces exactly proportional, we may assert 

 that when pure white light is either transmitted by, or reflected from, bodies 

 perfectly colourless, the transmitted and reflected pencil must be coloured, how- 

 ever inappreciable the colour may be.* 



* See " Memoirs of the Life and Writings of Sir Isaac Newton," vol. i. p. 163. Second edition. 



