428 Dr. E. L. Nichols's new Explanation of 



cnles per unit of volume ; in other words, the mass of the 

 particles is in direct proportion to the density of the substance. 

 If reflexion and refraction result from impact between the 

 particles of the gas and of the vibrating aether, or from equi- 

 valent interaction, we should expect to find, in the case of the 

 different gases, that their power to reflect, and consequently 

 (in accordance with the formulae 



R _ sin 2 (i-r) 

 p sin 2 (i + r) 



for the reflected component polarized parallel to the plane of 

 incidence, and 



-o _ tan 2 (i—r) 

 s tan 2 (i + r) 



for the component polarized perpendicularly to that plane) 

 their indices of refraction are in direct proportion to their 

 density. How far this is from the fact appears from the fol- 

 lowing results of Dulong : — 



Refraction-index 

 compared with 

 Gas. Density. air. 



Air 1-000 1-000 



O 1-1026 0-924 



H 0-0685 0-470 



N 0-976 1-020 



CI 2-470 2-623 



CO 0*972 1*157 



CG 2 1-524 1-526 



Cy 1-818 2-832 



C 2 H 4 0-980 2-302 



CH 4 0-559 1-504 



Under the second hypothesis, the analogy between the bit 

 of wood tossed by water-waves and the process taking place 

 when light is reflected vanishes completely. Any body, how- 

 ever large or small, which is capable of causing the aether 

 pervading it to differ in density froni the aether within the 

 adjacent medium, must reflect and refract light, irrespective 

 of the wave-length of the rays. 



That peculiarity of the human eye the influence of which 

 upon the theory of the colour of the sky it is proposed to dis- 

 cuss, is as follows : — 



According to Young* and Helmholtz's theory of colour- 

 impression, there are in the eye three sets of nerve-termini — 

 one set chiefly influenced by the red, another by the green, 



* Dr. Young, Works edited by Peacock, vol. i. p. 147. 



