Astronomical and Nautical Collection^, 213 



are substances in which the velocity of propagation varies 

 in the same medium, with the direction of the rays, and then 

 those which are thus affected are no longer refracted in the 

 same manner. 



The relation which we have just found, between the sines 

 of the angles of incidence and of refraction, agrees perfectly 

 with the experiment of Mr. Arago, which shows that the 

 lengths of the undulations of light in different media are to 

 each other as the sines of incidence and of refraction at the 

 passage of the light from one medium into the other : and this 

 relation explains, at the same time, why the plates of air 

 and of water, which reflect the rings of the same colours, are 

 to each other as the sines of incidence and refraction of the 

 light which passes from air into water. 



If we generalise the considerations which have been em- 

 ployed for explaining the ordinary law of refraction in the 

 particular case of a continued and extensive surface, we may 

 determine, by means of the same formulas which represent 

 the phenomena "of diffraction, the niuch more complicated 

 laws which are followed by the refracted rays, when the sur- 

 face is narrow or interrupted ; and we always obtain, in this 

 manner, results which are conformable to experience ; which 

 proves both the accuracy and the generality of the principles 

 of Huyg'ens, and of that of interferences, on which the whole 

 of this theory is founded. 



It is impossible to conclude this concise account of refrac- 

 tion, without offering some theoretical views of an optical 

 phenomenon which always accompanies it, which has been 

 much studied, but which is, perhaps, still very little under- 

 stood ; that is, the division which light undergoes in passing 

 through a prism, and to which the name of dispersion has 

 been given, because it separates, and in some measure dis- 

 perses the coloured rays, of which white light is composed, 

 and makes them follow different routes. It results from this 

 phenomenon, that the rays of different colours are not all 

 equally refracted, or that the undulations of different lengths 

 are not propagated with the same velocity in the same me- 

 diums : for it follows, from the explanation which has been 

 given of refraction, that the relation between the sines of inci- 



