814 FOURTH RKPORT — 1834. 



tliat of the surrounding ether. On this supposition Fresnel has 

 calcuhited the length of a wave in the moving medium, and 

 thence also the actual change in the direction of the refracted 

 ray produced by the earth's motion *. This change is found to 

 be opposite, and exactly equal to that produced by the same 

 cause in the apparent direction of the ray ; so that the ray is 

 actually seen in the same direction as if the earth were at rest, 

 and the apparent refraction is unaltered by the earth's motion. 

 These results, it may be observed, are precisely the same for 

 terrestrial objects, the velocity of wave-propagation being inde- 

 pendent of the motion of the luminous body. 



Newton tliought that the different refrangibility of the rays of 

 light could be explained by supposing simply that they were 

 bodies of different sizes, the red being greatest and the violet 

 least. It is obvious, however, that this supposition can have no 

 reference to the simple projectile hypothesis held by his followers, 

 or to the demonstration of the law of refraction given in the 

 Prmcijiia. It is connected with that more complex theory, in 

 which the molecules of light are supposed to excite the vibrations 

 of the ether in the bodies which they meet. 



M. de Courtivron and Mr. Melville proposed to account 

 for the dispersion of light by a difference in the initial velocity 

 of the molecules, the red being swiftest and the violet slowest. 

 But were such the cause of the phenomenon, the dispersion 

 should be proportionate to the mean refraction. Indeed the 

 hypothesis was abandoned almost as soon as proposed. Its 

 authors had foreseen the consequence that, in the eclipses of 

 Jupiter's satellites, the colour of the light should var}' just before 

 immersion, and after emersion ; and the existence of such an 

 effect, in the degree indicated by theory f, was completely dis- 

 proved by the observations of Mr. Short J. Another conse- 

 quence of such a difference in the initial velocities of the light of 

 different colours is, that the aberration of the fixed stars should 

 also vary with the nature of the light, and each star appear as a 



* The sine of the change is to the sine of the total deviation of the ray in the 

 ratio of the velocity of the earth to that of light. Fresnel's result is much more 

 complicated, but it will be easily seen to reduce itself to this. — " Sur I'lnfluence 

 du Mouvement terrestre dans quelques Phenomenes d'Optique," Annales de 

 Chimie, torn. ix. 



f The duration of this change, according to Mr. Melville, should amoiuit 

 to thirty-two seconds, the velocity of the light of different colours being in- 

 versely as their refractive indices. — {Phil. Trans. 1753.) This principle, how- 

 ever, as M. Clairaut has shown {Phil. Trans. 1754), is obviously incorrect. 

 It will easily appear that the initial velocities must vary inversely as 

 the quantity /^(^^ — 1, in order to account for dispersion ; and that the dura- 

 tion of the expected phenomenon must be even greater than that assigned 

 by Mr. Melville. 



: Phil. Trans. 1753. 



