VELOCITY OF LIGHT. 241 



parent bodj. If this velocity increase, the refraction 

 will be less, and, reciprocally, a diminution of velocity 



that it would, on any obvious a pricn-i grounds, enable us to predict 

 the result of such an experiment one way or the other. There is 

 indeed involved the difficult and complex consideration of the propa- 

 gation of vibrations through eether, while the earth and transparent 

 media upon it are moving through that cether; a problem which ex- 

 ercised the ingenuity of Fresnel, and which, after a long investigation, 

 he decided by concluding that the effects would be exactly the same 

 as if the earth were at rest. This, however, may be still regarded as 

 one of those points connected with what is the most difficult part of 

 the wave theory, viz : the primary conception of a3ther and its prop- 

 erties. 



But apart from this consideration, and looking only to the abstract 

 problem of light (suppose emitted on the surface of the earth) falling 

 on a refracting body with cUfftrent velocities, there is nothing appar- 

 ently in theory to determine whether the refraction will be affected, 

 or in what way, by this difference. 



On the undulatory principle, it is true, velocity is intimately con- 

 nected with refraction ; retardation and refraction being c'oextensive 

 and almost equivalent terms ; but it must be borne in mind that it is 

 not absolute but relative velocity which is thus connected with refrac- 

 tion. It is the relative retardation in the denser medium, whatever 

 the absolute velocity may be, which causes refraction. If in theory it 

 were shown that the ratio would be constant for all velocities, it 

 would give a constant refraction for the medium. But this is the 

 verj' point in question ; and there appears nothing antecedently to 

 show, on any distinct theory of the nature of asther or of waves, that 

 the relative velocities must necessarily be in a constant ratio. There 

 is, however, nothing in any conception of waves at variance with the 

 idea; and it must be admitted as in itself a rational and probable sup- 

 position, fairlj' admissible in the first instance to ground any reason- 

 ing upon. When therefore the fact was established by Arago's 

 experiment, while it completely subverted what was a necessary 

 consequence of the emission theory, it offered no contradiction to the 

 undulatory; but the proposition it established being one already />ro6- 

 Me, and consistent with that theory, was now to be recognized as an 

 essential part of- it. Yet the result of Arago's experiment has been 

 represented by some able wTiters as of a very startling and unex- 

 pected nature, and, at first sight, equally perplexing on either 

 hypothesis. 



SEC. SEE. 11 



