Or, lastly, according to a third hypothesis, which includes both the former ones, only a portion of the æther is free, the other portion being attached to the molecules of bodies and participating in their motion.
This latter hypothesis was proposed by Fresnel, and constructed for the purpose of equally satisfying the phænomena of aberration, and a celebrated experiment of M. Arago, by which it has been proved that the motion of the earth has no influence upon the refraction which the light of the stars suffers in a prism.
We may determine the value which in each of these hypotheses it is necessary to attribute to the velocity of light in bodies when the bodies are supposed to be in motion.
If the æther is supposed to be wholly carried along with the body in motion, the velocity of light ought to be increased by the whole velocity of the body, the ray being supposed to have the same direction as the motion.
If the æther is supposed to be free and independent, the velocity of light ought not to be changed at all.
Lastly, if only one part of the æther is carried along, the velocity of light would be increased, but only by a fraction of the velocity of the body, and not, as in the first hypothesis, by the whole velocity. This consequence is not so obvious as the former, but Fresnel has shown that it may be supported by mechanical arguments of great probability.
Although the velocity of light is enormous comparatively to such as we are able to impart to bodies, we are at the present time in possession of means of observation of such extreme delicacy, that it seems to me to be possible to determine by a direct experiment what is the real influence of the motion of bodies upon the velocity of light.
We are indebted to M. Arago for a method based upon the phænomena of interference, which is capable of indicating the most minute variations in the indexes of refraction of bodies. The experiments of MM. Arago and Fresnel upon the difference between the refractions of dry and moist air, have proved the extraordinary sensibility of that means of observation.
It is by adopting the same principle, and joining the double tube of M. Arago to the conjugate telescopes which I employed for determining the absolute velocity of light, that I have been able to study directly in two mediums the effects of the motion of a body upon the light which traverses it.
I will now attempt to describe, without the aid of a diagram, what was the course of the light in the experiment. From the focus of a cylindrical lens the solar rays penetrated almost immediately into the first telescope by a lateral opening very near to its focus. A transparent mirror, the plane of which made an angle of 45° with the axis of the telescope, reflected the rays in the direction of the object-glass.
On leaving the object-glass, the rays having become parallel among themselves, encountered a double chink, each opening of which corresponded to the mouth of one of the tubes. A very narrow bundle