118 PROCEEDINGS OF THE AMERICAN ACADEMY 



For example : the corpuscular theory of light requires that the veloci- 

 ties of light in different media should vary directly as the indices of 

 refraction, whereas the undulatory theory inverts this ratio. Arago 

 prepared for the trial by experiments on rapid rotations, the mechanical 

 difficulties to be overcome, and the comparative advantage of slower 

 rotations, assisted by several reflections, in place of a single mirror 

 turning with its maximum speed. Aided by the refined skill of Bre- 

 guet, he realized velocities in the mirror of 1000 turns a second, and 

 of the axis detached from the mirror of even 8000 turns. In the 

 mean while his eyesight began to fail him, and younger physicists 

 entered into the fruit of his labors. After Foucault and Fizeau by 

 separate efforts had decided the question in relation to the velocities of 

 light in air and in water in favor of the undulatory theory, and thus 

 confirmed a conclusion which Arago reached by diffraction in 1838, 

 and after Fizeau had studied the variation of the velocity of light in 

 running water, according as the motions agree or differ in direction, 

 Foucault was emboldened to attempt a measure of the absolute velocity 

 of light by an experiment which could be brought within the compass 

 of a single room. I translate his own account of the arrangements 

 made for this purpose : — 



" A pencil of solar light reflected into a horizontal direction by a 

 lieliostat, falls upon the micrometric mark, which consists of a series of 

 vertical lines distant from one another y\y of a millimetre. This mark, 

 which in the experiment is the real standard of measure, has been 

 divided very carefully by Froment. The rays which have traversed 

 this initial surface fall upon a plane rotating mirror at the distance of 

 a metre, where they suffer the first reflection, which sends them to a 

 concave mirror at the distance of 4 metres. Between these two mir- 

 rors, and as near as possible to the plane mirror, is placed an object- 

 glass, having in one of its conjugate foci the virtual image of the mark, 

 and in the other the surface of the concave mirror. These conditions 

 being fulfilled, the pencil of light, after traversing the lens, forms an 

 image of the mark on the surface of this concave mirror. 



" Thence the pencil is reflected a second time in a direction just 

 oblique enough to avoid the rotating mirror, an image of which it forms 

 in the air at a certain distance. At this place, a second concave mirror 

 is placed, facing so that the pencil, once more reflected, returns to the 

 neighborhood of the first concave mirror, forming a second image of the 

 mark. This is taken up by a third concave mirror, and so on to the 



