intelligence and Miscellaneous Articles, 79 



It remained, finally, to count the number of turns, or rather to 

 impart to the moveable body a determinate velocity. This problem 

 was completely solved in the following manner : — 



Between the microscope and the mirror for partial reflexion is 

 placed a circular disc, whose finely-toothed edge encroaches upon 

 the image observed, and partially cuts it off. The disc rotates uni- 

 formly on itself so that, if the image shone in a continuous manner, 

 the teeth which it carries on its circumference would escape from 

 sight by the rapidity of its motion. But the image is not perma- 

 nent, it arises from a series of discontinuous appearances, which are 

 equal in number to that of the revolutions of the mirror ; and in the 

 particular case in which the teeth of the screen also succeed in the 

 same number, an easily explicable optical illusion is produced, which 

 makes the teething appear as if the disc had no existence. Let us 

 suppose, then, that this disc, carrying n teeth on its circumference, 

 makes one turn in a second, and at the same time the turbine is 

 started ; if, in regulating the flow of the motive air, the apparent 

 fixity of the teeth can be maintained, it is certain that the mirror 

 makes exactly n turns in a second. 



M. Froment, who had made the turbine, has been good enough to 

 devise and construct a chronometric wheelwork to make the disc move; 

 it is a very remarkable piece of clockwork, which solves in an 

 elegant manner the problem of uniform motion in the particular case 

 in which there is no work to perform. The success is so complete, 

 that I am daily in the habit of starting the mirror at 400 turns in a 

 second, and of seeing the two apparatus work to almost ^-^ f° r 

 entire minutes. 



Yet, although I had obtained all certainty with regard to the 

 measurement of the time, I was surprised at finding in my results dis- 

 accordances which were not in any ratio with the precision of the 

 means of measurement. After long research, I ultimately found 

 that the source of error lay in the micrometer, which is by no means so 

 accurate as is usually thought. To obviate this difficulty, I introduced 

 into the system of observation a modification which ultimately comes 

 to a simple change in the variable. Instead of measuring the de- 

 viation micrometrically, I adopt for it a value defined beforehand, 

 either y 7 ^ of a millimetre, or 7 entire parts of the image, and I try, 

 experimentally, what distance there must be between the sight and 

 the rotating mirror to produce this deviation : as the measures refer 

 then to a length of about a metre, the last fractions still retain a 

 directly visible magnitude and leave no room for error. 



By this means the apparatus has been freed from the principal 

 source of uncertainty. Since then, the results agree within the limits 

 of the errors of observation, and the means are so fixed that I have 

 been able to give confidently the new number which it appears to 

 me expresses very nearly the velocity of light in space ; that is to 

 say, 298,000 kilometres (190249-16 miles) in a second of mean time. 

 ■ — Comptes Rendus, Nov. 24, 18G2. 



