PROGRESS IN PHYSICS. 323 



lark or elliptically polarized. Tliis siiuplo hypothesis has been found 

 quite adequate, and through its application to the various phenomena 

 of polarization, together with the application of Young's theory of 

 the interference of wa\'es to the production of color, the undulatory 

 theory of light was iirnily established before the middle of the cen- 

 tury. There were many noted philosophers, however, who stood out 

 long against it, notably Brewster, the most famous English student of 

 optics of the early part of the century, who declared that his "chief 

 objection to the undulatory theory was that he could not think the 

 Creator guilty of so clumsy a contrivance as the filling of space with 

 ether in order to produce light." 



In studying the nature of light it became very important to know 

 how fast a light wave traveled. A tolerably good measure of the 

 velocity of light had been made long before by means of the eclipses 

 of Jupiter's moons and by observations upon the positions of the 

 stars as influenced by the motion of the earth in its orbit. It was 

 found to be approximately 180,000 miles per second, a speed so great 

 that it seemed impossible that it should ever be measured by using 

 only terrestrial distances. This extremely difficult problem has been 

 solved, however, in a most satisfactory manner b}^ nineteenth century 

 physicists. Everybodj" knows that in a uniform motion velocity is 

 equal to space or distance divided l)y time. If, then, the time occu- 

 pied in passing through a given distance can be measured, the velocity 

 is at once known. As the velocity of light is very large, unless the 

 distance is enormously great the time will be extremely small, and 

 if moderate distances are to be used, the problem is to measure very 

 small intervals of time very accurately. Light will travel 1 mile in 

 about the one hundred and eighty-sixth thousandth part of a second, 

 and if by using a mile as the distance the velocity of light is to be 

 determined within 1 per cent, it is necessary to be able to detect dif- 

 ferences of time as small as about one-twenty -millionth of a second. 

 This has been made possible by the use of two distinct methods. 

 Foucault, on the suggestion of Arago, used a rapidly revolving mir- 

 ror, a method introduced by Wheatstone, the English electrician, who 

 used it in finding the duration of an electric spark. The essential 

 principle is that a mirror may be made to revolve so rapidly that it 

 will change its position by a measurable angle while light which has 

 been reflected from it passes to a somewhat distant fixed mirror and 

 returns to the moving reflector. In the other method a toothed 

 wheel is revolved so rapidly that a beam of light passing between two 

 consecutive teeth to a distant fixed mirror is cut off on its return to the 

 wheel by the tooth which has moved forward while the light has made 

 its journey. This method was first used by Fizeau. In either method, 

 if the speed of rotation is known, the time is readily found. In point 

 of time Fizeau was the first to attack the problem, which he did about 



