92 



SCIENCE. 



[N. S. Vol. XVI. No. 394. 



the angular motion of the mirror. The 

 mechanical oscillator of Mr. Tesla sug- 

 gested itself as capable of giving pos- 

 sibly sufficient speed. The variation in 

 the speed which would cause the bands 

 to be superimposed and thus obliterated 

 rendered this method impossible. The same 

 difficulty would be experienced with any 

 reciprocating means. The compensation 

 for angular movement of a disc did not 

 seem clear and its use was abandoned for 

 a time. Instead of this the system indicated 

 in Figures 4 and 5 was tried. L and L' are 



experiment was about thirty meters. With 

 sunlight and a much greater distance, dis- 

 tinct bands could easily be obtained with 

 only one fiftieth the intensity, which repre- 

 sents the fraction of the incident light avail- 

 able during one revolution. Here again we 

 are met mth the difficiilty of obtaining an 

 insufficient component velocity in the direc- 

 tion of the ray, which is the velocity of the 

 disc into the cosine of the angle of the prism 

 into the index of refraction less one. 



Through a fortunate idea the rotating 

 disc was made possible and the final and 



-?%. ^ 



two bisected lenses, P and P' two prisms, 

 one of which, P', is mounted on a rotating 

 disc so that the total thickness would be 

 increased or diminished by its rotation. 

 The split lens L forms two images of the 

 beam from the slit S and one half of its 

 aperture. In one focus the double prism is 

 placed. The split lens L' forms coincident 

 images on the distant mirror ill, which re- 

 flects each ray back over the path through 

 the opposite halves of the lenses to the 

 mirror m, which reflects the light to the eye. 

 If now the prism P' be moved in the in- 

 terval of transit of this optical circuit the 

 ray returning through it will be retarded 

 over the other. Knowing the constants of 

 the prism and the speed of the disc we can 

 calculate, in this way, the wave-velocity. 

 The interference bands in this system re- 

 mained distinct during the movement of the 

 prism over an arc of five to ten degrees. 

 The length of this system in the preliminary 



most serious difficulty was overcome. Fig- 

 ure 6 gives the arrangement with somewhat 

 distorted details to show the optical rela- 

 tions. The principle that an even number 

 of reflections in a rotating system does 

 not change the relation of the incident to 

 the transmitted ray was made use of. A 

 further difficulty had to be met in main- 

 taining the ray upon the reflecting ele- 

 ments, as the rotation of the reflected ray 

 is twice that of the reflector. If the radius 

 of rotation is reduced one half the linear 

 movement of any point is reduced one half. 

 These considerations applied to two mirrors 

 on the rotating disc met the required con- 

 ditions. Thus the mirror m is placed just 

 half the distance of the mirror m' from the 

 center. A ray incident upon m is reflected 

 to m' and thence reflected in some definite 

 direction. If the disc now rotate, the ray 

 will rotate through twice the angle, but still 

 strike the mirror m' in the same point if 



