322 



PROCEEDINGS OP THE AMERICAN ACADEMY. 



to the observer at T. Another part of the incident ray is reflected along 

 the other arm of the cross, is similarly passed to and fro, returned, and 

 at last transmitted to the observer. In the apparatus actually used, 

 mirror 5 lay above 3, rather than to one side of it ; Figure 2 shows this 

 arrangement. The whole path of the light along these mirrors 'was 

 enclosed and covered, to lessen the effect of air currents and other 

 local disturbances. An acetylene flame was carried as a source of 



light. A telescope magnify- 

 ing thirty-five diameters 

 gave distinct vision of mir- 

 ror 8, at whose surface the 

 interference fringes are ap- 

 parently localized. 



The mirrors, being sil- 

 vered and polished, were 

 put in place, and the lengths 

 of the two paths were meas- 

 ured with a split rod and 

 then made nearly equal. 

 Establishing interferences 

 in sodium light, we found 

 the central part of a series 

 of some seven hundred 

 interferences which are 

 brighter than the adjoining 

 three hundred. With no 

 long search, we could see 

 interferences in white light, 

 although we had provided 

 no screw for moving a mir- 

 ror with its surface always parallel to a given surface. This we had 

 avoided, in order to have everything about the two arms as symmetrical 

 as possible. 



We now computed the direction and velocity of the motion of the 

 centre of the apparatus by compounding the annual motion in the orbit 

 of the earth with the motion of the solar system towards a certain point 

 in the heavens. During part of August, the whole of September, and 

 nearly all of October, this motion never coincides with the plane of our 

 apparatus. For other dates, there are two hours in each day when the 

 motion is in the desired plane, except for two days when the two hours 



Figure 1. 



Figure 2. 





