334? Mr. R. Potter, Jun., on the Velocity with which 



same glass be placed as at Im n i, and inclined to the direction 

 of the incident light, then very frequently (that is, when the 

 apparatus used is properly adjusted and applied,) there is an- 

 other set of bands seen atf. In this case it will be seen that 

 the upper ray a q rf, will have passed through a greater di- 

 stance in air than the lower ray bstf But on account of the 

 glass Imni being more inclined to the ray passing through 

 it than the glass ghik to the ray a qrf it will be seen that 

 the lower ray will have passed through a greater distance in 

 glass. Or the sum of the" distances a q and rf is greater 

 than the sum b s and tf, and the distance in glass s t is greater 

 than qr. 



Now, putting d = the distances in air a q + rf 



and d / = bs + tf 



and putting D = the distance in glass q r 



D, = i st 



Then, when the interference at the point f is produced be- 

 tween rays which have set out simidtaneously from a and b, 

 putting T = the whole time of passing from a or b tof and 

 v = the velocity of light in air 

 iso = in glass, 



™ d D d, D, 



we have M = 1 = — - + — 



V w V w 



from which we find 



w D-D 



When, therefore, we can find by experiments the distances 

 in air and glass, we can also find the relative velocities of light 

 in them. This, however, it must be remembered, is on the 

 supposition that the velocity is the same in glass for all inci- 

 dences, and that the velocity in air is the same after having 

 passed through th| glass as before. The former of these points 

 we shall have to refer to again presently. 



In the apparatus I used for the experiments about to be de- 

 scribed, the luminous points a and b were produced by re- 

 flections of the image of the sun given by an equi-convex 

 lens of crown glass of T \jth inch focus, from two mirrors of 

 speculum metal. These mirrors were set in a frame, and so 

 that their contiguous edges could be adjusted by means of 

 four fine-threaded screws. The glasses g h i k, Imn i, were 

 produced by dividing a piece from one plate, which had been 

 examined very carefully to see that the surfaces were flat and 

 truly paralfel. The mode of examining the parallelism of the 

 surfaces which I adopted, was that of viewing the reflected 

 images they gave of a narrow line of light in the following 

 manner. — Standing in a room facing the window through which 



