AQS Prof. D. B. Brace or the Resolution of Light into 



the electromagnet, so that the lines of force were parallel 

 to the mica plate. By tilting the system about an axis 

 parallel to the lines of force, the number of passages of the 

 ray around the system could be varied from one up to five 

 or more, after which the image became too indistinct for 

 proper definition. The number regularly used was five, this 

 giving twenty internal reflexions, the path of the ray being 

 along the lines of force after every alternate reflexion, and 

 normal to them before these reflexions. 



In order to find the direction of the ray after reflexion on 

 either of the suppositions, namely, that there is a relative 

 change in velocity, or that there is a relative change in 

 the phase of the vibration, we may use the following 

 method. 



Fig.. 3. 







In fig. 3, if is any radiant point and I its image — i.e , 

 a point in which the various rays would meet in the same 

 phase, — produced by any arbitrary system of reflecting and 

 refracting surfaces s 1 . . . s n , and 7 . . . l n the successive 

 elements of the path of any ray, and N . . . N» the corre- 

 sponding number of wave-lengths A . . .A n in these elements, — 

 I and N being reckoned positive in the direction of pro- 

 pagation, then the condition that the phases shall be the 

 same (or differ by less than 1/4 >„) is 



8(1*0+]*!+ .. . N) =0or <1/4X, 

 where the variation is to be taken with respect to s 1 . . . s n . 



