Intelligence and Miscellaneous Articles. 



159 



principal section of the double-image prism is parallel to that of the 

 polarizing nicol N,. The two images furnished by this system are 





observed by means of a telescope V. By a suitable rotation of the 

 nicol N 2 , these two images can be brought to the same intensity ; 

 this rotation of the nicol measures the difference of mean phase 

 due to Kerr's phenomenon during the passage of the luminous flow 

 in the condenser. 



A continuous source of light (sun, electric arc, &c.) cannot be 

 employed ; the spark presents the advantage of being produced 

 only at the moment of the discharge and of having a duration 

 comparable to that of the disappearance of the birefringence. 

 The experiment furnishes therefore, together with the result 

 sought for, the evaluation of the order of the length of duration 

 of the spark. 



The light of the spark is brought into the condenser, at different 

 periods, by the following arrangement : — The converging lens L x 

 renders the luminous beam cylindrical to enable it to traverse 

 the interval between the plates of the condenser (mirror M x 

 removed). On the other hand, putting the mirror M L in place, 

 the system of two converging lenses, L 2 , L 3 , and four plane mirrors 

 M 2 , M 3 , M 4 , M L will bring the light of the spark to the condenser. 



By displacing the two mirrors M 2 , M 3 the length of path 

 EM_,M 3 M 4 M 1 which the light of the spark traverses before arriving 

 at the condenser K can be varied. 



The first measurement, made with the mirror M x removed, gives 

 the value of the phenomenon at the instant when the spark E 

 appears, or, at the latest, at the end of the time which light 

 takes to traverse the path EK, which is about 20 centimetres. 



In the following measurements, M x being in place, the time 

 which elapses between the production of the spark and the instant 

 when the light passes K can be increased at will ; it is only 

 necessary to draw back by degrees the mirrors M.„ M . In this 

 manner one can, after a fashion, construct the curve of the 

 phenomenon as a function of the time, a delay of one metre 

 corresponding to one three-hundred-millionth of a second. 



The results of one experiment are as follows : — 



Path traversed. Kotation of the nicol. 

 era. Q 



20 17-3 



100 8-7 



400 and over. not measurable. 



