-\'2 Dr. J. Kerr's Elect ro-cptic 



Let the plate K be now put in position as in the diagram, 

 with the line of compression either vertical or horizontal. 

 The light is restored in the polariscope, and appears (when 

 the eyepiece S is at rest) as a sharply defined and very narrow 

 horizontal band extending from ball to ball : but when the 

 eyepiece is set in motion, the band is transformed into a large 

 ring, narrow and bright at the top and bottom, wide and faintly 

 luminous at the sides. The plate K is finally removed : and 

 the extinction is seen to be pure. 



56. Experiment. — Things being thus prepared, the cell is 

 charged with clean nitrobenzol, and the balls D and F are 

 connected with the terminals of a large RuhmkorfPs coil by 

 unbroken wires. The coil is worked by one Grove's element, 

 and gives a half-inch spark with broken circuit. 



(1) The eyepiece S is left at rest. TThen the currents 

 pass, the luminous band is restored continually in the polari- 

 scope, not, however, as a steady light, but flickering or 

 quivering very rapidly. For our present purpose, this is a 

 much better form of effect than any thing that can be obtained 

 from the plate machine. 



(2) The eyepiece S is set in motion at a moderate speed. 

 "When the currents pass, the light is restored in detached 

 bands, which are distributed, of course, as generating elements 

 of the luminous ring observed immediately above (55). Xo 

 trace of the ring can be detected except the flashing bands 

 themselves, these being separated by wide and perfectly dark 

 intervals. 



(3 ) Observation (2) is carried forward, the eyepiece being 

 now driven round as rapidly as possible. There is a clear 

 change of optical effect, the bands restored by the currents 

 being enlarged into wide patches, which are much fainter in 

 their later than in their earlier parts. It appears thus that 

 each of the discharges of the coil through the liquid, considered 

 as a discharge optically effective, occupies a sensible interval 

 of time : it appears also that each discharge falls gradually in 

 power towards its termination. There was nothing like this 

 widening of the bands observed with the plate machine in any 

 case, with or without condenser. 



(4) The plate K is compressed, so as to give much the 

 same effect in the polariscope as the hand compensator when 

 well strained ; and it is then put in position with the line of 

 compression vertical (55). When the currents pass, and the 

 eyepiece revolves at a moderate rate, the flashing bands are 

 well restored on the faintly luminous ground of the ring ; and 

 when the eyepnece is driven very rapidly, the bands are widened 

 as in (3). The flashes are quite as distinct to the eye as in 



