240 Dr. J. Kerr's Electro-optic 



(5) When a la rge induction-coil worked by one Grove's 

 element was used as a source of electricity, I found it suffi- 

 cient to connect the balls D and F with the two ends of the 

 coil by unbroken wires. The light was then sustained conti- 

 nually from initial extinction in the polariscope, but always 

 with a very sensible nicker. The liquid required in this case 

 to be particularly clean ; for otherwise the effect was marred 

 by specks or chains of particles, as in the former experiments 

 with insulating liquids when these were not very clean. 



(6) The contrast between the central and the outer parts of 

 the field, with reference to intensity of optical effect, is as 

 marked here as in any of the insulating liquids. Returning 

 to the experiment of 51, I remove the screen P Q and turn 

 the flame L edgewise to the cell, so that the object in the 

 polariscope is a narrow vertical streak of light extending well 

 above and below the balls. The object now restored in the 

 polariscope by discharge is a comparatively short central seg- 

 ment of the flame, lying well within the cylinder which enve- 

 lops the two balls, and fading gradually to extinction at both 

 ends, but very well outlined laterally. 



53. Experiment. — The arrangements and procedure of 51 

 are retained with only one change ; the plane of polarization 

 of the lio-ht rendered by the first Nicol is horizontal or vertical — 

 that is, either parallel or perpendicular to the line of discharge 

 through the centre of the cell. The optical effect of discharge 

 in the polariscope is now evanescent. 



In most of my repetitions of this experiment, the restoration 

 from pure extinction was reduced to an extremely faint but 

 still perceptible flash at the instant of each spark ; but when 

 the cell was placed at a good distance from the eye, and great 

 care was taken otherwise with the observation, this faint effect 

 disappeared completely. The Nicols had only to be turned 

 then through half a right angle each, to give the strong and 

 clear effects already described in 51. These observations 

 were made with the longest spark from the plate machine. 



54. Special Eye-piece. — Many attempts were made to sup- 

 plement the information obtained in the last article, by bring- 

 ing the nitrobenzol-effect into closer comparison with the 

 effects formerly observed in CS 2 and other insulating liquids. 

 The results given by the hand compensator pointed certainly 

 to one conclusion ; but they were not very distinct, nor were 

 they even regular enough to be satisfactory until recourse was 

 had to a special eyepiece. 



This is a small piece made of brass, and shaped somewhat 

 like a common thread-bobbin with the shank prolonged a little 

 way through one of the flanges, this flange being grooved as 



