and Law in Electro-optics* 165 



motion, the band is displaced slowly (here to the right), and 

 can be driven easily out of the field of the instrument. 



Each measurement requires the conjoint and repeated action 

 of two observers — one at the electrometer, the other at the 

 second Nicol. The lower plate of the electrometer is fixed at 

 a chosen point of the scale ; and the screw of the compensator 

 is worked so as to displace the black band some distance to the 

 left. When the machine is set in motion, the potential of the 

 prime conductor rises slowly, and the observer at the electro- 

 meter gives a continued set of sharp audible signals, which 

 are intended partly to inform the observer at the polariscope, 

 and partly to guide the worker at the machine, whose task is 

 to keep the index of the electrometer as steadily as possible in 

 or near the sighted position. The present relation of actual 

 to cJiosen potential, whether of defect, or equality, or excess, 

 is indicated incessantly by the signals. 



Meanwhile the observer at the polariscope has a distinct 

 view of the black band, and sees whether it is midway between 

 the wires, or to the right or left, at the time of the critical 

 signal. The screw of the compensator is now worked in the 

 proper direction, and the observation is repeated. The screw 

 is again worked and the observation again repeated, and so for- 

 ward, until at last the black band lies exactly midway between 

 the reference-wires when the index of the electrometer is in 

 the sighted position — that is, when the potential of the upper 

 conductor has the value indicated on the scale of the electro- 

 meter. Two other observations are always made at the same 

 time, and the results recorded — one giving a superior limit of 

 the optical effect at the chosen potential, and the other an in- 

 ferior limit ; and the two limits are always drawn as close as 

 possible. 



9. Preliminary Work. — In mastering this method, as also 

 in applying it afterwards, I was able to keep the manipula- 

 tions, readings, and records all in my own hands; but for the 

 observations I had to depend partly on my two assistants. 

 The preliminary drill was found useful in respects not at first 

 contemplated; for in the course of it I discovered several 

 sources of error, and saw how to avoid them : I found also 

 that, when proper care was taken, the observer's error with 

 the compensator could be certainly limited within a definite 

 and small range of the scale. 



After more than a week of incessant exercise of myself and 

 assistants in this way in the practical management of the 

 instruments and method, I felt confident that good results 

 could be obtained in one day's work at least— one mounting of 

 the cell, and one setting of all the optical pieces, to be kept 



