2 Giisfaf JV. El men 



investigations of Kerr/ Quincke,- Lemoine," and Schmidt,* high 

 potentials were employed. The difference of phase was meas- 

 ured by means of a Babinet's compensator, and the potential, 

 supplied either by a static machine or an induction coil, by means 

 of an electrometer. The difference in the values of B obtained by 

 different observers and also of the values obtained by the same 

 observer, as in the case of CS.., wdiere the values of Quincke vary 

 by one part in five, Schmidt points out as probably due to the 

 difficulty in measuring high potentials with a sufficient degree of 

 accuracy. To eliminate this error he used two sets of electrodes 

 on the same axis of rotation as the beam of light and set at right 

 ang-les, so that the effect of one set was compensated by that of the 

 other. This, however, gives only relative values, and some other 

 method must be depended upon for absolute measurements. 



A sensitive system of determining the difference of phase has 

 been used in this investigation. In place of the Babinet's com- 

 pensator, an elliptic half-shade polarizing system^ was employed. 

 With this system, as used in this particular problem, values of 8 

 equal to .oooiA could be observed. In measurements taken the 

 lowest value recorded is .00025 X. This brought the potential 

 that could be used down to less than 100 volts per millimeter of 

 distance between electrodes of 47 cm. length. For CS2 with the 

 same length of electrodes the smallest value recorded by Quincke 

 is over 3,000 volts per millimeter of distance. The potential was 

 supplied by a storage battery of about 1,400 volts, and as the 

 smallest potential used was less than 200 volts, a range of seven 

 times this voltage was at the writer's disposal. The potential 

 was measured by means of a voltmeter which had been cali- 

 brated so that the possible errors for the lowest potentials were 

 less than t per cent. 



The half shade system consisted of two thin mica sections. 

 ■One, the "sensitive strip" 5 (fig. i), placed next to the polarizing 



1 T. G. Kerr, /%//. Ma^. (4), 50, p. 446, 1875; (5), 8, pp. 85 and 229, 1879; 

 l(5\ 9, p. 157, 1880, and (5), 13, pp. 153 and 248, 1882. 

 2G. Quincke, IVted. Ann., 10, p. 729, 1883. 



3 J. Lemoine, Compt. Rend., 122, p. 835, 1896. 



4 W. Schmidt, Ann. d. Physik., 7, p. 142, 1902. 



»D. B. Brace, Phys. Rev., vol. 18, p. 70, vol. 19, p. 218, 1904. 



