in Glass and Bisulphide of Carbon. 133 



became smaller (as in fig. he). At last these tails were com- 

 pletely squeezed against the electrodes; and with still stronger 

 pressure they disappeared. With a greater difference of po- 

 tential they emerge again, to again vanish in consequence of 

 stronger pressure. Both with horizontal and with vertical 

 pressure of the glass plate, a slight difference of potential with 

 gentle pressure has the same effect as a great potential-differ- 

 ence with stronger pressure. The latter agrees with what was 

 found by Eontgen. 



When I substituted for the compressed glass plate a plate 

 of calc spar cut perpendicularly to the axis, I convinced myself 

 that here also the light was elliptically polarized. Before the 

 machine worked, the coloured rings with the black cross ap- 



changed into a hyperbola ; the ends of the two branches 

 of the hyperbola receded further from each other as the 

 potential-difference increased. 



If the polarization-planes of the nicols were brought into a 

 horizontal and a vertical position, the phenomenon of fig. hf 

 showed itself already at a slight potential-difference ; at a 

 greater difference it became sharp and distinct. With this 

 position Kerr did not obtain any evident results, probably 

 because in his experiments the field of view was too small for 

 him to see the whole of the phenomenon. But how it was 

 that Eontgen* saw only the black horizontal line that issues 

 from the sphere is so much the more inexplicable, as he with 

 different electrodes, and with vessels of different widths, always 

 found the same results. I have repeated this experiment with 

 electrodes corresponding exactly in dimensions with those em- 

 ployed by Eontgen. Yet this did not alter the phenomenon, 

 any more than an alteration of the distance of the electrodes 

 from 2*7 to 5 millim. In frequent repetitions of this experi- 

 ment it was always seen by me unchanged ; and even the 

 results obtained with two spherical electrodes agree with this, 

 as will be subsequently shown. 



With horizontal and vertical pressure upon the glass plate 

 while it again occupied the previously mentioned position, the 

 phenomenon was unchanged if the glass was placed so that 

 the black cross was midway between the electrodes. 



When the direction of the pressure upon the glass plate 

 made an angle of 45° with the horizon, the figure was unsym- 

 metrical. This is also the case when the nicols are gradually 

 moved out of the second into the first position. The figures 

 then form a transition between fig. 5 e and fig. hf. The latter 

 also accords with Eontgen's results. 



* See fig 5 of his paper. 



