1887.] Rate at which Electricity leaks through Liquids. 415 



Experiments on Benzene. 



The first successf ill experiments were made with an impure sample 

 of benzene, and the method was shown to be practicable. But before 

 more careful readings were taken the cylinders had to be re-worked, 

 because of some irregularities on the surfaces. On setting the appa- 

 ratus up again, it was found that the impure benzene no longer insu- 

 lated well enough to allow us to get readings. The reason of the 

 change we have not been able to discover. 



A pure sample of benzene was got, and after the usual filtering 

 processes, the condenser was filled and charged, and readings were 

 taken at five seconds intervals. 



Tables I — IV give the electrometer deflections, and the ratios of 

 successive pairs for different sets of readings under varying con- 

 ditions, which are specified at the head of each table. 



Tables II and IV are shown graphically in fig. 3, which gives two 

 •curves of falling potential and two ratio curves. 



Table I. 



Experiment 12. Pure Benzene. \ inch up straight. 



Deflections. Eatios. 



Deflections. Eatios. 



Deflections. Eatios. 



440 -i .i-i 

 396 {. LL 

 357 H[ 

 821 {.{I 

 290 I .[[ 

 261 IH 

 235 1 , 10 



212 1-09 



174 i .12 

 155 112 

 140 l\l 



126 1 -09 

 115 HI 



103 



93 1 11 

 % 1-09 



f 7 let 



Table II (see Fig. 3). 

 Experiment 13. Pure Benzene. § inch up straight. 



Deflections. Eatios. 



Deflections. Eatios. 



Deflections. Eatios. 



S5 i« 



tit 1076 

 3?6 1-077 

 111 1 ' 0n 

 327 1-073 

 305 1,072 

 283 1,077 

 263 1,075 



245 1,073 

 ^ 1-069 



S 1>0 *° 



200 1,070 

 |2S 1-069 

 161 1,074 



150 i :° 78 



137 1 °95 

 130? 1,054 

 190 P 1 083 



112 1071 

 112 1 -056 



X 99 1,070 

 X? 1 -087 

 It 1 -070 



79 1,075 

 74 1,067 

 Aft 1 '088 



63 1,079 

 aa 1 '050 



56 1,071 

 52 1076 



VOL. XLII. 



