1887.] Rate at which Electricity leaks through Liquids. 421 



Fig. 6. 



Curves plotted for carbon disulphide from Tables X and XI. 



Upper curve from readings taken after a long period (10 minutes) of charging the 

 condensers ; lower curve after a short period (2 seconds) . Otherwise conditions 

 similar. 



The curves of ratios are on the same scale and have the same zero line about 3 inches 

 below the lower edge of the figure. 



We deal later with further experiments on these phenomena o£ 

 absorption and residual charge. 



Experiments on Olive Oil. 



Very little trouble was experienced with olive oil. The sample 

 used proved to be the best insulator of all the liquids we tried. After 

 filtering once through Swedish, paper and once througb a glass-wool 

 plug, it insulated as well as after numerous filterings. 



Readings of the deflections were taken, as before, every five seconds ; 

 but in Tables XII and XIII, only those taken at intervals of 

 25 seconds are recorded, the former being taken after a short time 

 of charging, the latter after a long time. Fig. 7 is plotted from 

 Table XIII, but is on a different scale from figs. 3 — 6, so far as the 

 time coordinate is concerned, in order to bring more of^the curve into 

 the plate. 



Table XII (see Fig. 7). 



Experiment 44. Olive oil. 1 inch up straight. Time of charge 



short. 



Deflections. Ratios. 



Deflections. Ratios. 



Deflections. Ratios. 



500 , ., q 



417 



' 1-19 



349 }.tZ 



295 



247 {.{I 



209 1 . 19 

 176 I'll 



127 lit 

 107 l.g 



58 1 14) 



