﻿338 Prof. Trouton and Mr. Searle on Leakage Currents 



(fig. 1), which represents a portion of such a plate ; the tinfoil 

 surface is shown in black. Such a plate will be called a 

 " grid." 



Fisr. 1. 



Let the grid be connected up in series with a galvanometer 

 and suitable E.M.F., say 100 volts. The current will assume 

 a definite value and will then fall, more or less rapidly at, 

 first but at a gradually decreasing rate, until a fairly steady 

 but relatively small value is attained. 



This apparently steady value is in reality not constant. 

 The current undoubtedly slowly falls, though at this stage 

 the time-current curve is almost parallel to the time axis. 



If, however, during the period of transition the current 

 through the grid be reversed, it will assume a new value 

 greater than before, and will begin more or less slowly to 

 further increase in amount, the duration of the first stage and 

 the rapidity of rising depending on the E.M.F., the previous 

 history of the grid, and on the hygrometric state of the sur- 

 rounding air. The initial value on reversal also depends on 

 these conditions. 



In attempting to explain this phenomenon two suggestions 

 presented themselves, and are as follows : — 



I. The fact that glass is hygroscopic leads us to suppose 

 that what current does flow is conducted across by the film 

 of moisture on the surface, possibly containing salts dissolved 

 out of the glass. Taking a section of the grid perpendicular to 

 the strips, the initial state before a current flows will be when 

 the moisture lies evenly dispersed over the surface. If now the 

 current be switched on, it is possible that a species of electric 

 endosmosis takes place whereby water is heaped up at, say, the 



