﻿iii the Moisture condensed on Glass Surfaces. 337 



in the other. It was found indeed possible to produce a 

 resistance which had the value of about 2 x 10 5 ohms in one 

 direction, but of 10 6 ohms for currents in the other direction. 



The exact conditions necessary for bringing a conducting 

 surface into this sensitive state of " unidirectional conduc- 

 tivity " are complicated. The effect was originally noticed 

 where the current was passed over a glass surface between 

 two platinum electrodes melted on to the glass. The effects 

 in this case appear to be very similar to those observed where 

 tinfoil electrodes cemented on to glass are used which are 

 described below. 



That the change in resistance observed on reversal cannot 

 be explained by the usual back-E.M.F. of chemical decom- 

 position, is proved by the effect appearing with differences of 

 potential ranging up to a3 much as 3000 volts. 



The effect was observed to be very sensitive to changes in 

 the hygrometric state of the atmosphere, as well as to be 

 dependent on the length of time for which the current was 

 allowed to run before reversal. 



The general character of the phenomenon is as follows. 

 On a difference of potential being continuously applied across 

 a surface, say of glass between two metallic electrodes, the 

 current is observed to gradually diminish with time. In 

 addition, if the current is reversed at any moment during the 

 interval, it is found to be different from the direct current — 

 sometimes being as much as six or seven times greater than 

 the direct current. The exact ratio of the currents for any 

 given hygrometric state depends on the length of time the 

 current has been allowed to run before reversal. 



With the view of elucidating the requirements necessary 

 for the appearance of these sensitive conditions of the surface, 

 it was thought advisable to investigate the following three 

 points: — (1) The rate at which the direct current diminishes 

 with time ; (2) the law connecting the value of the initial 

 current with the hygrometric state of the surrounding air ; 

 and (3) if in any given condition of sensitiveness Ohm's 

 law holds for currents sent in either direction over the 

 surface. 



The glass surfaces used in these experiments were prepared 

 in the following manner. A sheet of tinfoil is fixed over a 

 piece of ordinary window-glass about 30 cms. square by 

 means of shellac. Parallel cuts are now made on the surface 

 at a distance of about 2 mms. and alternate strips are removed. 

 The final condition of this plate is shown in the sketch 



Phil Mag. S. 6. Vol. 12. No. 70. Oct. 1906. Z 



