﻿in the Moisture condensed on Glass Surfaces. 347 



Several independent effects might conspire to cause this 

 diminution in resistance at higher voltages: — (1) I£ the 

 resistance is, as above suggested, nearly all in a film of oxide 

 on the metal, the heating by the larger current might sensibly 

 reduce its resistance. It is true the heat generated is a very 

 small quantity (about '000,02 calorie per centimetre run of 

 the metal electrode) ; but then the thickness of the oxide is 

 correspondingly minute. (2) There may be an increased 

 transference by osmosis of liquid produced by the larger 

 currents, which flooding over the oxide causes the electrode 

 to present a larger surface for the current. (3) Minute 

 discharges of the nature of "sparks" may occur intermittently 

 over the oxide * and be more pronounced with high voltages. 

 This discharge superadded to the ordinary current would 

 cause the apparent reduction in resistance. 



Apart from these effects the experiments suggest that there 

 may be a true want of constancy in the ohmic resistance of 

 the surface-layers of liquids, and that it might be desirable to 

 carefully examine for such an effect with very thin soap-films. 



Conclusions. 



(1) That the falling off observed in the leakage-currents 

 is due to the formation of a film of oxide on the elec- 

 trode, or in the case of platinum of gas. 



(2) The alteration in the amount of the current on reversal 

 is due to this oxide, or gas as the case may be, being 

 reduced through the action of the reversal current 

 itself. 



(3) The initial resistance of a " grid " varies very rapidly 

 for variations in the hygrometric state of the sur- 

 rounding atmosphere for values above 80 per cent, of 

 moisture, and suggests its applicability to measuring 

 the hygrometric state. 



(4) That the currents sent by different electromotive forces 

 do not follow Ohm's law, but are in a greater pro- 

 portion than given by it for high electromotive forces. 



* We can actually see the sparks formed by discharges of this character 

 ■when the grid is placed in a very damp atmosphere. A similar effect 

 can he easily produced in the air of an ordinary room if the grid is first 

 ■wiped over with a minute quantity of some slightly hygrometric material 

 such as glycerine. Such a grid presents a very beautiful appearance 

 under a difference of potential of even 100 volts. The whole plate 

 spangles with minute sparks which take place at points all over the grid. 

 The fact that sparks can under certain circumstances take place with such 

 low voltages may be of interest from an insurance point of view. 



