﻿Electric Currents in Air at Atmospheric Pressure. 751 



short time earth both plates, and then insulate A and connect 

 it with the electrometer, the plate B remaining earthed. A 

 will charge up negatively, behaving exactly as if it were a 

 plate o£ another metal electropositive to B when both are 

 immersed in ionized air. This charging up of A cannot be 

 accounted for by leakage through the insulating supports, 

 as, in that case, it would charge up positively. It is therefore 

 due to a contact difference and to the presence of carriers, 

 both produced by the previous current through the air. 

 There is no such effect when the plates are put into metallic 

 contact and a current allowed to pass from one to the other. 

 This induced contact difference is not permanent, but gra- 

 dually diminishes and disappears altera time, Its magnitude 

 and duration depend on the time during which B is in con- 

 nexion with the battery and on the potential to which it is 

 raised. If contact difference is to be explained by a differ- 

 ence in the amount of work done by an electron in escaping 

 from a metal, then it would follow that when negative 

 electrons have for some time been expelled from the surface 

 of a metal, the effect is to lessen this work, that is, to make 

 it easier for an electron to escape. This might arise through 

 the removal of a layer of such electrons from the surface the 

 presence of which would have the effect of preventing others 

 from escaping. A similar induced contact difference is 

 produced by the action of ultra-violet light, and may account 

 for at least part, if not the whole, of the so-called photo- 

 electric fatigue which metals exhibit when exposed to ultra- 

 violet rays. It may be remarked that removing electrons 

 from a metallic surface by means of a current, removing 

 electrons by ultra-violet light, and cleaning the surface have 

 similar effects on contact potential difference. 



But currents can be produced between two plates of 

 different metals under the action of their contact difference 

 alone. Let the upper plate A be zinc and the lower plate B 

 copper, and let the former be connected with one pair of 

 quadrants of an electrometer and the latter earthed. Then, 

 if the distance between the plates is about *005 cm., there 

 will be no charging up of A observable, when the pair of 

 quadrants with which it is in connexion is insulated. If the 

 plates are in contact the same is true, and there is again no 

 deflexion of the electrometer-needle. But if, starting from 

 the position in which the plates are in contact, the plate A 

 be gradually raised by means of the micrometer-screw a 

 position is soon reached when a very rapid charging up of A 

 takes place, the deflexion showing sometimes the whole 

 contact difference, corresponding to a deflexion of 37 cm. on 



