Atom and the Charge of Electricity carried by it. 539 



perature of the gas or metal, would increase this effect ; we 

 should therefore expect to find evidence of this effect in the 

 case of hot bodies. A very good instance of this effect is the 

 fact, discovered by Guthrie, that a hot ball of iron could re- 

 tain a negative charge when it could not retain a positive 

 one. The experiments of Elster and Geitel also illustrate 

 this point. Again I found (see Proc. Roy. Inst. April 13, 

 1894), that if in a vessel containing hydrogen (and probably 

 a little air) at a low pressure, we split up some of the mole- 

 cules of the gas so as to get a supply of positively and nega- 

 tively electrified atoms, a red-hot clean copper rod connected 

 with the earth would discharge the negative electricity and 

 leave the positive behind. The negative electricity escapes 

 from the gas, since the negatively electrified gaseous atoms 

 combine chemically with the positive atoms of the metal. 

 If, however, a red-hot copper rod thickly coated with oxide 

 is placed in the gas, the positive electricity escapes from 

 the gas through the rod to the earth. This is in accordance 

 with the theory ; the chemical process which now goes on is 

 the reduction of the oxide : and this is effected by the positive 

 hydrogen atoms combining with the negative oxygen atoms, 

 thus the process affords a means of escape to the positive 

 electricity in the gas, but not to the negative. 



The second process by which we have explained the escape 

 of electricity from a gas is that of the interchange of charges 

 between oppositely charged atoms. As the tendency for this 

 process to take place arises from the Volta effect, it will be 

 great when we have a number of strongly electropositive 

 atoms charged with negative electricity near a number of 

 electronegative ones charged with positive electricity. Thus 

 we should expect the effect to be great when one of the sub- 

 stances is an electropositive metal, as when this is in an 

 electrically neutral state half the strongly electropositive atoms 

 are charged with negative electricity, and on account of the 

 Volta effect there is a great tendency for them to lose their 

 charges. When, however, we have an electrolyte instead of 

 the metal, we should expect the effect to be much smaller, as 

 in an electrolyte the electropositive element has the positive 

 charge, the electronegative one the negative ; it might, 

 however, still exist to some extent if, for example, the elec- 

 trified gas in contact with the electrolyte were more electro- 

 negative than the negative element of the electrolyte, and if 

 the gas carried a positive charge. 



This difference between metals and electrolytes may perhaps 

 be the explanation of the fact that, whereas a sheet of metal 

 when illuminated by ultra-violet light rapidly loses a negative 

 charge while it can retain a positive one, this effect being 



