108 



Prof. J. J. Thomson. 



[Mar. **. 



I tried, therefore, the effect of covering the electrode with glass, 

 with mica, with ebonite, and sulphur. I found that, in all these 

 cases, the electrometer was deflected as long as the arc existed, and 

 that the deflection was in the direction corresponding to a positive 

 charge when the arc was in oxygen, and in that corresponding to a 

 negative one when the arc was in hydrogen. The deflection, though 

 not so large as when the electrode was bare, was quite unmistakable. 

 It disappeared almost entirely as soon as the arc stopped. 



Another experiment which I tried was to surround the arc by a 

 large glass tube, coated inside and out with a thin layer of sulphur to 

 prevent conduction over its surface. A ring of tin-foil was placed 

 outside the tube, so as to surround the place where the arc passed ; 

 this ring was connected with one of the quadrants of an electrometer. 

 As a further precaution against the creeping of the electricity over 

 the surface of the tube, two thin rings of tin-foil, connected to the 

 earth, were placed round the ends of the tube. In this case, when 

 the arc passed through oxygen the quadrants of the electrometer 

 connected with the central ring of tin-foil were 'positively charged by 

 induction, while when the arc passed through hydrogen these quad- 

 rants were negatively charged. These experiments show that the 

 oxygen in the arc behaves as if it had a charge of positive electricity, 

 while the hydrogen in the arc behaves as if it had a charge of negative 

 electricity. 



The electrodes in the preceding experiments were so large that 

 they were not heated sufficiently by the arc discharge to become 

 luminous. 



Elster and Greitel found that a metal plate placed near a red-hot 

 platinum wire became positively electrified if the plate and the wire 

 were surrounded by oxygen, negatively electrified if they were sur- 

 rounded by hydrogen. If we suppose that the effect of the hot wire 

 is to make the surrounding gas in a condition resembling the gas in 

 the arc, Bister and Geitel's results would be explained by the preced- 

 ing experiments, for these have shown that when this gas is oxygen 

 it is positively electrified, and when hydrogen negatively electrified. 



The following explanation of the results of the experiments on the 

 electrolysis of steam seems to be that which agrees best with the 

 preceding investigation. 



When an electric discharge passes through a gas the properties of 

 the gas in the neighbourhood of the line of discharge are modified. 

 Thus, as Hittorf and Schuster have shown, the gas in the neighbour- 

 hood of the discharge is no longer an insulator, but can transmit a 

 current under a very small potential difference. Faraday's remark, 

 that when once a spark has passed through a gas the passage of 

 another following it immediately afterwards is very much facilitated, 

 is another example of the same thing. We have thus good reasons 



