244 Prof. J. J. Thomson. [June 20, 



IX. " On the Electrolysis of Gases." By J. J. THOMSON, M.A., 

 F.R.S., Cavendish Professor of Experimental Physics, 

 Cambridge. Received June 17, 1895. 



In the experiments described in this paper I have used the spectro- 

 scope to detect the decomposition of gases by the electric discharge 

 and the movement of the ions in opposite directions along the dis- 

 charge-tube. 



The method consists in sending the electric discharge through a 

 tube so arranged that the spectra close to the positive and negative 

 electrodes can easily be compared ; thus the presence or absence of 

 certain ions at these electrodes can be ascertained. This method 

 is capable of much wider application than the one I previously 

 used in my experiments on the " Electrolysis of Steam " (' Proc. 

 Roy. Soc.,' vol. 52, p. 90), the use of which is attended with very 

 great difficulty for any substance other than steam. The earlier 

 method has, however, the advantage of being a quantitative method 

 the present one is only qualitative. 



In my former experiment with steam, when I worked at atmo- 

 spheric pressure and varied the length of the spark, I found that 

 when the spark-length exceeded a certain length, d h there was an 

 excess of hydrogen at the negative electrode and of oxygen at the 

 positive, equal in amount to the quantities of hydrogen and oxygen 

 liberated from a water voltameter placed in series with the steam- 

 tube. When the sparks were shorter than a certain length, d z , the 

 hydrogen appeared at the positive, the oxygen at the negative electrode, 

 but the quantity of these gases was again equal to the quantities 

 liberated in a water voltameter placed in series with the steam- tube. 



When the spark-length was between di and d 2 the effects were 

 irregular, and there seemed to be no connection between the amounts 

 of gases liberated in the steam-tubes and those liberated in the 

 voltameter. 



In the following experiments in which the sparks were of constant 

 length and the pressure was altered, corresponding effects were 

 observed. Within certain limits of pressure definite and perfectly 

 regular evidence of the separation of the ions of the gas sparked 

 through was obtained ; and the electrode at which a . given ion 

 appeared could be reversed by altering the pressure; there was, 

 however, a range of pressures in which the separation of the ions was 

 either not well marked or was irregular in character. 



I shall begin by describing a very simple method of showing the 

 separation of the ions produced by the discharge of electricity 

 through a compound gas such as hydrochloric acid gas, which is 



