1898.] 



The Electrolysis of Steam. 



9i> 



electrodes. I have never observed these spangles on the hydrogen 

 side. 



When a sufficient quantity of the residual gas had been collected, 

 which generally happened when the sparks had been passing for an 

 hour or an hour and a half, a considerable volume of the mixed gases 

 was allowed to accumulate, so as to make sure of an explosion when 

 the spark from a Wimshurst passed through them. The coil was 

 then stopped and the mixed gases exploded, and the quantity of 

 hydrogen in the water voltameter determined. 



The residual gases in the collecting tubes were then analysed ; the 

 first step was to cool these gases, which while the steam had been 

 rushing into the tubes had been at a temperature of more than 80° C, 

 down to the temperature of the room. After this had been done, the 

 nature of these gases was determined by adding known volumes of 

 oxygen and hydrogen, prepared electrolytically from water, and 

 observing the contraction which took place when a spark passed ; the 

 addition of the hydrogen or the oxygen, as the case might be, was 

 continued until no further contraction took place on sparking. 



The result of these analyses was that when the sparks were not too 

 long the residual gas in one tube was found to be pure hydrogen, 

 that in the other pure oxygen ; if any other gases were present their 

 volume was too small to be determined by my analysis. This result 

 was only attained after considerable experience with the experiments 

 and with the precautions necessary to obtain correct results ; in the 

 earlier experiments there was always a considerable quantity of some 

 other gas (which, I suppose, was nitrogen) present. 



When the sparks passing through the steam were very long I 

 never succeeded in getting rid of this nitrogen ; indeed, in some cases 

 it amounted to more than 30 per cent, of the oxygen. I am not sure 

 what the source of this nitrogen is ; it may have been absorbed by 

 the electrodes and given out when the sparks pass, or it may have 

 come from the walls of the discharge tube, as these long sparks have 

 a tendency to occasionally jump to the walls, and when they do so 

 they may liberate air which would otherwise adhere to the glass. 



Results. 



The results obtained by the preceding method varied greatly in 

 their character with the length of the spark ; I shall therefore con- 

 sider them under the heads " short sparks," " medium sparks," and 

 M long sparks." 



The lengths at which a spark changes from " short " to " medium," 

 and then again to " long," depend on the intensity of the current 

 passing through the steam, and therefore upon the size of the induc- 

 tion coil and the battery power used to drive it. The limits of 



H 2 



