100 



Prof. J. J. Thomson. 



[Mar. 9, 



" short," " medium," and " long sparks " given below must therefore 

 be understood to have reference to the particular coil and current 

 used in these experiments. With a larger coil and current these 

 limits would expand ; with a smaller one they would contract. 



Short Sparks. 



I shall begin by describing the experiments with short sparks, i.e., 

 sparks from 1*5 to 4 mm. long. Here the appearance of the spark 

 shows all the characteristics of the " arc " discharge. 



The discharge passes as a thickish column with ill-defined edges, 

 and when placed in a wind it is blown out to a broad flame-like 

 appearance. 



For these short sparks or " arcs," as I prefer to call them, two very 

 important laws were found to be true — 



1. That within the limit of error of the experiments the volumes of 



the excesses of hydrogen in the one tube, and of oxygen in the 

 other, which remain after the explosion of the mixed gases, 

 are, respectively, equal to the volumes of the hydrogen and 

 oxygen liberated in the water voltameter placed in series with 

 the steam tube. 



2. The excess of hydrogen appears in the tube which is in con- 



nexion with the positive electrode, the excess of oxygen in the 

 tube which is in connexion with the negative electrode. 



The second of these results surprised me very much when I first 

 observed it, as both Perrot and Ludeking had found that in the 

 electrolysis of steam, as in that of water, the excess of hydrogen was 

 at the negative and that of oxygen at the positive electrode. Accord- 

 ing to my experiments, however, the electrode at which the hydrogen 

 appears in the electrolysis of steam is of the opposite sign to that at 

 which it appears in the electrolysis of water. So that, if a water 

 voltameter is placed in series with the steam tube, the gases are 

 liberated in the way shown in the accompanying diagram (fig. 4), in 



Fig. 4. 



Water voltameter 



Steam tube 



