1892.] On the Electrolysis of Silver Nitrate in Vacuo. 351 



the later stages, and after a number of experiments there may be a 

 thin layer of silver, possibly an alloy of silver and platinum, which 

 resists the action of acid, and can only be scraped out with sand. It 

 is to be noted that Kohlrausch took his silver deposits on platinum 

 which had previously been covered with a layer of silver ; while in 

 Lord Rayleigh's experiments the silver deposits were removed from 

 the dish before a new experiment was made. The difference may 

 account for the somewhat greater equivalent found by Kohlrausch ; 

 but the concordance of the results shows that there can be no 

 systematic difference amounting to more than a few parts in ten 

 thousand. 



We turn now to the main part of the investigation, which is the 

 comparison of the deposits obtained in air and vacuo. The solution 

 nsed in the different voltameters was always taken out of the same 

 bottle. We had intended in this way to make sure that any difference 

 in the deposits was not due to some chemical difference in the solu- 

 tions. It did not occur to us at the time that the solution in one 

 voltameter being freed of air, we should gradually diminish the 

 amount of air also in the other voltameter, for the solutions were 

 kept in stoppered bottles, which did not allow of a ready re-absorption 

 of oxygen. It will be seen that the differences in the deposits, when 

 these were taken in air and vacuo, were larger and more regular in 

 the first experiments than later on, and this may have been due to 

 the gradual elimination of oxygen out of the solution. 



Our first experiments were made with the large bowl placed in 

 vacuo, and the small one in air. The results, to which a later one is 

 added for the sake of completeness, are embodied in Table IV. Ex- 

 periment 17 was not a satisfactory one, as will be explained later on, 

 and is therefore included in square brackets. 



The large difference between the result obtained in air and in vacuo 

 first drew our attention to a possible influence of the size of the bowl. 

 The experiments made to clear up this point have already been 

 described. A few deposits were taken with the small bowl in vacuo 

 and the large bowl in air ; although the two effects counteract each 

 other, the deposits in vacuo are larger than those in air, as is shown 

 by Table V. 



On the supposition that the effect due to the size of the bowl is the 

 same in air as it is in vacuo, we may combine the results of Tables 

 IV and V, and thus find that the deposits of silver in vacuo are about 

 one part in a thousand larger than those in air. The next two 

 experiments (Table VI) were a surprise. 



2 B 2 



