ON THE SILVER VOLTAMETER. 569 



atmosphere of nitrogen. RICHARDS and HEIMROD* verified these results. MERRILL! 

 varied the pressure from 1 to 103 atmospheres and found no pressure effect for this 

 range he did not make olwervatious at less than atmospheric pressure. 



Temperature Coefficient. 



In order to investigate the effect (if any) of temperature upon the silver deposits, 

 we first compared the deposit in two Rayleigh forms maintained at different tempera- 

 tures, but in the same circuit. The lower temperature was that of the room in which 

 the observations were made and averaged about 16 C. The higher temperature was 

 that of an electric oven and was varied from 40 C. to 95 C. Sufficient distilled 

 water was taken to fill the voltameters and was warmed until its temperature was 

 comparable with that of the electric oven. The solution was made and divided into 

 two parts, that portion for the voltameter at the normal temperature being cooled to 

 16 C. and the other portion inserted in the kathode vessel of a voltameter and the 

 whole placed in an electric oven. The results were as follows : 



grammes. * C. grammes. " C. 



7-00401 15 (70a) 7'00483 65 (70c) +2 4 xlO-", 



7-01085 16 (71a) 7'01210 92 (716) +2 4 xlO~ 4 , 



7-05431 16 (72a) 7'05500 46 (726) 



The values are fairly consistent, but we were not satisfied. The mean temperature 

 coefficient appears to be positive and about 2 or 3 parts in 1 ,000,000 per degree, but 

 it appeared to us that there were sufficient disturbing influences at work to account 

 for the higher deposits at the higher temperatures. The filter paper cup, the folds 

 of which were secured with platinum wire and not with wax, turned a very dark 

 brown colour on exposure to the atmosphere of the electric oven, and we felt that we 

 were not justified in using filter paper at temperatures much higher than that of the 

 room. The same thing happened to the exposed portions of a porous cup, and we 

 resolved, therefore, to use a syphon at both high and low temperatures. In addition, 

 we placed sheets of glass over the kathode bowl, so that, together with the limb of 

 the syphon which entered the vessel, they shielded the electrolyte from currents of 

 air. The following results were obtained : 



-* 



+ 0,xlO 



These indicate that the temperature coefficient over the range 15 C. to 92" C. is 

 either nil or negligibly small. Unfortunately, we could not pass a current of 1 ampere 



* RICHARDS and HEIMROD, 'Proc. Am. Ac.,' 37, p. 430, 1902. 

 t MERRILL, 'Phys. Rev.,' 10, p. 170, 1900. 

 VOL. OCVII. A. 4 D 



