562 



MR. F. E. SMITH, MR. T. MATHER, AND DR. T. M. LOWRY 



Discussion of the Results. 



Observations 1 to 13 were made with solutions which are now known to have been 

 impure. Apart from the electrolytes, the observations are also very unsatisfactory, 

 for in these early experiments the silver deposits were partly removed with the aid of 

 a platinum spatula, and the remainder with warm nitric acid ; the bowls were much 

 scratched in consequence, and many of the deposits were very loose. In addition, we 

 were not so expert in the manipulation of the voltameters as in the later experiments, 

 although the errors introduced on this account are regarded as very much smaller 

 than the differences in the electrochemical equivalent which we found from day to day. 

 As the electrolytes were not pure we have deferred the discussion of the first 13 

 sets of observations to Part II. (pp. 582, 595). 



The Rayleigh Form of Voltameter. 



Table II. summarises the results obtained with the liayleigh form. In all, there 

 are 52 independent observations ; the mean value of the electrochemical equivalent is 



TABLE II. Rayleigh Form. Absolute Determinations. 



1 '118273 milligrammes per coulomb, and the mean observational error is 2'4 parts in 

 100,000. The greatest differences from the mean occur with 18, 186, and 44r, and it 

 is possible that the solution used on the occasions 18a and 18ft was impure owing to 



