1884.] On the Electro-chemical Equivalent of Silver, $c. 143 



sections. Full particulars are given of the procedure adopted in the 

 reduction of the method to practice. 



The insulation of the small coil, which was wound upon a ring of 

 ebonite, was carefully tested with the induction balance after the 

 manner recommended by Graham Bell. The first attempt at 

 winding it proved a failure, several turns being short-circuited ; and 

 we are of opinion that no coil of fine wire can be thoroughly 

 depended upon which has not been tested by some such method. 



The calculation of the constant of the current weighing apparatus 

 is best made with the aid of elliptic functions. Both for our own 

 purposes and in order to facilitate the use of the method by others, we 

 have calculated a table of the function 



sin 7 { 2F y - (1 + sec 2 7 E v }, 



(see "Maxwell's Electricity," 2nd edition, 701), for values of 7 

 ranging from 55 to 70 J . 



For determining the electro-chemical equivalent of silver, the 

 current passes also through silver voltameters. The solution of 

 nitrate, or of chlorate, is contained in a platinum basin which serves 

 as the kathode. The anode is a flat piece of fine silver sheet, 

 wrapped in filter-paper, and suspended by platinum wire at the top 

 of the liquid. The duration of the current is determined by a 

 chronometer, and allowance is made for the small loss of time (about 

 one-tenth second), incurred at each reversal of the current in the 

 fixed coils of the measuring apparatus. 



In the preliminary notice of March, 1884, the troubles into which 

 we were led by the use of acetate of silver were referred to. With 

 pure nitrate the manipulations present no particular difficulty. We 

 were equally successful with chlorate, prepared for us by Mr. Scott ; 

 and the comparison of the results with nitrate and chlorate verify 

 Faraday's law to a high degree of accuracy. 



In the reduction of the current weighings, we found it necessary to 

 time all the observations, and to plot the readings obtained in the two 

 positions of the reversing key as separate curves. The difference of 

 ordinates then represents the double electromagnetic force, as it 

 would have been found were it possible to take both observations 

 simultaneously. What we require for comparison with the mean 

 rate of silver deposit is the mean square root of the difference of 

 weighings, and is easily obtained when once the curves are constructed. 

 Apart from errors relating to the constant of the apparatus, the 

 mean value of a tolerably steady current of half an hour's duration 

 should be obtainable to about 10 go'o' ^ n our experiments, the whole 

 change of weight on reversal was about 1 grm., and each single 

 observation was correct to half a milligram. In the passage from 



