ROSA, DORSEY AND MILLER: INTERNATIONAL AMPERE J 1 7 



scope and scale and a small mirror mounted upon the central 

 knife edge of the balance. The scale distance was about 2.5 

 meters. The accuracy of reading, therefore, was equivalent to 

 that which would be obtained with a balance pointer 5 meters 

 long. Thus, it was possible to have the balance very stable and 

 of short period without decreasing the accuracy of the weighing. 



Since the intensity of a current cannot be represented by a 

 concrete standard, two expedients have been adopted to pre- 

 serve the results of the absolute measurement of current. One 

 is to determine the mass of silver deposited per second in a silver 

 voltameter by a current of one absolute ampere, the other is to 

 determine the electromotive force of a standard cell in terms of 

 a standard of resistance and the absolute ampere. In this work 

 the electromotive force of the Weston Normal cell at 20° C. was 

 determined in terms of the international ohm and the absolute 

 ampere. The authors have called the unit of electromotive force a 

 semi-absolute volt. Also, the final result has been expressed in 

 terms of the electrochemical equivalent of silver as determined by 

 a large number of observations with two types of voltameters at 

 the Bureau of Standards. 



In the final work four combinations of coils were used, besides 

 one which was not used in taking the mean, owing to a very slight 

 uncertainty in its cross-sectional area, and the following results 

 obtained : 



il/2, L3, IA. 



M3, L3, IA. 



MZ, II, L2. 



MS, 51, 52. 



Mean. 



COMBINATION 



E.M.F. OF WESTON NOR- 

 MAL CELL AN 22° C. 



1.01822 8 

 1.01820 2 

 1.01822 6 

 1.01823 5 



1.01822, 



DEVIATION FROM MEAN 

 PARTS IN A MILLION 



5 

 21 



3 

 12 



10 



The conclusion is that the electromotive force of the Bureau 

 of Standards concrete realization of the mean Weston Normal 

 cell at 20° C. January 1, 1911, was 1.01822 semi-absolute volts. 

 We believe that it is a conservative estimate to assign t«> the value 

 given above a possible uncertainty due to all causes of two in 



