554 



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



adopted this method, but, owing to the ampere balance being less steady when a 

 current passes through it for more than half an hour, the measurement of the current 

 through the voltameters would have been subject to a comparatively large error. 

 By frequently standardising a combination of resistance coil and cadmium cell (as 

 already described) and employing these for the measurement of ciirrent through the 

 voltameters, the probable error was appreciably reduced. 



Mass of Depo * //*. 



The large bowls were weighed on a balance by L. OERTLING, sensitive to one-fiftieth 

 of a milligramme ; the small bowls were weighed on a more sensitive balance by 

 OERTLING, and very kindly lent by Mr. OERTLING for this work, or on a balance by 

 RUPRECHT, which was loaned to the National Physical Laboratory by Dr. SCOTT, of 

 the Davy-Faraday Laboratory. All weighings were made by the method of GAUSS, 

 a similar vessel being used as a tare. The mass of silver usually deposited was about 

 7 grammes in the large bowls and a single weight of 7 grammes was used to counter- 

 poise ; the difference was obtained by means of a rider. The weight employed was 

 standardised by comparison with a 100-gramme weight from Sevres. For difference 

 determinations the four bowls A, B, C, D were largely used. 



To reduce the error of weighing we have found it convenient to compare the 

 masses of the bowls when empty and again with the deposits of silver. The following 

 example is one with the bowls containing silver, and gives an idea of the error 

 introduced : 



