172 THE ROYAL SOCIETY OF CANADA 



normal Weston standard. Not only have a number of cells been con- 

 structed and compared, but the following series of direct comparisons 

 have made it possible to check with great accuracy for any given 

 batch of cells, the calculated value of the deviation in voltage from the 

 international mean value. 



1. Six Weston cells carried from Bureau of Standards, Washing- 

 ton, in 1908. 



2. Three Weston cells carried to Bureau of Standards, Washing- 

 ton, in 1909. 



3. Three Weston cells carried from Bureau of Standards, Wash- 

 ington, in 1909. 



4. Three Weston cells carried from the National Physical Labora- 

 tory, Teddington, in 1909. 



5. Three Weston cells mailed from the National Physical Labora- 

 tory, Teddington, in 1909. 



6. Four Weston cells carried to National Physical Laboratory, 

 Teddington, in 1911. 



7. Two Weston cells carried from the National Physical Labora- 

 tory, Teddington, in 1912. 



8. Two W^eston cells carried from the Bureau of Standards, in 

 1919. 



It is with much pleasure that we record our indebtedness and 

 thanks to Dr. Glazebrook and to Dr. Stratton for this kind help from 

 the National Physical Laboratory and the Bureau of Standards, and 

 especially to Dr. F. E. Smith and to Dr. Wolff who were respectively 

 in charge of the cell investigations at these institutions. 



§2. The Constancy of Normal and of Modified Standard 

 Weston Cells During a Period of Ten Years 



The following table shows the constancy of two of our standard 

 batches, one composed of our normal, and the other of our modified 

 cells, during a period of ten years. The results are expressed in micro- 

 volts, and with the exception of the lowest row of values, give the 

 differences from the international mean value as determined with 

 the aid of the various comparisons already enumerated. The last line 

 is expressed in terms of our own normal cells and illustrates the re- 

 markable confidence with which independent workers may rely upon 

 'Standards constructed in their own laboratory according to modified 

 specifications. An average drop of six microvolts per annum is as- 

 sumed in the latter estimation. 



