OF ARTS AND SCIENCES. 347 



precipitate by dissolving one to two grams in nitric acid and adding 

 excess of amnionic hydrate. In no case was the slightest blue color 

 noticed ; and as under the same circumstances a tenth of a milligram 

 of copper gave a distinct bluish tinge, it may be assumed that the 

 silver contained no copper. This method is a convenient one for the 

 preparation of chemically pure silver, and avoids the necessity of fusion 

 and the concomitant absorption of oxygen and possible admixture of 

 silicic anhydride. 



Although the value of the atomic weight of copper obtained, 63.44, 

 does not exactly coincide with Shaw's results, it is at least very much 

 nearer to them than is the old accepted value of 63.17 ; especially if we 

 take his result as corrected for current density. The new value is 

 very nearly that found by Dumas from cuprous sulphide, and it falls 

 within the limits assigned by L. Meyer and Seubert as the possible 

 error of the accepted value. 



