344 PROCEEDINGS OF THE AMERICAN ACADEMY 



tation of silver from a solution of a pure silver salt by means of pure 

 copper ; and this was the method adopted in the present determination. 

 Hampe, in his work on the atomic weight of copper, attempted the 

 same method, but rejected it for two reasons, — the first being that the 

 silver dissolved, or appeared to dissolve, to a slight extent in the hot 

 water used for washing ; and the second being that it was impossible 

 to prevent copper from coming down with the silver, no matter how 

 long the precipitate was digested with the argentic nitrate solution. 

 As will be seen, however, both of these difficulties have been entirely 

 overcome. 



The silver salt selected for the precipitation was the nitrate, on 

 account of its ready crystallization, its easy solubility in water, and 

 the facility with which it can be obtained pure. For the preparation 

 of the salt used in the work, ordinary pure argentic nitrate was crys- 

 tallized many times from hot water, and finally fused for two hours in 

 an air-bath kept at 205° C, — a few degrees above its melting point. 

 This preparation was a white translucent substance interspersed with 

 transparent crystals, dissolving completely in water and giving a color- 

 less solution which was wholly neutral. 



The copper used was prepared by electrolysis from cupric sulphate, 

 through the kindness of Mr. Wilson, of the University Press. It was 

 cut into small pieces, and these were digested in succession with weak 

 potassic hydrate, dilute sulphuric acid, and then a very large amount 

 of water. The copper was then boiled with water for about half an 

 hour and washed with a large amount of distilled water, then dried 

 and reduced at a low red heat by means of perfectly pure hydrogen, 

 and allowed to cool in a stream of the gas. The metal thus pre- 

 pared had a beautiful red metallic lustre, and showed no trace of 

 oxidation after keeping a month. Before use it was dried in an air- 

 bath at 110°, allowed to cool in a desiccator, and weighed by itself on 

 a balance which was distinctly sensible to a twentieth of a milli- 

 gram. In addition to receiving the treatment desci'ibed above, the 

 copper used in the fifth and sixth experiments was oxidized in a 

 stream of pure air for half an hour, and then again reduced by hy- 

 drogen ; but the concordance of those two results with the others 

 shows that this precaution was not necessary. 



In a few preliminary experiments it was found that on the tempera- 

 ture of the solution, and on the temperature alone, depended the regu- 

 larity with which the silver was precipitated by the copper. At 90° 

 the deposition is very rapid, nitrous fumes are evolved, and a large 

 amount of copper comes down with the silver. As the temperature 



