880 ON SILYEB. 



quently separated in the state of yellow nitrat, or nitra* 

 with an inferior portion of acid. With boiling water the 

 change proceeds farther* 



If a few drops of a somewhat concentrated solution, 

 be let fall into a glass of boiling distilled water, three 

 colours will be seen distinctly to succeed each other very 

 quickly, yellow, red, and black. If, when the mixture 

 is yellow, or red, a few drops of acid be added, the whole 

 grows clear, and the change is stopped. If this be done 

 the moment after the black colour has appeared, the acid 

 no longer restores the transparency of the mixture, be- 

 cause the black powder is not an oxide, like the two 

 DIsoxidates the preceding, but disoxidatcd silver. Now to dissolve this 

 iilver. ^ stronger acid is required. 



The black powder silvers the glass as it comes into 



contact with its sides. None of these effects will takUr 



place, if a few drops of nitric acid be added, before the 



solution is dropped into the boiling water. 



Kitrat of mcr- The yellow nitrat of mercury at a minimum exhibits 



«xry has simi- similar phenomena, though in a different manner. TKis 



njtrat, boiled with water in a retort, affords powder of 



mtercury, which passes into the receiver with the aqueous 



vapour ; that is to say, by the help of a high temperature 



part of the mercury has a tendency to complete its oxida« 



tionat the expense of the other. But in whatever cir- 



j^ J cumstances we discover a metal passing from one state of 



found at an oxidation to another, we shall never find it stopping at 



intermediate any intermediate term between the two extremes of oxida- 

 state of oxida- .. ... 



tion. tian proper to it. 



^ . An aqueous solution of minimum nitrat of silver does 



gen from the not superoxide itself very rapidly by contact with air. 



air but slowly. \ havQ formerly shewn, that, if copper have a greater 



Acids have affinity for oxigen than silver has, we must not thence 



more affinity conclude, that acids would have more affinity for oxide of 

 for oxides of ' ^ . , ^ ., ,./-,. t, 



silver thun of copper than for oxide of silver ; and in fact the nitrat and 



copper. sulphat of silver dissolved and kept on carbonat of cop- 



per do not attack the latter, or even become coloured. 

 The minimum nitrat of silver is equally void of action on 

 carbonat of copper. 

 ^ The yellow powder, or minimum nitrat with less acid, 



is equally void of action an the carbonat of copper. 



The 



