ON SILTEll. * 377 



to conccnirat^ it, pour it into a retort, at the bottom of 

 which a few bits of pure silver have been previously put; 

 if not, keep it in a bottle. 



But the parted silver must be cleansed before it is used. Parted sliver 

 to avoid any mixture of copper. For this purpose boil should be freed 

 the powder in a saturated solution of silver slightly di- ^^ 



luted, when the copper will be converted into nitrat, and 

 the silver will come out sufficiently pure to effect no 

 alteration in the colour of the fresh nitrat. 



The solution is invariably of a pale yellow colour; 

 and it may be concentrated much beyond the nitrat at a 

 maxi?ni(m, without fear of crystallization, as the nitrat it 

 affords is infinitely more soluble. When it is in the pro- Minimum ni- 

 portion of 240 to 100 of water it is still far enough from b^^' """^ •°^''' 

 crystallizing, and sometimes it remains fluid for several 

 days : but if it be poured into a phial, it congeals so 

 suddenly, that the last portions from the mouth of the Suddenly be- 

 r^tort become solid like icicles from the eaves of a house, comes solid. 

 and a great deal of heat is evolved. 



While this nitrat is concentrating, a little is always a little vola- 

 tolatilized. This passes from the minimum of oxidation tilized and 

 to the maximum ; and sometimes we discover in it a mix- in? evapora-' 

 ture of the two nitrats. In the first case animonia does tion. 

 not change the transparency of the product ; in the 

 second, the new nitrat is indicated by a black hue. 



It is difficult to bring it to a regular crystallization. Difficult to 

 because it has a much greater tendency to congeal, than crystallize. 

 to separate into crystals. Jf it congeal, it cannot be 

 redissolved, without the separation of a yellow powder, yellow pow^ 

 which obliges us to suspend the process, that this pow- der produced, 

 der may be allowed to subside, decant the clear fluid, and 

 return it into the retort. It is true that this inconveni- 

 ence may be avoided, by an addition of acid to the water 

 with which the mass is to be redissolved ; but by thia 

 addition we are liable to increase the oxidation of a por- 

 tion of the nitrat, and to convert the product into a 

 mixture of the two nitrats. 



The yellow precipitate we have just pointed out^ is a ^^j^i^^jj ^o^^gij^, 

 nitrat doubly at a minimum^ both of oxigen and of acid, a minimum 

 It forms because part of the nevjr nitrat cannot dissolve in ^°^^of ^acid^" 

 water, unless it takes from the other a little of its acid ; 



and 



