92 PROCEEDINGS OP THE AMERICAN ACADEMY 



volume) is poured into the crucible until it is two-thirds full, a second 

 piece of platinum foil united to a wire is then placed in the acid solu- 

 tion, care being taken that it does not touch the silver mixture. The 

 zinc pole of a two-cell Bunsen battery is connected with the platinum 

 foil that rests on the silver salts, and the carbon pole with the plati- 

 num foil just mentioned. The decomposition begins immediately, 

 chlorine and bromine being given off from the positive electrode, the 

 reduced silver remaining at the bottom of the crucible in the form of 

 a porous mass. The reaction requires from twelve to eighteen hours ; 

 with a weight of less than one and a half grammes of the mixed haloids 

 I have found twelve hours sufficient to produce complete reduction ; 

 with a weight exceeding that amount I prefer to allow eighteen. 

 When the argentic chloride and bromide are completely reduced, the 

 battery is disconnected, the electrodes taken out of the sulphuric acid 

 solution and washed with distilled water. The sulphuric is then 

 poured off from the silver, and the silver is washed by decantation 

 with distilled water, the decanted liquid being poured through a small 

 filter ; this is afterwards burnt, added to the silver sponge which still 

 remains in the crucible ; the crucible is then heated over a low free 

 flame to constant weight. The weight thus found, minus the weight of 

 the crucible and filter ash, is of course the weight of silver contained 

 in the argentic chloride and bromide. 



The only point that requires any great degree of care in this process 

 is the melting of the mixed haloids. The temperature at which the 

 fusion takes place must be as low as possible, so as to avoid any vola- 

 tilization, and the melted mass should be united as far as practicable 

 in one piece. 



In some of the following analyses I have used a platinum crucible. 

 When this has been the case, one wire has been wound around the 

 crucible, while the other, as before, has been merely allowed to dip 

 into the acid solution ; in this way the whole crucible serves as an 

 electrode, and there is no need of bringing the melted chloride and 

 bromide into one globule. With a platinum crucible, the washing of 

 the reduced silver must be continued until a few drops of the filtrate 

 gives no precipitate with baric chloride, and the drying should be done 

 in an air bath at a temperature of about 150° C. After weighing, the 

 silver, which always adheres to the crucible, can be dissolved out with 

 dilute nitric acid. 



The first series of analyses shows the accuracy of the process when 

 either argentic chloride or bromide is taken alone, the second series 

 when they occur together. 



