SEPAEATIIfa GOLD FKOM AEaENTIC CHLOEIDE. 69 



in a No. 18 boraxed French clay pot. Twenty ounces of soda 

 produce a gold button weighing about 35 ounces, assaying from 

 870 to 880, while the silver bullion produced will contain from 

 two to five parts of gold in 10,000 parts. 



The operation is very regular in its results, but as seen, not all 

 the gold is removed thereby ; in fact, in no case, even with an 

 increased quantity of soda, was the whole of the gold removed in 

 one operation. To produce silver bullion free from gold was, 

 however, always successful when the argentic chloride was 

 subjected to a second treatment, with about 3 ounces of 

 soda pro 200 ounces of argentic chloride. This second opera- 

 tion is done similar to the first, but in a new pot, also boraxed ; 

 it requires a short time,- — about one hour. The argentic chloride 

 containing only traces of gold from the previous treatment with 

 soda, fuses much more readily than when it contains much gold. 

 The time occupied by the first operation varies somewhat, accord- 

 ing to the heat of the melting furnace and the character of the 

 chloride. To fuse 230 ozs. argentic chloride, after having been 

 introduced into a red-hot pot, placed inside a guard, takes from 

 60 to 80 minutes ; the addition of the soda from 20 to 30 min- 

 utes, after which the pot is covered and the heat of the furnace 

 increased to get all in good fusion, which takes from 10 to 20 

 minutes. The pot is then removed from the fire, allowed to cool 

 sufiiciently for the gold to solidify, when the still liquid argentic 

 chloride is poured ofl" into iron pans and placed into the galvanic 

 battery, a description of which I gave in a paper read before this 

 Society in December, 1869. 



"While the soda is being added, the top layer is occasionally 

 gently dipped with a stirrer slightly underneath the molten 

 chloride, without stirring the same ; in fact, it is preferable not 

 to stir the fused chloride, but to let the gold collect at the bottom 

 of the pot, and to pour the chloride carefully off". 



The presence of a large proportion of copper in the chloride 

 has been found to prolong the operation considerably ; it is 

 therefore advisable to refine gold bullion containing much copper 

 by itself, and to free the resulting argentic chloride, which there- 

 fore contains much copper, by dissolving the same after being 

 reduced to the metallic state. 



It is remarkable how uniformly the gold is difi'used in the 

 argentic chloride. Any portion of a slab of this chloride, free from 

 borax, may be assayed for gold and will be found alike. This 

 offers a convenient means for ascertaining the result of the treat- 

 ment with soda before the argentic chloride is placed in the 

 battery for redaction. A small piece is broken oS" from one 

 corner of the slab of chloride, and reduced to fine powder in a 

 wedgewood mortar ; the powder is kept in a corked glass tube, 



