1887.] Bismuth on the ductility of Silver. 125 



is now precipitated by the addition of 2*5 cc. of hydrochloric acid, and 

 after vigorous shaking the supernatant fluid will be found perfectly 

 clear ; and it will remain so when the bottle is filled up with water, all 

 the bismuth present being in solution. Whenever samples of silver 

 now show the presence of bismuth during the assay, a fresh set is taken 

 up and worked by the modified process, the delay thus caused not 

 amounting to more than a few minutes. It may be mentioned here that 

 all our assays are reported to one-tenth of a millieme (0*1 per mille). 



Having thus ascertained the presence of bismuth in silver bullion 

 and put in practice a modification of the assay process which renders us 

 indifferent to its presence, it is still of importance to ascertain the 

 exact proportion of bismuth which is present in the bullion ; and, to be 

 of practical use for mint work, this determination must be effected 

 rapidly and as simply as possible. The ordinary directions given for 

 the separation of bismuth in the presence of silver, by first removing the 

 latter as chloride and then precipitating the bismuth as carbonate, do 

 not, I find, give accurate results when silver is present in such over- 

 whelming proportions as obtain in the cases under consideration. 



I have therefore adopted the following plan, which a number of syn- 

 thetically prepared solutions have proved to give quick and good results, 

 though sometimes the amount of bismuth present is very slightly un- 

 der-estimated. The ordinary silver assay having given a rough visual 

 estimate of the amount of bismuth likely to be present, enough of the 

 bullion is taken to yield a fairly weighable amount of bismuth oxide in 

 the final result. The bullion is dissolved in a small amount of nitric 

 acid, the solution carefully diluted, and an excess of ammonium carbo- 

 nate at once added, the precipitation being aided by heating. The 

 carbonates of silver and copper at first formed are re- dissolved, and the 

 carbonate of bismuth after a time settles completely at the bottom of 

 the beaker. The contents of the beaker are then passed through a 

 filter, of which the weight of ash yielded by incineration is known, and 

 the carbonate of bismuth on the filter washed quite free of all traces of 

 silver. The filter is then dried, its contents transferred to a porcelain 

 crucible for ignition, the filter paper being ignited separately, treated 

 with a drop or two of nitric acid to re-oxidise any bismuth oxide reduced 

 by contact with the carbon of the filter, and the ash added to the cruci- 

 ble. From the weight of bismuth oxide thus found, after deducting 

 the weight of the filter ash, the amount of metallic bismuth present in 

 the sample of bullion taken for analysis can be at once found. 



There are only two metals likely to iuterfere with accuracy of the 

 process here described, namely, cadmium and lead ; the carbonates of both 

 these metals being as insoluble in excess of the precipitant employed as 



