388 On the Jlethod of Assaying Silver. [No. 4, 



a vessel of water, and the granules so formed received on a 

 strainer, lifted out and perfectly dried.* The weight of this speci- 

 men representing each pot was at first fixed at 24 grains techni- 

 cally called the " assay lb" ; this in the case of pure silver yi elded 

 31.87 grains of chloride of silver, while the same quantity of 

 Indian Standard silver (which is -\]th. silver plus ^th copper = 

 916.66 in 1000) yielded one-twelfth less or 29.21 grains ;— on 

 the weight of chloride being ascertained in each case a table which 

 was calculated and prepared for the purpose was referred to and the 

 equivalent fineness assigned to the £ dwt., plus the odd grains, when 

 any. But when it became desirable to prepare for the decimal form 

 of notation, a number more convenient than 31.87 was looked for 

 to represent purity or 1000, and 25 was fixed on as a desirable start- 

 ing point, particularly as the quantity of pure silver yielding that 

 amount of chloride, viz. 18.825 grains, was quite large enough to 

 represent each pot.f 



The weight therefore of the " assay pound" in use at present is 

 18*825 grs. This produces (with chlorine) in the case of pure 

 silver 25 grains of chloride of silver. J 



But to obviate the necessity of constant reference to a calculated 

 table to find the equivalent in pure silver of the amount of chloride 

 of silver found in each case, it was ingeniously arranged to stamp 

 each of the assay weights not with its actual weight, but with the 



* The introduction into the Calcutta Mint, of this system of taking musters 

 is I find attributable to Dr. Boycott, late Assay Master, and to Dr. 

 Shekleton, who by a number of interesting experiments satisfied themselves 

 that samples so taken represent the mass of mixed metal to be valued much more 

 fairly than samples of the same mass cut or gouged from it after it has been 

 poured and allowed to cool in the ingot moulds where a partial separation of the 

 copper from the silver seems to take place; the result being, according to the above 

 experiments, that in the case of ingots cast in upright moulds, all the outside i3 

 much below the average fineness of the mass on assay, and the centre much above 

 it. This refers to alloys of silver and copper mixed in or about the proportion of 

 " standard." According to Monsieur L e v o 1, however, it would appear that when 

 an alloy of silver and copper in which the proportion of the latter is very high 

 (viz. over 28 per cent.) has been melted, poured, and allowed to cool, an opposite 

 result to the above is found, viz. the outside of the ingots is above the average 

 fineness. An assay therefore from a granulated sample must give a much nearer 

 approximation to truth, than one from a cut sample. 



f The average weight of the contents of each melting pot is 12,500 tolas or 

 about 390 pounds Troy, so that the specimen taken to represent this is but 

 about the 119,000th part ; each sample is assayed in duplicate. 



X The basis for these numbers was founded on the proportion in which, 

 according to Turn er, silver combines with chlorine viz. 100 parts with 32.80. 



