Bismuth will serve the same purpose as lead in cupellation; but, besides being dearer, it is found to carry down with it into the cupel somewhat more of the sil ver than the same quantity of lead does.
To estimate the quantity of alloy in silver, the ancient assayers used touch needles, or small slips of silver, alloyed with known proportions of copper, in a regularly increasing series, from the least to the greatest proportion ever used. The silver to be assayed was compared with these, and its alloy estimated by that of the needle, to which it skewed the closest resemblance. But an experienced assayer is at the present time able to judge of the alloy with sufficient exact ness, by the ease or difficulty with which the silver is cut, by the colour and grain of a fresh cut surface, the malleability, the change of surface when made red hot, and the general appearance.
The assay of gold is more complicated than that of silver. The baser metals may be separated from it by cupellation in the same manner as from silver, ex cept copper, which has so strong an affi nity for gold, that it can scarcely be over come by this method, unless silver is first combined with the mass ; and this makes the second operation necessary, mention ed before, namely, the parting of the gold from the silver..
The process of parting is performed by the aquafortis of commerce, which dis solves the silver, and leaves the gold un touched. But in this operation it is found, that when the gold exceeds a certain pro portion in the mixture, it so much pro tects the silver from the acid, as more or less to prevent its action. Therefore, when the gold is in excess, it becomes necessary to add so much silver as to give this metal the predominance. The pro portion of silver generally used is three parts to one of the gold, from whence the process obtained the name of quarts tion. Several good assayers think two parts of silver are sufficient. More than three parts also may be used, but then it will protract the process needlessly.
Though, when copper also as well as silver is present, the parting may be pro ceeded to, as this metal is likewise solu ble in asptafortis ; yet it is found to have some advantages to cupel the mixture first with lead ; and likewise even when no copper is combined with the gold.
The cupellation of gold is thus con ducted ; the portion of the alloy of silver being estimated by touch-needles, as much silver is added as will make the en tire quantity of this metal about thrice the weight of pure gold.
The proportion of lead to the alloy of copper, or other base metal, is nearly the same as for silver ; which will be shewn particularly in the annexed table. The heat necessary in the process is greater than for silver, and may be used with freedom, as none of the gold is lost by vo latilization. The lighting of the fused globule of gold takes place as in silver. The button is cooled, taken out and weighed, then hammered flat and anneal ed, and afterwards laminated between steel rollers to a thin plate about the sub stance of a wafer, again heated to red ness, and then coiled up into a spiral roll. The button of gold, when it lightens, still retains a minute portion of lead ; this may be got rid of by its being kept a little time in fusion in a clean vessel. The lead en tirely disappears after parting.
The spiral roll is called a cornet, and when prepared is put into a glass matrass, shaped like a pear, in order to part the silver from it, and about thrice its weight of pure nitric acid poured on it moderate ly diluted, (so as to be about 1.25 specific gravity.) The glass is set on a sand bath, or over charcoal, to boil. 'When warm, the acid dissolves the silver ; as long as it continues to act, the cornet is studded with minute bubbles : when these dis. continue, or are united in one large one, it is a sign that the acid has ceased to ope rate. About twenty minutes are requir ed for this process.
The cornet is now corroded through out, having lost its silver ; it retains the same form, but is very slender and brittle. It is of importance to the accucray of the assay that it should not be broken. The hot acid solution of silver is then poured off with great care, and fresh acid, rather stronger, is added, to clear away all re mains of the silver, and boiled as before, but only for five or six minutes. It is then decanted, and added to the first so lution, and the parting glass is filled with hot distilled water, to wash.off all remains of the solution. A small crucible is to be inverted over the glass while it is full of water, the latter is then nimbly turned, and the cornet falls gently into the cruci ble through the water ; which being poured off, the crucible is dried and heat ed to redness under a muffle, when the cornet shrinks extremely in every direc tion, becomes firm, and when cooled, re gains its metallic lustre, and is soft and flexible. It is then most accurately weighed, and the process is finished.