SILVER,.— Assay. C ufellation.

758

stances, as the resulting button of alloy should notbe too rich, since in that case a portion of the silveris lost in the slag; nor too poor, as the cupellationwould then occupy a long time, and a loss throughsublimation be entailed. In ordinary cases, if 400grains of ore be the quantity operated on, a buttonof 200 grains will be a very convenient amount forcupellation ; and this may be obtained by the addi-tion of 300 grains of litharge, and from 7 to 8grains of finely-powdered charcoal. The whole isto be well mixed with 200 grains of carbonate ofsoda on a sheet of highly-glazed paper, and after-wards introduced into an earthen crucible, of whichit should not fill more than two-thirds the capacity.It is now covered with a thin layer of borax, andheated in an ordinary assay furnace—care being takento withdraw it from the fire as soon as a liquid andperfectly homogeneous slag has been obtained, asthe unreduced litharge would otherwise be liable toeat through the pot and spoil the experiment. Whenit has sufficiently cooled, the crucible is broken, andthe button of alloy obtained is cupelled. Whenother minerals than oxides or carbonates are to beexamined, the addition of charcoal, or any similarreducing agent, becomes in many instances unneces-sary, as litharge readily attacks all the sulphides,arsenio-sulpliides, &e., and oxidises nearly the wholeof their constituents, with the exception of silver,whilst a proportionate quantity of metallic lead is atthe same time set free. The slags formed in thisway contain the whole of the excess of lithargeadded, and the button of alloy produced is subjectedto cupellation in the usual manner.

All the directions given with regard to the assayof gold ores (see vol. ii. page 114), and the methodthere described for diminishing the amount ofreduced lead which has to be subsequently treatedby cupellation, apply with equal force to the assayof silver ores, and need not here be repeated. Thetreatment to which an ore is subjected dependingto a great extent on its nature, it is well to make apreliminary assay, in order that in the actual assayonly a certain amount of lead alloy may be submittedto cupellation. Mitchell recommends that all oresbe subdivided into three classes, indicating the specialtreatment to which they should be subjected. Theseclasses are—(1) Ores which, on fusion with excess oflitharge, give no metallic lead, or less than their ownweight. (2) Those which give their own weight, ornearly so, of metallic lead; and (3) those which givemore than their own weight of metallic lead. For thispreliminary assay 20 grains of very finely pulverizedore are mixed with 500 grains of litharge, and placedin a crucible of such size that it is only half full.After being carefully warmed it is rapidly heated ina bright fire, and when the mixture is thoroughlyfused the pot is removed, cooled, and broken. Theore belongs to the first, second, or third of theabove classes, according as the resulting buttonweighs less than, the same as, or more than 20grains, the weight of ore originally taken. Now200 grains of lead alloy is a convenient quantity tocupel; hence, in the actual assay, bodies of the

second class require simple fusion with litharge anda suitable flux; those of the first class require theaddition of such a quantity of a reducing agent thatthe final button may weigh 200 grains; and thoseof the third class, such a proportion of an oxidisingagent as shall oxidise all lead in excess of 200 grains.An excellent reducing agent is argol (commercialcream of tartar) ; and nitrate of potash is the oxidis-ing agent usually employed. The former is capableof reducing about 6 times its weight of litharge, andthe latter of oxidising about 4 times its weight.These data suffice for determining the quantitiesrequired for assaying ores of the first and thirdclasses respectively, when the results of the pre-liminary assays are known.

Scorification .—This process has been sufficientlydescribed under the Assay of Gold: it is only neces-sary to add that it is applicable to every kind ofargentiferous matter, and is both the most conve-nient and the most exact method of assay when largenumbers of assays are required to be done.

Cupellation .—This process has also been describedin considerable detail in the Assay of Gold; but itwill be necessary here to give some account of theprecautions to be taken when silver is underexamination.

In the assay of alloys of silver with copper it is, ofcourse, necessary to add the requisite amount of lead;whereas when ores are being assayed, the preliminaryfusion or scorifi'”'tion has rendered this unnecessary.If the composition of an alloy be not approximatelyknown, a preliminary assay will therefore be needed.An experienced assayer will frequently be able tojudge with sufficient accuracy from the colour andhardness of the metal; but when cupellation mustbe resorted to, 50 grains of pure lead and 2 grainsof the alloy may be introduced into the cupel, andworked in the manner about to be described. Orthe composition may be roughly ascertained by com-paring the appearance of a streak made by the alloyon a touchstone with those of a series of “testneedles,” or by noting the amount of oxidation pro-duced on a clean piece of the alloy when introducedinto the muffle for a definite period. Gay-Lussac himself suggested that a galvanic couple, formed ofthe unknown alloy and pure silver, should be em-ployed as a means of determining its approximatecomposition ; or else, that a series of standard alloysbe used, and successively associated with the alloy untilno deflection of the galvanometer needle is observed.

The amount of lead present should be greater inproportion as the quantity of copper is more con-siderable ; but in cupelling an alloy of silver andcopper, the same amount of lead cannot safely beadded as in treating an alloy of copper and gold, asthe greater volatility of the silver would cause aconsiderable loss of that metal before the whole ofthe lead could be absorbed. In this case, therefore,the following table, showing, according to D’Aucet,the amount of lead necessary to effect the propercupellation of various alloys of silver and copper,must be substituted for that given at the place abovereferred to:—