792

ASSAY OF SILVER BY CUPELLATION.

long pair of tongs. It quickly fuses; fumes of oxide of lead are scen rising from the cupels, but the greater part of the oxide is absorbed by the cupel, and the silver is left behind in a state of purity. At the moment that the last portion of lead undergoes oxidation, the surface of the silver flashes, or lightens as it is technically termed, owing to the cause already explained (892). This phenomenon indicates that the process is completed. The button is allowed to cool very gradually, to prevent the loss of silver by dispersion from spitting (931); it is then detached from the cupel, brushed, and accurately weighed. If the piece of alloy originally taken weighed 10 grains, the weight of the button in hundredths of a grain gives the number of parts of silver in 1000 parts of alloy. A minute quantity of silver always passes into the cupel during the process, for which an allowance must be made in weighing the button; and if the proportion of lead be too great this loss is increased, but if too little be used, part of the copper is left in the bead. Upon an alloy which contains 925 parts of silver to 75 of copper, the loss is about 4 per 1000; and upon silver which contains 900 parts in 1000, the loss on the button is about 5 parts in 1000. In order to be able to estimate the amount of this loss in each operation, the best plan is to pass three or four proofs, with each set of assays. These proofs consist of pieces of fine silver of known weight, which are mixed with quantities of lead and copper, approximatively of the same amount as those present in the assays under trial. The loss experienced by these proofs affords a method of checking the results of the assay. The amount of this loss varies with the temperature.

The most convenient system of reporting the fineness of silver is the decimal method, which is employed in most countries with the exception of England. The practice of reporting both gold and silver decimally was introduced a few years ago by Sir J. Herschel into the Mint of this country, and it probably will gradually supersede the cumbrous and artificial method which is still generally employed by the English assayers. Upon the decimal system, fine silver is termed 1000'0, and the report upon any sample of alloy simply indicates the number of parts of pure silver in 1000 which it contains. Thus English standard silver contains 925 parts of silver, and 75 of copper in 1000 of the alloy. French standard contains 900 parts of silver, and 100 of copper in 1000 of alloy. English standard would therefore be reported as 925; French standard as 900.

The proportions of lead which are generally employed for the cupellation of different alloys are the following:

A skilful assayer will generally be able at once to determine the comparative fineness of an article from its mere appearance, and will judge accordingly of the proportion of lead which it will require. Great care is needful in regulating the temperature of the furnace during the cupellation; if too high, a part of the silver will be lost by volatilization; if too low, portions of lead and copper are liable to be retained. When the assay is properly performed, the button is brilliant, well rounded, free from irregularities, and somewhat granular upon its surface: it is readily detached from the cupel. If the assay adheres strongly to the cupel, or is irregular in its outline, it retains a portion of alloy.

(939) Assay of Silver by the Humid Process.-The results of the process of the assay by cupellation, even in experienced hands, may vary as much as 2 parts in 1000: this circumstance induced Gay-Lussac to contrive a different method, which is now adopted not only in the French Mint, but is employed in the Mints of Great Britain and the United States, as well as in almost all the Mints of Europe: it admits of an accurate estimate of the value of an alloy to within o'5 in 1000. This process depends upon the solution of the alloy in nitric acid, the precipitation of the silver from the nitrate in the form of an insoluble chloride, and the measurement of the amount of a standard solution of chloride of sodium which is required to effect the complete precipitation of the silver in a given weight of the alloy. Chloride of silver easily collects into dense flocculi by agitation in a solution which is acidulated with nitric acid, and which contains no excess of soluble chlorides; so that the exact point at which the precipitate ceases to be formed is readily perceived.

A solution of common salt is prepared of such a strength that 1000 grains of it are exactly sufficient to precipitate 10 grains of pure silver. 10 grains of the alloy for examination are placed in a stoppered bottle capable of holding about 6 oz. of water, and by

794

ASSAY OF SILVER BY THE HUMID PROCESS.

the aid of a gentle heat, are dissolved in 2 drachms of nitric acid of specific gravity 125: the solution of salt is then placed in a

FIG. 360.

burette (fig. 360) capable of holding rather more than 1000 grains. The burette, when filled with the solution, is weighed before being used, and the liquid is added to the nitrate of silver in the bottle; when it is supposed that the silver is nearly all precipitated, the liquor is briskly agitated in the bottle, and the precipitate is allowed to subside; a drop or two more of the solution of salt is then added: if a precipitate be produced, the liquid is again agitated; and when clear, more of the solution is added, as before, so long as any turbidity is produced by the addition. When a cloud ceases to be formed, the proportion of solution of salt which has been added is ascertained by weighing the burette a second time. The number of grains of the solution employed indicates the degree of fineness of the alloy.*

When, as in the assay of bars for coin or for jeweller's work, a large number of assays must be executed, all very nearly of uniform fineness, the operation may be reduced to a system by which its precision may be increased, at the same time that it is rendered much more easy of execution. For this purpose, two solutions of salt are employed: one, the standard solution, containing in 1000 grains a sufficient quantity of commercial chloride of sodium to precipitate 10 grains of silver;t the second solution, the decimal solution, having one-tenth of the strength of the first, and being prepared by diluting 1 pint of the standard solution with 9 pints of water. These solutions are to be preserved in well-closed bottles. The standard solution is prepared in large quantities at a time, and kept in stoneware jars, A, fig. 361, capable of containing 20 or 25 gallons: b is a tube open at both ends, which passes nearly to the bottom of the jar, to admit air, whilst the liquid is drawn off by the stopcock, c, without allowing any loss by evapo

In the Calcutta Mint this precipitate is washed by subsidence in the vessel in which it is formed, and is then collected in a small porcelain crucible, as in the process of collecting gold, in the operation of parting (958). The chloride is dried, and then weighed, and the corresponding value of the silver is calculated.

This solution contains approximatively 380 grains of chloride of sodium in a gallon: but as the commercial salt contains chloride of magnesium, the exact strength must be determined by dissolving 10 grains of fine silver in acid, and precipitating it by the addition of 1000 grains of the solution, ascer taining the amount of the excess or deficiency of chloride in the manner about to be detailed, and then adding water or salt as may be needed.

ASSAY OF SILVER BY THE HUMID PROCESS.

:

FIG. 361.

795

ration; d is a gauge by which the quantity of liquid within is indicated. A series of bottles, capable of containing about 6 fluid ounces each, is fitted with ground stoppers, numbered consecutively from 1 upwards into each bottle 10 grains of the alloy for assay are weighed; 2 drachms of nitric acid are added to each bottle, which is placed in a shallow vessel containing water, and gradually raised to the boiling point; in ten minutes the alloy is completely dissolved.

The precipitation of the silver in the form of chloride is then effected by the aid of the apparatus now to be described:-g, fig. 361, is a glass pipette which can be filled with the standard solution. The quantity of liquid introduced into the pipette is regulated by means of the stopcock, e f, the peculiar construction of

which is shown on a larger scale in fig. 362, in which e represents an ordinary stopcock (constructed of silver to resist the action of the solution), terminating below in a long tube, c; at f is an opening for the escape of air, which can be closed at pleasure by the plug, a. Suppose it be desired to fill the pipette, g, fig. 361; the lower opening of the pipette is closed by the forefinger, the solution is admitted by opening the stopcock, e, whilst the air escapes at f, which is open; as soon as the liquid has risen a little above the mark, n, both the stopcock, e, and the plug at ƒ are closed, and the finger is withdrawn. In this position the pipette will retain its charge for an indefinite time. The apparatus re

796

FIG. 362.

ASSAY OF SILVER BY THE HUMID PROCESS.

presented at ml is intended to facilitate the exact emptying of the pipette; the tray, h i, slides easily between two grooves, in which its motion is limited by the stops and m; h is a compartment for the reception of the assay bottle, so adjusted that when the tray rests against the stop m, the pipette shall empty itself into the bottle without wetting its neck; i is another compartment for receiving the superfluous solution of salt, and k represents a piece of sponge, the object of which is to remove the drop which hangs to the lower end of the pipette; the stop / is so placed, that when the slide rests against it, the sponge shall just touch the lower extremity of the pipette. The sponge, k, having been brought to touch the lower end of the pipette, the plug at f is slightly relaxed to allow the air to enter, and a portion of the liquid gradually to escape, until it has fallen exactly to the mark n. The slide is now moved until the bottle, h, is directly underneath

Fig. 363.

the pipette, and on opening the plug at f to its full extent, the charge flows freely into the bottle.

Suppose now the object of the assay be to ascertain whether a certain number of bars be of the fineness of English standard silver, or if not, what is the amount of their variation from standard. The pipette, g, is so graduated that when filled up up to the mark n, it shall deliver exactly 922 grains of the standard solution, which will contain a sufficient amount of common salt to precipitate 9*22 grains of silver; a