ASSAYING

Process of Assaying.

We shall now proceed to a description of the process of assaying, as performed by the assayers of the Royal Mint and Goldsmiths' Hall; and shall then state the facility afforded by the furnace of Messrs Anfrye and D'Arcet in conducting this operation upon a smaller scale and reduced expence.

Some preliminary observations may be requisite in regard to the muffle and cupels, to the proportioning of lead in assaying, &c., before the operation of the assay commences.

In the furnace above described, the number of assays that can be made at one time is 45. The same number of cupels are put into the muffle. The furnace is then filled with charcoal to the top, and upon this are laid a few pieces already ignited. In the course of three hours, a little more or less, ac-

cording to circumstances, the whole is ignited; during which period the muffle, which is made of fire clay, is gradually heated to redness, and is prevented from cracking, which a less regular or more sudden increase of temperature would not fail to do. The cupels also become properly annealed; all moisture being dispelled, they are in a fit state to receive the piece of silver or gold to be assayed.

The greater care that is exercised in this operation, the less liable is the assayer to accidents from the breaking of the muffle, which it is both expensive and troublesome to fit properly into the furnace.

The cupels used in the assay process are made of the ashes of burnt bones (phosphat of lime). In the Royal Mint, the cores of oxhorn are selected for this purpose, and the ashes produced are about four times the expence of the bone ash used in the process of cupellation upon the large scale. So much depends upon the accuracy of an assay of gold or silver, where a mass of 15 lbs. Troy, in the first, and 60 lbs. Troy in the second instance, is determined by the analysis of a portion not exceeding 20 Troy grains, that every precaution which the longest experience has suggested is used to obtain an accurate result; hence the attention paid to the selection of the most proper materials for making the cupels.

The cupels are formed in a circular mould made of cast steel, very nicely turned, and by which means they are easily freed from the mould when struck. The bone ash is used moistened with a quantity of water sufficient to make the particles adhere firmly together. The circular mould is filled and pressed level with its surface; after which a pestle or rammer, having its end nicely turned, of a globular or convex shape, and its size, equal to the degree of concavity wished to be made in the cupel for the reception of the assay, is placed upon the ashes in the mould, and struck with a hammer until the cupel is properly formed. These cupels are allowed to dry in the air for some time before they are used. If the weather is dry, a fortnight will be sufficient.

The greatest possible attention should be paid to the quality of the lead used in assaying. If it contain silver, it will be easy to perceive a source of material error in the delicate operations of the assayer. Lead revived from litharge contains only about half a grain in the pound weight, and is preferred, on that account, to lead immediately revived from the ore, which usually contains a larger quantity.

The proportion of lead used in an assay of silver varies according as the external character of the silver to be assayed indicates a comparative state of fineness or coarseness to standard metal; of which an expert assayer may pretty accurately determine by the eye; but his opinion will also in some measure be regulated by the comparative ease or difficulty of flattening upon an anvil the piece of silver to be assayed;—if coarse, the metal is harder than standard, and of a brilliant glossy appearance; but, if soft and easily flattened, and of a dead white colour, it will indicate a state approaching to purity. The quantity of lead is then proportioned by the opinion of the assayer, and varies from 10 to 20 times the weight of the silver used. It should be observed, that a cupel is capable of absorbing only its own

Assaying. weight of litharge, and attention should accordingly be paid to the size of the cupel, when any silver is to be assayed, which requires a great quantity of lead.

As it is always requisite to proportion the lead to the estimated quantity of alloy in the silver before cupellation, the ancient assayers made use of touch-needles, which were bars or slips of metal, made with pure silver, alloyed with definite proportions of copper, in a regularly increasing series, from the least to the greatest proportion, which may ever be required. The silver to be assayed was examined in comparison with the touch-needles, in colour, tenacity, and other external characters, and its alloy was estimated by that of the needle, to which it showed the closest resemblance. These needles are seldom or never used now; and the external character of the metal is sufficient to guide an experienced assayer in the proportioning of the lead, to the estimated alloy in the silver, which he has to assay.

In Aikin's Dictionary of Chemistry and Mineralogy, under the article ASSAYING (to which we here acknowledge our obligations), there is a table of the proportions of lead to the estimated alloy in fine silver, founded upon the experiments of Messrs Tillet, Hellot, and Macquer, which were the basis of a regulation subsequently adopted by an edict of the late French government. The great uncertainty of the use of the touch-needles, probably suggested these experiments to the

French Chemists; and as this table may be extremely useful to unexperienced assayers, we shall insert it here, together with the observations accompanying it in the above work.

"Copper, the usual alloy of the fine metals, when taken singly, is found to require from ten to fourteen times its weight of lead for complete scorification on the cupel. Now, all admixtures of fine metal tend to protect the copper from the action of the litharge, and the more obstinately, the greater the proportion of fine metal. So that copper, with three times its weight of silver (or 9 oz. fine), requires forty times as much lead as copper; with eleven parts of silver it requires seventy-two parts of lead, and the like in an increasing ratio. The following is the table of the proportions of lead required to different alloys of copper; of which a few points are founded on the above mentioned experiments, and the rest filled up according to the estimated ratio of increase (being multiples of the assay integer 24 in arithmetical progression). In the three first columns is shown the absolute increase of the quantity of lead in alloys of decreasing fineness; in the three last columns will be seen the gradual diminution of the protecting power of fine metal against scorification, in proportion to the increase of alloy, shown by the decreasing quantity of lead, required for the same weight of copper, under different mixtures." (Aikin's Dictionary.)

In the article just referred to, it is remarked, that many assayers of good authority use proportions of lead to alloy, considerably different from the above table, and that the whole of the numbers here given may be considered as rather high in regard to the quantity of lead. The German assayers, it is added, observe the following rule:

In proportioning the lead to the alloy supposed to

exist in the silver to be assayed, care must be taken not unnecessarily to increase the quantity; though it would be all oxidated or absorbed by the cupel sooner or later; which is proved by the cupellation of lead per se, in order to ascertain the portion of silver it contains; the latter being always found in a globular shape on the cupel, and in a state of purity.

In the process of cupellation with lead, however, there is always a loss of silver. Mr Tillet found, by experiments which he made with pure silver and lead, whose retent of silver was known, that after the process of cupellation, the button of silver was never precisely of the same weight as before; but was always a portion lighter, even when the heat of the assay furnace was not sufficient to drive off any of the silver. The conclusion was obvious,—a part of the silver was carried into the cupel by the lead;

Assaying. this was proved by reviving the oxide of lead from the cupel, and cupelling the lead by itself, when the quantity of silver left upon the test was found to be ten times as great as the natural proportion of this metal in the lead, and very nearly corresponded with the loss of silver in the first instance. It will be obvious, then, that the assayer's report of the title or purity of any sample of silver (unless corrected) would make the metal somewhat less pure than it actually is, because all loss is put to the account of alloy. Mr Tillet calculates, when no more lead is used than is necessary for the entire separation of the alloy, that it carries down into the cupel as much silver as, when the whole is again reduced, would make the noble metal $\backslash frac\{1\}\{128\}$ of the mass, when the natural admixture of the silver is only about $\backslash frac\{1\}\{152\}$. But if an excess of lead is employed for cupellation, this loss of silver is somewhat greater, though it does not increase in the ratio of the excess of the lead; for 10 parts of lead to a given alloy will not carry down twice as much silver as 5 parts, though the difference of loss will be very sensible.

Assay Weights. The weights used in assaying gold and silver are peculiar to the profession. In the assaying of silver, a given number of grains are taken, which is called the assay pound. This imaginary pound is subdivided into ounces and pennyweights, and the latter into half-pennyweights, which is the lowest term used in reporting assays of silver; so that there are 480 different reports for silver (this being the number of half-pennyweights in the pound); and therefore each nominal half-pennyweight weighs $\backslash frac\{1\}\{480\}$ of a Troy grain, when the entire assay pound is 24 grains.

The report of an assay of silver is made according to the proportion of pure metal which it is found to contain. The legal standard of Sterling money of silver is 11 oz. and 2 dwts. fine, and 18 dwts. alloy. If an assay of silver was found to contain 11 oz. only of pure silver, it would be reported worse 2 dwts. meaning worse than standard silver by 2 dwts. or 48 grains in the pound Troy. If an assay, on the contrary, contained 11 oz. and 6 dwts. pure silver, it would be reported better than standard by $\backslash frac\{1\}\{4\}$ dwts. in the pound Troy; because 18 dwts. being the standard proportion of alloy, it was found that it only contained 14 dwts. alloy. When bullion thus assayed and reported is for sale, its value is calculated by reducing the bar or ingot of silver into standard metal. In the first example which we have given, lbs. Troy. oz. if the ingot of silver assayed weighed 50 0 there would be deducted from the weight 2 dwts. per lb. or 5 which 5 oz. is the excess of alloy above the proportion of 18 dwts. to the 11 oz. 2 dwts. of fine metal, and the bar of silver would be, in standard weight,

lbs. oz.
50 0

On the contrary, if the ingot weighed and the alloy were deficient, which is the case when the metal is reported better than standard by 4 dwts. in the pound

Troy, there would be added to the 50 lbs. 4 dwts. per pound, which would be equal to

lbs. oz.
0 10

Making the standard weight of the ingot 50 10

The gold assay pound, which is from 10 to 20 grains Troy, is subdivided into 24 carats, and each carat into 4 assay grains, and each grain into quarters; so that there are 384 separate reports for gold, each equal to 15 Troy grains, or what is termed a quarter carat grain. An accurate assayer, however, can ascertain, in an assay of gold, to 3 grains Troy; but it is the custom of the trade not to report less than a quarter carat, or 15 grains Troy. A substantial reason is given for this rule, to justify the practice of it. An ingot of gold generally weighs a journey weight, which is 15 lbs. Troy; from a sample cut from the two opposite ends, weighing from 10 to 20 grains, the value of the mass of 15 lbs. is to be determined; if this ingot had been imperfectly melted, the mass would not be homogeneous, and a difference might exist in it of several Troy grains; and the allowance between the quarters given in the assay report is an indemnity to the purchaser. Indeed, so particular are many in the bullion trade, that they will not purchase any foreign gold bullion until it has been remelted by refiners or melters in whose integrity they repose confidence. This, we believe, is generally the case with the Bank of England, in all her purchases of foreign gold bullion.

The assay report of gold is made according as it is better or worse than standard. The standard of our gold coin is 22 carats fine, and 2 carats alloy. If, by assay, an ingot of gold was found to contain 21 carats fine gold, it would be reported worse 1 carat, the mass containing a carat of alloy more than the proportion of 2 carats to 22 carats fine. If the ingot weighed 15 lbs. Troy, there would be deducted from the gross weight 1 carat, or 240 grains Troy, reducing the standard of the mass to 14 lbs. 11 oz. 10 dwts. If, on the contrary, the mass was found to contain 23 carats fine gold, it would be reported 1 carat better than standard; and this carat would be added to the gross weight of the ingot, which we have supposed to weigh 15 lbs. Troy, and would be called 15 lbs. 0. oz. 10 dwts. of standard gold. When the gold assay pound or integer is only 12 grains, the quarter assay grain weighs only $\backslash frac\{1\}\{48\}$ part of a Troy grain. This will show how delicate the scales must be by which the assayer works in order to obtain accuracy. In the Royal Mint, the scales of the assayers will be sensibly affected even with the $\backslash frac\{1\}\{1000\}$th part of a Troy grain. When the Emperor of Russia lately visited the Mint, he was particularly struck with the extreme delicacy of the assay scales of Mr Bingley, the King's Assay-Master. That gentleman requested the favour of his Imperial Majesty to put one of the hairs of his head into the scale, which he did, and, to the great satisfaction of his Majesty, it very sensibly affected the equilibrium of the beam.

When the assay pound is subdivided, as for silver, in the same manner as the Troy pound, it is obvious

Assaying. that all the lower denominations bear the same relation to each other; which is some little advantage in transferring the assay reports to real mixtures for use. On the contrary, the carat subdivision for gold is confined to assaying, but its fractions being aliquot parts of the pound Troy, the calculation for real use is very easy. As the pound Troy contains 5760 grains, the carat corresponds with 240 grains or 10 dwts; the assay grain, or fourth part of a carat, with 60 Troy grains; and the assay quarter-grain with 15 Troy grains; to which report, when the assayer has separated the gold (4 oz. for example), he adds 4 oz. gold in a pound Troy. Whereas in gold parting he takes two equal pieces, treats one as a silver assay, and the other as a gold assay, to find the absolute quantity of each metal, after which the report is made on gold singly, to which is added the report of the silver separately. Thus, if he finds 4 oz. of gold, and 3 oz. of silver, he reports tworse 14 carats (2 carats being equivalent to an assay ounce, and consequently the 4 oz. of gold equal to 8 carats, which subtracted from 22 carats, the gold standard, leaves 14), to which report he adds, fine silver 3 oz. But when the mixed metal contains more than half alloy, it is called metal for gold and silver, and the absolute quantity of each reported separately. (Aikin's Dictionary.)

Having made the reader acquainted with these details, we shall now proceed to explain the process of assaying silver, commonly known by the name of cupellation.

Assay of Silver.

When we have an assay of silver to make, we flat the portion of metal upon a polished anvil; the face of the flattening hammer is also highly polished; that the metal may receive no extraneous matter whatever. The piece of metal is flattened to about the thinness of a sixpence, and an assay pound is cut from it, and most accurately weighed in such scales as we have already noticed. This assay pound is then enveloped in a sheet of lead, which is flattened from a lead bullet, and circular, but made into a funnel shape, in order to contain the silver; and the more nicely to prevent any portion of the silver from being lost, the corners of this leaden funnel are closely and firmly folded down. If the Assay-Master has 45, or, indeed, any number short of 45, they are ranged according to their number upon the table, fig. 13. When the furnace and cupels have been prepared according to the number of assays to be made, and when the proper degree of heat has been attained, the assays are charged into the cupels; and the following method is followed in this part of the process: In the first instance, a ball of lead is charged into each cupel, with the charging tongs (fig. 12.), and its weight is according to the quality of the silver to be assayed; the assayer keeping a stock of leaden bullets of different weights for the purpose. As soon as this lead is melted, which is instantaneous, the assays of silver enveloped in lead are also charged into the cupels. The mass is very soon in complete fusion. The mouth of the muffle, which had before been partially closed with cylinders of charcoal about 6 or 7 inches long, and of different diameters suitable to the convenience

of the assayers, as represented in fig. 8., is now nearly closed by smaller cylinders of charcoal. The Assaying. object of this precaution is, that the stream of air admitted to pass over the surface of the cupels, and which is indispensably necessary for the oxidation of the lead in the process of cupellation, may not chill the muffle, and retard the progress of the assay. The oxidation of the metal will proceed with more or less rapidity, according as the stream of air admitted is great or small, and which the assayer has it always in his power to regulate at pleasure. The work already referred to in this article, has so beautifully and accurately described the progressive appearance of the assay process of silver, that we cannot do better than quote the description: "The melted metal begins to send off dense fumes, and a minute stream of red fused matter is seen perpetually flowing from the top of the globule down its sides to the surface of the cupel, through which it sinks and is lost to view. This fume and the stream of melted matter consists of the lead oxidated by the heat and air, in one case volatilized, in the other vitrified; and in sinking through the cupel it carries down with it the copper or other alloy of the silver. In proportion to the violence of the heat, is the density of the fume, the violence with which it is given off, the convexity of the surface of the globule of melted matter, and the rapidity with which the vitrified oxide circulates (as it is termed), or falls down the sides of the metal. As the cupellation advances, the melted button becomes rounder, its surface becomes streaky with large bright points of the fused oxide, which moves with increased rapidity, till at last the globule, being now freed from all the lead and other alloy, suddenly lightens; the last portions of litharge on the surface disappear with great rapidity; showing the melted metal bright with iridescent colours, which directly after becomes opaque, and suddenly appears brilliant, clean, and white, as if a curtain had been withdrawn from it. The operation being now finished, and the silver left pure, the cupel is allowed to cool gradually, till the globule of silver is fixed, after which it is taken out of the cupel while still hot, and when cold weighed with as much accuracy as at first. The difference between the globule and the silver at first put in, shows the quantity of alloy, the globule being now perfectly pure silver, if the operation has been well performed. The reason of cooling the globule or button gradually is, that pure silver, when congealing, assumes a crystalline texture, and if the outer surface is too suddenly fixed, it forcibly contracts on the still fluid part in the centre, causing it to spurt out in arborescent shoots, by which some minute portions are often thrown out of the cupel, and the assay spoiled." (Aikin's Dictionary.)

The assaying of gold, preparatory to the parting process, which we are about to describe, is exactly the same as in the case of silver; the object in the process of cupellation being to destroy the base metal or alloy contained in the gold. If gold contained only copper as alloy, the assaying of gold would be as simple and expeditious as that of silver; but all gold contains a portion of silver, which, though reckoned as alloy, cannot, as we have already seen, be destroyed

Assay of Gold.

Assaying. by cupellation. Recourse is had to the process, commonly called the parting process, to get rid of the silver contained in the gold. This is done by means of nitric acid, which entirely dissolves the silver, and leaves the gold perfectly pure. The quantity of silver which gold generally contains, is too small to allow the nitric acid to act upon it without addition; and the general allowance by assayers is two or three parts of silver to one of gold. If the quantity of silver greatly exceeded these proportions, the operation would not succeed so well; the fine gold would be obtained in the state of brown powder, the particles having been too minutely divided by the excess of silver.

When assays of gold have passed the test, by which all the alloy, excepting silver, has been destroyed, it is in this process that the additional quantity of silver is added. Suppose, for example, that a gold assay is made from the integer, or pound, weighing 12 grains Troy, an addition of from 24 to 36 grains of pure silver is made in addition to the small portion already supposed to exist in the mass. This becomes thoroughly incorporated with the gold in the process of cupellation. The globule, or button, as soon as it is taken from the furnace, is passed between a pair of polished steel rollers, and drawn out into a thin lamina, or plate, of the thickness of a sixpence, and returned into the furnace to be annealed. After being kept in a red heat for some time, it is taken out, and suffered to cool. It is then wound up into a cornet. This is put into a glass matrass, of the shape of an inverted cone, and with about twice or thrice its weight of very pure nitric acid. M. Vauquelin recommends it to be 1.25 specific gravity. But the true test of its strength is in the working of the process. The assayer's attention being directed to the point of strength that will maintain the gold when the silver is extracted in the spiral form, if the acid were too strong, or the silver in too great excess, the gold, as we have already mentioned, would be reduced to powder, and considerable danger exists that it would not be accurately collected, by which an imperfect result would be obtained. The glass matrass is placed upon a sand-heat, or bath, which is generally a square or oblong pan of copper, with from one to two inches of dry sand in the bottom. The pan is placed over a small square furnace, in which is burning charcoal or coke. As soon as the acid is warm, it begins to act upon the silver, and a dense stream of nitrous gas is disengaged. As long as the acid continues to act, the metal appears everywhere to be studded with very minute bubbles, which issue in jets. The disappearance of these, or their uniting into a few large ones, is a sign or mark that the acid has ceased to act. The disappearance also of the nitrous fumes is an indication that the acid has no silver to act upon. In the course of fifteen or twenty minutes, the process is finished. But, in order to extend the last portions of silver which the mixture may contain, a small quantity of highly concentrated acid is poured upon the cornet, and boiled, by which the last portions of the silver are extracted. The cornets of gold are thoroughly corroded, but

retain the same form, having lost all the silver, to two thirds or three fourths of their weight; they are slender and brittle, as we observed before. It is an object of considerable importance to prevent the cornets from being broken, the result being more likely to be accurate than having the gold in fragments; and to prevent this, the quantity of silver used is no more than is absolutely necessary, it being obvious that the less the quantity of gold, compared to the silver, used in the assay, the more likely is the gold to be broken into pieces.

The hot acid is poured very carefully from the matrass, and warm water is added to wash any remain of silver from the gold, and the addition repeated until the water comes off perfectly clear. The cornets of gold, which are of a dull brown colour, and unmetallic appearance, are then put according to their numbers into small clay crucibles, into which they are allowed gently to fall by inverting the matrass, with a portion of water in it, which breaks their fall, and also collects any grains of gold that may be in the matrass. The water is then poured off, and they are put into the furnace, and annealed under a bright cherry heat. When cooled, the pieces of gold have regained their beautiful metallic lustre, and possess all the softness and flexibility of this truly noble metal.

The pieces of gold, thus thoroughly purified, are carefully and accurately weighed, the absolute loss in weight indicating the purity of the metal assayed.

It is a matter of the greatest importance, that the silver used in this process should contain no gold, otherwise a source of very material error would arise in the delicate operations of the assayer. Silver generally contains a small portion of gold. Spanish dollars, for example, are found to contain about 4 Troy grains in the pound, and are generally preferred in the parting process upon the large scale; but assayers in general use silver revived from a precipitation of the nitrate of silver, which they are sure contains no gold.

The nitrate of silver is precipitated by immersing in it plates of copper, which throw down the silver in the metallic state. It may also be recovered by a solution of common salt, which converts the silver into luna cornea, of which, when washed and well dried, 100 parts contain 75 silver. The accuracy of the assay may also be proved by this process. The luna cornea, however, is more difficult to reduce to the metallic state, and the mode of recovery by plates of copper is always preferred.

It remains for us now briefly to mention the process of assaying by the Petit Fourneau à Coupelle of Messrs Anfrye and D'Arcey. This process is the same in principle, in all respects, as that which we have been detailing. The only difference consists in the greater facility of the process, and the comparative diminution in the fuel used. The furnace first used by these gentlemen had a small pair of bellows attached to it; see fig. 16.; and after the furnace was brought to a proper degree of heat, which required two hours and a half, the following was the result of the experiments made: