placed in the crucible in which the ore or regulus was roasted, and covered by a thin stratum of fused borax. Instead of the powdered charcoal above recommended, from 20 to 30 per cent of crude tartar may be employed.

The crucible is now introduced into the fire and exposed to a strong heat during a quarter of an hour, at the expiration of which time the bubbling of the assay will have entirely ceased, and it must now be covered and exposed for a short time to a strong red heat, bordering on whiteness. The button may be obtained either by pouring into a mould, or by cooling and breaking the pot. If required, the button may be refined according to the Cornish method.

ORES OF THE THIRD CLASS.-Minerals belonging to this class must be treated in a similar way to those of the second, except that the preliminary roasting will, from the great fusibility of these ores, require to be conducted with more caution, and the button obtained from the calcined ore or regulus, instead of being, as in the former case, nearly pure copper, will consist of an alloy of copper with other metals.

When the ore contains lead, the roasting must be conducted with the greatest care, as it then becomes extremely difficult so to moderate the heat as to expel the arsenic and sulphur, and at the same time avoid the agglomeration of the powder.

In all cases the assay of minerals of this class should be commenced by a fusion for regulus, as, by operating in this way, the sulphides are not only reduced to a lower state of sulphuration, but a considerable portion of the associated arsenic is eliminated. The purification of the button of alloy obtained is now to be effected, either by the common Cornish method, or by the humid process, which will be shortly described.

CORNISH METHOD OF ASSAYING COPPER ORES.

The ore to be treated is first pounded and sifted, and a portion of it subsequently washed on a shovel, and carefully examined as to its quality and the amount of sulphur, arsenic, and other substances with which it may be contaminated. By a little practice in this operation, it becomes easy to judge of the amount of nitre necessary so to attack the combined sulphur, &c., as to yield a regulus of the requisite degree of fineness.

Two hundred grains of the mineral to be operated on are now weighed out, and intimately blended with a flux consisting of a mixture of nitre, borax, lime, and fluor-spar, and the fusion for regulus is commenced. The amount of nitre employed will, of course, vary with the quantity of sulphur and arsenic present; but

the other fluxes can be used in the following proportions:-Borax, 5 dwts.; lime, 1 ladleful; fluor-spar, 1 ladleful. When placed in the pot, the whole may be covered by a thin stratum of salt. The crucible is now strongly heated in a wind furnace for about 15 minutes, when, its contents being in a state of complete fusion, it may be withdrawn, and the melted matter poured into a cast iron mould. The button of matt thus obtained is subsequently examined, in order to determine whether a proper proportion of nitre has been employed. When a proper amount has been used, the button obtained should, on being broken, present a somewhat steely fracture, and yield from 8 to 12 for 20 of copper. A sample, however rank, is never mixed with more than 9 or 9 dwts. of nitre, and if the amount of sulphur is small, 3 dwts. are often sufficient. Some of the grey sulphides, the black oxides and carbonates, have sulphur added to them, as not having enough of that substance in them for the purposes of assay.

Very highly sulphurised or rank samples, requiring more than 9 dwts. of nitre, are sometimes treated by a somewhat different method. In this case the ores are first carefully roasted, and subsequently fused for copper with about 5 dwts. nitre, 9 dwts. tartar, and 3 dwts. of borax.

The roasting of the regulus obtained by the first operation is performed in a smaller crucible than that used in fusion for matt. This button of regulus, after being carefully pounded in an iron mortar, is exposed to the action of a slow fire, for the purpose of eliminating the sulphur. During the first 15 minutes a very low temperature is sufficient; the heat is then gradually raised for about 20 minutes more: during the former period it should be constantly stirred with a steel rod. After the first period of 15 minutes, an occasional stirring will be sufficient. At the close of the operation the assay must be subjected to a high degree of heat for a few minutes, and then withdrawn and allowed to cool. The subsequent treatment for copper is made in the crucible in which the roasting has been effected.

The relative quantities of the fluxes to be employed for this operation are regulated by the weight of the button of regulus previously obtained. A flux consisting of 2 dwts. nitre, 71⁄2 dwts. tartar, and 1 dwt. borax, is sufficient for the reduction of a calcined regulus that weighed previous to roasting from 48 to 50 grains. For a button weighing from 90 to 100 grains, 3 dwts. of nitre, 9 dwts. of tartar, and 2 dwts. of borax should be employed: these quantities are, however, seldom weighed; since, with a little practice, it becomes easy to guess with sufficient accuracy the relative amounts necessary.

*The ladle employed for this purpose is three-quarters of an inch wide, and half-an-inch in depth.

The button of copper obtained by the process of reduction is seldom sufficiently fine, and is therefore subjected to an operation called refining.

pre

The same crucible being heated to a full red heat, the button is thrown in the bottom of it, and at the same time some refining flux and salt are placed in the scoop for immediate use.* In a few minutes after the introduction of the flux a bright surface is sented. The crucible is then lifted a little from the fire with a pair of tongs, and a gentle agitation given to it. An appearance now presents itself similar to that observed in silver previous to its separation from the last traces of lead which it contains. Before this ceases, the contents of the scoop are introduced and the crucible returned to the fire for about 4 minutes. The crucible is now withdrawn, and its contents rapidly poured into a mould; when the button will be found quite fine and presenting a slight depression on its upper surface.

The slags from the reducing and refining operations are afterwards treated by fluxing them with a couple of spoonfuls of tartar and a little powdered charcoal. On withdrawal from the fire, the copper retained by the slags will assume the form of a small button or prill. The weight of this is to be added to that of the principal button.

HUMID METHOD OF ASSAYING COPPER ORES.

This method of estimating the produce of copper ores consists in attacking the sample to be operated on by a mixture of nitric and hydrochloric acids, the subsequent elimination of the nitric acid, and the precipitation of the copper by metallic iron. The whole of the ores belonging to the three classes before described may be treated by this process.

The sample to be attacked must first be pulverized, and passed through a fine sieve. Of this powder weigh 100 grains and introduce it into a long narrow-necked flask, which, if obtainable, should be made of hard glass. Wash down with sufficient water to thoroughly moisten the whole mass, and to remove any portions of the substance that may have remained attached to the sides of the vessel. Nitric acid is now to be cautiously added, and the flask gently heated in a sand-bath; since, if too large a quantity of acid be added at a time, or too strong a heat applied, a violent ebullition might ensue, and a loss on the assay be the result.

When at this stage of the operation the evolution of nitrous vapours entirely ceases, or has become much reduced, add gradually

* The refining flux is composed of two parts of nitre and one of white tartar, used together and subsequently pounded.

hydrochloric acid, place the flask in an inclined position on the sand-bath, and gently boil the mixture. This is to be continued until the residue, if any remain, appears by its colour to be free from all metallic stains.

The contents of the flask are now to be transferred to a porcelain dish, and evaporated to dryness, taking care to stir continually towards the close of the operation, in order to prevent the mass from spirting. When sufficiently cool, moisten with a little hydrochloric acid, heat gently, and subsequently add water; boil, and filter into a beaker.

A piece of bright wrought iron, about 2 inches in length, threequarters of an inch in width, and a quarter of an inch in thickness, is now attached to a piece of string, in order that it may be readily removed from the solution and lowered to the bottom of the beaker. It is essential to the success of this operation, that the piece of iron employed should be entirely immersed in the liquid; since, if this be not attended to, the precipitate becomes rapidly oxidised and the results consequently rendered inaccurate. The whole must now be kept constantly hot until all the copper is thrown down, which may be known by the liquid becoming colourless. This may also be confirmed by trying a drop of the solution on a piece of clean zinc, or by the blue colour which is produced by solutions containing copper when added to an excess of ammonia.

Having ascertained that the whole of the copper has been precipitated, carefully decant off the supernatant liquor, a small glass syphon being employed for this purpose, and repeatedly wash with hot water, until the metallic precipitate is entirely freed from any traces of chloride of iron.

The washing water is finally decanted off, leaving the metallic copper in the bottom of the beaker, which is now placed in a water bath, or in a warm place near the furnace, until it has become completely dried. The heat employed for this purpose should be extremely gentle, in order to avoid oxidising the product, and thereby vitiating the result. The copper thus obtained is subsequently brushed out into a watch-glass and weighed; on deducting from this weight the tare of the glass the per-centage amount of copper will be obtained.

In case the mineral operated on belongs to the third class and contains either lead, tin, or antimony, minute traces only of those metals will be found associated with the precipitated copper. When lead is present in large quantities, it is, however, best to add a few drops of sulphuric acid during the progress of the attack. By this means any lead that may exist in the solution will be precipitated. When a button of alloy has been obtained from ores of this class, by the method already described, the per-centage of tin

and lead may be readily determined by the following process:Attack the button by strong nitric acid, evaporate to dryness, take up with hydrochloric acid, add water, and filter. Then wash, dry, and calcine the filter, and each 100 parts of stannic acid found represents 78.61 of metallic tin. To the filtrate from the stannic acid add a little dilute sulphuric acid, boil and filter. After calcination each 100 parts of sulphate of lead represents 68.28 of metallic lead.

From the filtrate obtained from sulphate of lead precipitate the copper by means of iron as before described.

When antimony and tin are both present, their separation is only to be effected in the hands of a skilful operator and by a complicated process of analysis. If, however, antimony be alone present, its separation may be accomplished, although less completely than in the case of alloys of tin and copper, by the process recommended for the separation of those metals. In this instance every 100 parts of residue obtained corresponds to about 76 of metallic antimony.

ASSAY OF TIN ORES.

Oxide of tin, although completely reduced to the metallic state by charcoal at a white heat, has such a strong affinity for silica, that the highest temperature of a wind furnace is often necessary for their separation. For this reason, when great accuracy is required, tin assays should be made in lined crucibles and at the temperature employed for those of iron ores.

The nature and proportions of the fluxes employed will necessarily depend on the amount and composition of the gangue or matrix, with which the ore is associated; but as this is in most instances highly siliceous, carbonate of soda and fused borax may always be employed with advantage.

For ores of moderate purity from 30 to 40 per cent. of a mixture of equal parts of fused borax and carbonate of soda will be found to answer extremely well; and equally good results are obtained by the addition of 40 per cent. of a flux composed of three parts of carbonate of soda and one part of finely-powdered lime.

In Cornwall, where dressed tin ores are alone operated on, the following method is employed for making an assay:

Two ounces of black tin are intimately mixed with about onethird of that weight of finely-powdered culm or anthracite coal, and a quarter of an ounce of fused borax. The charge thus prepared is now introduced into an ordinary black-lead pot and placed in a small wind furnace, in which a strong heat has been already got up. At the expiration of a quarter of an hour, the contents