bottom, communicating with the space under the false floor, is opened, and the solution is drawn off and conveyed to the precipitation vat. This is a wooden tub or cistern, but without a false bottom. Küstel recommends a rectangular form, with a half-round, somewhat inclined bottom, and a lining of sheet lead. The precipitant employed is a solution of sulphate of the protoxide of iron, (copperas, or green vitriol,) which is usually prepared fresh at the works from wrought iron and sulphuric acid. This part of the process is so simple as not to require detailed description here. The mixture generally remains undisturbed over night, giving the gold precipitate an opportunity to settle. The supernatant liquid is then carefully removed by tapping the tank at successively lower levels, until little is left with the precipitate. The latter is dipped out with care, by means of a dipper or scoop, into a clean porcelain or enameled dish, the final residuum being washed out through the lowest stop-cock of the vat, and the vat is cleansed from adhering particles with a stream of water, in the same manner as precipitates are washed upon filters, or breakers are cleaned of adhering portions of precipitate in the chemist's laboratory.

The gold obtained is in the form of a brown powder or "cement." This is filtered upon paper, and subsequently dried in an iron or porcelain vessel. Finally, it is smelted to a metallic regulus in clay crucibles, a little salt, borax, and nitrate of potash (saltpeter) being used as fluxes. Küstel gives the following expense of the cost of treatment, for a capacity of three tons daily from a double furnace. His figures refer to Grass Valley in 1867, since which time some items of expense have been reduced in that locality. The results obtained in this table are, however, in my opinion, not far from the present cost, since Mr. Küstel has made no allowance for incidental outlays, which are inevitable. I have added a column, giving estimates (of less authority) for the same items in Arizona, where expenses are much higher:

Thirty pounds manganese, at 61 cents..

Forty pounds salt, at a cent...

Seventy-five pounds sulphuric acid, at 23 cents.

One man at the vats two days, at $3 50..

Sulphate of iron....

Total for three tons....

Or, in Grass Valley, $14 55 as the average cost per ton, and $20 33 in Arizona.

Much more could be written in detail concerning the manipulations of this process, and the combinations of other kindred processes, such as the plan of Calvert, who recommends the production of "nascent chlorine" in the chlorination tank, instead of the introduction of ready-made chlorine gas; of Roeszner, who employs a salt solution saturated with chlorine; and of Patera and others. Many of these processes are intended to save the silver as well as the gold. But I must refer the reader to the books on the subject for all these matters. None of the processes, save simple chlorination, is, so far as I know, employed in the United States. I have confined myself, therefore, to a brief and general description of what is universally known as the Plattner process.

There is no doubt of the metallurgical perfection of this process. The reasons why it is not more frequently employed in this country

are

1. The cost of treatment per ton, excluding all low-grade ores from profitable reduction by it.

2. The real scarcity, except in one or two mining districts, of ores suitable for chlorination. Even perfectly effective chlorinating works suffer almost everywhere from lack of ore, and scarcely any in the country are run continuously at full capacity. Yet this "full capacity" would require but from one to three tons of ore daily.

3. The lack of metallurgical skill in the construction and operation of furnaces for the preliminary roasting. On this everything depends, and it is perhaps in this department that several failures in Colorado have occurred. It certainly seems strange that the Territory offering apparently most suitable conditions for the process should witness so many failures in it. The works of Cash & Co. at Central City are not here referred to; they are reported to be successful; but the owners are reticent as to their methods and results, and there is consequently nothing certain to be said of them.

Chlorination, in its present form, cannot supersede amalgamation for ordinary milling ores. It can compete with smelting where nothing is to be extracted but gold, (or, by Roeszner's method, gold and silver,) and in any case it is quite likely that the process will be best administered by custom works, receiving the ore from different mines, and running continuously.

But, since the cost of roasting is more than half the total cost, it is possible that improvements in the mechanical means of roasting, such as the introduction of the O'Hara, the Gerstenhöfer, the Brückner, or the Stetefeldt patents, may considerably reduce the expense, and thus enlarge the field of the Plattner chlorination. The capacity of the Stetefeldt and the Gerstenhöfer furnaces is very great, and the evil of a short supply of ore might be aggravated by their employment; but, on the other hand, the reduction of the cost of treatment by a little would increase the amount of ore chlorinated by a great deal.

GOLD REFINING BY CHLORINE GAS.

The following interesting paper, read before the Royal Society of Victoria, by F. B. Miller, F. C. S., Assayer in the Sydney Branch of the Royal Mint, describes a new method of refining gold, which, it is reported, will be tried by the Mint of the United States, Mr. Miller having visited this country for that purpose:

There is no recorded instance of gold having been found in an absolutely pure state. Every natural alloy of gold (or native gold, as it is called by mineralogists) contains more or less silver; and in almost all bullion resulting from the melting of Australian alluvial gold, the portion that is not gold consists chiefly of silver, with only a very small proportion of foreign metals, usually copper and iron, with occasionally a little lead or antimony, and sometimes a trace of tin, iridium, etc. This, however, though true generally, is not always the case with gold obtained from quartz veins by amalgamation, as the mercury occasionally reduces aad takes up other metals as well as the gold, which appear in the bullion on melting. The accompanying table will give some idea of the proportion of the precious metals contained in the gold from the various districts of New South Wales, after melting. It will be seen that the most argentiferous is that from Boonoo Boonoo, in the north, containing as much as 34 per cent. of silver. This approaches in composition the gold from the productive Thames district of New Zealand; while the gold from Nerrigundah, in the south, only contains 1.5 per cent. of silver, the remaining 981 per cent. being gold with a trace of copper:

Table showing the proportion of gold and silver in characteristic samples of gold dust from various localities in New South Wales, (after melting.)

An interesting, and as yet unanswered question here arises: Is this argentiferous character in any way connected with the geological structure of the district?

It is a fact, and certainly a very curious one, whether it arises from accidental causes, or whether it may hereafter be traced to peculiarity in the rocks whence the gold of the different districts is derived, that its quality or fineness deteriorates the further north we go; in other words, it contains more silver and less gold.

Thus, the average fineness of Victorian gold is about 23 carats; that is to say, it contains about 96 per cent. of gold and 34 of silver, with a per cent. of base metals; while, on passing north, we find the average fineness of New South Wales gold to be only 22 carats 13 grain, or to contain 934 per cent. of gold and 6 per cent. of silver. On going still further north, to the colony of Queensland, the average fineness is little more than 21 carats, (considerably below standard,) or it contains 87 per cent. of gold and 12 per cent. of silver; that from Maryborough containing as much as 14 per cent. of silver and only 85 per cent. of gold.

These are averages only. It is not to be supposed that there is a regular and consecutive diminution in fineness with every degree of latitude we go north. There are exceptional localities in the north of this colony, where the gold found is of a high degree of purity, as at Rocky River, where it is over 23 carats fine, or 96 per cent. Possibly at a future time our geologists may be able to throw some light on these curious facts, and the exceptional cases may then even help in explaining the apparently general rule. The point, however, of principal interest, as far as regards the subject of this paper, consists in the fact that, as the alloy obtained by the gold miner is poorer in gold, it is proportionally richer in silver.

According to the published returns, 6,820,198 ounces of gold have been received for coinage in the Sydney Mint between its establishment, in May, 1855, and December 31, 1838. The average assay of this quantity would be about 943; in other words, it contained 94 per cent. of gold, 5 per cent. of silver, and per cent. of base metals. Allowing an average loss of 2 per cent. in melting the gold dust, there would remain, after smelting, 6,683,795 ounces of gold bullion; and as the silver it contained amounted to 5 per cent. of this quantity, the gross amount of silver in the gold received for coinage was 334,190 ounces; being at the rate of 24,720 ounces per annum.

The average proportional quantity of silver contained in the gold arriving in

Sydney is at present very much greater than that given above, owing to the large amount of silvery gold now being found, especially in the neighboring colony of Queensland, and for the year 1868 was not less than 36,000 ounces, (£9,150,) and was probably (including that in the gold shipped direct as bullion by the banks) nearer 42,000 ounces. Most of the silver thus naturally present in the gold has hitherto been lost to the colony, owing to the expense, in Sydney, of the acids, etc., necessary for its extraction by any of the usual methods of refining, which left little, if any, margin of profit on the operation. It therefore seemed desirable that some easy and economical process should be contrived for refining in Australia, without the aid of costly plant and chemicals.

Twelve months ago a paper of mine, describing a new process for refining and toughening gold by means of chlorine gas, was read before the Chemical Society, London. As, since the publication of that paper, the method of refining therein proposed has been successfully brought into practical operation on a large scale, both here and in New Zeland, and there is a probability that its adoption will, before long, become more general, I lay before the members of this society a somewhat detailed account of the process, and some of its more striking results. I shall, as far as possible, avoid giving the details of the preliminary experiments which lead to the practical application of the process, and which have already been published in the Journal of the Chemical Society; but, in order to render myself intelligible, some repetition of what is therein contained will be necessary.

Most people at all interested in the matter are aware that the ordinary method of separating silver from natural alloys of that metal and gold, is a complicated and expensive process, and that the end is attained by melting the gold with at least two and a half times its own weight of silver, and then again separating, by the action of acids, the silver thus added, and also, at the same time, the small quantity originally contained in the gold, thus leaving as a residue fine gold assaying from 990 to 993; the rationale of the operation being this: If the natural alloy were simply placed in the acid, the very large excess of gold in the alloy would completely protect the silver it contained from the action of the acid; but if the gold is melted with a large excess of silver, so that the silver greatly preponderates over the gold in the alloy treated, then the acid is able to exert its solvent action not only on the silver thus added, but also on that originally contained in the gold. To arrive at this end, a complicated and very costly plant is required, besides large quantities of expensive acids; and several days are occupied in the operation. It is evident, then, that if all this complicated process can be avoided, and the silver simply and completely separated in one operation at the time the gold is being melted, a very great saving of time, of material, of plant, and of the interest involved in all these will be effected.

Such an end is attained in the plan now being adopted for effecting this operation. It is well known that chlorine readily enters into combination with almost every known metal, the action in some cases being so violent as to be attended with vivid combustion. Many metals, such as lead, tin, zinc, and antimony, when introduced into this gas, even at ordinary temperatures, combine with it, forming highly volatile chlorides. The two latter, if in a state of fine division, burst into flame on being placed in an atmosphere of chlorine. Copper also exhibits spontaneous combustion under similar circumstances, but the resulting chloride formed is only slightly volatile. Silver immersed in chlorine gas at ordinary temperatures slowly unites with it, forming chloride of silver; but if the gas be passed over it while red-hot, the action is much more energetic, the compound formed being more volatile than the chloride of copper, but much less so than those of lead, tin, zinc, or antimony.

The method of refining now to be described is based upon these facts: It consists simply in passing a current of chlorine gas through the gold while in a melted state, which is easily done by thrusting into the molten metal a small clay tube connected with a stone-ware vessel in which chlorine is generated. The chlorine on coming in contact with the silver in the molten alloy at once combines with it, forming chloride of silver, which, being of less specific gravity, rises to the surface of the melted gold, while this latter remains in a purified condition beneath. Chloride of silver has always been considered a somewhat volatile substance, and under circumstances such as those here described, it was naturally supposed that it would either be sublimed in the flue or escape entirely up the chimney; but in practice it is found that the volatility of the chloride is not nearly so great as might have been anticipated, and that if its surface is coated with a layer of fused borax it may be kept melted at a high temperature without any very material loss. The furnace required for the operation is the ordinary 12-inch square gold-melting furnace, the principal points to attend to in its construction being: 1. That the flue should be as near the top as possible, so as to allow of the crucible standing high up in it without being cooled by the draught; and, 2. That the furnace itself should not be too deep, so that when the pot is placed in the fire the bottom of it may not be more than 3 inches above the bars. The covering of the furnace should consist of two fire-tiles, 7 inches wide and 15 inches long, one of which should have a long slot or hole in its center for the clay chlorine pipes (which

I shall describe presently) to pass through. An iron cover will not answer, as it soon becomes much too hot for convenient working.

The crucibles in which the refinage is performed should be French white fluxingpots, (creusets de Paris, made by De Ruelle, late Payen, Paris;) ordinary black-lead pots will not answer, owing to the reducing action they exert on the compounds formed. To prevent the infiltration of the very fluid chloride of silver into the pores of the clay pots, (which would otherwise occur, and necessarily entail loss,) they are prepared by filling them with a boiling saturated solution of borax in water, which is allowed to stand in them for ten minutes, and is then poured off, the crucibles being afterward set aside to dry; the borax forms glaze on the inner surface of the crucibles when they become hot in the furnace.

When used for refining these French clay crucibles are placed within black-lead pots, as a precaution against loss, should the former crack, which, however, seldom happens. The crucibles are covered with loosely-fitting lids with the requisite holes bored through them for the passage of the clay chlorine pipes, etc. Ordinary clay tobacco-pipe stems, from 17 to 22 inches long, have been found to answer well for the purpose of passing the chlorine gas through the melted gold. Of late, a pipe made in London to order, inch in diameter, 22 inches long, and inch bore, has been found to answer all requirements. The chlorine generators should consist of the best glazed stone-ware acid jars, capable of holding from ten to fifteen gallons, and furnished with two necks. One of these openings should be stopped with a sound cork (or vulcanized India-rubber plug, if obtainable,) through which should pass tightly two glass tubesthe eduction tube and the safety or pressure tube, the length of the latter being a few inches, and the former 8 or 10 feet, spliced, where necessary, by means of vulcanized India-rubber tubing. The other opening, intended for introducing the oxide of manganese, etc., should be closed with a leaden plug, covered with a short piece of India-rubber tube by way of a washer, and well secured.

Each generator should be charged with a draining layer of small quartz pebbles, down nearly to the bottom of which the pressure tube should extend. On this layer should be placed from 70 to 100 pounds weight of binoxide of manganese in grains about -inch cube, sifted from powder; this quantity will be sufficient to effect many refining operations, and will obviate the necessity of repeated dismantling of the apparatus. Each generator should be suspended to about half its height in a galvanized iron water-bath.

The chlorine gas is produced when required by pouring common hydrochloric acid (sp. gr. 1.15) down the safety-tube, the apparatus being warmed by means of gas-burners beneath the water-baths. The gas is conveyed from the generators by means of a leaden pipe fitted with branches to supply the several furnaces, all intermediate connections being formed by means of vulcanized India-rubber tubing which, if screened from the direct radiation from the fire, stands the heat well, even immediately over the furnaces. All joints between the various pipes and India-rubber tubes are easily secured, and rendered perfectly gas-tight with a cement consisting of a thin solution of India-rubber in chloroform.

Screw compression-clamps on the India-rubber tubes give the means of regulating the supply of gas as required, and enable the operator to shut it off entirely as soon as the refining is over. The chlorine then having no means of escape accumulates in the generator, and soon forces all the acid up the safety-tube into a vessel placed above to receive it, and the acid no longer acting on the oxide of manganese, the supply of gas of course ceases.

These generators are very convenient and manageable, and it is questionable whether a gas-holder for the chlorine (even if the practical difficulties in its use could be overcome) would be at all preferable. Two such generators as are here described, and three ordinary gold-melting furnaces, have been found capable of refining daily about 2,000 ounces of gold, containing about 10 per cent. of silver, between 9 a. m. and 2 p. m.

Very many thousand ounces (upwards of 200,000 ounces) have now been refined by this process; and the mode of operation which has in practice been found the most advantageous has been as follows:

The French crucibles, (say, size 17 or 18,) duly prepared with borax, having been placed in the cold furnace, and slowly and carefully heated to dull redness, the gold (from 600 to 700 ounces to each crucible) is introduced, and the fire urged until the metal is melted, the necessary generation of chlorine having meantime been commenced by the introduction of a little hydrochloric acid poured down the safety-tube. into the generators.

In order to fill the pots, and avoid the risk of splitting them by the wedging of the ingots at their contracted bottom, the gold for refining is cast in molds of a peculiar form. Two inches from the end, the sides and bottom of the iron ingot-molds converge so as to produce a slipper-shaped ingot, two of which, placed face to face, fit conveniently into the pot.

As soon as the gold is melted, from 2 to 3 ounces of borax in a state of fusion is