This would cause a loss; for calculating 4 per cent. loss on gold and 10 per cent. on silver in the original ore, we have only $7.50 which can be extracted from the matte. The loss would therefore be $3.38, and unless the matte were worked for other products, as for copper, it probably could not be utilized at present, though in many cases it would be required as a basic flux in the first fusion. Considering the present state of the West and proportion of copper in the ore, the process would probably consist of three operations-first, roasting the ore; second, fusion; and third, cupellation, the copper matte being sold.

Accepting the Burroughs milling ore as an average of the secondclass ore, we have for this, one ounce gold, worth $20.67, and 4.5 ounces silver, worth $5.81; total, $25.48. The cost of treating it would be:

This would leave a matte containing $1.86 in gold and silver, and perhaps some copper.

The above calculations are of course theoretical, so far as they relate to works which have never yet been established in the Territory. There may be errors in the prices assumed for labor and materials; but there is no reason why the amount of labor and material expended per ton should be more than at Lend; that part of the calculations is not theoretical. Undoubtedly in establishing such works some difficulties would be experienced, but with a railroad to the foot of the mountains, and the improved facilities for communication, the difficulties in the way

* A certain correction has to be applied, because the amount of matte is taken as larger than at Lend. I have assumed it to be 50 per cent. more.

Fine ore requires a more expensive roasting than coarse, for which reason I have added 50 per cent. to the cost. Roasting in furnaces, Mr. Hague says, costs $5 in Colorado.

cannot compare with those which have been overcome in establishing the milling system.

A chief drawback to extracting the precious metals in the Territory, instead of concentrating them in a matte to be exported, is thought to be the lack of lead-ores, since the Georgetown mines have not fulfilled their promise as lead mines. Let us see how much is required for works. treating 25 tons of ore a day, a capacity which is considered to be quite respectable for a mill; and it is to be remembered that these 25 tons are concentrated ore representing several times that quantity of ore as it came from the mine. The loss amounts to about 17.7 pounds of lead per ton, or less than one per cent. Of galena-ore yielding, say, 70 per cent. lead, one ton daily suffices for, say, 70 tons pyrites, or 21,000 tons yearly; two tons daily suffice for, say, 140 tons pyrites, or 42,000 tons yearly; five tons daily suffice for, say, 350 tons pyrites, or 105,000 tons yearly. If each ton of smelting ore represents 6 tons of ore from the mine, we have more than 600,000 tons of ore treated with 1,500 tons of galena-ore. Even if the mines of Georgetown and Argenta are unable to supply this amount, it could easily be bought in, and brought from Utah, at rates which would at least pay its own cost. My object, however, is not to urge any process upon the attention of western miners, or prove by full figures its applicability. I offer the Lend process as one which deals with ores precisely similar to those of Colorado, and leave it to those who are interested in the mines of that territory to work out its adaptability.

CHAPTER XII.

THE AMALGAMATION OF GOLD-ORES.

This chapter constitutes a supplement to the preceding one on economical results in the treatment of gold and silver ores by fusion, and was likewise furnished to me by John A. Church, E. M., of New York. I give the chapter without change or comment; but I do not fully concur in the theory of amalgamation which it presents.

It is commonly supposed that mercury takes up gold by reason of an affinity which causes the union of the metals whenever they are brought in contact, and in the use of amalgamated copper plates for catching the gold, the Americans have trusted the success of their gold-mills entirely to this action. In Austria they proceed on a different basis. There they acknowledge the affinity of gold for mercury, but confine it within small limits. The gold which is dissolved by the mercury, and which passes with it through the filter, is that which has a chemical union with the mercury; while that which remains in the filter, and after distillation forms the "retort," is merely particles of gold which have mechanically sunk into the mercury by force of gravity. Their surfaces are attacked by the fluid metal, which acts as a cement to bind them together; but in no sense do they form a definite amalgam. I will not discuss this point thoroughly here, but merely point out some facts in relation to Colorado ores which, on this hypothesis, give a ready explanation for the poor yield of those ores in the mill.

*The principles on which the separation of gold from its ores is effected by mechanical means are easily explained. If we have a substance composed of two elements, one having a specific gravity of 10 and the other of 5, it is clear that if we can provide a liquid having_a density of, say, 7, the former can sink in it, and the latter cannot. To accomplish the separation of the two we have only to crush the substance to a certain fineness and place it in a bath of the liquid. As soon as each particle of gravity 10 comes in contact with the fluid, it sinks, and we have only to agitate the sand and bring every particle in contact with it to produce perfect separation. We have then the two elements, one at the bottom and the other on the top of the liquid.

This is precisely what takes place in the so-called amalgamation of gold-ores. Gold has a specific gravity of 19.33, and mercury of 13.60. The iron pyrites in which the gold of Colorado is found has a gravity of about 5, and quartz, another constituent of those ores, has a gravity of 2.6. It would appear, then, that in a mixture composed of gold, specific gravity 19.33, and pyrites, specific gravity 5, there should be no difficulty in effecting the separation when the gold in a finely divided state is passed over mercury in which the gold can and the pyrites cannot sink. The Austrian gold-mill was devised to satisfy these conditions, and it works perfectly. In it mechanical contact between the gold and mercury is effected in the most perfect way, and the mercury lying in a bath 1 to 13 inches deep, is in a condition to act either by affinity or merely as a fluid of medium density. And yet this apparatus fails to extract the gold from most of its ores, and the tailings are sent to the smelting

*What follows is partly taken from an article by me in the Scientific American of October 7, 1871. In that article an error was made in putting the "normal alloy" at 35 silver and 65 gold. It should have been the reverse, or 35 gold and 65 silver. The error, however, leaves the argument unaffected.

works, if they can be made to pay the cost of treatment.

In some cases,

as for instance at Zell, spoken of in the beginning of this paper, the ore, worth only $2 and less a ton, is unable to bear any expense but amalgamation, and it could not bear even this were it not for the fact that its gold is fine and contains little silver.

There is a difficulty in treating gold-ores with mercury, in the explanation of which we may perhaps account for the trouble experienced in Colorado. Native gold is rarely or never pure. It is alloyed with silver, which has a specific gravity of 10.56. An alloy of the two metals, therefore, has a specific gravity between 19.33 and 10.56, depending upon the proportion of the two metals. With gold 35 and silver 65 parts, the specific gravity of the alloy is about the same as that of mercury, and it cannot sink in that fluid; that is, it will not "amalgamate." The question is then, do the ores of Colorado contain more than 65 of silver to 35 of gold? Let us calculate the assays given above, and we have the following table, the 35 gold and 65 silver being taken as the normal alloy :

These are fair specimens of Colorado ores, and we see that the gold they yield will not sink in mercury. And yet those who adhere to the milling process say it does amalgamate. That is true to a certain extent. Part of it amalgamates, and in that respect it exactly resembles the Lend ores, in which part of the gold amalgamates and part will not. The explanation is that Colorado ore contains 1 free gold, 2 gold alloyed with silver, and perhaps 3 silver not alloyed with gold. Mr. Hague thinks that the mills extract about 55 per cent. of the gold in the first operation, and 15 per cent. more by a repetition. If we construct a table for Colorado ores such as I have given for the Lend ores we shall have something like this:

Thus we see from this table that of the above Colorado ores only 57.7 per cent. of the gold and 2.9 per cent of the silver is extracted by amalgamation. These proportions are, of course, hypothetical; but we may regard them as near the truth.

The Burroughs milling ore contains 1 ounce gold and 4.5 ounces silver. At the same rate of yield the proportions would be:

The value of the Burroughs milling ore is therefore $12.15 in gold that will amalgamate, and $14.42 alloy that will not amalgamate; or 45.5 of the former and 54.5 of the latter in 100 of value. Thus we see that to its great fault of not extracting more than 70 per cent. of the gold amalgamation adds the loss of nearly all the silver; so that the real saving, even by the best work, including a repetition of the milling, is under 60 per cent. of the value.

I judge that the Colorado ores contain silver not alloyed with gold, from the fact that, although a great deal of the gold has been removed by the mercury, which leaves nearly all the silver, the tailings show no proportionate increase of silver to gold. Silver has therefore been removed as well as gold, and in about the same proportion. In three tables, giving assays of tailings, which Mr. Hague publishes, we have the following proportions:

Compared with 23 gold and 77 silver,. which is the average of the ores, these figures show that both silver and gold have disappeared, and about equally, in the process of milling. Though hardly necessary, I will say that nothing in the bullion explains this fact, for that is composed of 845 gold to 155 silver, on an average. The cause of this loss is undoubtedly defective concentration. The ores probably contain proper silverminerals, which are very brittle, reduce to a fine powder in crushing, and are easily carried off on the stream. It may be, too, that the small proportion of galena found in the ore is highly argentiferous, but contains little or no gold. This would partly account for the loss of silver, for when galena is stamped through a mesh of 25 to the inch, and then concentrated in a buddle we may be sure that very much of it goes in the water.

It must not be supposed that, the above table, in which the gold of Colorado ores is divided into free gold and auriferous silver, is correct in its proportions. In the ore there are probably an unknown number of distinct alloys, and the gold we obtain comes (1) from fine gold, (2) from those alloys which contain more than 35 per cent. of gold. We know from the bullion that about one-sixth of the amounts which I have put down to fine gold is really silver. The great fact remains that, if