maintained about midway up the furnace. The cold charge above the roasting zone would catch most of the material that tends to volatilize. The fuel consumption would be low, not more than 2 or 3 per cent of coal dust, or 4 per cent of sulphur being needed.

This roaster was thoroughly tested at the Ontario mill. The results showed that from the oxidized, siliceous material tested an extraction of 90 per cent of the silver, gold, and copper content was possible when a leach containing 20 per cent sodium chloride was used. Such a brine is not a good solvent for lead, although considerable lead was recovered along with the cement copper from the scrap-iron precipitation launders of the mill. The method proposed by the writers involves the use of brine containing 26.8 per cent sodium chloride, which would be a better solvent for the lead.

TESTS AT KNIGHT-CHRISTENSEN MILL.

The other mill in which brine-leaching tests were made was known as the Knight-Christensen mill, in Silver City, Utah, in the Tintic district. The procedure was similar to that at the Ontario mill, except that a down-draft roaster with a shallow bed of ore was used for the chloridizing roast. This roaster, patented by Christensen," resembles a circular Dwight-Lloyd sintering machine, with a suction box beneath a porous annular grate. The experimental mill was practically completed when it was burned down. Various parts of the mill had been installed and thoroughly tested and the problem under way at the time it burned was precipitation of the lead, copper, gold, and silver in metallic form, ready for marketing, from the solutions obtained by leaching. The feasibility of extracting all four metals from the ores seemed assured, as the mill leaching solutions had been tested a number of times with favorable results.

BUREAU OF MINES TESTS.

With this much data in hand, tests were made at the Salt Lake City station to determine whether the saturated brines used for the leaching of lead alone could extract silver in the same proportion from an ore that had been properly prepared by roasting in a chloridizing atmosphere.

Three types of material were tested-material from the Colorado mine in the Tintic district, material from the American Flag mine in the Park City district, and tailing from the Wilbert dump in Arco, Idaho.

PROCEDURE IN TESTS.

A Holt-Dern roaster of laboratory size was used in these tests. This roaster (fig. 10) was constructed of reinforced concrete in two sections-a top or shaft and a base. The shaft was filled with a

a Christensen, N. C., jr., Method of treating ores and the like: U. S. patent 1075011, Mar. 26, 1913.

charge of ore and fuel which rested on a piece of screen over a wind box in the base. The charge was ignited at the bottom by means of a gas burner, placed in the wind box, that was lighted before the charge was dumped in. As a rule it was also necessary to form a thin bed. of glowing coals above the screen before dumping in the charge. After the charge had ignited the gas was turned off, as there was usually no difficulty in keeping up combustion if sufficient air was supplied.

EN

VERTICAL SECTION

VERTICAL SECTION

O

GAS

The method of mixing the charge was important. The material had to be mixed dry with powdered coal or coke and then enough water was added to the mixture so that some of it would just hold together when pressed in the hand. When tumbled in a box or passed through a coarse screen, the moistened mass tends to form into little balls about one-fourth inch in diameter. In blast roasting such a mixture will become light and fluffy as it dries out, and most materials will not roast well unless previously prepared in this way. Even a wellrecognized "slime" can be balled in this way to permit successful roasting without too great a loss of metal in the dust. Oil may be used instead of solid fuel but

PLAN FURNACE
SHAFT

PLAN FURNACE BASE

THERMOCOUPLE

TU

AIR

must be added after the FIGURE 10.-Details of laboratory roaster of Holt-Dern type. moisture has been stirred in, as the dry material would absorb the oil so quickly that a few oily lumps would form and the rest of the material would contain no oil.

In the large roasters at Park City material crushed to pass a onehalf inch ring was used, but ores that will give good extraction on such coarse crushing are rare. A laboratory roaster will usually require material fine enough to pass through 2-mm. or 1-mm. mesh sieve. The salt (sodium chloride) must be finely granulated in order that good chloridizing roasting can take place. Also, the presence of sulphides of some kind, or of sulphur, is usually required for good chloridizing, as has been mentioned in nearly every textbook on the subject. Pyrite is added to the carbonate ores whenever they

contain no sulphides. The pressure of air applied will determine the temperature of the roast for any given amount of fuel. Temperatures below 600° C. are difficult to maintain. In a laboratory roaster of the size used the roasting lasts one to two hours before fire appears at the top and dies out.

The laboratory roaster had been checked against the large-scale practice in the Ontario mill with somewhat unfavorable results. However, several tests were made with this roaster.

TESTS OF MATERIAL FROM MINE IN TINTIC DISTRICT.

Ore from the Colorado mine in the Tintic district was tested in three different ways. The data of these tests, made by C. L. Larson, are shown in Table 23. Only enough ore for three 20-pound roasts was available.

TABLE 23.—Results of shaft roasting of argentiferous lead carbonate ore from the Colorado mine.

[Tests by C. L. Larson.]

Assay of original material: Lead, 5.5 per cent; silver, 7.9 ounces per ton; quantity used per charge, 11

In the first test only pyrite, and no sodium chloride, was added to the ore to determine the effect of the presence of sulphides on the solubility of the lead in the calcines. Supposedly the sulphur in the pyrite would form sulphur dioxide and trioxide and these would react with the lead to form lead sulphite and lead sulphate, both of which are soluble in neutral brine. In order to differentiate between lead oxide or lead carbonate, which is soluble in acidified brine, and lead chloride, which is soluble in hot water, each of the calcines was given three leaches-one with hot water, one with neutral brine, and one with acidified brine. The proportion of sulphate or sulphite lead formed was 42 per cent of the total lead content, and all of the lead was soluble in acidified brine.

In the second test only sodium chloride was added to the ore to determine how efficient it would be in chloridizing the lead without the addition of sulphide to the ore. Seemingly roasting with salt leaves some of the lead in a form insoluble in acidified brine.

The mixture used in the third test, containing both salt and pyrite, fulfills the conditions for good chloridizing of silver as stated by Holt and by Christensen, and mentioned in textbooks on the hydrometal

lurgy of silver. A fairly good extraction of silver was obtained under these conditions and a 93 per cent extraction of the lead. The results of these three tests hence lead to the conclusion that, under proper conditions for chloridizing the silver in this ore, the lead can be extracted by leaching with acidified brine. Holt found that when he was able to get as high as 70 per cent extraction of the silver in the testing roaster used by him he was usually able to get a somewhat higher extraction in the large roaster, the difference being about 10 per cent. This difference is probably due to the longer length of time the ore was in the large roaster, this time often being 8 to 24 hours.

TESTS OF MATERIAL FROM MINE IN PARK CITY DISTRICT.

Material from the American Flag mine, in the Park City district, Utah, was next tested by the same methods. This material was not much different from that treated by the Ontario mill, except that it was higher in gold. A mill known as the Park City mill was built to treat this material with the intent of using the same kind of machinery and the process used at the Ontario mill. Trouble from the lime content, about 10 per cent CaO in the material finally sent to the mill, and difficult extraction of the gold caused this plant to close down after running one month. To date it has not resumed operations.

Results of the laboratory tests on this material are shown in Table 24. The analysis of the iron sulphide used for aiding chloridization shows 4.4 per cent zinc, 3 per cent lead, and 0.78 ounce silver per ton. The leaches of the unroasted ore and the unroasted sulphide show that practically all of the silver and all of the zinc in both ore and sulphide, and 74 per cent of the lead in the ore and 95 per cent of the lead in the sulphide, were soluble in acidified brine without roasting.

TABLE 24.-Results of shaft roasting of argentiferous lead carbonate ore from the American Flag mine.

[Tests by C. L. Larson.]

Assay of original material: Lead, 4.16 per cent; zinc, 6.93 per cent; silver, 15.94 ounces per ton. Assay of sulphide used: Lead, 3.0 per cent; zinc, 4.4 per cent; silver, 0.78 ounce per ton.

The extractions of every metal except lead unexpectedly decreased with the addition of sulphide to the charge before roasting. Adding a small proportion of sulphide seemed to make the lead much less soluble, but increasing proportions up to 40 per cent rendered the lead entirely soluble. With proportions of sulphide of more than 20 per cent the extraction of the silver fell. In every test the zinc in the roasted material was less soluble than in the unroasted material. As a whole, chloridizing roasting failed to improve the extractions of metals from this material and was generally a hindrance to good extraction. It was thought that the lead was converted to basic sulphate to a greater extent in the presence of sulphides during the roasting, and results of later tests confirmed that view. This accounts for the higher extractions of lead, but the causes of the lower extractions of the silver are not known. The fact that the silver in the raw material was soluble would seem to indicate that it was present as chloride of silver, and prolonged heating of silver chloride at the roasting temperatures could cause thermal decomposition of silver chloride into metallic silver and chlorine. Good chloridizing roasting, as mentioned by Hoffmann," has always been carried to the point where silver chloride has been formed and no further.

TESTS OF MATERIAL FROM TAILINGS DUMP.

The strange behavior of the lead in these two ores during roasting, led to further experiments to determine the effect of sodium chloride and of sulphide, individually and together, on the roasting of material containing lead carbonate in the shaft roaster. A large supply of the material from the Wilbert tailings dump was at hand and some of it was used for the purpose. The absence of sulphur and of other contaminants made this material desirable for this purpose. Two series of roasts were made, the results are shown in Table 25. In the first series of roasts, common salt was added and no sulphide; in the second series both sulphide and salt were added to the charge, except in two tests in which no salt was used.

Roasting this material at temperatures between 500° and 800° C., with salt added and no sulphide present, did not convert a high percentage of the lead content into chloride of lead. Chloride of lead is soluble in hot water, hence the results of the hot-water leaches indicate the efficiency of the roasting. Unexpectedly, more lead was soluble in hot brine than in hot water after such a roasting, as is shown by tests 1, 2, and 3. However, the efficiency of the process is very low. By using a considerable amount of salt and roasting the mixture twice, in order to give a longer time of treatment, the content of water-soluble lead chloride was increased to 45.9 per cent. As can be seen from the later tests, it was found that the extraction could

• Hofmann, Ottokar, Hydrometallurgy of silver, 1907, p. 20.