Each furnace it closed at the top by a hopper with 2 doors, the upper of which has a water joint ; 12 charges are made a day, and care is taken to open the upper door as little as possible. At the bottom of the furnace is a grating with movable bars. The charge consists of 2 cwt. of ore, mixed with 1/4, the qnantity of charcoal. The charge is only lowered into the furnace by opening the lower door, while the upper one is closed; and a fire is kept burning at the upper part of the furnace for preventing the escape of the gases. The mercurial vapours and gases of combustion are led by a chimney into condensing chambers. From 150 to 160 tons of ore are extracted from the Siele mines every month, yielding about 22 tons of metallic mercury, which is packed in about 300 cans with screw stoppers.
This distillation with lime is a generally adopted method for poor ores, as it is effective, though troublesome. Fig. 153 shows the arrangement of apparatus a lopted on the Rhine. About .'10 pear-shaped cast-iron retorts a are placed in a gallery-furnace fr, treated by the vapours from a single coal fire running throughout the whole length, but not in contact with them. Each retort is 2/3 filled with about 70 lb. of a mixture of 4/5 ore and 1/5 quicklime, and its mouth opens into a stoneware vessel c containing water. The " black mercury," consisting mainly of finely divided metal and undecomposed sulphide, collected here, is dried and re-distilled, The action of the lime (calcium oxide) is mutual decomposition with the cinnabar (mercury sulphide), yielding metallic mercury in vapour, and leaving behind a deposit of calcium sulphide, partially converted into sulphate by excess of oxygen. The process is wasteful and inconvenient, from the numerous joints and small working capacity. The pear-shaped retorts may be advantageously replaced by gas retorts of 10 times the capacity.
A very crude and wasteful practice is current in Hungary for recovering the mercury from grey copper ore by preliminary roasting before smelting for copper. The ore is burned in mounds 40 ft. long, 20 ft. wide, and 4 ft. 6 in. high, built as follows: a patch of ground is covered with a bed of small ore, followed by a layer of larger lumps whieh have already been once roasted; next comes a spreading of fuel (con) and wood), and then more ore, finishing at top with a covering of finely powdered ore to check the draught and condense the mercurial vapour- Combustion is supported by channel. left in building the mounds; and the heat generated by the fuel ignites the sulphur in the ore, and thus hastens the decomposition. When a mound has burned in this way for about 3 weeks, the upper layer,containing the condensed mercurial vapours, is washed to re-cover the metal.
The condenser of a mercury - roasting furnace forms of course an exceedingly important portion of the structure, and no absolutely satisfactory construction has thus far been hit upon. Brick is a porous material, and mercury both filters through it and is absorbed into it by the almost resistless force of capillary attraction. Some years ago, in pulling down an old furnace at New Almaden, it was found that the metal had penetrated the foundation and the earth for over 30 ft., and down to the bed-rock. Since then the furnaces have been built on arches, and in the 2 C 2 pillars on which the arches rest sheet-iron plates have been placed to intercept the metal in its downward course; much metal, however, undoubtedly penetrates the outer walls of the condenser, and is evaporated from the surface, for mercury evaporates, more or less, at all temperatures above the freezing-point of water, and the condensers, though built of a very considerable thickness, are of course somewhat warm on the outside. On the other hand, the thicker the condenser the less does the vapour inside cool; and as it must cool to condense, the necessity for thick walls renders a large number of condensing chambers indispensable.
It is natural, therefore, to seek some material better adapted to the work required than brick. Cast iron and wrought iron answer very well for condensers, so long as the gas is hot, but when the temperature falls below the boiling-point of water, a very disagreeable action sets in. Sulphurous acid, heated with air in contact with highly porous substances, is partially converted into sulphuric acid; this condenses with the water, and attacks and soon destroys the iron. Wood withstands this action tolerably well, and does not let mercury through like brick, and it has consequently been much used for the later condensers of the series; but wood is a non-conductor of heat, and hence, properly speaking, ill-adapted for cooling apparatus. At New Almaden glass is found to answer excellently, being a much better conductor of heat than wood, though not so good a one as iron, and being of course quite unaffected by the acid liquors. The condensers assume a great variety of forms, such as chambers of various shapes, and tubes. Columns of coke are also used with some success as a final conductor.
The draught is sometimes kept up by the aid of a chimney, with or without the help of a fire, to give the air in the stack additional buoyancy, and sometimes by a suction-fan, or a "water-drum," an apparatus which sucks in air by means of the friction of a jet of water. So far as the condensation is concerned, an artificial blast would be preferable, but the danger of blowing jets of the metallic vapour through fine cracks in the apparatus would be too great to admit of its employment.
The losses in mercury-smelting are various and very important, though they are not all well understood. The whole bulk of gas passing out of the last condenser is saturated with mercurial vapour. The amount of metal contained in this vapour depends on the temperature, and for low temperatures is very small per cub. ft., but its aggregate amount per day is by no means inconsiderable. Another source of loss is the fine dust of mercury suspended in the gas, minute globules so small that they fall very slowly through the air, and are therefore carried away in the current of gas. The rate at which a spherical body falls through the air depends upon its specific gravity and its size, and we have therefore only to make a globule small enough to give it a permanent velocity of 10 ft. or of 1 ft. per second. As the mercurial vapour is cooled in the condensers, the mercury is separated out like mist, and the individual globules, which are too small to fall rapidly, and which do not collide and aggregate with others, are, to a great extent, carried off as a dust. Nothing is more difficult than to filter very fine dust effectually on a large scale from a current of gas, as has been proved over and over again in the attempts made to catch the fume from silver-smelting furnaces.
How much such fume is formed in mercury-smelting must depend on various circumstances, but the formation of soot in the conductors is sure to be accompanied by a loss in this way, for the metal which is separated out at the same time as the soot, will to some extent be coated by non-metallic impurities, and rendered incapable of running together into larger drops.
As a general rule, mercury in its crude imported condition needs refining for use. This may be known by its "tail"-ing, or leaving streaks where the globules are rolled over a glass surface. The impurities (lead, bismuth, and zinc) are sometimes eliminated by re-distillation. A better and simpler plan for small quantities is to shake it up with powdered sugar in the presence of air, whereby the foreign metals are oxidized to a grey powder, and removed by filtration through perforated paper. The best refining process is washing the metal in a thin layer with dilute nitric acid (1'15 sp. gr.) for a couple of days, with frequent agitation. The acid dissolves out the undesirable metals as nitrates, and may be used several times before its energy is exhausted. The cleansed mercury is washed (with clean water) free from acid, and dried, first with bibulous paper and then by gentle heating. Mechanical impurities can be separated by filtration through perforated paper.