Assay of Ilercury.—The only ore of mercury likely to come under the assayer's notice is cinnabar, which is always assayed by distillation. The ore is finely powdered and mixed with half its weight of iron filings and some slaked lime or soda in an iron retort. The retort is then raised slowly to a strong red heat ; in its neck is placed an iron pipe about 2 ft. long and I of an in. wide ; this pipe is surrounded by another pipe, made of tin, so as to leave a space between the two. In this space flows a current of cold water, entering at the lower end of the pipe and passing out at the other. As the retort is heated, the mercury in the ore is volatilized, and condenses as it passes through the cooled tube ; it is then collected under water contained in an iron and porcelain basin. The retort should never be more than one-third full of ore, and great care must be taken to keep it entirely covered with fire, so as to prevent the fumes of mercury ham condensing at the top. The iron pipe should be quite smooth on the inside in order that the metal may not adhere to it. Failing an iron retort, a stoneware bottle, or an earthenware jug may be employed. The mercury collected in the basin is put into a tared bottle and weighed ; it is not always perfectly pure, but sufficiently so to be depended upon, the amount of impurity never being very large.
Assay of Silver.—Silver ores are assayed in very much the same way as gold ores, differing, however, in the degree of heat used, silver requiring a higher temperature to smelt than gold does. If the ore be rich in silver, a quantity of litharge should be added to it ; the quantity may be optional, but should always be inversely proportional to tLe amount of lead already contained in the ore. One ounce naay be considered enough in all cases for a good assay. Half an ounce of black flux, or, better still, of powdered hard-wood charcoal, is added to this, together with oz. to 1 oz. of dried bollix. The mixture is thrown loosely into a dry crucible, which should not be more than one-third full. A layer of common salt is spread on the top and the crucible is covered and exposed to a strong and rapid beat.
If the ore contain nothing but sulphides, the smelting is comparatively easy. It is finely pul verized, mixed with an equal weight of litharge and nearly half its weight of saltpetre. If it contain much iron or copper pyrites, the amount of litharge and saltpetre must be proportionately increased. Iron pyrites needs four times, and copper pyrites three times, the above weight of saltpetre. The result of the smelting is an alloy of lead and silver which is afterwards submitted to the process of cupellati" in the manner described under the assay of gold.
Assay of Tin.—The assay of tin ores is a difficult and rather unreliable process. An absolutely correct assay can be made only by the wet methods, or, no some prefer, by a partially wet and dry operation. The safest method of aacertaining the exact amount of tin in any ore is to submit it to a chemical analysis, made by an experienced chemist. We shall describe here a rough and practical method employed at the mines in Swansea. Take 100 grains of the ore, which must be very care fully aamplod, and mix it with twice its weight of powdered anthracite in a large blue pot, or black lead crucible, covering the mixture with a this layer of the powdered coal. Place the crucible in a
furnace, in the moat intense white heat that can poaaibly be produced, for a period of fifteeu or twenty minutes, or until the masa has ceased to boil, and has settled qnietly down in the crucible. Then remove ; stir the crucible well with an iron rod, and heat again for the apace of five or aix minutes, after which the crucible may be removed from the furnace, and the contents poured into a metal mould. The slag should bb scraped out into a basin as direfully as possible, pounded, and washed ; by this operation much tin may be saved and added to the ingot. In this way the ore will yield nearly all the tin that it contains, but not quito. The metallic tiu made by this assay is never pure; it often contains 20 per cent. or more of other metals, which mud be subtracted from the actual yield of the ore. In pure orea, this loss is found to he only 8 or 10 per cent. If powdered anthracite cannot be obtained, dry sawdust may be used inatead.
In employing the above method, great care must be taken to have the ore very finely powdered, the furnace at a strong white heat, and to pour the molten metal very carefully. The beat way to perform this ia to push back the slag from the surface with a piece of wood, to place the mould acrosa the mouth of a large iron mortar, and then to pour out as much tie possible. All that does not go into the mould falls into the mortar.
The following method ia proposed by Domcyko :—Treat 100 grains of the ore with aqua regia. Wash and calcine tho residue, and mix with 20 grains of charcoal of white augar. Put the mixture in a email porcelain crucible, and cover over with I00 grains of the name charcoal. Heat in a good muffin furnace for a quarter of an hour. Remove the crucible, and treat the reaiduo when cool with aqua regia and precipitate the tin from the solution by mane of pure zinc.
Assay of Zinc.—This metal ia far too volatile to admit of ita being aaaayed iu the usual way. Tho beat plan, if the oro bo blend°, is to roaat it thoroughly with powdered charcoal. The oxides do not require ronating, but muat be heated in order to drive off the water and carbonic acid contained in them. Thia done; the ore must be powdered finely, mixed intimately with 15 or 20 per cent. of charcoal powder, and exposed to a atrong heat in a black-lead crucible. It should be perfectly dry, and packed tightly into the crucible, which must be covered over with another, inverted. The zinc is entirely volatilized if the heat has been sufficient, and the remaining matter to scraped out of the crucible, and calcined in an open dish to remove the carbon. The weight of the residue from this operation, deducted from the weight of ore originally taken, gives the amount of oxide of zinc contained in the roasted ore, of which every 100 parta correspond to 81 parts of metallic zinc. Small quantities of other metala are always volatilized along with the zinc in this operation, ao that it can only serve as a very rough guide.