paste and turned round and round, the excess escaping over the edge of the crucible. A few light taps with a mallet help to make the lining more compact. The wooden plug has a mark, and when this reaches the upper rim of the crucible the plug is withdrawn with the same rotary motion with which it was forced in. It has the form of a long cone, the point being slightly flattened. Its size is such that when forced into the crucible down to the mark, the lining at the bottom will be inch thick. This thickness decreases until near the upper edge of the crucible, where it is only inch. It is then rounded off until even with the inner edge of the crucible, and the sides are polished with a strong test-tube to prevent any particles of tin from adhering, and to obstruct the filtering of fluxes through it. In drying, it is best to place the crucible upside-down. Otherwise the lining is liable to separate from the wall, especially if it dries too rapidly. When a little skill has been acquired, it takes four and a half hours, including the preparation of the mixture, to line fifty Battersea crucibles, half of size F and half of size C. It will be seen that this lining is quickly made; it dries quickly without cracking and is hard when dry. After being in the fire, it is readily removed from the crucible, which can be used again, and is so compact and hard that it is difficult to break the thin upper edge with the fingers, while the thick lower part has to be broken with the hammer. In summing up the subject of brasqued crucibles, it may be said that although charcoal linings are perfect from a chemical point of view, their use has practical objections for the separating of fine tin from fine slag in the pan. The results to be described will show that crucible-linings are in fact unnecessary. The plumbago crucible stands between the naked and the brasqued crucibles, for while the action of the fluxes on the clay is weakened by the plumbago, the neutral character of the latter is slightly modified by the clay that binds it together. Plumbago crucibles are frequently used in the Cornish assay, but we shall see that this method is unreliable, and as the plumbago crucible is expensive, more need not be said about it. Finally the porcelain crucible used by Levol has the same disadvantages as the naked clay crucible, although in a less marked degree. II. PREPARATION OF THE ORE FOR THE EXPERIMENTS. The more representative lode-tin not being available in sufficient quantities to be made the basis of the following experiments, stream tin from the Nigger Hill district had to be resorted to. The cassiterite of the Bismarck district would have been the most desirable of all, as it is the purest. It is found in quartz and has comparatively no deleterious minerals occurring with it. This makes the separation of black tin from the gangue simple and perfect. The other lode-tins, as well as all the stream-tins of the Black Hills, have a number of minerals occurring with them which are difficult of removal, for instance, garnets. The first pulverizing of the ore was performed with the distance of inch between the faces of the sample-grinder. The crushed ore was then screened through a 40-mesh sieve. At the second grinding the faces of the machine were brought somewhat nearer together and the sifting afterward repeated. The operation was continued until the ore had all become fine enough to pass through the sieve with the exception of some mica, which was thrown out. This gradual method was used in order to prevent the cassiterite from becoming too fine, as a large amount might thus have been lost in the washing of the ore. The pulp obtained was then roasted in a muffle to decompose any sulphides and arsenides that might be present. After roasting, and while still hot, it was thrown into cold water, to make silicates, not otherwise attacked by acids, decomposable. It was also hoped that the cassiterite might thus be made more readily reducible, an expectation which the event fulfilled. The ore was now boiled with nitro-hydrochloric acid to remove all soluble metallic compounds; this was continued until no more iron was perceived to go into solution. The ore was then washed with hot water and transferred to the gold pan, where the panning was carried on irrespective of the loss of some fine tin which floated off with the tailings, giving them a chocolate-brown color. As the main object was to obtain a pure black tin, the loss was unavoidable. The panning was continued until no more impurities were visible to the naked eye; under the microscope, however, particles of garnet with a little quartz were still visible. The black tin obtained was dried and intimately mixed by passing it repeatedly through a 20mesh screen and rolling it on glazed paper. The now uniform ore was sampled down, and the final average sample pulverized in an agate mortar and put aside for chemical analysis. The weight of the purified black tin obtained, which was to form the basis of the experiments, was 7968 grammes. To find the size of the different particles, 500 grammes were taken and screened through different sized sieves. The result of the mechanical analysis is given in the following table: The very brittle character of the cassiterite will be seen from this table, as 53 per cent. of the entire ore passed through a sieve of 100 meshes to the linear inch, although, as has been said, the greatest care was taken to grind in such a manner as to produce as small an amount of slimes as possible. III. ANALYSIS OF THE PURIFIED BLACK TIN. In making the analysis of the average sample of the purified black tin, the Rose method was followed, as modified by Chauvenet, who substitutes potassium carbonate for the sodium carbonate commonly used. Two separate determinations were made, taking 0.5 gramme in each case with 3 grammes of a mixture consisting of equal parts of potassium carbonate and sulphur, and fusing in a No. 1 Royal Berlin porcelain crucible. The first sample was heated three-quarters of an hour over a Berzelius alcohol lamp, and then for the same length of time over a gasolene lamp, after which the fused mass was treated with water in a beaker. Brown, gritty cassiterite now became visible at the bottom, showing that the decomposition was incomplete. The soluble potassium sulphostannate was then filtered off and the filtrate acidulated with sulphuric acid and put aside; the residue on the filter was treated with hot dilute nitric acid, and the iron, manganese and lime solution also kept. The residue was again fused with potassium carbonate and sulphur, but in a different way. The No 1 porcelain crucible was placed within a No. 3 porcelain crucible, and this within a No. 6 plumbago crucible, the bottom of which had been filled with refractory material, that the porcelain crucible might be nearly as high as the plumbago crucible. The three crucibles were then each covered with a lid, and the fusion performed in a pot-furnace, heated with anthracite coal at a moderate red heat. This lasted one hour. The fused mass was then treated as before, and the resulting decomposition found to be perfect. The fusion for the second analysis was made directly in the three crucibles, and was complete in one operation. To make sure of this, the fusion was repeated, and on acidulating the aqueous solution with sulphuric acid, the precipitate was found to have the milky-white color of sulphur. The reason that the first fusion over the gasolene lamp was not successful, although prolonged until the fused mass became dry, must be sought in the fact that the air was imperfectly excluded. By fusing in a crucible which is inclosed in others, this exclusion can be entirely accomplished; in fact, upon removing the outer lid, when the crucible had been taken from the fire, but had not yet cooled, a strong odor of sulphurous acid was perceived, showing that not only had the air been excluded, but that the fusion had taken place in an atmosphere of sulphur.* The analyses were both made by the usual methods for separating iron, manganese, and lime, and the results corresponded so closely that they can be expressed by a common average: If this be compared with the analysis of the purest crystal of Nigger Hill stream-tint that could be found: SnO2 93.06 3.08 3.90 100.04 it will be seen that the cleaning of the stream-tin, as described above, *This use of the three crucibles was suggested by Dr. W. P. Headden, Professor of Chemistry at the Dakota School of Mines. † Report of the Dakota School of Mines, 1888, p. 142. was very successful, the ore sample showing only 2.4 times as much insoluble residue as the crystal. IV. THE ASSAY IN DETAIL. The different methods of assaying black tin in the dry way may be grouped under two heads: A. Those which aim at finding the actual amount of tin contained in the ore. B. Those which determine approximately, how much metallic tin can be recovered by treatment on a large scale. The first class divides, according to the character of the resulting metal, into methods (a) yielding a button of metallic tin, and those (b) yielding an alloy of tin with copper or iron. Class A, a. 1. The German Method of Assay.—This method as given by Kerl* and by Ballingt is as follows: Five grammes of ore are intimately mixed with 0.75 to 1 gramme of charcoal dust and charged into a clay crucible; on top are placed 12.5 to 15 grammes black flux (or the substitute: 2 parts of potassiumsodium-carbonate, and 1 part of flour), with 1 to 1.25 grammes borax glass, then a salt cover, and finally a piece of charcoal. The crucible is covered, heated in a muffle or a pot-furnace at a moderate, gradually increasing, temperature, until the boiling has ceased, and then for from half to three-quarters of an hour at a white heat. The crucible is removed from the fire, broken, when cool, and the tin button weighed. The process that takes place is simple. With the gradually rising temperature the tin becomes reduced to the metallic state by the charcoal with which it has been intimately mixed, while any ferric oxide contained in the cassiterite will be reduced only to ferrous oxide and taken up by the slag. At a certain stage the black flux, or the flour of the substitute, becomes decomposed, the result being that finely divided carbon is uniformly distributed through the flux. This hinders any particles of stannic oxide from combining with the alkali when fusion begins, and assists the reduction of particles of ore that have not been been completely converted into metal by the charcoal. The active fluxes-potash, soda, borax,-combine with the * Metallurgische Probirkunst, Leipsic, 1882, p. 412. |