It will be seen from these figures that black tin is a very variable substance; and that the specific gravity is largely influenced by the impurities; hence, it is only an indication of the percentage of metal when the same kind of ore is dealt with.

As already pointed out, the object of vanning is to determine the proportion of black tin in the lode stuff. The relation between the actual content in oxide of tin and the produce got by vanning has been tested on several occasions with results which show a fair degree of approximation.

The following are some published results of assays of the same batch of ore. The vanning results were obtained by a Cornish vanner of recognised ability, and the wet assays by two London firms of the highest standing:—

Vanning results: (Average) 91 lbs. of "black tin." Wet Assay results: A 83.7 lbs. of stannic oxide. B 79.7 lbs. "

The vanner reported his black tin as containing 70 per cent. of tin. This will bring his result, if calculated as stannic oxide, to 80.9 lbs. to the ton; which agrees with the others.

According to our experience the "van" assay agrees fairly well with the "wet" one, if the black tin is assumed to contain 92.5 per cent. of stannic oxide (SnO_{2}).

Vanners are, as a rule, skilful men, and show remarkable dexterity in separating the black tin, with the help of their apparatus, which consists simply of a shovel and a kieve of water. An account of the process is given below. But different vanners, all good men, will get different results working on material new to them. The black tin weighed by the vanner is supposed to correspond in quality with the black tin returned from the floors of the mine for which he is assaying, but this differs materially in different mines with the nature of the gangue. The process leaves too much to the judgment of the vanner. It is more than probable that in practice the returns from the dressing-floors check the assayer, instead of, as should properly be the case, the assayer checking the returns. It is only when this last is done that any control is had over the system of dressing. A correct assay of this ore is a matter of some importance, because of the high price of the metal.

The method of assaying the black tin is a dry one, and consists of mixing it with "culm," and submitting it in a black-lead crucible to the highest temperature of a wind furnace. The sample is taken wet as it arrives at the smelting house, and is assayed direct. The product of the assay is examined, and a deduction of a considerable percentage is very properly made for impurities, since the assay really determines the percentage, not merely of tin, but of the bodies present which are reducible at a white heat. The judgment as to how much is to be deducted is assisted partly by an examination of the metal got from the assay, and partly by the experience acquired in smelting similar ores. The produce, which is that of the impure tin, is stated in parts in twenty; thus a produce of 14 is equivalent to 70 per cent., or to 14 cwt. per ton.



MECHANICAL SEPARATION.—VANNING.

This process, which has already been referred to, is carried out as follows:—After sampling the ore in the ordinary way, a quantity (varying with its richness) is weighed out. Special weights are generally used. The standard weight, marked 200, weighs about an ounce; with poor ores this quantity is taken for an assay, but with richer ores 100 or even 50 is sufficient. The unit of weight has no special name, but the parts in 200 are spoken of as the produce; thus, if 200 of ore were taken and 9.5 of black tin were separated, the produce would be 9-1/2: obviously half the "produce" will give the percentage. The weighed portion of the ore is placed on the vanning shovel. The vanner stands in front of a tub of water (kieve) and allows 30 or 40 c.c. of water to flow on to the ore. He then raises the shovel a little above the surface of the water, and, holding it nearly horizontal, briskly rotates the water by imparting to the shovel a slight circular motion, passing into an elliptical one (front to back). This causes the finer mud to be suspended in the liquid, which is then run off, leaving the body of the ore in the centre of the shovel. This is repeated until the water after standing a moment is fairly clear. About half as much water as before is brought on; then, with a motion which is similar to the previous one, but with a jerk added in one direction, the heavier minerals are thrown up, and the stony matter brought back. The jerk is produced just as the wave of water is returning. The descending wave of water draws with it the bulkier and lighter particles of the ore, whilst the heavier matter lying on the bottom is scarcely affected by it. The jerky motion, however, carries it to the front of the shovel. The lighter stuff is washed off, and the residue dried by holding the shovel over the furnace. It now corresponds, more or less, to the stuff which on the mine is sent to the calciner. It is swept from the shovel into a scoop, and transferred to a hot crucible; in which it is calcined until free from sulphur. Some vanners calcine their samples before commencing to van. The calcined ore is shaken out of the crucible on to the shovel; rubbed up with a hammer; and washed (as at first) to get rid of the finer and lighter "waste." The separating motions are again gone through; and the "head" of the best of the black tin is thrown well up on one side of the shovel in the form of a crescent, so as to leave room on the shovel to work with the "tailings." The quantity of water used is kept low, to prevent this "crop" tin from being washed back again. The tailings are then crushed to free the tin from adherent oxide of iron; and again washed to throw up the remaining tin ore. As this tin is finely divided, it is more difficult to bring it up, so that a vigorous and rapid motion is required. The tailings are now washed off, and the whole of the black tin is brought into the centre of the shovel. It requires two or three washings more to free it from the waste it contains. Very small quantities of water are used. The purity of the black tin can be seen by its appearance on the shovel. The cleaned ore is dried as before, freed from particles of iron with the aid of a magnet, and weighed. The weighings are carried to 1/8th of the unit used. The following example illustrates the method of calculation adopted on the mine. A parcel of 1 ton 2 cwt. 3 qrs. of tin ore with a produce of 45 (equal to 22-1/2 per cent.) contains 5 cwt. 0 qrs. 12 lbs. of black tin. This result is obtained as follows:—

ton cwt. qrs. 1 2 3 9 } ————————- } 10 4 3 } equivalent to multiplying by 45. 5 } ———————— } 5.1 3 3 strike off the first figure to the right. 4 multiply by 4 to reduce to quarters. ————- 4 12 3 ————- 4 15 28 multiply by 28 to reduce to pounds. ——- 112 15 ——- 12.7 strike off the first figure to the right.

Similarly, a parcel of 20 tons 10 cwt. with a produce of 9-1/2 contains 19 cwt. 1 qr. 25 lbs. of black tin. For the following information, as well as for much of that already given about vanning, we are indebted to Captain Reynolds, of Cook's Kitchen Mine. "To have a complete set of tools for all vanning purposes, it will be necessary to get the following:—A vanning shovel 14 inches long and 13 inches wide, weighing not over 2-3/4 pounds. It is made of hammered sheet iron of the shape shown in fig. 57. It must have a light wooden handle (preferably of deal) 3 feet long. A bruising hammer, weighing 2-1/2 pounds, with a handle 1 foot long. A pair of tongs (furnace) 2-1/2 feet long, made of 1/2-inch round iron. And a set of ordinary clay crucibles for calcining. There ought to be two sets of scales and weights: the first should be confined to weighing the powdered tin stuff, and the second ought to be a much higher class one, for weighing the black tin obtained. The furnace for roasting the sample should be 10 inches square and 12 inches deep, with the fire-bars at the bottom three-quarters of an inch apart. The water-box for vanning in should be at least 4 feet long, 2 feet 6 inches wide, and 8 inches deep."

DRY METHODS.

For the following description of the process adopted in Cornwall we are indebted to Mr. A.K. Barnett, F.G.S., of Chyandour.

Cornish Method.—Tin Ore Assay.—The ore to be smelted or assayed should be concentrated to say not less than 50 per cent. of metallic tin; though to obtain satisfactory results it should be brought nearer 70 per cent., as with ore containing less than 40 to 50 per cent. of metal there will be a considerable loss both in the assaying and in the smelting. If the ore to be operated on does not contain this quantity of metal, then the sample (if coarse) must be reduced to a fine state, the gangue being removed by vanning, and the ore saved for the fire assay.

The method adopted for the determination of tin in the ore is as follows:—About 2-1/2 ounces troy (1200 grains, or about 80 grams) of the ore to be assayed is weighed out and mixed on a flat copper pan (shaped with a long lip) with one-fifth of its weight (240 grains, or 15.5 grams) of powdered culm (anthracite). The mixture of ore and culm is either transferred to a black-lead crucible before the latter is put into the furnace, or, as some prefer, it is carefully swept into a crucible which has been imbedded in the fire. Some assayers cover their pots with a flat cover placed loosely on, while others leave the mixture in the open pot. The furnace, which has been previously fired to a strong heat, is then covered, and the sample is subjected to a sharp fire for a period of from twelve to twenty minutes. No definite time can be stated, as, besides the strength of the fire, the quality and condition of the ore, and the impurities associated with it, greatly affects the time required for the complete reduction of the ore. As soon as the mixture in the crucible has settled down to a uniform white heat, and any very slight ebullition which may have taken place has subsided, the crucible is gently shaken, removed from the fire (the culm-ash or slag which covers the metal being carefully drawn aside with an iron scraper), and the metal is poured quickly into an iron ingot-mould, which is usually placed on a copper pan to save the culm-slag and the adherent metal which comes out with it. The crucible is then carefully scraped, and the scrapings, together with the contents of the mould and pan, are transferred to a mortar. There the ingot of tin is freed from slag and then taken to the scales. The rest, after being finely powdered, is passed through a sieve. The flattened particles of tin which remain on the sieve are weighed with the ingot (the lump, as it is called); whilst the siftings are vanned on a shovel, and (the slag being washed off) the fine tin is collected, dried, and weighed with the rest: the whole gives the produce or percentage of metal in the ore. The results of the assays are expressed in cwts. of metal in the ton of ore. The percentage is rarely given and never used in Cornwall. Thus—"13-1/2 Produce" would mean that the assay yielded results at the rate of 13-1/2 cwts. of metal for one ton of the ore. Some assayers use a little powdered fluor-spar to assist the fusion of refractory slags. A small quantity of borax will also occasionally be of service for ores containing silica in excess of any iron that may be present. The borax renders the slag more fusible, and assists the formation of a larger lump (with less fine tin in the slag) than would be obtained by the use of culm alone.

The quality and the percentage of pure tin in the metal will vary considerably, according to the impurities that are associated with the ore to be assayed.

The crude lump is then remelted in a small iron ladle at as low a temperature as possible, and the fused metal is poured into a shallow trench about 4 inches long by 3/4 of an inch wide cut in a block of white marble. The metal will be silvery-white if the temperature employed be correct; if too hot, the surface will show a yellow, red, or blue colour (according to the heat employed); in such case the metal should be remelted at a lower temperature. If the metal on cooling remains perfectly clear and bright, then it may be assumed that the tin is of good quality and commercially pure. A crystallised or frosted appearance of the metal indicates the presence of some alloy, say of iron, copper, zinc, lead, antimony, &c. The assayer who has had much practice can readily distinguish the metal or metals that are associated with the ore by noting the appearance of the tin on cooling; and can fairly judge the quantity of impurity present by the amount of the crystallisation or stain.

Whilst the foregoing method of assaying cannot lay claim to scientific accuracy, it is by no means so imperfect as some writers would have us believe, who state that a loss of 5 to 10 per cent. arises in the operation. It is certainly the most ready and expeditious mode of determining the commercial value of a parcel of tin ore, which, after all, is the main object of all assaying operations.

The difficulty which beginners find in obtaining satisfactory results, and any loss of metal which those not accustomed to the process may incur, will invariably occur in the vanning of the powdered slag for the fine tin, the rest of the operations being easy of execution, and requiring only the ordinary care necessary for all metallurgical work.

There is no doubt that if low percentage ores containing silica are assayed in this manner, low results are obtained, as it is impossible to reduce the whole of the tin in the presence of free silica; with this class of ores, care should be taken to remove some of the silica by preliminary vanning, or some flux should be added which will combine with the silica, and so prevent its entering into combination with the tin. Low quality tin ores containing iron, copper, lead, zinc, antimony, etc., combined with arsenic, sulphur, or oxygen, will give very much higher results than the actual percentage of tin in the sample. The other metals (being readily reduced in the presence of tin) alloy with it, and give a hard lump difficult to fuse in the iron ladle; where the quantity of foreign metals is large, the metal can only be melted to a stiff pasty mass; so that (in determining the value of a ton of tin ore, or even reporting on the percentage of tin it contains) not only must the weight of the assay be the basis for calculation, but the quality and character of the metal obtained must also be considered. Thus two ores of tin might be assayed both yielding a similar produce, say 13-1/2 (67-1/2 per cent.), and yet one might contain 5 per cent. less tin than the other.

If it be required to obtain the pure metal from tin ores containing the ores of other metals associated with them, the latter must be removed by digesting in strong hydrochloric acid, and washing. The assay may then be conducted in the usual way, and a fairly pure lump will be obtained.

If wolfram be present in any appreciable quantity in the ore, it considerably reduces the proportion of lump, and at the same time it increases the fine tin (or prillion, as it is termed) in the assay. This may be got rid of by boiling in aqua regia, and dissolving out the tungstic acid which has been liberated by means of ammonia.

It will be seen that this method of assaying tin has its advantages and its drawbacks. It is quickly performed; with ores of good quality it gives results not to be excelled by any other process; and it gives the smelter the actual alloy and quality of metal he may expect to get in the smelting of the ore, which no other mode of assaying will do: against which may be set the skill required to obtain accurate results with the vanning shovel; the loss of metal in poor ores containing an excess of silica; and the high results from ores containing a large quantity of metallic impurities.

Cyanide Method.—Weigh up 20 grams of the ore and dry it on a scoop over the Bunsen flame. When dry, weigh, and calculate the percentage of water from the loss in weight. Transfer the dried ore to an evaporating dish, and cover with 30 c.c. of hydrochloric acid; boil for 10 or 12 minutes, and then add 5 c.c. of nitric acid and boil again. Dilute with water, and filter. Transfer the filter and its contents to an E Battersea crucible, and calcine it for a few minutes. Cool, and weigh the residue. The loss equals the oxides soluble in acid. Transfer the residue to the crucible and mix it with its own weight of cyanide of potassium; add a similar amount of "cyanide" as a cover. Place in the furnace, and when the charge has attained the temperature of the furnace (in from 3 to 6 minutes), remove it at once; tap the pot vigorously several times, and then pour its contents quietly into a mould. Dissolve the slag in water, clean, dry, and weigh the button of tin.

WET METHODS.

Detection.—Tin ore is detected by its insolubility in acids, high specific gravity, and characteristic appearance in water. The powder is separated from the lighter gangue by washing. It is fused in a Berlin crucible with five times its weight of potassic cyanide at a moderately high temperature in a muffle, or over the blowpipe. The slag is washed off with water, and the metallic buttons or residue treated with hydrochloric acid (not aqua regia), for some time. One portion of the solution strikes a purple colour with chloride of gold, another portion gives a white or grey precipitate or cloudiness with mercuric chloride. These reactions are characteristic of tin as stannous chloride.

Metallic tin treated with nitric acid becomes converted into a white insoluble powder (metastannic acid). Aqua regia dissolves tin readily, forming stannic chloride, and in this solution the metal is detected by precipitation with sulphuretted hydrogen, which gives a yellow precipitate. Tin in solution as stannic or stannous chloride is precipitated as metal by means of zinc.

The fact that tin forms two well-defined series of compounds is taken advantage of in assaying (just as in the case of iron), by determining how much of an oxidising agent is required to convert it from the stannous into the stannic state. For example, on the addition of a solution of permanganate of potash to a solution of stannous chloride the oxidation goes on rapidly, and the finishing point is sharp and distinct; but acid solutions of stannous chloride quickly take up oxygen from that dissolved in the water used and from the air. Unfortunately, there is no obvious sign that such oxidation has taken place, except that (fatal to the assay) a smaller volume of the permanganate is required. Great care is required with such solutions, both before and during titration. The addition of an excess of ferric chloride to the stannous solution, as soon as the whole of the tin has been dissolved, will lessen this liability to oxidation.

Separation.—If the tin is present in an alloy, the substance is boiled in an evaporating dish with dilute nitric acid until the whole of the material is attacked. Evaporate nearly to dryness, dilute, boil for a few minutes, and filter off the white insoluble residue. Under certain circumstances this residue will be nearly free from other metals, in which case it is ignited and weighed. If not known to be pure it must be ignited, reduced in a current of hydrogen, and treated as subsequently described.

When the tin is present as insoluble oxide in an ore, the substance is finely powdered, and from 1 to 5 grams of it (according to its richness) boiled with 30 c.c. of hydrochloric acid in an evaporating dish till the oxide of iron is seen to be dissolved. Then add 1 c.c. of nitric acid (or more if much pyrites, &c., is present) and continue the boiling till these are decomposed; dilute and filter off, washing first with dilute acid and afterwards with a little dilute ammonia, dry, ignite, and place in a combustion tube (together with the filter-ash) and heat to redness for about thirty minutes in a current of dried hydrogen.



The oxide of tin is placed in a porcelain boat (fig. 58), which is then introduced into a piece of combustion tube. The latter, wrapped in a piece of wire gauze, is supported on a couple of iron rings, and heated by one or two Bunsen burners in a furnace fitted up with loose fire-brick tiles, as shown in fig. 59.



When the reduction is complete the tube is allowed to cool; the boat is removed and the tin dissolved. Add a rod of zinc to the freely-acid hot solution, and in a few minutes decant through a filter and wash with water, after having removed the zinc. Wash the precipitated metal back into the beaker, and dissolve in 10 c.c. of dilute nitric acid, evaporate off the excess of acid; dilute, boil, and filter. Wash, dry, ignite strongly in a porcelain crucible, and weigh.

In the absence of antimony the above separation works very well, but if this metal is present in quantity the metals precipitated on the zinc must be covered with hydrochloric acid and treated with a few drops of nitric. It is then warmed with iron wire until no more of the latter dissolves. The antimony is precipitated as metal, and the tin remains in solution as stannous chloride. The antimony is filtered off, and may be washed with alcohol, and weighed, whilst the tin in the filtrate is precipitated with zinc, and treated as already described.

GRAVIMETRIC METHOD.

If the tin is not already in the metallic state it is reduced to this condition by the method given (precipitation by zinc). Treat the finely-divided metal (washed free from chlorides) in a four-inch evaporating dish with 10 c.c. of dilute nitric acid, cover with a clock-glass, and apply a gentle heat until the precipitate appears of a white colour and the metal is completely attacked. Evaporate nearly to dryness on a water-bath; then add 50 c.c. of water, heat to boiling, and filter. Wash with hot water, dry, transfer to a weighed porcelain crucible, add the filter-ash, ignite strongly, and weigh. The precipitate after ignition is stannic oxide (SnO_{2}). It is a yellowish-white powder (darker whilst hot), insoluble in acids, and contains 78.67 per cent. of tin. Cold dilute nitric acid dissolves tin to a clear solution, which becomes a white enamel-like jelly on heating; this (filtered off, washed, and dried) forms an opal-like substance, which is converted on ignition into stannic oxide with evolution of nitrous fumes. Stannic oxide when ignited with chlorides is more or less completely converted into stannic chloride, which volatilises. The presence of chlorides during the evaporation with nitric acid causes a similar loss.

Determination of Tin in an Alloy.—(Bronze.)—Take 2 grams, and attack with 20 c.c. of dilute nitric acid in a covered beaker with the aid of heat. Boil till the bulk is reduced by one-half, dilute with 50 c.c. of water, allow to settle for a few minutes, and filter; wash well first with water acidulated with a little nitric acid, and afterwards with water; dry, ignite, and weigh as stannic oxide.

Determination of Tin in Tin Ore.—Treat 5 grams of the dried and finely-powdered ore with 30 c.c. of hydrochloric acid in a four-inch evaporating dish. After the soluble oxides have been dissolved add 1 or 2 c.c. of nitric acid, boil off nitrous fumes, dilute, and filter. Dry the filter, transfer the cleaned ore to a piece of combustion tube ten or twelve inches long and narrowed at one end. Pass a current of hydrogen through the tube and heat to redness for 30 minutes; cool whilst the gas is still passing. Dissolve in 20 c.c. of dilute hydrochloric acid and keep the solution tinted with permanganate of potassium. When the colour of the permanganate becomes permanent dilute to a bulk of 50 c.c. with water, filter, and wash. Heat; add a rod of zinc weighing about 3 grams; allow to stand for a few minutes; decant through a filter; and wash, removing the remaining zinc and returning the tin to the beaker. Treat with 5 c.c. of dilute nitric acid, boil for some time, take up with water, filter, wash, dry, ignite, and weigh as stannic oxide.

VOLUMETRIC METHOD.

Titration with Solution of Permanganate of Potassium.—This titration may be made either directly on the solution of stannous chloride (prepared by dissolving the precipitated metal in hydrochloric acid), or indirectly, on a solution of ferrous chloride (produced by the reducing action of the precipitated metal on ferric chloride). The standard solution of permanganate of potassium is made by dissolving 5.356 grams of the salt in water and diluting to one litre. 100 c.c. are equivalent to 1.00 gram of tin.

The precipitated tin is transferred to a flask; and dissolved in 10 c.c. hydrochloric acid, with the aid of heat and in an atmosphere of carbonic acid. The acid and metal are placed in the flask; which is then filled with the gas, and stopped with a cork provided with a rubber valve. When solution is complete the flask is again filled with carbonic acid. Fifty c.c. of water freed from air and saturated with carbonic acid are then added. This water is made by adding a gram of bicarbonate of soda and 2 c.c. of hydrochloric acid to 100 c.c. of water: the effervescence sweeps out the dissolved oxygen. The permanganate of potassium solution is then run in from a stop-cock burette in the usual way until a faint pink tinge is obtained.

The following experiments show the effect of variations in the conditions of the assay. A solution of stannous chloride equivalent in strength to the "permanganate" was made by dissolving 19.06 grams of the crystallised salt (SnCl{2}.2H{2}O.) in 50 c.c. of water and 10 c.c. of hydrochloric acid and diluting to 1 litre with water freed from dissolved oxygen. 100 c.c. contain 1 gram of tin. In the first experiments tap water was used and no precautions were taken for excluding air. Except when otherwise stated, 20 c.c. of the stannous chloride were used in each experiment with 10 c.c. of hydrochloric acid, and were diluted to 100 c.c. with water before titration.

Effect of Varying Hydrochloric Acid.

Acid added 1.0 c.c. 10.0 c.c. 20.0 c.c. 50.0 c.c. "Permanganate" required 18.8 " 18.9 " 18.8 " 18.8 "

The only effect of the increase in quantity of acid was to give the brown of perchloride of manganese instead of the pink of permanganic acid to mark the finishing point.

Effect of Varying Temperature.

Temperature 15 C. 50 C. 70 C. 100 C. "Permanganate" required 18.8 c.c. 18.7 c.c. 18.6 c.c. 18.4 c.c.

Rate of Atmospheric Oxidation.—Solutions ready for titration were exposed to air at the ordinary temperature for varying lengths of time and then titrated.

Time exposed 0 min. 5 min. 10 min. 20 min. 60 min. "Permanganate" required 18.8 c.c. 18.8 c.c. 18.8 c.c. 18.8 c.c. 18.6 c.c.

It is best to titrate at once, although the loss by oxidation is only small after one hour's exposure.

Effect of Varying Tin.

Stannous chloride added 1.0 c.c. 10.0 c.c. 20.0 c.c. 50.0 c.c. 100.0 c.c. "Permanganate" required 0.7 " 8.8 " 18.0 " 47.4 " 95.4 "

Effect of Varying Bulk.

Bulk 50.0 c.c. 100.0 c.c. 200.0 c.c. 500.0 c.c. "Permanganate" required. 9.0 " 18.3 " 17.4 " 15.1 "

The two last series show an interference, which is due to the oxygen dissolved in the water, as may be seen from the following similar experiments, which were, however, performed with water freed from oxygen and in which the titrations were effected in an atmosphere of carbonic acid.

Effect of Varying Tin.—A new solution of stannous chloride was used.

Stannous chloride added 1.0 c.c. 10.0 c.c. 20.0 c.c. 50.0 c.c. 100.0 c.c. "Permanganate" required 1.0 " 10.0 " 19.8 " 49.6 " 99.3 "

Effect of Varying Bulk.

Bulk 30.0 c.c. 50.0 c.c. 100.0 c.c. 200.0 c.c. 500.0 c.c. "Permanganate" required 19.8 " 19.8 " 19.8 " 19.8 " 19.8 "

It will be seen that in working under these conditions the results are proportional and the method satisfactory.

Examination of Tin Phosphide.—(Phosphor Tin.)—This substance is used in the manufacture of "phosphor bronze" and similar alloys. It is a crystalline, imperfectly-malleable, metallic substance. It is soluble in hydrochloric acid with effervescence; phosphoretted hydrogen, which inflames on the addition of a drop or two of nitric acid, being evolved. It is attacked by nitric acid, yielding a white powder of stannic phosphate; this is not easily decomposed by ammonium sulphide or readily soluble in hydrochloric acid.

"Phosphor-tin" is made up only of tin and phosphorus. For the estimation weigh up 1 gram. Place in a weighed Berlin dish; and cover with 10 c.c. of nitric acid and 3 or 4 c.c. of water. Let the reaction proceed (under a clock-glass) on the water-bath till complete. Remove the glass; evaporate to dryness, and ignite, at first gently over a Bunsen burner, and afterwards in the muffle at a red heat. Cool in the desiccator, and weigh as quickly as possible when cold. The substance contains the tin as stannic oxide, SnO_{2}, and the phosphorus as phosphoric oxide, P_{2}O_{5}. The increase in weight on the gram of substance taken gives the weight of the oxygen taken up by the phosphorus and tin, and since 1 gram of tin takes up only 0.271 gram of oxygen, and 1 gram of phosphorus takes up 1.29 gram, the proportion of tin to phosphorus can be calculated from the increase in weight. For example, 1 gram of a sample gave 1.3410 gram of mixed oxides, which is 0.070 gram in excess of that which would be got with pure tin. If the substance was all phosphorus the excess would be 1.0190 gram; consequently the proportion of phosphorus in the substance is 0.070 / 1.019, or 6.87 per cent. The tin is calculated by difference, 93.13 per cent.

Another method of separating and determining the phosphorus is as follows:—Take 1 gram of the substance and add to it 15 c.c. of hot aqua regia. Boil till dissolved, dilute, and precipitate the tin with sulphuretted hydrogen. To the filtrate add ammonia and "magnesia mixture." Filter; wash the precipitate with dilute ammonia; dry, ignite, and weigh as magnesic pyrophosphate. Calculate the phosphorus, and take the tin by difference.

A sample of phosphor tin gave—

Tin 93.1 per cent. (by difference) Phosphorus 6.9 " ——- 100.0

Tin Arsenide.—This is met with in tin-smelting; it closely resembles the phosphide, but the crystals have a duller grey appearance. It contains simply tin and arsenic. The determination is made by treating 1 gram of the substance with nitric acid and weighing the mixed oxides of tin and arsenic in the same manner as in the case of the phosphide. One gram of arsenic will give 1.533 gram of arsenic oxide, As{2}O{5}; consequently the excess of weight of the mixed oxides over 1.271 gram must be divided by 0.262; the result multiplied by 100 gives the percentage of arsenic. In consequence of the higher atomic weight of arsenic the results by this method are not so close as with the phosphide. Each milligram of excess weight (over 1.271) represents 0.38 per cent. of arsenic, As. Both in this and in the corresponding phosphide determination care must be taken to avoid absorption of moisture, by allowing the oxides to cool in a desiccator and weighing quickly.

The percentage of arsenic is better determined as follows:—Weigh up 1 gram of the substance, dissolve in aqua regia, dilute, and pass sulphuretted hydrogen. Render alkaline with ammonia, and add ammonium sulphide till the precipitate is dissolved. Add "magnesia mixture." Filter off the precipitate, wash with dilute ammonia, ignite with a few drops of nitric acid, and weigh as magnesic pyrarsenate. Calculate the arsenic and take the tin by difference. A sample treated in this way gave—

Tin 96.8 per cent. by difference Arsenic 3.2 " ——- 100.0

Examination of Black Tin.—Dry the ore, and reduce it to a fine powder. Weigh up 2 grams, and boil with 20 c.c. of hydrochloric acid and 2 c.c. of nitric for ten or fifteen minutes. Filter, and reserve the filtrate.

Tungstic Acid.—Digest the residue with about 50 c.c. of water and a few c.c. of dilute ammonia for a few minutes, and filter; collect the filtrate in a weighed porcelain dish, evaporate to dryness, ignite, and weigh as tungstic acid, WO_{3}.

Stannic Oxide.—Dry, ignite, and weigh the insoluble residue. Transfer to a porcelain boat, and reduce in a current of hydrogen at a red heat for half an hour. Allow to cool whilst the hydrogen is still passing. Transfer the boat to a beaker, and dissolve up the tin in 10 c.c. of hydrochloric acid and a c.c. or so of nitric. Wash out the combustion tube with some acid and add the washing to the contents of the beaker. Warm gently, dilute with water, and filter. Collect, dry, ignite, and weigh the insoluble residue. Through the filtrate pass a rapid current of sulphuretted hydrogen, allow to settle, and filter. Wash the precipitate with hot water, dry, calcine gently; ignite with ammonium carbonate, and weigh as stannic oxide, SnO_{2}. The insoluble residue will in most cases retain some tin. Fuse it with fusion mixture, take up with hydrochloric acid, filter, pass sulphuretted hydrogen through the filtrate, collect and wash the sulphide of tin. Ignite and weigh as stannic oxide, and add it to that previously obtained.

Copper.—Pass sulphuretted hydrogen through the acid filtrate obtained in the first cleaning of the ore, collect the precipitate, and wash first with soda solution and then with hot water. Dry, ignite, and weigh as cupric oxide, CuO. Mix the filtrate with that from the main portion of the sulphide of tin.

Ferric Oxide.—Boil off the sulphuretted hydrogen from the mixed filtrates and peroxidise with nitric acid. Add ammonia in slight excess, boil, filter, dry, ignite, and weigh the precipitate as ferric oxide. This will be practically pure, but the iron in it must be determined by dissolving and titrating. The filtrate from the iron may contain zinc, lime, and magnesia, but rarely in quantities sufficient to be determined.

Silica, &c.—-The silica may be calculated from the weight of the residue insoluble in acid, after the reduction of the tin in hydrogen, by deducting from it the weight of the oxide of tin subsequently found. Or it may be determined as follows:—The insoluble portion is fused with fusion mixture, and taken up with hydrochloric acid, as already described. On filtering, the filter will retain a portion of the silica. The rest is recovered, after the removal of the stannous sulphide, by evaporating to dryness, taking up with hydrochloric acid, and filtering through the same filter. It is washed, dried, ignited, and weighed as silica. The filtrate from the silica is boiled with a little nitric acid and precipitated with ammonia. The precipitate is collected, washed, ignited, and weighed as ferric oxide and alumina (but it frequently contains oxide of titanium). When the last is present it is determined by fusing with bisulphate of potash and extracting with cold water. The solution is nearly neutralised with ammonia, charged with sulphurous acid, and boiled. The precipitate is collected, washed, dried, ignited, and weighed as oxide of titanium, TiO_{2}. The difference between this weight and that of the combined oxides gives the ferric oxide and alumina. The filtrate from the mixed oxides is examined for lime and magnesia.

Sulphur.—Rub up 5 grams of the ore with 5 grams of nitre, transfer to a porcelain dish, and fuse over a Bunsen burner for fifteen minutes. When cold, extract with water, and determine the sulphur volumetrically with standard barium chloride. The sulphur may be present as sulphide or sulphate.

Arsenic.—Take 5 grams, and evaporate with nitric acid; dilute, add ammonia, pass sulphuretted hydrogen, and filter.

To the filtrate add "magnesia mixture." Collect the precipitate, ignite with nitric acid, and weigh as magnesic pyrarsenate.

The following may be taken as an example of the composition of an impure black tin:—

Tungstic acid 1.8 % Stannic oxide 79.0 Silica 2.6 Titanic oxide 0.8 Copper oxide 0.9 Ferric oxide 13.4 Sulphur 0.4 Arsenic 0.3 —— 99.2

Examination of Hardhead.—In the smelting of tin ores a quantity of speise, known as "hardhead," is produced. It is essentially an arsenide of iron, carrying a considerable quantity of tin. Much of this last is present in the form of small buttons of metal distributed through the mass. The buttons can be seen on careful inspection, and become evident on powdering.

In assaying the substance, a variation in the usual method of sampling is required, because of the quantity of metal present which cannot be powdered. After powdering as finely as possible, the coarse particles are sifted off and weighed. The weight of the powder is also taken. The method of working is best illustrated by an example. A sample of hardhead weighed 155.1 grams, and gave 21.0 grams of coarse particles, equivalent to 13.5 per cent. of the whole. The fine portion weighed 134 grams, which is equivalent to 86.5 per cent.

Thirteen and a half grams of the coarse material were dissolved in aqua regia, and diluted with water to 1 litre. Ten c.c. of this contain 0.135 gram of the metallic portion, which is the amount contained in 1 gram of the original hardhead. If, in a determination, 1 gram of the substance is wanted, weigh up 0.865 gram of the powdered portion, and add to it 10 c.c. of the solution. It will be seen that these together make up 1 gram of the original sample. The solution of the metallic portion must be saved until the analysis is finished.

Tin and Copper.—Weigh up the portion of the powdered stuff equivalent to 1 gram of the sample. Transfer to a flask, and cover with 10 c.c. of the solution of the metallic portion and 10 c.c. of aqua regia. Boil gently till oxidation is complete and the nitric acid for the greater part driven off. Dilute to 100 c.c. with water, and pass sulphuretted hydrogen for some time. Filter, wash with hot water, and rinse through the funnel back into the flask. Digest with yellow sodium sulphide until only a light, flocculent, black precipitate is left. Filter this off, wash with hot water, dry, calcine, treat with a little nitric acid, ignite, and weigh as copper oxide, CuO. The weight multiplied by 0.7983 gives the weight of copper.

The filtrate containing the tin is rendered acid with hydrochloric acid, and filtered. The precipitate is rinsed into a half-pint beaker, covered with 20 c.c. of hydrochloric acid, and boiled down to about 20 c.c. The solution is filtered off from the sulphur and sulphide of arsenic, which, after washing with hot water, is transferred to a flask labelled "arsenic." A strip of sheet zinc (2 in. by 1 in.) is placed in the solution. The evolution of hydrogen should be brisk. In five or ten minutes decant off a few c.c. of the liquid, and test with sulphuretted hydrogen for tin. If no yellowish precipitate is formed, decant off the rest of the liquid, and wash the precipitated metal with hot water two or three times by decantation. The metal should be in a lump; if there are any floating particles they must be made to sink by compression with a glass rod. Transfer the washed metal to an evaporating dish 3 or 4 in. across, and cover with a few c.c. of hot water. Add nitric acid drop by drop till the tin is completely attacked. Evaporate nearly to dryness, and add a drop or two more of nitric acid and 20 c.c. of water. Boil and filter. Wash with hot water, dry, ignite, and weigh as stannic oxide, SnO_{2}. Calculate to metallic tin by multiplying by 0.7867.[76]

The filtrate from the first treatment with sulphuretted hydrogen will probably no longer smell of the gas. Warm and pass the gas for a few minutes longer. Filter off any precipitate of sulphide of arsenic, and transfer it to the flask for "arsenic." Boil the filtrate (ignoring any signs of a further precipitation of arsenic) with a few c.c. of nitric acid, and separate the iron as basic acetate. Wash; reserve the filtrate for cobalt.

Iron.—Rinse back the "basic acetate," precipitate into the flask, add ammonia, dilute with water to about 100 c.c., and pass sulphuretted hydrogen for a few minutes. Filter, and wash with hot water. Collect the filtrate in the flask labelled "arsenic." Boil the precipitate with dilute sulphuric acid, filter, and titrate the filtrate with the permanganate of potassium solution after boiling off the sulphuretted hydrogen. Report the result as iron. The sulphuric acid will not effect complete solution, a light black residue will remain, chiefly sulphur; this must be rinsed into the filtrate from the acetate separation. It contains cobalt.

Cobalt.—The filtrate from the acetate separation will have a pink colour. Render it ammoniacal and pass sulphuretted hydrogen. Collect the precipitate on a filter, dry, and ignite. Dissolve in hydrochloric acid, and evaporate nearly to dryness with an excess of nitric acid. Dilute with 10 or 20 c.c. of water and add potash solution in slight excess. Add acetic acid until the solution is acid and the precipitate is quite dissolved. Add 20 or 30 c.c. of a strong solution of potassium nitrite, and determine the cobalt, as described on pp. 254, 256. Boil the filtrate from the cobalt, precipitate with hydrochloric acid, render ammoniacal, and test for zinc, nickel, and manganese.

_The remainder of the tin_ will be contained in the flask labelled "arsenic." Acidify with hydrochloric acid and filter. Rinse into a beaker, and evaporate to a small bulk with 10 c.c. of nitric acid. Dilute and filter. Dry the precipitate, consisting of stannic arsenate (2SnO_{2}.As_{2}O_{5}), ignite, and weigh. Calculate the tin it contains by multiplying by 0.4453, and add to that already found.

Arsenic.—This is determined in a separate portion. Weigh up a portion of the powder equivalent to 1 gram of the hardhead, place in a pint flask, and boil with 10 c.c. of nitric acid. When action has ceased add 10 c.c. of the solution of the metallic portion and then hydrochloric acid (a few drops at a time) till solution is complete. Warm gently in dissolving, but do not boil. Dilute to about 100 c.c., render alkaline with ammonia, and add 20 c.c. of yellow ammonium sulphide. Digest at a gentle heat for about thirty minutes, filter, and wash. Add 50 c.c. of magnesia mixture, shake well, allow to stand for an hour, filter, and wash with dilute ammonia. The precipitate is dissolved and then titrated with uranium acetate, or it is evaporated with nitric acid, ignited, and weighed as pyrarsenate of magnesia. Calculate the result to arsenic, As.

Sulphur.—Weigh up a portion of the powder equivalent to 2 or 3 grams of the hardhead. Rub up in a mortar with 5 grams of nitre and fuse in a porcelain dish for ten minutes. Extract with water, add 20 or 30 c.c. (as the case may be) of the solution of the "metallics." Add 10 grams of sodic acetate, and ferric chloride until the precipitate turns brown; dilute with water to half a litre, boil, and titrate with standard baric chloride, as described under Sulphur. Report as sulphur.

A sample of hardhead examined in this way gave—

Sulphur 3.00% Arsenic 27.10 Tin 22.2 Copper 1.64 Iron 43.2 Cobalt 2.6 ——— 99.74

Examination of Tin Slags.—In tin smelting works the term "slag" is applied to the unfused portion of the charge. It is made up of unburnt anthracite and small lumps of slag proper together with some buttons of metallic tin. This is rarely, if ever, assayed. The slag proper (or, as it is generally called, "glass") is a silicate of iron, alumina, and lime, containing from 3 to 7 per cent. of tin. It is thus examined:—The sample after bruising on an iron plate, is reduced to a very fine powder by grinding in an agate mortar. In this state it is in most cases readily decomposed by hydrochloric acid.

Determination of Tin.—Where the percentage of tin only is required, take 2 grams of the powdered slag and well mix with it 20 c.c. of hydrochloric acid, and heat to boiling. Add 1 c.c. of nitric acid, allow to stand for fifteen minutes, dilute with water, and filter. Pass a rapid current of sulphuretted hydrogen for some time. Allow to settle, and filter. The precipitate, after washing with hot water, is dried, and gently calcined until the greater part of the sulphur is burnt off. It is then strongly ignited in the muffle (or over the blowpipe) with the addition of a small lump of ammonic carbonate. The residue is weighed as stannic oxide (SnO_{2}); and is calculated to metallic tin by multiplying by 0.787. The percentage on the slag is calculated in the usual way.

The tin is always best determined in the examination of slags by a separate assay carried out in this way. The determination of the other constituents is thus made:—

Silica.—Take 2 grams of the powdered slag and cover them, in a small evaporating dish, with 20 c.c. of hydrochloric acid; mix well by stirring with a glass rod; and evaporate to dryness. If (as is generally the case) tungsten is present the solution will be blue. Take up with 20 c.c. of hydrochloric acid. Add 1 c.c. of nitric acid; and reduce by boiling to about half the bulk. Add about 20 c.c. of water, boil, and filter. Wash the residue with hot dilute hydrochloric acid. It consists of silica with the tungstic acid. Wash it back into the dish; and digest with 5 or 10 c.c. of a cold solution of ammonic carbonate. Filter; and collect the filtrate and washings in a weighed porcelain dish. Dry the residue, ignite strongly, and weigh as silica, SiO_{2}. In certain exceptional cases this may contain some unaltered cassiterite, which is easily recognised by its appearance.

Tungsten.—The ammonic carbonate filtrate from the silica is evaporated to dryness, ignited strongly over the blowpipe, and weighed. The residue is tungstic acid, WO_{3}. The tungsten may be conveniently reported in this form, although it is probably present as a lower oxide.

Tin.—The acid filtrate from the silica and tungstic acid is treated with sulphuretted hydrogen. The sulphide of tin is filtered off. Since the percentage of tin has been already determined, this precipitate may be neglected; or may be treated in the same way as the previous one, so as to check the result. Since some stannic chloride will have been lost in the evaporation, a low result may be expected. The tin should be reported as stannous oxide; and is calculated by multiplying the percentage of tin by 1.136.

The filtrate from the tin is boiled rapidly down to remove sulphuretted hydrogen; and then peroxidised with 1 or 2 c.c. of nitric acid. It is cooled, transferred to a graduated flask, and diluted with water to 200 c.c.

Ferrous Oxide and Alumina.—Half the filtrate from the tin (that is, 100 c.c.) is taken, nearly neutralised with soda, and treated with sodium acetate. The basic acetate precipitate obtained on boiling is filtered off and washed. Reserve the filtrate. The precipitate is dissolved off the filter with hot dilute hydrochloric acid; and the solution thus formed is treated with a slight excess of ammonia, and boiled. The precipitate is filtered off, washed with hot water, dried, ignited, and weighed as mixed ferric oxide and alumina. The ignited precipitate is then dissolved with sulphuric and hydrochloric acids; and the iron determined in the solution by titration with the solution of stannous chloride. The iron found is calculated to and reported as ferrous oxide, FeO (factor = 1.286). To find the alumina, which is best estimated by difference, multiply the iron by 1.428 to get the weight of ferric oxide, and deduct this from the weight of alumina and ferric oxide found. This, of course, gives the alumina. A direct determination may be made by removing the tin from the titrated solution with sulphuretted hydrogen, filtering, nearly neutralising with ammonia, and boiling with a few grams of hyposulphite of soda. The precipitate, filtered, washed, and ignited, is the alumina, which is weighed. The direct determination gives a slightly low result.

Oxides of Zinc and Manganese.—These are determined in the filtrate from the basic acetate precipitate by rendering alkaline with ammonia, and passing a current of sulphuretted hydrogen. Generally a small, but decided, precipitate of alumina comes down, together with sulphides of any zinc or manganese which is present. The precipitate is allowed to settle, dried, ignited, and weighed. The metals are separately determined in it; and the residue is counted as alumina, and added to that already found. The mixed precipitate amounts to from 1 to 2 per cent. of the sample.

Lime.—The filtrate from the last is treated with ammonic oxalate, boiled for a few minutes, allowed to settle, and filtered. The precipitate is washed with hot water; dried; ignited; and weighed as carbonate, after gentle ignition; or as lime, after strong ignition in the muffle.

Magnesia.—The filtrate from the lime is treated with sodic phosphate and ammonia. It is well mixed by stirring, and allowed to stand overnight. The precipitate is washed with dilute ammonia, dried, ignited, and weighed as pyrophosphate.

Soda and Potash.—These are determined in the remaining half of the filtrate from the tin. The solution is rendered ammoniacal with ammonia; and treated, first with sulphuretted hydrogen, and then with ammonium oxalate. The precipitate is filtered off and rejected. The filtrate is evaporated in a small porcelain dish over a Bunsen burner, or on the sand bath; and towards the close (or earlier if the evaporation is not proceeding well) nitric acid is added. The evaporation is carried to dryness; and the residue heated nearly to redness. The residue, which consists of magnesia with carbonates and chlorides of the alkalies, is extracted with water; and filtered. The filtrate is evaporated with hydrochloric acid in a weighed platinum dish, ignited gently, and weighed. This gives the weight of the mixed chlorides of sodium and potassium; which are then separated and determined as described under Potash.

It must be remembered when calculating the percentage that (with the exception of the silica, tungstic acid, and tin) the determinations have been made on 1 gram of the sample.

The following analysis will illustrate the composition of such a slag:—

Tungstic acid 1.3% Silica 39.4 Stannous oxide 8.1 Ferrous oxide 26.2 Alumina 14.8 Oxide of manganese traces Lime 7.9 Magnesia 0.5 Alkalies calculated as soda 1.7 ——- 99.9

TITANIUM.

Titanium only occurs as a mineral in its oxidised state, or as titanic oxide (TiO{2}). It is a substance which has little commercial value, and is generally recognised as one of the rare bodies; although, in small quantities, it is widely disseminated. It occurs in granite, basalt, and other igneous rocks in quantities up to as much as 1 per cent. It is also met with in clays and iron ores, and in river sands, in which it is often associated with stream tin. The proper minerals of titanium are rutile (TiO{2}), titaniferous iron (titanate of iron), and sphene (titanate and silicate of lime).

The oxide of titanium (like cassiterite and quartz) is undecomposed by hydrochloric or nitric acid; so that it is generally found in the residue insoluble in acids. The titanates, however, are attacked, and a portion of the titanium dissolves; so that it must be looked for in both the filtrate and residue. Oxide of titanium in its native form, or after ignition, may be made soluble by fusing the finely-divided substance with fusion mixture in a platinum dish. The resulting titanate is dissolved out of the "melt" by cold hydrochloric acid.

The method most commonly used is fusion with bisulphate of potash. This renders the oxide of titanium soluble in cold water. The process is as follows:—The substance is extracted with hydrochloric and nitric acids, and the solution reserved for further treatment; the residue is dried, moistened with sulphuric acid, and evaporated once or twice to dryness with hydrofluoric acid. It is then fused with bisulphate of potash, and the "melt" extracted with cold water until all soluble matter is removed. The solution is filtered. The residue may consist of unremoved silica, and oxides of tantalum, niobium, and, perhaps, chromium. On the prolonged boiling of the filtrate, the oxide of titanium (and oxide of zirconium, if any) is precipitated.

Any titanium dissolved by the first extraction with acids is recovered in the following way:—Sulphuretted hydrogen is passed into the acid solution, and any precipitate that may be formed is filtered off. The filtrate is oxidised, and the iron, aluminium, and titanium are separated as basic acetates (see under Iron). The precipitate is dried and fused with bisulphate of potash. The "melt" is extracted with cold water, filtered if necessary, and the solution rendered first faintly alkaline with ammonia, then very slightly acid with sulphuric acid. 30 or 40 c.c. of a saturated solution of sulphurous acid is added, and the oxide of titanium precipitated by prolonged boiling. It is filtered off, added to the precipitate previously got, ignited with ammonic carbonate towards the end, and then weighed.

Detection.—Titanium is detected in an insoluble residue by fusing the residue for some time in a bead of microcosmic salt. In the reducing flame it gives a violet colour, which becomes reddish-brown if much iron is present. In the oxidising flame it gives a colourless or whitish bead. It is best detected in acid solutions by the deep brown or iodine colour developed on adding hydroxyl. A solution of this can be prepared by pouring peroxide of barium (BaO_{2}) diffused in water into dilute hydrochloric acid (a little at a time), and keeping the acid in excess.

Separation.—In the usual course of an analytical separation the hydrate of titanium will be thrown down with ferric hydrate, &c., on the addition of ammonic chloride and ammonia. It is best separated from this precipitate by fusion with bisulphate of potash, as already described, but it must be remembered that the presence of much mineral acid prevents complete precipitation when the solution is boiled. Further, if phosphates are present, the precipitate will contain phosphoric oxide; it may be freed from this by fusion with sodium carbonate. A very good method of separating titanium from iron is to add tartaric acid and ammonia to the solution, and then precipitate the iron (as sulphide) with sulphuretted hydrogen. The filtrate contains the titanium, which is recovered by evaporating and igniting. It may be separated from zirconia by the action of sodium carbonate, which precipitates both; but when concentrated, redissolves the zirconia. The separation from large quantities of silica is best effected by evaporating with hydrofluoric acid, which volatilises the silicon; but sulphuric acid must be present, otherwise some titanium also will be lost, as may be seen from the following experiments,[77] in which oxide of titanium (pure, ignited) was evaporated to dryness with a quantity of hydrofluoric acid known by experiment to be sufficient to volatilise 1 gram of silica.

Without sulphuric acid, 0.0466 gram of titanic oxide left 0.0340 gram, showing a loss of about 25 per cent.

With sulphuric acid the following results were obtained:—

Oxide taken. Left after Evaporation and Ignition. 0.0340 gram 0.0340 gram 0.0414 " 0.0413 " 0.0520 " 0.0520 " 0.0352 " 0.0352 "

GRAVIMETRIC DETERMINATION.

The titanic hydrate thrown down by ammonia (or on boiling the solution from the bisulphate) is collected, washed, dried, ignited strongly with the addition of a little ammonic carbonate, and weighed. The substance is titanic oxide (TiO_{2}), and is generally reported as such. It contains 60.98 per cent. of titanium. It should be white, if pure (Holland), white, yellow, or brown (Fresenius), or black (Tidy).

VOLUMETRIC METHOD.

A method has been proposed based on the reduction of titanic oxide by zinc in hydrochloric acid solutions to the sesquioxide. The reduction is marked by the development of a violet or green colour, the former with chlorides and the latter when fluorides are present. The quantity of titanium reduced is measured by titrating with permanganate of potassium solution. The water used must be free from dissolved oxygen.

TUNGSTEN AND TUNGSTATES.

Tungsten occurs in nature only in the oxidised state, or as tungstic acid (WO_{3}), either free, as in wolframine, or combined with oxides of manganese and iron, as in wolfram, or with lime, as in scheelite. Wolfram occurs associated with tin ores, the value of which is consequently lowered. Both wolfram and scheelite are of considerable importance as a source of tungstic acid for the manufacture of sodium tungstate, which is used as a mordant and for some other purposes, and as a source of metallic tungsten, which is used in steel-making.

The tungsten minerals have a high specific gravity (6 to 7.5). On treatment with hydrochloric acid or aqua regia they are decomposed; the yellow tungstic acid separates and remains insoluble.

Tungsten itself is insoluble in nitric acid or aqua regia; but is converted into tungstic acid (WO_{3}) by prolonged and strong ignition in air. Alloys containing tungsten leave tungstic acid after treatment with nitric acid or aqua regia. Tungstic acid may be got into solution after fusion with alkalies or alkaline carbonates. This solution gives with hydrochloric acid a white precipitate of tungstic acid, which becomes yellow on boiling, but the separation is not complete. Fusion with bisulphate of potash gives a residue, which does not dissolve in water, but is soluble in ammonic carbonate. For the assay of minerals containing tungsten these reactions are only occasionally taken advantage of for testing or purifying the separated tungstic acid.

Detection.—The minerals are easily recognised by their physical characters, and the yellow tungstic acid separated by boiling with acids is the best test for its presence; this, after decanting and washing, immediately dissolves in a few drops of dilute ammonia. A solution of tungstate acidulated with hydrochloric acid becomes intensely blue on the addition of stannous chloride and warming. Fused in a bead of microcosmic salt it gives a clear blue colour (reddish-brown if iron is also present) in the reducing flame, but is colourless in the oxidising flame.

Solution and Separation.—The decomposition and solution of natural tungstates is difficult to effect owing to the separation of tungstic acid; the method of treatment is as follows:—Boil the finely-powdered substance with hydrochloric acid or aqua regia till it apparently ceases to be attacked; dilute, filter, and wash with dilute hydrochloric acid. Cover with dilute ammonia, and filter the solution, which contains ammonic tungstate, into an evaporating dish. Treat the residue again with acid, and again dissolve out the separated tungstic acid with ammonia, and repeat this operation until decomposition is complete. By this means there will be obtained—(1) a solution containing tungstate of ammonia; (2) an insoluble residue with silicates, and oxides of tin, niobium, tantalum, &c.; and (3) an acid solution containing the soluble bases. The tungstate of ammonia requires simple evaporation on the water-bath and gentle ignition in order to cause the tungstic acid to be left in an almost pure state; possibly, it may carry a little silica.

GRAVIMETRIC DETERMINATION.

The tungstic acid is dissolved, and separated as ammonic tungstate, and, after evaporation, is gently ignited, the heat being increased towards the end. The residual tungstic acid is fixed, so that when the ammonia has been driven off it may be strongly heated without loss. It is a dark yellow or brown powder whilst hot, which becomes a light yellow on cooling. If any reduction has taken place it will be more or less greenish. It is weighed when cold, and is the trioxide or "tungstic acid" (WO_{3}), which contains 79.31 per cent. of tungsten. After its weight has been taken its purity is checked by fusing with hydric potassic sulphate, extracting with water, and treating the residue with ammonic carbonate. Any silica present will be left undissolved; it should be separated and weighed, and its weight deducted from that of the tungstic acid found.

Determination of Tungstic Acid in Wolfram.—Take 2 grams of the finely-powdered sample and boil with 50 c.c. of hydrochloric acid for half an hour, adding 5 c.c. of nitric acid towards the end. Allow to stand overnight and boil again for 15 or 20 minutes; dilute with an equal volume of water, and filter. Wash with dilute hydrochloric acid, dissolve in a few c.c. of warm dilute ammonia, and dilute to 200 c.c. with distilled water; allow to settle, and filter. Evaporate in a weighed dish, ignite, and weigh.

The following analysis will illustrate the composition of a sample of Cornish wolfram as brought into the market:—

Tungstic acid 50.1% Cassiterite 10.9 Ferrous oxide 24.6 Manganous oxide 5.4 Niobic oxide, alumina, &c. 3.5 Silica 1.2 Copper oxide 2.7 Zinc oxide 0.22 Arsenic 0.51 Sulphur 0.20 ——- 99.33

NIOBIC AND TANTALIC OXIDES.

These oxides are commonly met with in samples of wolfram and tinstone, especially niobic. They are probably present in the form of columbite, a niobate of iron and manganese; and tantalite, a tantalate of the same metals.

On boiling with hydrochloric acid they are both liberated, and remain for the greater part (all the niobic) in the insoluble residue with the tungstic acid. On removing the latter with dilute ammonia they remain as a white insoluble precipitate, very prone to run through the filter on washing. They may be dissolved in hydrofluoric acid either at once or after fusion with bisulphate of potash, and extraction with cold water. To the solution in hydrofluoric acid gradually add a boiling solution of acid potassium fluoride (HF, KF.). Potassic fluotantalate (soluble in 200 parts of water) separates out first, and afterwards potassic fluoniobate (soluble in 12 parts of water). The separated salts (after heating with sulphuric acid and washing out the potassium sulphate formed) are ignited with ammonic carbonate, and weighed as tantalic oxide (Ta{2}O{5}) and niobic oxide (Nb{2}O{5}) respectively.

They are both white powders. The oxide of niobium dissolved in a bead of microcosmic salt gives a bluish colour in the reducing flame. The oxide of tantalum dissolves in the bead, but gives no colour.

FOOTNOTES:

[76] This will give almost the whole of the tin; a further portion will be got in subsequent work, and must be added to this result.

[77] Published by P. Holland, in the Chemical News, vol. lix. p. 27.



CHAPTER XIII.

MANGANESE, CHROMIUM, &c.

MANGANESE.

Manganese occurs mainly as black oxide (MnO{2}) in the mineral pyrolusite; and, in a less pure form, in psilomelane and wad. The value of the ore depends rather on the percentage of available oxygen than on the proportion of metal present. The results of assays are generally reported as so much per cent. of the dioxide (MnO{2}). In smaller quantities it is very widely distributed. Manganese itself has a value for steel-making; or, rather, for the making of spiegeleisen and ferro-manganese, which are used in the Bessemer and Siemens processes. For this purpose the percentage of the metal (Mn) is required. Consequently the minerals of manganese may be considered in two aspects—(1) as a source of oxygen; and (2) as a source of manganese. These will require separate consideration.

The black oxide is mainly used in the preparation of chlorine, liberation of which it brings about when treated with hot hydrochloric acid, or with a mixture of common salt and sulphuric acid. The quantity of chlorine which is obtained depends upon the proportion of dioxide present;[78] and in assaying may either be measured by its equivalent of iodine liberated, or by the oxidising effect on an acid solution of ferrous sulphate. When the ore also carries substances which have a reducing effect (such as ferrous compounds), such assays will give, not the total dioxide (MnO_{2}), but less, by the amount required to oxidise these impurities; and this is exactly what is required in valuing such an ore for commercial purposes. Manganese compounds are characterised by the readiness with which they may be converted into highly-oxidised bodies. Solution of manganese in hydrochloric acid, rendered alkaline with ammonia, yields a clear solution,[79] which rapidly takes up oxygen from the air, forming a brown precipitate of the oxide (Mn_{2}O_{3}). The addition of bromine or chlorine to such a solution determines the precipitation of a still higher oxide (approximately MnO_{2}). On treating a compound containing manganese with nitric acid and dioxide of lead (PbO_{2}), the oxidation is carried still further, a purple-coloured solution of permanganic acid (HMnO_{4} or H_{2}O.Mn_{2}O_{7}) being formed. On fusing minerals containing (even traces of) manganese with sodium carbonate in an open crucible, a green "melt" is obtained which owes its colour to sodium manganate (Na_{2}MnO_{4} or Na_{2}O.MnO_{3}). This salt is soluble in water, forming a green solution; which, when rendered acid, rapidly changes into the permanganate with the characteristic purple colour. Permanganate of potash is a salt much used in assaying, with some properties of which the student will have already become familiar.

Compounds of manganese, on boiling with strong hydrochloric acid, yield manganous chloride[80] (MnCl_{2}).

The properties given above serve for the detection of manganese; the higher oxides are distinguished by causing the evolution of chlorine (with its peculiarly suffocating smell) when acted on with hydrochloric acid; while the green "melt," with sodium carbonate, can be relied on for the recognition of manganese itself. There is no dry assay of manganese ores.

WET METHODS.

Strong hydrochloric acid is the best solvent for ores of manganese; but where the proportion of dioxide (MnO_{2}) is required, the solution is effected during the assay. The ore should be in a very fine state of division before treatment with acids.

The separation of manganese from other metals is thus effected: Ignite, in order to destroy any organic matter which may be present; dissolve in hydrochloric acid, and evaporate to dryness, to separate silica. Take up with hydrochloric acid, dilute, pass sulphuretted hydrogen, and filter. Boil off the excess of gas, peroxidise the iron with a drop or two of nitric acid, and separate the iron as basic acetate (as described under Iron).[81] If the iron precipitate is bulky, it is dissolved in a little hydrochloric acid, reprecipitated, and the filtrate added to the original one. Neutralise with soda, and add bromine in excess; heat gradually to boiling, allow to settle, and filter. The precipitate is impure dioxide of manganese (containing alkalies and, possibly, cobalt or nickel).

GRAVIMETRIC DETERMINATION.

Dissolve the precipitate in hydrochloric acid, and boil; add a slight excess of carbonate of soda, warm, and filter. Wash with hot water, dry, carefully ignite in an open Berlin crucible, and weigh. The substance is the brown oxide (Mn{3}O{4}), and contains 72.05 per cent. of manganese. If the percentage of dioxide is required it may be calculated by multiplying the percentage of manganese by 1.582. It must be borne in mind that the manganese should never be calculated to dioxide except when it is known to exist in the ore only in that form.

VOLUMETRIC METHODS.

The two methods are based on the oxidising effect of manganese dioxide; and if the metal does not already exist in this form it will require a preliminary treatment to convert it. The following method due to Mr. J. Pattinson[82] effects this: A quantity of the ore containing not more than .25 grams of the metal (Mn), is dissolved in hydrochloric acid in a pint beaker, and, if necessary, 3 or 4 c.c. of nitric acid are added to peroxidise the iron, and ferric chloride is added if required, so that there may be at least as much iron as manganese. Calcium carbonate is added till the solution is slightly red; and next the redness is removed by the cautious addition of acid; 30 c.c. of zinc chloride solution (containing 15 grams of zinc per litre) are added, the liquid is brought to boil and diluted to about 300 c.c. with boiling water; 60 c.c. of a solution of bleaching powder (33 grams to the litre and filtered), rendered slightly greenish by acid, are then run in and are followed by 3 grams of calcium carbonate suspended in 15 c.c. of boiling water. During effervescence the beaker is covered, the precipitate is stirred, and 2 c.c. of methylated spirit are mixed in. The precipitate is collected on a large filter, washed with cold water, and then with hot, till free from chlorine, which is tested for with starch and potassium iodide. The acid ferrous sulphate solution (presently described) is then measured into the beaker, and the precipitate, still in the paper, added; more acid is added (if necessary), and the solution is diluted and titrated. In place of bleaching powder solution, 90 c.c. of bromine water (containing 22 grams per litre) may be used.

FERROUS SULPHATE ASSAY.

This method, which is the one commonly used, is based on the determination of the amount of ferrous iron oxidised by a known weight of the ore. It is known that 87 parts of the dioxide will oxidise 112 parts of ferrous iron;[83] therefore 1 gram will oxidise 1.287 gram of ferrous iron, or 1 gram of ferrous iron oxidised will be equivalent to 0.7768 gram of the dioxide. The finely-divided substance containing the dioxide is digested in a solution of a known quantity of iron in sulphuric acid. The iron, of course, must be in excess, which excess is determined when the ore is dissolved by titrating with standard permanganate or bichromate of potash solution. The assay resolves itself into one for the determination of ferrous iron, for which the standard solutions and method of working described under Iron are used.

The assay is as follows:—For rich ores, 2 grams of clean soft iron wire are treated, in a pint flask, with 100 c.c. of dilute sulphuric acid and warmed till dissolved. Carefully sample the ore, and in one portion determine the "moisture at 100 C.;" grind the rest in a Wedgwood mortar with a little pure alcohol until free from grit. This reduces the substance to a finely-divided state and assists solution. Evaporate off the alcohol and dry at 100 C., mix well, and keep in a weighing-bottle. Weigh up 2 grams and add them to the solution of iron in the flask; carefully wash it all down into the acid liquid. On rotating the flask the ore will rapidly dissolve, but gentle heat may be used towards the end to complete the solution. When the residue is clean and sandy-looking, and free from black particles, the flask is cooled, and the residual ferrous iron is determined by titration with "permanganate." The iron thus found, deducted from the 2 grams taken, will give the amount of iron peroxidised by the dioxide contained in the 2 grams of ore. This divided by 2 and multiplied by 77.68 will give the percentage of dioxide in the sample, or multiplied by 49.41 will give that of metallic manganese.

When the quantity of manganese or of the dioxide to be determined is small, it is not necessary to use 2 grams of iron; 1 gram, or even less, may be taken. The iron may be used in the form of a standard solution of ferrous sulphate and portions measured off, thus saving the labour of weighing.

Determination of Dioxide in a Manganese Ore.—Weigh up 1 or 2 grams of the finely-powdered ore[84] and an equal weight of pure iron wire, dissolve the wire in 50 or 100 c.c. of dilute sulphuric acid, and, when solution is complete, add the ore and warm till it too is dissolved. Cool and titrate the remaining ferrous iron with the permanganate or bichromate of potassium solution.

For example, 0.7560 gram of pyrolusite and 1.000 gram of iron were taken and treated as above; 13.9 c.c. of "permanganate" (standard 100 c.c. = 0.4920 gram iron) were required; this indicates that 0.0684 gram of iron was left unoxidised by the ore. The iron oxidised, then, was 0.9316 gram (1.000 - 0.0684); multiplying this by 0.7768, we find that 0.7237 gram is the quantity of manganese dioxide which was present. This is equivalent to 95.77 per cent.;

0.7560 : 0.7237 :: 100 : 95.77.

IODINE METHOD.

It has been already stated that when dioxide of manganese is boiled with strong hydrochloric acid chlorine is given off, and that the amount of chlorine so liberated is a measure of the dioxide present. If the chlorine is passed into a solution of potassium iodide, an equivalent of iodine will be set free.[85] This is apparently a very indirect way of determining how much of the dioxide is present; but the reactions are very sharp, and the final determination of the iodine is an easy one.



The finely-powdered sample of dioxide is placed in a small flask provided with an exit tube leading into a solution of potassic iodide (fig. 60). On adding hydrochloric acid and boiling, the chlorine evolved is driven into the iodide solution and there absorbed; the boiling is continued till the steam and hydrochloric acid fumes have driven the last portions of the chlorine out of the flask and into the solution. In this experiment there is a strong tendency for the iodide solution to rush back into the flask. This tendency is overcome by avoiding draughts and regulating the heat; or by placing a lump of magnesite in the flask, which acts by evolving carbonic acid and so producing a steady outward pressure. When the distillation is finished the tube containing the iodine is detached and washed out into a beaker. If the solution is strongly acid it should be almost neutralised by the cautious addition of dilute ammonia. If crystals of iodine have separated, potassium iodide must be added in quantity sufficient to dissolve them. The condenser must be kept cool whilst the chlorine is passing into it.

The solution, transferred to a beaker, is titrated with a standard solution of sodic hyposulphite (100 c.c. = 1.27 gram iodine or 0.435 gram of dioxide of manganese). In titrating, the solution should be cold, or not warmer than 30 C. The bulk may vary from 100 to 200 c.c.; but it is best always to work with the same volume. The "hypo" is run in with constant agitation until the brown colour has been reduced to a light yellow; 5 c.c. of starch solution are then added and the titration cautiously continued until the end is reached; the finish is indicated by a change from blue to colourless.

The assay solution may be acidified with acetic, sulphuric, or hydrochloric acid before titrating with "hypo;" but it must be only faintly so. An excess of acid may be nearly neutralised with ammonia without interference, but excess of alkali is fatal. Bicarbonate of soda must not be used in excess; it is best to avoid it altogether. The assay solution should be titrated at once, as it weakens on standing; and the "hypo" solution should be standardised every two or three days, as its strength is not constant.

The standard solution of hyposulphite of soda is made by dissolving 25 grams of the salt (Na{2}S{2}O{3}.5H{2}O) in water and diluting to 1 litre. 100 c.c. are equivalent to 1.27 gram of iodine.

This solution is standardised by weighing, in a small beaker, about half a gram of iodine, to which is added a crystal or two of potassium iodide and a few drops of water. When dissolved, the solution is diluted to 100 c.c., and titrated in the manner described. The starch solution is made in the manner described under the iodide copper assay. 5 c.c. are used for each titration.

In determining the effects of variations in the condition of the assay a solution of iodine was used, which was equivalent in strength to the "hypo" solution. It was made by dissolving 12.7 grams of iodine with 25 grams of potassium iodide in a little water and diluting to 1 litre. 100 c.c. of this solution were found (at the time of the experiments) to be equivalent to 102.0 c.c. of the "hypo."

Effect of Varying Temperature.—The bulk of the solution was 100 c.c.; 20 c.c. of iodine were taken, and 5 c.c. of starch solution were added towards the end as indicator. These conditions are also those of the other experiments, except where otherwise stated. Iodine being volatile, it is to be expected that with hot solutions low results will be obtained.

Temperature 15 20 40 60 80 "Hypo" required 20.4 c.c. 20.4 c.c. 20.1 c.c. 19.2 c.c. 15.5 c.c.

These show that the temperature should not much exceed 20.

Effect of Exposure of the Iodine Solution.—Twenty c.c. of the iodine were diluted to 100 c.c., and exposed for varying lengths of time in open beakers at the ordinary temperature, and then titrated.

Time exposed — 1 day 2 days 3 days "Hypo" required 20.4 c.c. 16.1 c.c. 13.6 c.c. 9.4 c.c.

Effect of Varying Bulk.—These experiments were carried out in the usual way, bulk only varying.

Bulk 100.0 c.c. 200.0 c.c. 300.0 c.c. 500.0 c.c. "Hypo" required 20.4 " 20.4 " 20.4 " 20.4 "

Effect of Varying Acid.—These experiments were under the usual conditions, the bulk being 100 c.c. The results were—

Acetic acid — 1.5 c.c. 30.0 c.c. "Hypo" required 20.4 c.c. 20.7 " 20.7 "

Hydrochloric acid — 1.5 c.c. 15.0 c.c. "Hypo" required 20.4 c.c. 20.6 " 20.9 "

Sulphuric acid — 0.5 c.c. 20.0 c.c. "Hypo" required 20.4 c.c. 20.7 " 15.2 "[86]

Nitric acid — 0.5 c.c. 10.0 c.c. "Hypo" required 20.4 c.c. 21.5 " could not be titrated.

In the application of this titration to the assay of manganese ores, hydrochloric and hydriodic acids are the only ones likely to be present.

Effect of Alkalies.—On theoretical grounds the presence of these is known to be inadmissible. A solution rendered faintly alkaline with ammonia required only 11.2 c.c. of "hypo;" and another, with 0.5 gram of caustic soda, required 4.0 c.c. instead of 20.4 c.c. as in neutral solutions.

Effect of nearly Neutralising Hydrochloric Acid Solutions with Ammonia.—Provided care is taken not to add excess of ammonia, this has a good effect, counteracting the interference of excess of acid. Thus 20 c.c. of iodine (as before) required 20.4 c.c. of "hypo;" with 15 c.c. of hydrochloric acid 20.7 c.c. were required, but with 15 c.c. of acid, nearly neutralised with dilute ammonia 20.4 c.c. were used.

Effect of the Addition of Starch.—The addition of varying quantities of starch has no effect, provided it is added when the titration is nearly finished, as the following experiments show:—

Starch added 1.0 c.c. 5.0 c.c. 10.0 c.c. 50.0 c.c. "Hypo" required 20.4 " 20.4 " 20.4 " 20.5 "

But if the starch is added before the titration, the results are liable to error.

Starch added 1.0 c.c. 50.0 c.c. "Hypo" required 20.4 " 24.0 "

The starch should be used fresh, and is best made on the day it is used; after four days the finishing point is not so good.

Effect of Varying Potassium Iodide.—An excess of iodide is always required to keep the iodine in solution; a larger excess has little effect.

Iodide added — 1 gram 20 grams "Hypo" required 20.4 c.c. 20.5 c.c. 20.6 c.c.

The 20 c.c. of iodine used, itself contained 0.5 gram of potassium iodide.

Effect of Foreign Salts.—

Bicarbonate of soda added — 0.5 gram 1.5 gram 5.0 grams "Hypo" required 20.4 c.c. 18.2 c.c. 17.1 c.c. 16.0 c.c.

The solution obviously must be free from bicarbonate of soda. This should be remembered, since when titrating arsenic assays with iodine it must be present; and students must avoid confounding the two titrations.

In some other experiments, in which 10 grams each of the salts were taken, the following results were obtained:—

Salt added — AmCl AmNO_{3} Am_{2}SO_{4} "Hypo" required 20.4 c.c. 20.5 c.c. 20.3 c.c. 20.2 c.c.

Salt added NaCl NaNO_{3} Na_{2}SO_{4} "Hypo" required 20.3 c.c. 20.4 c.c. 20.4 c.c.

Effect of Varying Iodine.—

Iodine added 1.0 c.c. 10.0 c.c. 20.0 c.c. 50.0 c.c. 100.0 c.c. "Hypo" required 1.3 " 10.2 " 20.4 " 51.0 " 102.0 "

Determination of Dioxide in a Manganese Ore.—Weigh up 0.25 to 0.3 gram of the powdered ore; place in a flask, cover with 10 c.c. of hydrochloric acid, and close the flask with a paraffined cork, and bulbs (as shown in fig. 60), having previously charged the bulb with 5 grams of potassium iodide in strong solution. Heat the flask, and boil cautiously for about fifteen minutes. Wash the contents of the bulbs into a large beaker, nearly (but not quite) neutralise with dilute ammonia, and titrate with the standard "hypo."

As an example, 0.2675 gram of pyrolusite was taken, and required 60.3 c.c. of standard "hypo" (100 c.c. equal 1.185 gram iodine, or 0.4042 gram MnO_{2}), which equals 0.2437 gram of the dioxide or 91.1 per cent.

COLORIMETRIC METHOD.

When compounds of manganese free from chlorides are boiled with nitric acid and dioxide of lead,[87] the manganese is converted into permanganic acid, which is soluble and tints the solution violet. The depth of colour depends on the amount of manganese present, and this should not much exceed 10 milligrams. A quantity of substance containing not more than this amount of manganese should be boiled for a few minutes with 25 c.c. of a solution containing 5 c.c. of nitric acid, and 10 or 20 c.c. of dilute sulphuric acid, with 2 or 3 grams of lead dioxide. Filter through asbestos, wash by decantation with dilute sulphuric acid, make up with distilled water[88] to a definite bulk, and take a measured portion for the colorimetric determination.

The standard solution of manganese is made by dissolving 0.1435 gram of permanganate of potash (KMnO_{4}) in a little water acidulated with nitric acid, and diluting to 1 litre. One c.c. will contain 0.05 milligram of manganese.

PRACTICAL EXERCISES.

1. What percentage of manganese (Mn) is contained in permanganate of potash (KMnO_{4})?

2. Ten c.c. of a solution of permanganate of potash is found to oxidise 10 c.c. of an acid solution of ferrous sulphate. The manganese is determined in the titrated solution by precipitation as dioxide and titrating. How much of the ferrous solution will be oxidised in the second titration?

3. What weight of potassium iodide would be just sufficient to absorb the chlorine evolved by 0.5 gram of pure dioxide of manganese?

4. What weight of iron must be dissolved up so as to have an excess of 0.25 gram after oxidation by 1 gram of pure dioxide?

5. What weight of the brown oxide, Mn{2}O{4} will be left on igniting 1 gram of the pure dioxide?

CHROMIUM.

Chromium occurs in nature chiefly as chromite or chrome iron ore (FeO{2}Cr{2}O{3}, with more or less MgO and Al{2}O{3}), which is the chief ore. It is a constituent of some silicates, and is frequently met with in very small quantities in iron ores. It occurs as chromate in crocoisite (PbCrO{4}), and some other rare minerals.

The metal is used in steel-making. Steel containing about 0.5 per cent. of it is rendered very hard; but its chief value is in its salts, the chromates. These are highly-coloured compounds, generally red or yellow. Some of the insoluble chromates are used as pigments; chromate of lead or chrome-yellow is the most important. The soluble chromates, those of soda and potash, are valuable chemicals, and are largely used in the preparation of pigments, dyeing and tanning, and as oxidising agents.

Chromium forms two important classes of compounds—chromic salts, corresponding to the oxide Cr{2}O{3}, and chromates, which contain the trioxide CrO{3}. Solutions of chromic salts are green, whilst those of the chromates are yellow. Chromates are reduced to chromic salts by the action of most reducing agents in the presence of an acid; and this property is used in assaying for the volumetric determination of ferrous iron, &c. The chromates in solution are more stable than other similar oxidising agents, and consequently are generally used in the laboratory as one of the standard oxidising agents for volumetric analysis. They have the disadvantage of requiring an outside indicator. Bichromate of potash (K{2}Cr{2}O{7}) is the salt generally used for this purpose.

Chromic salts are oxidised to chromate by fusion with "fusion mixture" and nitre, or by treating with chlorine in an alkaline solution.

Chromic salts closely resemble those of ferric iron, and in the ordinary course of analysis chromic hydrate (green) is precipitated together with ferric hydrate, alumina, &c., on the addition of ammonic chloride and ammonia. The ignited oxide, Cr{2}O{3}, however, is not reduced on heating to redness in a current of hydrogen.

Detection.—Chromium is detected by fusing the powdered substance with "fusion mixture" and nitre. The melt is extracted with water and filtered. The filtrate is acidified with acetic acid, and treated with a few drops of a solution of lead acetate. A yellow precipitate indicates chromium. Substances containing chromium impart a green colour to the borax bead in both flames. Small quantities of chromate in neutral solution can be found by the dark or violet-red colouration imparted thereto on boiling with a dilute decoction of logwood.

Solution and Separation.—Chromates and chromic salts are generally soluble in water or dilute acids. Chrome iron ore, however, and ignited chromic oxide are insoluble; and the former presents considerable difficulty on attempting to open up by the usual methods. A large number of mixtures have been tried in order to get all the chromium in a soluble form. Among these are the following. One part of the very finely-powdered ore is fused with any of these mixtures.

(1) 10 parts of bisulphate of potash. (2) 5 parts of bisulphate of potash and 5 parts of potassium fluoride. (3) 5 parts of hydric potassic fluoride. (4) 12 parts of bisulphate of potash; and, afterwards, with 6 parts of carbonate of soda and 6 parts of nitre. (5) 8 parts of borax; afterwards, with carbonate of soda till it ceases to effervesce; then, with 3 parts of carbonate of soda and 3 of nitre. (6) 4 parts of borax and 6 parts of fusion mixture. (7) 12 parts of caustic potash. (8) 10 parts of caustic soda and 30 of magnesia. (9) 5 parts of caustic soda and 3 of magnesia. (10) 2 parts of carbonate of soda and 1 of lime. (11) 6 parts of soda-lime and 2 of chlorate of potash. (12) Sodium peroxide.

Of these, numbers 1, 2, and 3 yield the chromium in a form soluble in dilute acids, as chromic salt. The rest in a form soluble in water, as potassium or sodium chromate.

On boiling an insoluble chromium compound with chlorate of potash and nitric acid, the chromium passes into solution as chromate. This method, however, does not answer for chrome iron ore. In the fusion methods the ore must be very finely powdered, well mixed with the fluxes, and subjected to a prolonged fusion in a platinum vessel at a high temperature. Undecomposed particles require re-fusion.

The aqueous extract containing the chromate is ready for volumetric work, except in those cases where nitre has been used. For gravimetric work the solution is acidified with hydrochloric acid, then mixed with ammonia in slight excess, boiled, and filtered. The filtrate is acidified with hydrochloric acid, and treated with sulphuretted hydrogen, warmed, rendered slightly alkaline with ammonia, and the gas again passed. The chromium is precipitated as chromic hydrate mixed with sulphur from the reduction with sulphuretted hydrogen. It is filtered off, washed with hot water, and ignited. It is weighed as chromic oxide.

GRAVIMETRIC DETERMINATION.

The solution containing the chromium, freed from other metals and earths and in the form of (green) chromic salt, is heated to boiling. If any chromate is present reduce it with sodium sulphite or sulphuretted hydrogen. Add ammonia in slight excess, boil till the liquid is free from a red tint, and allow to settle for a few minutes. Filter, wash with hot water, dry, and ignite strongly in a loosely-covered crucible. Cool, and weigh. The substance is chromic oxide, Cr{2}O{3}, and contains 68.62 per cent. of chromium. It is a dark-green powder insoluble in acids.

When, as is generally the case, the chromium exists altogether as chromate (phosphates and arsenates being absent) it is best to proceed as follows:—Render the solution acid with acetic acid, then add sodium acetate to the solution and heat nearly to boiling; next treat with a slight excess of acetate of lead, and boil. Allow to settle, and filter. Wash the precipitate with hot water, dry in the water-oven or at a low temperature. Transfer the precipitate to a weighed Berlin crucible, burn the filter separately, ignite below redness, cool in the desiccator, and weigh. The substance is lead chromate, PbCrO_{4}, and contains 16.1 per cent. of chromium, or 23.53 per cent. of chromic oxide (Cr_{2}O_{3}).

VOLUMETRIC METHOD.

This is based on the oxidation of ferrous iron by the solution containing the chromium as chromate. A known weight of iron (0.5, 1, or 1.5 gram, according to the quantity of chromate) is dissolved in 50 c.c. of dilute sulphuric acid. The solution containing the chromate is added, and the remaining ferrous iron titrated with the permanganate or bichromate of potassium solution, as described under Iron. The iron thus found is deducted from that taken, and the difference gives the iron oxidised by the chromate. This multiplied by 0.3101 gives the chromium, Cr, and when multiplied by 0.4529 gives the chromic oxide, Cr{2}O{3}.

COLORIMETRIC METHOD.

Small quantities of chromium may be determined, after conversion into chromate, colorimetrically. The solution, which should not contain more than a few milligrams in 100 c.c., is acidified with acetic acid and compared against an equal volume of water rendered acid with acetic acid and tinted with a standard bichromate of potassium solution. This standard bichromate is made by dissolving 2.827 grams of the salt in water and diluting to 1 litre. One c.c. will contain 1 milligram of chromium, Cr. The manner of working this assay is the same as that adopted in the other colorimetric processes.

Determination of Chromium in Steel.[89]—Weigh up 2.4 grams, dissolve in hydrochloric acid, and evaporate to dryness. Fuse with sodium carbonate and nitre, extract with water, and make up to 301 c.c. Take 250 c.c. of the clear liquor, boil with hydrochloric acid, add sodium phosphate, and then ammonia in slight excess. Heat till clear. Filter off the precipitate, dissolve it in hydrochloric acid, and evaporate to dryness. Take up with a little acid, filter, and precipitate with a slight excess of ammonia. Wash, ignite, and weigh as chromium phosphate (3Cr{2}O{3},2P{2}O{5}), which contains 42.2 per cent. of chromium.

VANADIUM.

Vanadium occurs in certain rare minerals, such as vanadinite (3Pb{3}(VO{4}){2}.PbCl{2}), a vanadate of lead; mottramite, a vanadate of copper and lead; and dechenite, a vanadate of lead and zinc. It is occasionally found in iron and copper ores and in the slags from them. In Spanish copper-precipitates it is found along with chromium, and is probably derived from the iron used for precipitating. The vanadates, like the chromates, are coloured compounds, generally yellow or red. On reduction, blue solutions are got. In their general reactions vanadates resemble phosphates.

Vanadium is detected by the red colouration produced by passing sulphuretted hydrogen into ammoniacal solutions for some time. On adding an acid to the filtered solution a brown precipitate of the sulphide is produced. This gives with borax a colourless bead in the oxidising, and a green one in the reducing, flame.

It is separated by fusing the ore with potassic nitrate, extracting with water and precipitating with baric chloride. The precipitate is boiled with dilute sulphuric acid, filtered, neutralised with ammonia, and saturated with ammonic chloride. Ammonium vanadate separates out. It is filtered off, ignited, and weighed as vanadic oxide, V{2}O{5}, containing 56.18 per cent. of vanadium.

MOLYBDENUM.

Molybdenum occurs in nature chiefly as molybdenite (MoS_{2}); it also occurs in wulfenite, a molybdate of lead (PbMoO_{4}), and in molybdic ochre (MoO_{3}).

Molybdate of ammonia is an important reagent in the determination of phosphates, the manufacture of which compound is the chief purpose to which molybdenum is applied.

Iron and copper ores frequently contain molybdenum, sometimes in quantity; consequently it is met with in slags and pig-iron.

Molybdenum forms several series of salts. In those corresponding to the lower oxides it is basic; but the trioxide (MoO_{3}) is the acid oxide which forms a series of salts called the molybdates. All molybdenum compounds are converted into the trioxide by boiling with nitric acid. The trioxide is a white powder readily dissolved by ammonia. It fuses at a red heat, and volatilises freely in contact with air. It is slightly soluble in water.

Molybdates are easily reduced, with the production of coloured solutions, by most reducing agents. Sulphuretted hydrogen first produces a blue tint, and then precipitates a brown sulphide. The precipitation as sulphide is only complete on prolonged treatment; a green colour indicates that some molybdenum still remains in solution. The precipitated sulphide is soluble in ammonium sulphide.

Detection.—Molybdenum is detected by its behaviour with sulphuretted hydrogen. Molybdenite can only be mistaken for graphite, from which it is easily distinguished by yielding sulphur dioxide on roasting, and by giving, on charcoal, a yellowish white incrustation, which becomes blue on touching it for a moment with the reducing flame. The borax-bead is colourless in the oxidising, and dark-brown in the reducing, flame.