occur and causo error) being thus avoided ; whilst they also find that a modification of a method originally due to Guyard (precipita tion of manganese as Mn0 2 by the addition of permanganate) will give good results provided that a zinc salt be added to ensure the formation of Mii0 2 only, and that the amount of free acid be not too great. When it is required to determine the alumina dissolved by tho acid employed to act on the original ore, the phosphoric acid in tho total precipitate thrown down by the acetate treatmentfor the estima tion of manganese gravimetrically may be determined ; subtracting this and the Fe 2 3 from the weight of tho precipitate, the A1 2 3 is approximately known ; or the alumina may be separated by other processes, e.g., use of caustic soda, &c. Sulphuric acid, if present in the ore, is precipitated as barium sulphate from the hydrochloric acid solution of the ore ; sulphur in the form of pyrites is deter mined by fusing the ore with sodium carbonate and nitrate in a gold crucible, and determining the total sulphate formed, the sul phate existing as such in the ore being subtracted. Phosphorus may be determined by dissolving the ore in aqua regia (usually the phosphorus exists as phosphate, and is wholly dissolved by hydrochloric acid), precipitating the phosphoric acid (best after separation of dissolved silica by evaporation to dryness and re-solu tion in dilute acid) in combination with part of the iron, by reduc ing most but not all the iron to the ferrous state, and then precipitating the ferric iron and phosphoric acid by boiling with an acetate ; the precipitate is finally converted into magnesium pyrophosphate by solution in hydrochloric aciJ, addition of citric acid (less conveniently tartaric acid), ammonia, and magnesia liquor, and ignition of the precipitate collected after standing twenty-four hours ; Eggertz s method of determining phosphoric acid is, however, more suitable for the estimation of minute quantities, this depending on the precipitation by molybdic acid of a peculiar yellow crystalline phosphomolybdate of ammonium on bringing together the phosphoric acid solution (from which dissolved silica has been removed by evaporation to dryness) and excess of molybdate of ammonium solu tion supersaturated with nitric acid. Calcium and magnesium are conveniently determined in the filtrate from the basic acetate and phosphate of iron and alumina thrown down in the separation of manganese gravimetrically, the filtrate from the precipitated man ganese dioxide being employed, the calcium being first precipitated as oxalate, and then the magnesium as ammonio-phosphate ; or the ferric oxide and alumina may be thrown down by ammonia free from carbonate, and the filtrate employed. Hygroscopic water and ordi nary moisture are determined by drying at 100, and noting the loss of weight ; whilst combined water is subsequently determined by heating to redness in a tube through which dry air is aspirated, the issuing gases passing through a drying tube to absorb the water evolved ; if nothing but water is lost on ignition, the weight so lost may be directly determined without collecting the water. Titanic oxide, chrome ironstone, complex silicates, &c. , are often contained in the substance left undissolved by acid ; for the modes of determination and analysis of these, and for the precautions in the determination of the soluble constituents should titanium be present and partly dissolved by the acid, &c., the reader is referred to larger treatises, in which also are to be found numerous methods of analysis other than those briefly indicated above. The analysis of iron and steel is carried out on much the same lines as that of iron ores. The metal being dissolved in nitric acid or aqua regia, phosphoric acid is separated as above described, usually by Eggertz s process, the acid solution being previously evaporated to dryness and treated with dilute hydrochloric acid, which leaves behind silica formed from the silicon present, graphite, and slag; after ignition to burn off graphite, the silica is dissolved out by sodium carbonate solution and the residual slag weighed ; in this way, however, more silica is generally obtained than repre sents the silicon originally present, as the silicates of the slag are apt to be more or less attacked by the acid ; a better method for the determination of the slag is to dissolve the iron in bromine or iodine water, or by means of copper chloride (or mixed copper sul phate and ammonium chloride solution), which gives rise to cuprous chloride, dissolved out by heating ; the slag is thus left undissolved, and may be weighed after boiling with sodium carbonate ; the silica in the united sodium carbonate and bromine solutions being deter mined, the silicon is readily calculable. Manganese is determined just as in the case of ores. Sulphur is conveniently determined by treating with hydrochloric acid, and leading the evolved gases through a solution of lead or silver or some analogous metal, and finally converting the precipitated sulphide into barium sulphate ; or by dissolving in aqua regia, evaporating, and converting the sulphuric acid found into barium sulphate : this method usually gives lower values than the others, barium sulphate not precipi tating readily from very dilute highly acid solutions. The so- called "combined" carbon is found by determining the graphite left undissolved during the treatment of the metal with hydro chloric acid (by collection and burning to C0 2 in oxygen, and absorption in caustic potash), and subtracting the amount from the total carbon found by digesting with copper sulphate or chloride, or with bromine water, collection of the undissolved mass on an 281) asbestos filter, and burning in oxygen, preferably with lead chromate iu the front of the tube, to prevent chlorine or bromine vapours, &c. , passing over, should the precipitate contain (through insufficient washing, &c.) substances which may evolve chlorine or bromine. Copper sulphate leaves behind an amount of copper equivalent to the iron dissolved ; tins covers up the iinely divided particles of carbon, and diminishes the chance of pyrophoric oxidation and con sequent loss of carbon during drying, which may otherwise occa sionally take place, especially when the filter is dried by the aid of heat ; but the solution of the iron is less rapid, and it is difficult to see or feel with a glass rod when all the iron is dissolved. Weyl dissolves the iron by making it the positive pole of a weak galvanic current passing through hydrochloric acid. Fresenius determines the "combined" carbon directly by dissolving in hydrochloric acid, passing the evolved hydrogen and carburetted hydrogen over red-hot copper oxide, and determining the carbon dioxide formed by absorption in potash as usual ; if much sulphur is present, lead chromate should be employed to avoid errors due to formation of sulphur dioxide ; if the amount of "combined" carbon is large, liquid non-volatile hydrocarbons are apt to be formed, which causes the method to yield too low a result. Ullgren determines the total carbon by oxidation to C0 2 in the wet way with chromic and sul phuric acids of the residue left after treatment with copper chloride or bromine ; the results are apt to be too low, owing to incomplete oxidation of the graphite. Kegnault determines the total carbon by heating the iinely powdered metal with copper oxide or lead chromate, and absorbing the C0 2 produced by potash ; by passing air over the finely divided metal at a low red heat, and when the oxidation is nearly complete finishing the operation in oxygen at a somewhat higher temperature, the use of copper oxide or lead chromate is rendered unnecessary ; if too high a temperature and oxygen be employed at first, there is risk of forming fusible Fe.,0 4 and of enclosing portions of carbonized unoxidized metal within a coating of that substance, which more or less protects it from the action of the oxygen, and tends to decrease the amount of C0 2 collected. Eggertz determines the "combined" carbon in steel by solution of a known amount of borings or filings in a known amount of nitric acid, and comparison of the fluid as regards its colour with a simi lar solution prepared from steel of ti known carbon percentage, cr with a series of solutions of caramel made so as to exhibit the same tints as those yielded by steels of known carbon percentage when treated in this way. The principle of the method depends on the formation of soluble humus-like carbon compounds by the action of the nitric acid, probably analogous to the vegetable colouring matter of peaty water. Opinions differ widely amongst chemists as to the absolute accuracy of the method for general analytical operations, especially where nothing is known of the precise details of the mode of manufacture of the steel ; but for a works laboratory, where speed is essential, and where it is only required to compare one specimen of steel with another one prepared in the same way but harder (the hardest steels being taken for the preparation of the standards, and the solutions representing the lower carbon percentages being obtained by diluting the fluid proportionately), the method is invaluable. For the sake of saving time under analogous circum stances, Eggertz somewhat modifies the above-described methods for the determination of sulphur and phosphorus, the amount of sulphur present being estimated by noting the discoloration produced on a plate of silver exposed to the gases evolved on solution in hydro chloric acid, or digestion with sulphuric acid, and comparing it with that produced under the same conditions from a metal of a known degree of sulphurization, and the amount of phosphorus being esti mated by transferring the phospho-molybdic precipitate into a narrow measuring tube, and observing the volume occupied by it com paratively with that occupied by the precipitate similarly produced from metal containing a known amount of phosphorus. Sir J. Alleyne has described a method of determining approximately the quantity of phosphorus present in iron and steel by means of tho spectroscope (Journal I. and S. Inst., 1875, 62). Nitrogen is deter mined by solution in hydrochloric acid free from ammonia, and titration by Nessler s test of the ammonia formed (A. H. Allen, Chemical News, xli. 231, 1880). The less commonly occurring substances, copper, chromium, arsenic, cobalt, nickel, zinc, alumi nium, vanadium, titanium, tungsten, molybdenum, &c., are sought for and separated by special methods for which the larger text-books must be consulted. 1 III. EXTRACTION OF IRON FROM ITS ORES. 7. History oftlw Mamifacture of Iron and Steel, Neither the period when malleable iron was first prepared from its ores nor the precise mode of manipulation then adopted is known with certainty, although the remains of iron 1 Parry and Tucker have employed the spectroscope successfully iu this direction, see Journal I. and S. Inst., 1880, 163. XIII. 37
