potassa, was recovered in the usual manner, by evaporation with hydrochloric acid; the residue was digested with water collected, washed, dried, and weighed. The amount of the silicon in the iron was calculated from the silicic acid obtained.

MANGANESE. The hydrochloric acid solution, separated from the silicic acid and graphite, was divided into two equal portions, one of which, representing 50 grains of iron, was always sufficient for the estimation of the manganese. The iron in the liquid having been per-oxidised by boiling the hydrochloric acid solution, and adding occasionally a little chlorate of potassa, the acid was to a great extent neutralised by addition of carbonate of ammonia. Sufficient acetate of ammonia was afterwards added for the conversion of the chloride of iron into acetate, and the liquid was boiled, when the iron was completely separated as insoluble basic acetate. The filtrate containing the manganese was rendered alkaline with ammonia, and, after the addition of a few drops of bromine, set aside for about eighteen hours. The hydrated binoxide of manganese which had separated from the liquid, was afterwards collected, washed, dried, and ignited at a high temperature, when it was weighed as manganoso-manganic oxide (Mn,O,), which furnished, by calculation, the quantity of manganese.

'PHOSPHORUS. For the estimation of phosphorus, 50 grains of the iron borings were acted upon with warm nitrohydrochloric acid, in a flask with a long neck, and, after complete solution of the metal, the contents of the flask were transferred to a porcelain basin and evaporated to dryness; the residue was moistened with concentrated hydrochloric acid, and again evaporated, so as thoroughly to expel nitric acid. The residue then obtained was dissolved in hydrochloric acid, the solution diluted, filtered, nearly neutralised with carbonate of ammonia, and the iron in solution reduced to protoxide by the addition of sulphite of ammonia to the gently heated liquid, and the subsequent careful addition of dilute sulphuric acid to expel excess of sulphurous acid. Acetate of ammonia and a few drops of

solution of sesquichloride of iron were then added, and the liquid boiled, when the phosphoric acid was precipitated as basic phosphate of sesquioxide of iron, with some basic acetate. The liquid was rapidly filtered, with as little exposure to the air as possible, the precipitate was slightly washed and dissolved in hydrochloric acid, the solution neutralised with carbonate of ammonia, and a mixture of ammonia and sulphide of ammonia added; it was then gently heated, to ensure the conversion of the phosphate into sulphide of iron. The latter was afterwards removed by filtration, washed with dilute sulphide of ammonium, and the phosphoric acid was precipitated from the solution in the usual manner as ammonio-phosphate of magnesia, and weighed as pyro-phosphate of magnesia, from the amount of which the phosphorus was calculated.

COMBINED CARBON.-After numerous comparative trials of the several methods in common use for determining the total amount of carbon in cast iron, that which was ultimately adopted (after necessary experiments had fully established its accuracy) consisted in dissolving the metal in an acid solution of chloride of copper, collecting and washing the insoluble residue which remained after the complete action of this solvent, and submitting it, when dry, to combustion with oxide of copper in a current of oxygen, the source of heat employed being the gas combustion furnace. The total amount of carbon in the iron was then calculated from the weight of carbonic acid absorbed by solution of potassa in the usual manner. The carbon, existing in a state of combination with the iron, was represented by the excess which this process afforded over that of the direct estimation of the carbon as graphite, in the manner already described.

'MINUTE PROPORTIONS OF FOREIGN METALS.-About 400 or 500 grains of the iron were employed in the examination for metals precipitated by sulphuretted hydrogen, e.g., lead, copper, arsenic, &c. The iron was dissolved in hydrochloric acid, and the solution, diluted and partly neutralised with carbonate of soda, was submitted to the action of sulphur

etted hydrogen. After saturation with the gas, the liquid was allowed to stand at rest for several hours, and the small quantity of sediment which had subsided was examined for metals by the ordinary analytical processes.

II.—Analysis of the Iron Ores.

'The analytical processes employed for the separation of the various constituents occurring in iron ores were, in great measure, identical with those employed in the examination of metallic iron. Thus, the estimation of oxide of manganese was conducted in a precisely similar manner; and, with the exception that no process of reduction was required in the case of clay ironstones and other ores containing the iron already in a state of protoxide, the phosphoric acid was determined by the same process as that employed for the estimation of phosphorus in pig iron. The amount of metallic iron, and its condition of oxidation in the ore, were determined by Marguerite's volumetrical method, with standard solution of permanganate of potassa; while the proportions of lime and magnesia, carbonic acid, water, hygroscopic and combined, insoluble residue, and the nature of this latter, were determined by following the analytical processes invariably employed in mineral analyses of this description.

'Sulphur was estimated by fusion of the ore (or, in the case of clay ironstone of the clay only) with a mixture of pure carbonate of soda and nitre; the sulphuric acid being precipitated by chloride of barium, from the acidified solution of the fused mass, and the sulphate of baryta collected, burnt, and weighed as usual. The hydrochloric acid solution of the ironstone was examined for sulphuric acid, but it was seldom that more than a trace of sulphur was detected in that form.

III. Analysis of the Samples of Fluxes.

'These materials, consisting of limestone, burnt shale, &c. were analysed by a method precisely similar to that em

ployed in the examination of the ores. In the tabulated statement showing the composition of the limestones, the amount of carbonate of lime is, in some few instances, represented by the difference, after the whole of the other constituents had been determined. In such cases the sumtotal of constituents is necessarily expressed by 100 exactly.'

B. THE ASSAY OF IRON IN THE WET WAY.

Fuchs' Method.-A suitable quantity (10 grs.) of the finely pulverised iron-ore (pig-iron, &c.), is dissolved in an excess of concentrated hydrochloric acid, and the resulting protochloride of iron is changed into sesquichloride of iron, by the addition of chlorate of potash; the chlorine is then to be expelled by heating the liquid. The latter is diluted with water, and perfectly clean strips of electrotyped copper (15-20 grs.) previously exactly weighed, are put into it. These operations must be done in a glass flask, tightly closed by a cork, through which a narrow glass tube passes. The liquid is now heated to the boiling-point, and kept at this temperature till the original dark-brown or yellow colour of the liquid changes to a light-yellow green or light blue-green colour. Then all the sesquichloride of iron will be reduced to protochloride in consequence of the formation of subchloride of copper. The open part of the glass-tube is now hermetically closed, and the liquid allowed to cool. The flask is then filled with hot water, and the liquid decanted from the undissolved copper. The latter is washed, first with diluted hydrochloric acid, and afterwards repeatedly with water; it is then dried and weighed. The amount of iron contained in the assayed substance may be calculated by the loss of copper, that amount of iron standing in the same proportion to the dissolved quantity of copper as the equivalent of iron (28) to the equivalent of copper (317). 28 parts of iron correspond to 36 parts of protoxide of iron and to 40 parts peroxide

of iron.

Some iron-ores are very difficultly or only incompletely

dissolved in hydrochloric acid; they must then be previously fused with carbonate of soda and potash.

The above method is based upon the fact that chemically pure hydrochloric acid is not able to dissolve copper without the presence of atmospheric air, and that this acid dissolves copper if peroxide of iron is present. The quantity of copper dissolved is then proportional to the amount of peroxide of iron present. In this In this process protochloride of iron and subchloride of copper are formed, FeCl3+2 Cu 2FeCl + Cu,Cl. Two equivalents of copper (63,4 parts) therefore reduce 1 equivalent of peroxide of iron (80 parts); the latter containing 2 equivalents of iron (56 parts).

=

In case the iron ore contains arsenic, the latter must be previously removed by fusing the iron ore with 3 parts dried carbonate of soda, and by lixiviating the fused mass.

If the iron ore contains titanium it will be necessary to decompose the chloride of iron by copper at a lower temperature than the boiling-point.

If the iron ore contains peroxide of iron together with protoxide of iron, the same process may be used to determine the quantities of both combinations. The process is then to be performed twice, once before oxidation and the second time after it. The first operation will show the quantity of peroxide of iron which the iron-ore originally contained, as the copper only acts upon peroxide of iron, and not upon protoxide of iron.

Marguerite's Process.-This method of analysis is based on the reciprocal action of salts of the protoxide of iron and mineral chameleon (permanganate of potash-KO, Mn,O;), whereby a quantity of the mineral chameleon is decomposed exactly proportionate to the quantity of iron.

Thus, in any given solution of iron at its maximum of oxidation, such as it more commonly exists in the mineral it is only necessary to bring it to the minimum of oxidation, and then to add gradually a solution of permanganate of potash of a known strength. As long as a trace of protoxide remains to be peroxidised the colour of the chameleon is destroyed; but it is at length noticed that the colour of