The ammonic carbonate filtrate from the silica is evaporated to dryness, ignited strongly over the blowpipe, and weighed. The residue is tungstic acid, WO3. The tungsten may be conveniently reported in this form, although it is probably present as a lower oxide.

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:—