bisulphate of potassa.* It should be kept in a state of fusion for a long time, without the disengagement of too much sulphuric acid. The mass is then dissolved in water containing a little hydrochloric acid, chloride of ammonium added, and the alumina precipitated by ammonia. In order to free it from any magnesia, the fluid is heated to boiling until no more ammonia is given off. The gelatinous alumina is filtered, and allowed to partially dry upon the funnel when it may be completely washed. It is ignited and weighed. The magnesia is precipitated by phosphate of soda and ammonia. Many specimens of spinel contain a little protoxide of iron and silica.

The red spinel contains sesquioxide of chromium, which may be separated from the alumina as in No. 18.

21. ALUMINA AND SESQUIOXIDE OF IRON.

The mixture of the two is dissolved in hydrochloric acid, the greater part of the excess of acid evaporated, the solution mixed with an excess of pure solution of potassa and heated nearly to the boiling point. The alumina is thus dissolved, the sesquioxide of iron being left behind of a dark brown color. The solution is filtered off, acidulated with hydrochloric acid, and the alumina precipitated by sulphide of ammonium.

The sesquioxide of iron, which contains some potassa, is dissolved in hydrochloric acid, re-precipitated by ammonia, and ignited.

This method of separation is unsafe, and unless repeated more than once, incomplete. It is better to heat

* Prepared by heating equal parts of neutral sulphate and concentrated sulphuric acid to a dull red heat until the the mixture flows quietly.

The same process as in the separation of alumina and lime.

the acid solution to ebullition, to add sulphite of soda, in order to reduce the sesquioxide of iron to the state of proto-sesquioxide, replace the solution over the lamp, boil for some time and then neutralize with carbonate of soda, and afterwards boil with excess of caustic soda until the precipitate is black and pulverulent.

The tendency to bumping preceding the actual ebullition of the fluid, may be guarded against by means of a spiral coil of platinum wire placed in the liquid, or by constant agitation of the latter; when ebullition has once set in, there is no further need of these precautions. Remove the liquid now from the gas, allow to deposit, pass the clear fluid through a filter, which must not be over-porous, boil the precipitate again with a fresh quantity of solution of soda, then wash it by decantation and afterwards on the filter with hot water. Acidify the alkaline filtrate with hydrochloric acid, boil with some chlorate of potassa (to destroy any traces of organic matter), concentrate by evaporation, and precipitate the alumina by sulphide of ammonium or ammonia. The boiling of the precipitated oxides with the solution of soda is best effected in a large silver or platinum dish. The soda must be free from alumina and silica. Or the very dilute solution of both bases may be neutralized with carbonate of soda, mixed with sulphite of soda and heated until no sulphurous acid is given off. All the alumina is precipitated, while the iron remains in solution. The precipitate is ignited. The solution of iron is concentrated, mixed with some chlorate of potassa and hydrochloric acid and heated. After the sulphur has been filtered off the iron is precipitated by ammonia.

The separation may be obtained by placing the mixed precipitate, ignited in a porcelain boat, which is placed in a tube of the same material heated to redness, through which a current of dry hydrogen is

passed, and which it is necessary to maintain until it is completely cool. Then there should be substituted in place of the current of hydrogen, a current of hydrochloric acid gas, which transforms the iron into volatile chloride and leaves the alumina, which is weighed. A complete description of this process of separation of iron and alumina may be found in the article on silicates. The apparatus is arranged as follows: L is a gas furnace upon which is placed a salt bath I. In the bottle H there is placed some fresh chloride of sodium, on which is poured concentrated hydrochloric acid.

Then the bottle H is placed in cold water and sulphuric acid gradually poured upon it, taking care that the mixture does not become heated, and stopping when vapors of hydrochloric acid begin to form. It

is sufficient to heat this mixture to 50° or 60° to evolve a steady current of hydrochloric acid gas.

The amount of sesquioxide of iron may be inferred from the loss of weight, and the result controlled by collecting the chloride of iron which passes over and weighing it.

22. PHOSPHORIC ACID AND SESQUIOXIDE OF IRON.*

In order to separate phosphoric acid from sesquioxide of iron, the compound is ignited with at least an equal weight of carbonate of potassa and soda (No. 10), the resulting mass exhausted with water, the solution supersaturated with hydrochloric acid and then with ammonia, and the phosphoric acid precipitated by sulphate of magnesia.

The residual sesquioxide of iron retains some alkali. Or the sesquioxide of iron containing phosphoric acid is dissolved in hydrochloric acid, precipitated by ammonia, and digested with excess of sulphide of ammonium (without previous filtration), until all the sesquioxide is converted into sulphide of iron. When the liquid is no longer green, but of a pure yellow color, it is filtered off, and the phosphoric acid immediately precipitated by sulphate of magnesia.

For the accurate quantitative separation of a small quantity of phosphoric acid from a large quantity of sesquioxide of iron, the latter is dissolved in hydrochloric acid, and the solution heated to ebullition with sulphite of soda till its color has changed to a bright green, when all the sesquioxide of iron is converted

For analyses for practice, the phosphate of sesquioxide of iron is prepared by precipitating sesquichloride of iron with phosphate of soda. Or a mixture of phosphates may be prepared by precipitating a solution containing sesquichloride of iron, chloride of calcium, chloride of magnesium, and chloride of manganese.

into protoxide. The solution is boiled till it no longer smells of sulphurous acid, neutralized with carbonate of soda, and, in order to produce a little sesquioxide of iron, mixed with a very little chlorine-water, the quantity of which must be regulated according to the amount of phosphoric acid which is present. The solution must now be mixed with an excess of acetate of soda, when phosphate of sesquioxide of iron separates as a white precipitate. Chlorine-water is then added, drop by drop, until the liquid has assumed a reddish color, when it is boiled, so that the precipitate may collect, and be filtered. From this precipitate the phosphoric acid is separated by sulphide of ammonium, as directed above.

Or it may be dissolved in hydrochloric acid, boiled with sulphite of soda, and afterwards with excess of caustic soda, till the precipitate is converted into black proto-sesquioxide of iron, which is filtered off. The solution is acidified, and the phosphoric acid precipi tated as above.

23. HEMATITE, Fe,O,, AND LIMONITE, Fe,О, 3 HO.

For the determination of the water, weighed fragments of the ore are heated to redness, for a long time, in a platinum crucible. If the mineral decrepitates, it must first be finely powdered.

In order to determine the oxygen, the fragments of ignited limonite or of hematite are heated to redness in a weighed bulb-tube-of very infusible glass (the bulb being as small as possible), through which a stream of dry hydrogen, free from arsenic, is transmitted as long as any water is formed.

In order to purify hydrogen it is passed through the U tubes containing pumice or fragments of porcelain