181. According to the chemical view of the production of chlorine from salt and manganese, under the action of sulphuric acid, 26.8 pounds of pure salt is sufficient for 20 pounds of pure manganese binoxide. But, in this connection, two points must be considered. Firstly, the commercial salt and manganese are never pure, the former not often containing more than 90 per cent of sodium chloride, while the latter is consid ered to be a good article if it contains 70 per cent binoxide. Secondly, the theoretical reaction is not perfectly realized in practice.

If the proportion of salt is too small, as compared with that of the manganese and acid used, the latter two will react, giving off oxygen, and forming manganese sulphate, so that the subsequent addition of more salt will not repair the error. A mistake in the direction of an excess of salt can be rectified by an addition of acid and manganese, but as salt is the cheapest of the three substances, this is less important. In the directions given (94) for making chlorine, the proportions are those in common use, and the salt is intended to be somewhat in excess of the other substances.

In case the quality of the salt or manganese should differ much from the assumed average, there might be a considerable waste of one or the other, whence it seems desirable that the reader should be informed as to the manner in which these substances may be assayed, in order that they may be proportioned correctly.

Assay Of Salt

182. Assay Of Salt. Take a fair sample of the salt to be assayed. It must not be dried, because what is required is its value for the production of chlorine in the condition in which it is to be weighed for working, but, if too coarse, it should be crushed. Mix thoroughly. Dissolve 10 grains in about 50 grains, or rather less than 1/8 fluid-ounce, of distilled or rain water, or at least water which gives no precipitate on addition of a drop of solution of silver in nitric acid. Filter the solution through a small wetted filter of fine bibulous paper. The filter must be well washed with water to completely remove the salt solution, and the filtrate and washings received in a flask.

To the filtered solution in the flask add, drop by drop, a solution of silver nitrate until no further precipitation of silver chloride takes place. Warm slightly, and shake vigorously, stopping the mouth of the flask with a greased cork, until the precipitate is found to settle readily, leaving the liquid clear. Now transfer to a small filter, the weight of which has been previously ascertained after drying on a water bath until it no longer lost weight. Wash the precipitate and the filter with hot water until the washings are tasteless. Dry the filter, containing the precipitate, on a water bath until two successive weighings give the same result. From the weight obtained deduct that of the filter; the remainder will be the weight of silver chloride produced by 10 grains of the salt.

If the salt were pure and dry, 10 grains would produce 24.53 grains of silver chloride, but if we say 24.5 grains it will be sufficiently accurate. Multiply the weight of silver chloride found by 100, and divide by 24.5; the result will be the percentage of pure salt in the sample. Example: - The silver chloride obtained from 10 grains of the sample weighs net 19.3 grains; then the sample contains 1930%24.5 = 78.77 per cent of pure salt.

The silver chloride should be dried without much exposure to light. For a water bath, the cover of a vessel in which some water is boiling answers very well. The pulp scales used in assaying ore are sufficiently accurate for this work, if sensitive to 1/10 grain.

183. The salt sometimes contains a considerable quantity of sodium carbonate, which neutralizes an equivalent quantity of acid. In the interior of the country, where this is most liable to occur, acid is very expensive on account of the cost of freight. Such salt should therefore be purified, by dissolving it in water, and exposing the solution to evaporation in shallow vats, when the purified salt will be deposited.