Another defect of the ochres prepared by this method resides in the excess of lime present, it being impracticable to measure out the quantity of lime used with such accuracy that only just enough is taken to precipitate the ferrous hydroxide, there being always a slight excess. This lime is transformed into calcium carbonate on the mass being exposed to the air, just as in the preparation of Vienna white; but as the saturation with carbon dioxide takes a considerable time, some of the lime remains in the caustic state and is liable to affect other colours that may be mixed with the ochre. An artificial ochre uniting in itself all the qualities of the natural product, and also capable of being shaded by burning, can be prepared in the following manner. An accurately weighed quantity of pure crystallised ferrous sulphate is dissolved in a definite amount of water, and the solution is treated with successive small portions of crude nitric acid, until all the ferrous oxide has been changed into the ferric state. The change can be detected by a very decisive test. If a liquid containing ferric oxide in solution is brought into contact with a solution of red prussiate of potash (potassium ferricyanide), no precipitate is formed in the absence of ferrous oxide, but only a brown coloration; whereas, if ferrous oxide is present, a beautiful blue precipitate is formed at once, the colour of which is so intense that very small quantities of ferrous oxide can be detected by this means. For the purpose now under consideration, the presence of small amounts of ferrous oxide in the solution is immaterial, because they are soon changed into ferric oxide on exposure to the air. It might, therefore, be asked, why take the trouble to oxidise the ferrous oxide by means of an agent involving expense, which could be saved by allowing the oxidation to take place in the air? The advantage, however, of the direct employment of a solution of ferric oxide is that it gives at once a colour that can be dried straight away; whilst at the same time the colour undergoes no change in drying, whereas it does when ferrous oxide solution is used. The method of producing ochres from this ferric solution varies according as the product is to be used without any further treatment than drying, or is to be modified by firing. In the former event, caustic lime is again used as the precipitant, but in only just sufficient quantity to throw down all the ferric oxide in the solution. This amount can be calculated exactly, 36·84 parts by weight of pure burnt lime being required for every 100 parts of pure ferrous sulphate taken. The actual quantity, whether larger or smaller, will depend on the relative purity of the sulphate and lime; and this can readily be ascertained by a simple trial.