According to this hypothesis, therefore, the addition of common salt in the calcination of colcothar is less for the purpose of maintaining a uniform temperature within certain limits than for decomposing the basic ferric sulphate present and inducing the formation of a product consisting entirely of pure ferric oxide. The various tones obtained are due to the varying length of exposure to the heat. The following method is pursued in the conversion of colcothar into iron pigments on a manufacturing scale. The crude colcothar from the sulphuric acid plant is ground, as finely as possible, in ordinary mills, and the resulting soft powder is intimately mixed with salt, 2, 4 or 6% being the usual proportions added. The calcination is ordinarily continued for six hours in the case of the mixture containing the largest amount of salt; but only two hours, or even one, for the other mixtures. The operation is carried on in earthenware pipes, a large number of which (up to sixty) are built into a furnace. The latter is fired very carefully, the temperature being raised only very gradually, since experience has shown that much better coloured products are obtained in this way than by raising the mass quickly to a high temperature. When incandescent ferric oxide is allowed to cool down with unrestricted access of air, the colour is not nearly so bright as when air is excluded during the cooling. Since air has no action on ferric oxide, this remarkable phenomenon cannot be due to the presence of the air, but probably to the influence exerted by the rapid change of temperature on the arrangement of the finest particles of the oxide. Nevertheless, some manufacturers hold that rapid cooling, with restricted access of air, improves the colour. To exclude air from the ferric oxide during calcination, the open ends of the pipes are flanged and covered with close-fitting plates, which are luted with clay. The expansion of the internal air as it grows hot would burst the pipes unless a means of escape were provided, which consists in leaving small vent holes in the cover plates. As previously mentioned, calcined ferric oxide is very inert, chemically, so that, when the calcination has been strong, prolonged boiling with the most powerful acids is needed to bring the oxide into solution. If the heating has been continued up to the strongest white heat, and the ferric oxide maintained in that condition for several hours, even hot sulphuric acid will have only a slight effect on the oxide, and the only way to make it more readily soluble is by fusion with potassium bisulphate. Now indifference to chemical action is just the property required of a pigment for fine work; and in this respect, the ferric oxide colours are superior to all others. The gradations of tone that can be obtained from ferric oxide by varying the calcination are very numerous, comprising all between iron red, red-brown and pure violet.