Manganese, of which the salts of this type are very unstable, has very slight tanning power, while titanium, which in many ways is allied to the group, though it does not strictly belong to it, has recently been patented as a tanning agent. For the present, however, we may limit our attention to the three metals first named. Alumina and its salts demand the first attention, not only as having been used for leather manufacture in very early times, but as being still important commercially. The metal aluminium is now well known, and its oxide, alumina, Al₂O₃ is abundant in nature, combined with silica in the form of clay and bauxite, as fluoride in combination with sodium fluoride in cryolite, and in some cases as a native sulphate. Alum-shale, which was formerly the principal source of alum, is a bituminous clay containing much iron sulphide, and which when calcined yields aluminium sulphate. As aluminium sulphate does not crystallise readily, and was difficult to free from iron, potassium sulphate was added to the liquor obtained by leaching the calcined shale, from which, after concentration by boiling, potash-alum, a double sulphate of potassium and aluminium, Al₂(SO₄)₃,K₂SO₄,24Aq, was easily crystallised out. Alum is now usually made by decomposing clay or bauxite with sulphuric acid, and ammonium sulphate is generally substituted for the potassium salt, yielding ammonia-alum, a double sulphate of aluminium and ammonium of similar constitution to potash-alum. Ammonium alum is easily distinguished from the potassium salt, by the strong smell of ammonia which it evolves on the addition of caustic soda or lime. So far as is known, there is no practical difference in tanning effect between the two salts, and ammonium alum is cheaper, and slightly stronger, its molecular weight being 906, as against 948 for the potassium salt. Either alum dissolves readily in cold water to the extent of about nine parts in 100 of water, and more easily, and to a much larger extent in hot water, from which the excess crystallises on cooling. It is said that for purposes of leather manufacture, alum solutions should not be boiled, and, though it is improbable that this produces any change, it must be remembered that chrome alum on boiling really does undergo decomposition to free acid and a more basic salt, indicated by change of colour from violet to green, from which it slowly returns to the violet form on cooling.