Another method which is frequently used to brighten the colour of extracts, is treatment with sulphurous acid. Dilute sulphurous acid solution may be used for extraction, but a more common method is to pass sulphur dioxide gas into the liquor before concentration. Sulphurous acid acts partly as a weak acid, in decomposing compounds of the tannins and colouring matters with bases, such as lime, iron, copper, but more actively by reducing oxygen compounds and preventing oxidation. Bleaching in this way does not actually destroy or remove the colouring matters, which are apt to reappear on exposure to the air, either in the liquor, or perhaps more often in the leather tanned with it, so that the gain is frequently more apparent than real. If present in any considerable quantities, sulphurous acid may also cause inconvenience by its swelling action on the pelt, but is mostly expelled in concentration. Another process should perhaps also be mentioned here, though not strictly a means of bleaching. Several tanning materials, and notably quebracho and hemlock, contain large quantities of “difficultly soluble tannins,” which render the liquors made from their extracts turbid on cooling. These tannins form soluble compounds with alkalis and with alkaline sulphites, in the latter case probably setting free the sulphurous acid and combining with the base. This has been taken advantage of in a recent patent[158] in which quebracho and other extracts are rendered soluble by heating in closed vessels with bisulphites, sulphites, sulphides, or even caustic alkalis; and many “soluble quebracho extracts” made on this principle are now on the market. In this case, even where bisulphites are used, the greater part of the sulphurous acid, after serving its purpose in preventing oxidation, escapes in course of manufacture, and the extracts remain neutral or alkaline. There is no reason that such extracts should not prove serviceable in tanning, but it has recently been shown by Paessler that the alkaline tannin is not absorbed by neutral hide-powder, and it therefore may lead, not only to discrepancies in analysis, but in case of drum-tannage, where no acid is naturally present, to failure to utilise the whole of the tannin, though, when added to ordinary liquors, the acids contained in the latter will set free the tannins. [158] Lepetit, Dollfus, and Gansser, Eng. Pat. 8582, 1896.