To illustrate the matter, the case of a basic chrome alum liquor will be considered. The chromium hydrate sol is primarily a positive sol, just like ferric and aluminium hydrate sols: _i.e._ in water they are somewhat exceptional in that they adsorb H+ rather than OH-. To cause precipitation therefore it is necessary to make the sol less positive and more negative. The positive charge of the sol, however, is greater than in water, because of the free acid formed in the hydrolysis, which results in the adsorption of more hydrions by the sol. Hence to ensure precipitation steps must be taken to reduce the adsorption of hydrions by the chromium hydrate sol. In practice such steps are taken, and to such an extent that there can be little doubt that the chrome sol is not far from its isoelectric point. Amongst these "steps" are (1) making the liquor "basic," _i.e._ adding alkali to neutralize much of the free acid, which involves a considerable reduction in the stabilizing effect of the hydrions; (2) the adsorption of hydrions by the hide gel when first immersed in approximately neutral condition; (3) the operation of the "valency rule" that the predominant ionic effect in discharging is due to the multivalent anions. In this case the divalent SO{4}-- ions assist materially in discharging the positive charge on the chrome sol; (4) the final process of neutralization in which still more alkali is added. The operation of the valency rule is the most complex of these factors, for there is also to be considered the stabilizing effect of the kations, especially of the trivalent kation Cr+++ from the unhydrolyzed chromium sulphate. It is quite possible also that in the last stages of chrome tanning there are "zones of non-precipitation" due to the total effect of multivalent ions, and it is quite conceivable that the chrome sol may change its sign, _i.e._ become a negative sol and thus give also a mutual precipitation with the hide-gel. This is particularly probable where a local excess of alkali occurs in neutralization. However that may be, it is probable that most of the tannage is accomplished by chromium hydrate in acid solution, and it is therefore legitimate to conclude that adsorption and gelation have a relatively greater part in chrome tannage. The operation of the valency rule makes it easy to understand why basic chlorides do not tan so well as sulphates; the precipitating anion is only monovalent (Cl-) and chromic chloride contains no substance analogous to the potassium sulphate of chrome alum and hence contains a less concentration of the precipitating anion. Hence also the stabilizing influence of common salt added to a basic alum liquor, the effect being to replace partially the divalent SO{4}-- by the monovalent Cl-. Lyotrope influence, however, may be here at work.