In a paper on the aluminium anode by one of us, in conjunction with Inglis, it was pointed out that if the suggested theory of the aluminium anode is correct, a precipitation film of aluminium hydroxide should dimmish the conductivity of salt solutions to very different extents depending upon the ions present, and that, e.g., the diminution in the case of potassium chloride might be expected to be slight, in the case of potassium bromide somewhat greater, and in the case of potassium sulphate to be very considerable. Preliminary experiments to measure these differences directly were not successful, almost certainly, as was stated at the time, because of the difficulty of freeing the precipitation membrane from the concomitant soluble impurities.