Table XII includes three experiments at a density of 0.013 ampère per square centimetre. The deposits show no difference in weight.

It would thus appear that, at current densities above 0:01 ampère per square centimetre, there is practically no difference in weight of the deposits obtained from solutions of copper sulphate (made acid), one of which is under ordinary atmospheric pressure, and the other under reduced pressure; but at densities below this there is a very appreciable difference, the deposit in a partial vacuum being always greater than the corresponding deposit in air. No experiment with acid solutions ever gave the vacuum deposit less than the air deposit. Why this difference should exist at certain densities and not at others is not very clear. The chemical corrosion above referred to will be more marked in the air solutions at weak current densities, where the duration of the experiments is three to four hours, but it would not be expected that it would be altogether missing at densities where the duration of the experiments is less than forty minutes and less.

My experiments do not completely reveal the cause of the relations existing between the deposit and current density. They clearly show, however, that the oxygen present in the solution has a certain diminutive action on the weight of the deposit, and that this action is prevented (partly or wholly) by conducting the electrolysis in The removal of the oxygen of the solution prevents the copper plates, and more particularly the copper deposit, from dissolving in the solution; but it does not prevent (apparently) what some physicists speak of as an electrical corrosion, which occurs only when the current is passing through the solution.

vacuo.

It is of interest to draw a curve representing the deposits obtained in vacuo with different current densities. This I have done by