In 1996 Kelsall et al. [5] reported electrophoretic experiments with oxygen bubbles in dilute solutions of several salts that were remarkably free of surfactants, but the experimental results agree with neither their own theory nor anyone else's known to the present author. This paper assumes a double-layer thickness much smaller than the bubble radius, as it was in the experiments. It redevelops the theory on various hypotheses about the precise location of the free surface charge, and both with and without van der Weg's recently-suggested modification to the electrochemical potential. The results suggest that the free surface charge is at or on the gas side of the change in permittivity at a bubble's surface in a surfactant-free liquid, and that the classical theory for the speed of rise of a bubble is not quite correct in an electrolyte solution, but the correction is too small to measure. However, there are still unresolved difficulties with both theory and experiments, which van der Weg's suggestion does not remove. Nor does it improve the fit between theory and Grahame's [13] experiments on the differential capacitance of a double layer. (C) 2010 Elsevier Inc. All rights reserved.