Remarks on the theory of diffusion in electrolyte solutions