The Faraday laws of electrolysis state (in modern terms) that  the charge transported by a solvated ion between two electrodes is an integer multiple of the elementary  charge e.  But a liquid is an assembly of nuclei and elec- trons: while the nuclear charges can be considered as point-like, the electronic charge is delocalized all over the cell. The moving ions drag “some” charge, whose integrated value is nonetheless ill defined. Ionic charges in solution can only be computed via some approximate formula, and are noninteger.  So why is the electrode-to-electrode transported  charge quantized? The answer (Thouless, 1983) is in topology: I will give a simplified view of this.