A forgotten experiment by Andre Blondel (1914) proves, as held on the basis of theoretical arguments in a previous paper, that the time variation of the magnetic flux is not the cause of the induced emf; the physical agent is instead the vector potential through the term [equation omitted] (when the induced circuit is at rest). The "good electromagnetic potentials" are determined by the Lorenz condition and retarded formulae. Other pairs of potentials derived through appropriate gauge functions are only mathematical devices for calculating the fields; they are not physically related to the sources. The physical meaning of a theoretical term relies, primarily, on theoretical grounds; a theoretical term has a physical meaning if it cannot be withdrawn without reducing the predictive power of a theory or, in a weaker sense, if it cannot be withdrawn without reducing the descriptive proficiency of a theory. (Contains 8 footnotes and 2 figures.)