Ampere, Electrodynamics, and Experimental Evidence

Abstract

IN AUGUST 1831 MICHAEL FARADAY discovered that he could produce a temporary electric current in a coiled wire by either starting or stopping another current in a nearby coil. Because of his detection of this induced current and his careful specification of the circumstances in which it had been produced, Faraday soon was credited with the discovery of electromagnetic induction. However, almost nine years earlier Andre-Marie Ampere had produced a surprising new electrodynamic phenomenon that he attributed to electric current produced by the influence of a nearby current in a coiled wire. It is not my intention to claim priority for Ampere. Although Ampere produced and detected the induction effect in 1822, he did not understand how it was related to the change in the intensity of his primary current. Ampere's belief that some form of induction persisted as long as he maintained a current in his primary circuit is enough to disqualify him as a claimant to Faraday's discovery. But, given Ampere's keen interest in electrodynamics at that time, it seems odd that he did not follow his initial observation with the detailed research that would have clarified the phenomenon. Ampere's failure to investigate induction was not simply an unreflective oversight. His reaction to the induction experiment was shaped by the strategy that governed all his most creative research in electrodynamics between 1820 and 1823. This strategy lay behind both his limited and inaccurate study of induction and his far more enthusiastic response to two other experiments he performed on the same day in 1822. To place these experiments in context, I begin with a general overview of the relevant aspects of Ampere's early research in electrodynamics. I then argue that this activity was guided by a carefully designed research program that was itself stipulated by a set of fundamental assumptions and a clear-cut heuristic procedure for theory revision. When Ampere argued that his program was superior to those of rival theorists, he particularly emphasized a specific methodological criterion: he insisted that, although theories obviously had to be based upon prior experimental information, the strongest possible confirmatory evidence was the derivation or prediction of additional experimental phenomena that had not been used to create the theory in question. I next explain the theoretical reasoning and prior experimental circumstances that motivated the three experiments Ampere performed in Geneva with Auguste de La Rive in September 1822. Each of the experiments is intimately tied to