Neglected Experiment of Ampere. 537 



in the Ampere manuscripts belonging to the Academie des 

 Sciences until 1885. It reads as follows : — 



" La troisieme experience avait pour objet de savoir si Pon 

 peut produire un courant electrique par Pinfluence d'un autre 

 courant. II y a plus d'un an que je Favais tentee, sans 

 succes, avec l'appareil represents [fig. 1]; mais j'ai reussi 

 avec un appareil absolument semblable, en employant Pexcel- 

 lent aimant en fer a cheval du musee de Geneve, que up avait 

 procure M. le professeur Pictet. Voici la description de cette 

 experience. - " [Here follows the description of the apparatus, 

 word for word almost the same as that above in the descrip- 

 tion of the unsuccessful experiment of July 1821.] " Le 

 circuit ferme place sous Pinfluence du courant electrique 

 dedouble, mais sans aucune communication avec lui, a ete 

 attire et repousse alternativement par Paimant, et cette ex- 

 perience ne laisserait, par consequent, aucun doute sur la 

 production des courants electriques par influence, si l'on ne 

 pouvait soupconner la presence d'un peu de fer dans le cuivre 

 dont a ete forme le circuit mobile. II n'y avait cependant 

 aucune action entre ce circuit et Paimant avant que le courant 

 electrique parcourut la spirale dont il etait entoure ; c'est 

 pourquoi je regards cette experience comme suffisante pour 

 prouver cette production ; je me propose, neanmoins, pour 

 prevenir toute objection, de la repeter incessament, avec un 

 circuit forme d'un metal nonmagnetique tres pur. Ce fait 

 de la production de courants electriques par influence, tres 

 interessant par lui-meme, est d'ailleurs independant de la 

 theorie generale de Paction electrodynamique. ; '' 



There is a discrepancy between the first and second versions, 

 in that while De la Rive says that the attractions and repul- 

 sions were in correspondence with the sense of the current 

 in the coil (which we know could not be the case), Ampere 

 says nothing about reversal of the current. As he attributed 

 the effect to the possible presence of iron, it is clear he was 

 expecting an effect due to the continued pressure of the cur- 

 rent, not an instantaneous effect at make or break. We know 

 now that this must have been so, and that an unvarying 

 current could have produced no such effects as those described. 

 That this should have escaped the notice not only of Ampere, 

 but also of so skilled an experimenter as De la Rive, is 

 remarkable. But the explanation of the phenomenon, as 

 being due to a temporary magnetization acquired by the non- 

 magnetic metal, given by De la Rive, and accepted by Ampere, 

 is still more strange. Had no suggestion ever been made that 

 there might possibly be an induction of currents by the action 

 of a magnet, an explanation which ascribed the effects to a 



