The experiment referred to was performed in 1823 by Ampère in the presence of Mr A. De la Rive. It is intended to show that if one portion of a rectilinear conductor is made moveable, it will be repelled from the portions which form its continuation. The arrangement employed was a copper wire-float in a double trough of mercury, which cannot easily be explained without a figure.

This experiment, though quoted by most writers on electricity, is understood to be one not easily repeated. Indeed, the author has not found an indication of its ever having been repeated with success. The author conceiving that perhaps this repulsion described by Ampère might assist in explaining the vibrations of rocking bodies through which an electrical current passes (as described in a communication by him to this Society in January 1859), devised another, and, as he conceives, a more delicate method of testing the supposed repulsion of a rectilinear current on itself.