near which it passes in motion in the same direction._" M. Ampère tells me
in a letter which I have just received from him, that he carefully avoided,
when describing the experiment, any reference to the direction of the
induced current; and on looking at the passages he quotes to me, I find
that to be the case. I have therefore done him injustice in the above
statements, and am anxious to correct my error.

But that it may not be supposed I lightly wrote those passages, I will
briefly refer to my reasons for understanding them in the sense I did. At
first the experiment failed. When re-made successfully about a year
afterwards, it was at Geneva in company with M.A. De la Rive: the latter
philosopher described the results[A], and says that the plate of copper
bent into a circle which was used as the mobile conductor "sometimes
advanced between the two branches of the (horse-shoe) magnet, and sometimes
was repelled, _according_ to the direction of the current in the
surrounding conductors."

[A] Bibliothèque Universelle, xxi. p. 48.

I have been in the habit of referring to Demonferrand's _Manuel
d'Electricité Dynamique_, as a book of authority in France; containing the
general results and laws of this branch of science, up to the time of its
publication, in a well arranged form. At p. 173, the author, when
describing this experiment, says, "The mobile circle turns to take a
position of equilibrium as a conductor would do in which the current moved
in the _same direction_ as in the spiral;" and in the same paragraph he
adds, "It is therefore proved _that a current of electricity tends to put
the electricity of conductors, near which it passes, in motion in the same
direction._" These are the words I quoted in my paper (78.).