87.  The galvanometer was roughly made, yet sufficiently delicate in its indications.  The wire was of copper covered with silk, and made sixteen or eighteen convolutions.  Two sewing-needles were magnetized and fixed on to a stem of dried grass parallel to each other, but in opposite directions, and about half an inch apart; this system was suspended by a fibre of unspun silk, so that the lower needle should be between the convolutions of the multiplier, and the upper above them.  The latter was by much the most powerful magnet, and gave terrestrial direction to the whole; fig. 8. represents the direction of the wire and of the needles when the instrument was placed in the magnetic meridian:  the ends of the wires are marked A and B for convenient reference hereafter.  The letters S and N designate the south and north ends of the needle when affected merely by terrestrial magnetism; the end N is therefore the marked pole (44.).  The whole instrument was protected by a glass jar, and stood, as to position and distance relative to the large magnet, under the same circumstances as before (45.).

88.  All these arrangements being made, the copper disc was adjusted as in fig. 7, the small magnetic poles being about half an inch apart, and the edge of the plate inserted about half their width between them.  One of the galvanometer wires was passed twice or thrice loosely round the brass axis of the plate, and the other attached to a conductor (86.), which itself was retained by the hand in contact with the amalgamated edge of the disc at the part immediately between the magnetic poles.  Under these circumstances all was quiescent, and the galvanometer exhibited no effect.  But the instant the plate moved, the galvanometer was influenced, and by revolving the plate quickly the needle could be deflected 90 deg. or more.

89.  It was difficult under the circumstances to make the contact between the conductor and the edge of the revolving disc uniformly good and extensive; it was also difficult in the first experiments to obtain a regular velocity of rotation:  both these causes tended to retain the needle in a continual state of vibration; but no difficulty existed in ascertaining to which side it was deflected, or generally, about what line it vibrated.  Afterwards, when the experiments were made more carefully, a permanent deflection of the needle of nearly 45 deg. could be sustained.

90.  Here therefore was demonstrated the production of a permanent current of electricity by ordinary magnets (57.).

91.  When the motion of the disc was reversed, every other circumstance remaining the same, the galvanometer needle was deflected with equal power as before; but the deflection was on the opposite side, and the current of electricity evolved, therefore, the reverse of the former.