hy M. Ampere m his Electro-Magnetic researches. 409 
duced. The moment the circuit was shut, the moveable part 
of the apparatus revolved hy thfe mutual action of that part and 
of the fixed conductor AB, till the currents which were at first in 
an opposite direction, became parallel, and in the same direction. 
It passed this last position by its acquired velocity, but soon 
settled in it after a few oscillations. 
The idea of M. Ampere that loadstones are an assemblage of 
electrical currents, led him to conceive the plan of imitating their 
action by spiral cbnductors. After some failures he at last suc- 
ceeded by means of the following apparatus. 
A spiral brass-wire. Fig. 3. PI. VIII. surrounds two glass 
tubes ACD, BEF, and being pi'olonged on both sides, return 
by the interior of the tubeSi The two extremities emerging at 
D and F, one of them, DG, descends vertically, while the other, 
FHK, is bent at H and K. Both these wires are terminated 
by steel points, which are plunged in the mercury contained in 
the cups M, N, which communicate with the two extremities of the 
pile, the upper point resting solely on the bottom of the cup N. 
The extremity of the spiral needle which is to the right of the 
currents j is that which presents, with regard to a magnetised bar, 
the phenomena of the north pole of a compass-needle, and the 
other that of the south pole. 
M. Ampere afterwai^ds constructed an apparatus similar to 
Fig. 1. in which AB and CD were replaced by brass spirals, 
surrounding tubes of glass, but whose prolongations, instead of 
returning by these tubes, were put in communication with the 
two extremities of the pile, like AB, CD. 
In using this apparatus, M. Ampere discovered the following 
law. Let AD be the resultant of two 
forces AB, AC, and let these lines re- 
present three infinitely small portions 
of electrical currents, whose intensities 
are proportional to their lengths. Then 
the 'portion AD of the electrical current wilt exert on another 
current.) or on a loadstone^ an attractive or repulsive action^ 
equal to that which would result in the same direction from the 
reunion of the two portiems AB, AC. Hence, it follows, that 
in spiral wires, the action produced by the current of each coil 
is composed of two, one of which would be produced by a cur- 
