by M. Ampere \n his Electro-Magnetic researches. 41S 
ted on the two equal and opposite halves of the moveable con« 
ductor. 
One of the instruments which M. Ampere succeeded in 
moving by the action of the terrestrial globe, is shewn in Fig. 7., 
where the conducting wire has the form of a semicircle ABCD, 
whose radius is nearly 8 inches. The two ends of the brass- 
wire out of which it is formed, are soldered to two copper-boxes 
E, F, attached to a glass tube Q, and carrying two steel points 
M and N, plunged in mercury in the two platinum cups 
0, P, the upper one, N only, reaching the bottom of the cup P. 
These cups are carried by the copper-boxes G, H, which com^ 
municate with the two extremities of the pile by means of two 
brass- wire conductors, one of which is contained in a tube of 
glass carrying these two last boxes, and forming the foot of the 
instrument ; while the other forms around that tube a spiral, 
whose coils have a sufficiently great height relatively to the 
diameter of the pile, in order that the actions of the two por« 
tions of the currents, which run through these conductors in 
opposite directions, may nearly neutralise each other. 
Under the glass cage which covers the instrument, M. Am- 
pere placed another circle he oi brass- wire, whose diameter was a 
little greater, and which was fixed and supported by a foot 
similar to that of the moveable circle. This circle communis 
cates also with two conductors disposed in a similar manner, 
and which serve also to convey the electric current, when, in 
place of observing the action of the globe upon the moveable 
circle, we wish to see the action of two circular currents on one 
another. The circle h c indicates the vertical plane perpendicu- 
lar to the magnetic meridian,' into which BC ought to be drawn 
by the action of the earth. The circle h c being placed in this 
plane by means of a compass, and BC in another situation, for 
example in that of the magnetic meridian, then, when an 
electric current is made to pass it, BC will move into the plane 
indicated by h c, and passing a little beyond it by its acquired 
velocity, will return to it after a few oscillations. 
The other apparatus shewn in Fig. 5., consists of a brass^r 
wire ABCUEFG, soldered at A to a piece of similar wire 
HAK, carried by the glass tube XY by means of the copper 
