S72 Mr Barlow’s Electro-Magnetic Experiments y 
magnet, is a thin cylinder, as a piece of quill, forming a cup or 
reservoir Z, to receive a small quantity of quicksilver ; and into 
this is inserted the wire z, amalgamated at its lowest point, and 
C c is a stout wire passing through' the side of the cup into the. 
quicksilver. Then, the contact being made at C and Z, the 
magnet will begin to revolve on its axis, with a very astonishing 
velocity, and continue in motion while the power of the battery 
lasts. ^ 
This pleasing experiment is due to M. Ampere, who employs 
only a piece of platinum attached to the magnet, to produce, by 
its superior gravity, a vertical position of the latter in the mer- 
cury ; the upper wire being then inserted into the quicksilver in 
the cylinder s, and the other wire into the cup C, the motion is 
produced exactly as above described: the greatest freedom of 
motion is, however, given by the apparatus shown in the figure. 
The explanation of this rotation is very obvious, according to the 
hypothesis we have adopted ; for the tangential force of the wire 
acting upon the magnetic particles on the surface of the magnet, 
must necessarily produce the rotation in question, on precisely 
the same principles as the magnet is made to revolve about the 
wire in the fifth experiment 
Exp. III. €( To exhibit the rotation of a galvanic wire on its 
axis by the action of a magnet. 
“ Let MS, Fig. 4, be a magnet, represented as broken in the 
figure, but which is fixed, in the experiment, in a foot, in order 
to keep it vertical, and let abed be a light hollow copper or 
brass cylinder, having a steel point passing downwards into the 
agate cup fixed to the upper end of the magnet, and let e be 
a small tube or quill fixed on the wire passing through the top 
of the cylinder, holding a little quicksilver, and receiving into it 
the descending conducting wire Z. AB is a piece of wood 
turned to fit on the cylindrical magnet NS, which has a hollow 
groove on its upper surface, to receive a quantity of quicksilver, 
into which the lower edge of the cylinder ad is slightly immersed, 
the surface being covered with weak dilute nitric acid. AC is 
a wire passing into the quicksilver. It is obvious that thus 
(the contact being made at Z and C) the galvanic circuit is tar- 
ried from Z through the cylinder aoed , thence to the quicksil- 
ver, and hence again through the wire AC to the other extre- 
