PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 727 
with their poles in dissimilar directions, while the other short wires, and the longer 
steel terminals, were left as free from magnetism as possible. The magnetizing helix 
was then remov^ed, and the compound conductor was made into a flat coil on a 
wooden core (2 inches broad and |^-inch thick), by bending the short copper wires, 
and arranging the 2-inch steel wires alternately on the two sides of the wood. The 
terminals were joined, with the usual precautions (§ 92), to the galvanometer elec- 
trodes, and one edge of the coil was immersed nearly an inch below the surface of a 
vessel of oil at the temperature of about 100° Cent. Immediately a strong deflection 
of the needle showed a current, of which the direction in the coil from unmag- 
netized to magnetized through hot. When the other edge of the coil was similarly 
heated, a contrary deflection of the needle as decidedly showed the same thermo- 
electric difference of quality between the magnetized and the unmagnetized steel 
wires. 
142. The object of the peculiar arrangement just described, was to prevent the 
magnetism from spreading to those of the steel portions of the circuit which were 
to be kept as free from magnetism as possible in order to be compared with those 
which were magnetized. The introduction of the connecting pieces of a different 
metal from steel into the circuit cannot give rise to any thermo-electric disturb- 
ances *, provided the two ends of each are at the same temperature, a condition 
which was nearly enough fulfilled in the way the experiment was made, and which 
was very much favoured by the shortness and the high thermal conductivity of the 
little copper arcs. 
The same result was demonstrated in an experiment made with a homogeneous 
coil of steel wire, of which parts had been magnetized, by ordinary steel magnets, 
before it was bent on the core. 
[§ 143. Received May 10, 1856.] 
§ 143. Experiment. — On the Effect of Magnetization on the Thermo-electric Quality 
of Nickel. 
Through the kindness of Dr. George Wilson, I have been able to experiment on 
a bar of nickel, about ^ an inch in diameter and about 8 inches long, in the form of 
a horse-shoe magnet, belonging to the Industrial Museum of Edinburgh. The accom- 
panying sketch and description show the plan of the experiment. 
Description of Sketch. 
N, nickel horse-shoe. 
B B, double tubes of sheet copper,, electrically connected with one another by a 
copper band, and insulated from the nickel by silk paper, laid on with shell- 
lac varnish ; serving to drain all electrical leakage from the magnetizing 
* Dynamical Theory of Heat, § 138, Cor. 1. 
