, 
eee ee 
TO THE MOVEMENT OF MACHINES. 529 
tion is not in perfect equilibrium, being stronger on one side than on 
the other, there will be a deviation of the needle proportional to the 
difference of the currents which traverse the wires ab, ed. This 
agrees with the experiments which Mr. Faraday has related at the be- 
ginning of his Eighth Series, on the subject of decompositions produced 
by a single pair of voltaic plates. In short what is termed tension is 
the effect of forces equal and contrary in direction. In mechanics 
such forces destroy themselves, their sum being zero; but in physics it 
is different. 
With regard to the direction of the magneto-electric current which 
occasions the shock, it is the same as that of the voltaic current. This 
was proved by a galvanometer, the multiplying wire of which ter- 
minated at the points e, f. There was a deviation on a part of the 
voltaic current traversing the secondary branch e, f. On applying the 
armature, the needle of this galvanometer advanced, at the same time 
that the needle m receded. The contrary effect might be observed on 
removing the armature by the blows of a hammer. 
21. 
The following are some further experiments relative to this subject. 
The extremities of the bar were surrounded with a thin plate of copper, 
fig. 6, in the circuit of which was placed a galvanometer. On applying 
the armature, the needle was unaffected by it; but after having wound 
the ends of the multiplying wire around the points e, f, and the cir- 
cuit being thus closed, a considerable deviation took place. 
An analogous result is shown in the following experiment. On 
plunging two thin plates of copper, held firmly in the hands, in the cups 
a, b, or c, d, of the bar, fig. 4, there was no shock when the circuit was 
broken by the separation of the wires ab or ed; for the human body 
formed part of a circuit, in which equal excitations took place on two 
opposite sides. The thin plates being plunged into the cups ¢ and 6, a 
violent action took place at the instant of disjunction. 
I formed a thermo-electric circuit of bismuth and antimony, in 
which was interposed a galvanometer: after having heated the two 
solderings to the same degree, there was no deviation of the needle; but 
the multiplying wire having been placed so as to form an interme- 
diary branch, and the solderings being on opposite sides, there was a 
considerable deviation. This would not have taken place if the circuit 
of bismuth and antimony had been in its normal state, for then it 
would have had to conduct the greatest part of the thermo-electric 
current, provided that the multiplying wire was sufficiently long and 
thin to intercept only an extremely feeble part of it. 
It seems to me that there are circumstances which cause metals to 
lose their conducting power, and that these same circumstances on the 
