PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 723 
was slipped upon a piece of thick straight iron wire about 2 feet long, which was 
supported in a horizontal position by its ends, and through them put in the circuit 
of a galvanometer. A spirit-lamp was held under the middle of the wire so as to 
raise it to a high temperature, and then a current from a few of the iron cells was 
sent through the helix, which was kept a little on one side of the middle of the wire. 
Immediately the galvanometer needle, which was not at first disturbed by the appli- 
cation of the spirit-lamp, experienced Fig. 30. 
a deflection. The little helix was 
slipped rapidly through the flame of 
the spirit-lamp to the other side of the 
hot part of the wire, and a reverse 
deflection was immediately produced. 
It was easy, by moving the helix alternately to the two sides of the hot middle of 
the wire, to make the needle of the galvanometer to swing through an arc of 10° or 
more. When the needle was brought to rest there was always a most sensible per- 
manent deflection, on one side or the other, according as the helix was left on one 
side or other of the heated parts. When the circuit of the galvanometer was broken, 
none of these effects followed from the motions of the helix. They were therefore 
not due to the direct force of the magnetism in the helix and iron wire, but to that 
of a current through the galvanometer coil. This always took place in such directions 
as to indicate a current from unmagnetized to magnetized through hot. 
135. The decided character of the result of this experiment established it beyond 
doubt, that the thermo-electric quality of iron is altered by magnetization. Imme- 
diately the question arose (from the general considerations referred to above, §§ 104 
and 105), are the thermo-electric qualities equally or even similarly ciffected in all direc- 
tions ? and the crystalline hypothesis suggested the answer : — no ; probably even the 
reverse thermo-electric effect may be found across their lines of magnetization. As 
theory could give no more than a conjectural answer, I tried to find the truth by 
experiment ; and, after various fruitless operations, obtained a very decided result, 
in the following way. 
136. A piece of thin sheet iron was cut into the shape shown in the diagram, the 
breadth everywhere being about ^ of an inch, the length of the longer branch 
45 inches, and that of the shorter 6 inches. The longer 
branch was rolled into a plane spiral, on a cylindrical core 
^ an inch diameter, the different successive turns being 
prevented from touching one another by a piece of narrow 
tape wound on along with the iron slip. The shorter 
branch, which stood out from the inner end of the coil at 
right angles to the plane of the spiral, was bent round into this plane, and carried 
out along one side of the spiral several inches beyond its circumference. 
5 c 
Fig. 31. 
I' 
I 
Along 
MDCCCLVI. 
