748 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
taken, and a multiplying branch of one yard No. 18 copper wire was soldered by 
its ends to them. Before, however, the effect of the magnetism could be decidedly 
tested, the zero-point had moved off the multiplying branch, which had accordingly 
to be shifted along the testing conductor to get into range again. The same pro- 
cess had to be gone through a great many times, and at last, after the current had 
been flowing continuously through the two squares and the divided copper channel 
for about five hours, the zero-point became sufficiently steady to remain on the mul- 
tiplying branch when fixed at the right place on the testing conductor, and to allow 
a decisive experiment to be made. The result was a very slight effect, proving a 
diminution of 1 'esistance in the iron square. 
1/4. The cause of the long-continued variation in the conditions of electric equi- 
librium between the testing conductor and the fixed point on the edge of the lower 
square, was clearly the gradual warming of the long copper wires extending up 
from this point, due to the conduction of heat generated in the iron squares by the 
electric current ; and it would obviously be much diminished by using a simple form 
of conductor with only one iron square at a time, and with the reference conductor 
kept near it, so as to acquire quickly whatever temperature it would rest with during 
the flow of the current. I accordingly made the following experiment, choosing 
first the effect of transverse magnetization, as the experiment just described had 
not been of a satisfactory kind, although apparently conclusive, while the first expe- 
riment of the series (Exp. 1. 155) had been less unsatisfactory in point of steadi- 
ness, and had led decisively to a conclusion regarding the effect of longitudinal 
magnetization on the resistance of a conductor. 
175 . Exp. 7 . A square of sheet iron like those used in the last experiment (four 
square inches, weighing 103 grains, and consequently about ^th of an inch thick,) 
was soldered along one edge to a slip of lead of the same widtli, about twice as thick 
and about one-half longer. To the opposite edge of the iron square was soldered a 
stout copper slip an inch broad and equal in length 
to the side. The piece of lead was bent round, so 
as to give a straight part lying about ^ of an inch 
from the plane of the iron, and to extend about as 
far as the copper slip soldered to the other edge of 
the square. A current from an arrangement of the 
cells (§§ 63 and 64) constituting a powerful single 
element of Daniell’s was sent through the iron 
square and the lead band, by electrodes clamped to 
one end of the lead and to the copper slip fixed to the other edge of the iron. A 
point in the lead slip having been found, such that the galvanic resistance between 
it and the edge next the iron was nearly equal to the resistance in the iron square 
itself, a testing conductor (two yards of No. 18 copper) was soldered by one end to 
that point in the lead, and by its other end to the middle of the edge of the iron 
