742 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
seven porous cells, contained in four large wooden cells, and exposing in all 8'75 
square feet of zinc surface to 15*3 square feet of copper, was then used to send a 
current through the iron square, insulated between the poles of the electro-magnet, 
in the manner described. 
163. The neutral point on the testing branch Fig- 46. 
being got by trial, it was found to remain 
tolerably steady, although no doubt during the 
first minutes of the flow of the current it may 
have varied much, as the iron got heated, which 
it soon did to a degree very sensible to the 
touch. Moving the electrode along the testing 
branch through a quarter of an inch on either 
side of the neutral point, gave a very marked 
deflection of the galvanometer. The galvano- 
meter circuit was then broken, and a current 
from six of the small iron cells was started 
through the coils of the electro-magnet. When 
the galvanometer circuit was again, after a few seconds, closed, with its electrode on 
the same point of the multiplying branch as before, a very considerable deflection was 
observed in the needle. To correct this deflection and bring the needle to zero, the 
testing electrode had to be moved to a position 2 or 3 inches nearer D on the testing 
branch. 
164. The new neutral point was unchanged when the electro-magnet was reversed, 
and when the magnetizing current was broken there was a permanent deflection in 
the galvanometer the reverse of that observed when the current was started in either 
direction. If the galvanometer circuit was completed within a second or two of 
any of the changes in the magnetizing current, the needle experienced, obviously 
from induced currents, powerful impulses in one direction or the other, according to 
the direction of the current made or unmade through coils of the electro-magnet. 
But in every case, although from various disturbing causes the neutral points gradu- 
ally shifted largely along the testing branch, the permanent effects of making and of 
unmaking the electro-magnet were most marked, and were uniformly as stated 
above. 
165. Thus it appears that magnetization shifts the equipotential Fig. 47. 
line through C from its position running across to the opposite 
corner, to a position (dotted in the diagram) a little nearer CB ; so 
much so that its end is shifted about or -^i^th of an inch 
from E towards D. This shows that the passage of electricity in 
the directions AE and CB has become less resisted than it was, 
relatively to the passage in the directions AC, DB ; and it therefore follows that the 
electric conductivity of magnetized iron is greater across than along the lines of 
magnetization. 
