PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 741 
the first place the testing electrode was applied at different places along the edges 
BD, EA of the square till a point was found which gave no deflection of the galvano- 
meter. If this was in BD, the plate had to be thinned in its middle parts parallel to 
CA and BD, or else to be thinned along the edges CB, AD, so as to increase the 
resistance to conduction parallel to the last-mentioned edges. Or if the neutral point 
was in EA, the plate had to be thinned in its middle parts parallel to CB and AD, or 
along its edges BD, CA. By using the file according to these directions, after a few 
trials the neutral point was brought upon the testing branch ; that is to say, the resist- 
ance was so adjusted in the square that the line from C cutting right across the lines 
of conduction, or which is the same thing, the equi- Fig. 44. 
potential line through C, passed between D and E. 
A piece of sheet copper as broad as the iron square, 
but rather longer, was bent as shown in the diagram, so as to give a depressed space 
in which the iron, insulated from the copper 
simply by a piece of writing-paper, could rest 
steadily. This copper cradle was placed resting 
on the flat poles of a Ruhmkorff electro-magnet, 
which were pushed together so as to hold it 
firmly. Any leakage of electric currents from 
the coils of the electro-magnet was thus effectu- 
ally drained by the copper, so that a simple 
sheet of paper was quite enough to do away with 
all sensible indications of currents in the iron 
acquired otherwise than through the electrodes A and B. [This electrical drainage 
would be made more nearly perfect by using paper or some other non-conductor to 
separate the cradle from the poles of the magnet.] 
162. A large single element of Daniell’s 63), consisting of seven zinc plates in 
5 E 2 
