746 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
being then guarded against irregular contacts by a little square of pasteboard pressed 
between the iron squares, a half-square of pasteboard between the first-mentioned 
iron square and the portion FCH of the connecting copper (see fig. 49), and frag- 
ments of paper and pasteboard elsewhere, the whole was placed, with the second- 
mentioned square lowest, in a copper cradle lined with paper, and resting between 
the horizontal edges of the flat poles of the Ruhmkorff electro-magnet used in the 
preceding experiment. 
170 . The positions of the magnetic poles of the squares, of the bent connecting 
piece of copper, of the testing conductor, and of the galvanometer electrodes are 
indicated in fig. 55, but, to avoid confusion, the principal electrodes are not shown. 
Fig. 55. 
A current from the four large double cells, connected so as to constitute in all a 
single element of Daniell’s, exposing 10 square feet of zinc surface to 17^^ square 
feet of copper, was then introduced by the principal electrode soldered to the edge 
MD of the upper square, and drawn off by the other principal electrode, namely, that 
soldered to the edge of the lower square lying exactly below the edge MH of the 
upper. The course of the current into the principal channel between these electrodes 
would be across the upper square from MD to HC, and across the lower square from 
the edge below CD to that below HM ; also, in the secondary channel between the 
same electrodes, from T soldered to the first through the testing conductor, to its other 
end U soldered to the second. 
171 . A fixed galvanometer electrode being (§ 168) soldered to the middle point, N, 
of the connecting-copper, the other electrode of the galvanometer was moved along 
the testing conductor till a point, O, was found at which it might be applied without 
giving any deflection. By moving it -^th of an inch on either side of O very sen- 
sible deflections were obtained, and therefore a yard of copper wire was soldered by 
its ends to points S and Q a quarter of an inch on each side of O, and was used 
instead of the “scale” of the testing conductor described as used in the first three 
experiments. The neutral point. O', on this multiplying branch having been found, 
the galvanometer circuit was broken, and the electro-magnet was excited by six of 
the small iron cells. On closing the galvanometer circuit again immediately, a con- 
siderable deflection was observed, to correct which the moveable electrode had to be 
