714 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
the different turns of the wire were kept from touching one another, by little notches 
cut in the edges of the pieces of wood. The whole coil was made firm, and its extreme 
turns tied down to the wood to prevent slipping. The ends of the wire, extending a foot 
or two on each side of the coil, were connected in the usual way (§ 92) with a galva- 
nometer. The bar bearing the coil was laid with its two wooden faces horizontal, 
and one of them supported on a thin piece of hard wood lying on the stage of a 
Bramah’s press. Another thin piece of hard wood was laid upon the top of the coil, 
to prevent the upper part of it (when, in 
the course of the experiment, it is forced 
upwards,) from touching the roof of the 
press. Blocks of iron were placed on the 
ends of the bar, so that when the stage is 
pushed up they may be resisted by the 
roof, cause a heavy stress to act on the 
bar, and press the lower horizontal parts 
of the wire coil between the two pieces of 
hard wood touching them above and be- 
low. The same blocks are afterwards 
shifted to rest on the stage and bear the 
ends of the bar upon them, so that, when 
the stage is forced up, the upper parts of 
the wire coil may be pressed against the piece of hard wood above them, which will 
then be resisted by the roof of the press. 
Pieces of plate glass highly heated were applied to the vertical parts of the wire on 
one side of the bar, those on the other side being left cool, and the galvanometer 
was observed. Some slight deviation of the needle was generally noticed. 7’hen the 
press was worked, and immediately a strong deflection took place, indicating a current 
in the iron coil, from the uncompressed portions through the heated vertical portions, 
Fig. 22. 
into the compressed portions. The pressure was 
relieved, and the galvanometer needle returned 
nearly to zero. It was reapplied, and the same 
powerful deflection was observed. The glass 
heaters were shifted to the other side, and, the 
pressure being continued, the deflection of the 
needle became reversed. The pressure was removed, and by shifting the iron 
blocks, and working the press again, was applied on the other horizontal side of the 
coil. The heating being kept unchanged, a reverse deflection was observed, powerful 
as at first. The current indicated w'as in every case from free iron wire to pressed 
iron wire through hot, as is illustrated in the diagram, for a case in which the upper 
parts of the wire are compressed. 
115. From this, in conjunction with the result regarding the effect of longitudinal 
stress previously obtained, we may nearly conclude that a longitudinal strain in iron 
