PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 719 
each, and the other ends of the rows were connected with the electrodes of a galva- 
nometer by slips of iron touching them. Each row was firmly wedged up between 
its terminal iron slips to ensure metallic Fig. 26. 
contact; but after several attempts, and 
with all care in cleaning the surfaces meant — 
to touch, no sufficient completeness of -- 
contact throughout the circuit could be 
obtained until mercury was introduced as a liquid solder to connect the pieces of 
iron. This was done simply by pressing them together as at first, pasting paper 
round the junctions, and pushing little drops of liquid mercury or small quantities 
of soft mercurial amalgam into apertures in the tops of these paper coverings. 
Twelve hollows were cut in the board under and round the junction of the iron bars, 
each except the last including a pair of ends of the bars in contact in each row, and the 
last including the ends of the extreme bars on that side and the slip of iron by which 
they are connected. These hollows were filled alternately with hot sand and cold 
sand, which was everywhere piled over the junctions ; and the galvanometer gave 
slight indications of a current, the direction of which through the iron appeared to 
be generally from uncompressed to compressed through hot. 
124. The result, however, was not satisfactory; and it was obvious that the plan 
which had been adopted for heating and cooling was quite insufficient to sustain the 
required differences of temperature through so considerable masses of iron ; I there- 
fore had an apparatus constructed for the purpose, consisting of two 
main pipes of tin-plate, each carrying six smaller pipes and leading 
to small cells, also of tin-plate, with cylindrical passages through 
them to admit the iron bars, and with short discharge pipes attached 
to them on the other side from that by which the former enters. 
These cells were fitted into the hollows cut for the sand in the board 
formerly used, the main pipes occupying parallel positions above 
them on each side several inches from one another. The iron bars, 
each coated with paper and united as before one to another with 
mercury solder, were pushed through the hollows of the cells, and were fixed in two 
rows, with a junction in the centre of each of these hollows, and with the terminals 
adj usted as before. Cold water from 
the town supply-pipes was then run 
into one of the main pipes, so as to 
flow through the branch pipes and 
cells connected with it ; and steam 
from a boiler heated by an ordi- 
nary wire-gauze gas-burner was sent 
through the other system, so as to 
cool and heat alternately in their 
- <-s-\rjnr x n a h s b j, h .s ir^ 
S) 
