PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 661 
pleted experiments, were all reduced, on the plan of the Tables given (§§ 47 and 56) 
below for subsequent experiments, and, when thus tested, they appeared to contain 
some indications of the effect looked for. Several more sets of observations of the 
same kind were therefore executed, but with various modifications of details. Still 
no decided result could be obtained, and I concluded from all the experiments 
which had been made, that the anticipated effect must be too small to be discovered 
without either increasing the sensibility of the test or diminishing the irregularities. 
I therefore prepared new apparatus, by which the former, and as much as possible of 
the latter object, would be attained. 
30 to 34. Improvements and Modifications of Apparatus. 
30. Instead of increasing the power of the battery, which I reserved as a later 
resource, if necessary, or of increasing the length of the conductor between the heater 
and the coolers on each side, which, while it would increase in the same ratio the 
amount of the effect looked for, would increase in a duplicate ratio the time that 
would have to be given to allow it to reach a stated proportion of its limiting value, 
I had conductors made of about the same length as the others, but of considerably 
less section. 
31. With a view to perfecting and testing the action of the heater and coolers, 
each conductor was made up of a number of slips of flat sheet metal, bent and placed 
together, as shown in the accompanying diagram (fig. 2). The slips were held firmly 
together by a vice, while collars of sheet copper, separated from them by vulcanized 
india-rubber, were soldered round them in the places for the sides of the heater and 
coolers. Tin-plate vessels, as shown in the diagram, were then put together, and 
soldered to these collars. The interstices between the slips and the india-rubber, and 
4 s 2 
