PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 663 
.35 to 38. First Experiments with Multiple Sheet Copper Conductor. 
35. The copper conductor on the new plan was first used in an experiment on the 
28th of October, 1853; with the central vessel heated by steam, and a current from 
the eight large iron cells kept flowing for seventy-two minutes, alternately in con- 
trary directions, six times six minutes each way. The thermometers were noted every 
half-minute. The observations thus recorded, when thoroughly examined, indicated 
a slight differential cooling effect in the part of the conductor in which the nominal 
current was from cold to hot, and a heating effect where it passed from hot to cold ; 
that is to say, a convection of heat in the nominal direction of the current, or as I 
shall call it to avoid circumlocution, a convection of heat by vitreous electricity. 
36. A second experiment with the same conductor was made on the 2nd of 
November, 1853, in which the current was kept flowing for ninety-six minutes, eight 
times six minutes each way, and the thermometers were noted every quarter-minute. 
An examination of the recorded results indicated still the same kind of effect, but to a 
much smaller extent. Thus the final average, for the alteration of difference, between 
the temperatures at A and B due to the flow of the current for six minutes in one 
direction, after it had been flowing for six minutes in the contrary direction, amounted 
to "039° Cent, in the first experiment, and to only ’0143° in the second experiment. 
A full analysis of the progress of the differential variation of temperature during the 
flow of the current is given in Tables I. and II., § 56 below, and shows through what 
fluctuations the final alterations are reached. The temperatures, at the ends of the 
successive times of flow in one direction or the other, and the evaluation of the mean 
final effect, ai-e shown, for each experiment, in the following abridged Tables. 
37 . The observations made during the first period (that is the time from starting 
till the second reversal of the current) are rejected from the average in every case of 
experiments on the new conductors, because they were found to show so great 
absolute elevations of temperature (due to the frictional generation of heat by the 
current) that no alteration of difference between the thermometer observed during 
them could be relied on as an effect depending on the direction of the current. 
