696 PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 
After a few minutes the break was raised further so as to reduce the current very 
much, and the liquid began to rise in the stem of each of the glass tubes, showing 
that both air-thermometers at first acted perfectly. One of the thermometers was 
then steadied with great ease to a small fraction of a scale-division by using the 
regulator. The liquid in the other thermometer was observed, and its position occa- 
sionally noted. The direction of the current was reversed every few minutes, as 
before, by means of the ordinary commutator. 
Fig. 11. 
73 . Slight differences were observed in the free thermometer after the reversals, 
but as yet no very decisive indications of the looked-for effect appeared. The mercu- 
rial thermometer beside the central conductor indicated less than 80° Fahr. (27° 
Cent.), its column of mercury having not yet become visible, after the experiment 
had been continued in this way for several reversals. 
74 . The regulating break was then pushed down until a somewhat further elevation 
in the temperature of the platinum was indicated by a considerable escape of air in 
bubbles from the open ends of the thermometer-tube. The break was again drawn 
up until the liquids again mounted in the stems. One of the thermometers was again 
steadied by the regulator, and, the other being observed, the experiment was continued 
as before. A decided effect now appeared almost immediately after each reversal. 
The free thermometer regularly indicated a higher temperature when the current 
nominally entered by the end next it, and a lower temperature when the current 
nominally entered by the remote end. After four reversals this part of the experiment 
had lasted about twenty minutes, and the mercury thermometer beside its middle 
showed 104° Fahr. (40° Cent.). 
75 . The break was again pushed down for some time, and again raised till the 
liquid rose in each thermometer-tube, and the experiment was continued as before 
