PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 697 
for four reversals, the central mercury thermometer rising to about 150° Fahr. (66° 
Cent.). The free thermometer rose and fell alternately through several scale-divisions 
almost immediately after each reversal, and showed the same convective effect as had 
previously been observed by smaller indications. 
76. The bridge was again pushed down and air again escaped copiously from the 
thermometers, but very soon beads of liquid began to appear following one another 
rapidly down the capillary tubes from the interior of the conductor. 
As the spirits of wine had not once been allowed to run up into the bulb of either 
thermometer, these beads of liquid could be nothing but products of the distillation 
of the oil which had been used in the luting of the central plug ; and on taking away 
the cups of spirits of wine from below the tubes, the smell and taste of the small 
quantities of liquid which continued to descend gave unmistakeable evidence of their 
origin. After this it was scarcely possible to get any satisfactory indication from 
either of the air-thermometers ; but the experiment was continued, and one or other 
of them, when by any means the beads of disturbing liquid could be sufficiently got 
rid of for a time, was steadied to a constant temperature ; the other thermometer 
being observed when possible, and the reversals repeated as before. 
The same result was still obtained ; and on the whole, notwithstanding the defect 
which caused so much inconvenience, it was very decidedly established by the expe- 
riment that the Resinous Electricity carries heat with it in 'platinum. 
77 - \_Added Dec. 1856.] — After many unsuccessful trials on short brass tubes, first 
with air-thermometers of the metal itself and capillary glass tubes arranged as in the 
platinum tube (§ 69), and latterly with glass air-thermometers (§ 62) having very 
small cylindrical bulbs, the following conclusive experiment was made a few days 
ago. Four of the large double cells, connected to form a single Daniell’s element, 
exposing 10 square feet of zinc to 17 square feet of copper, were used to send a 
current through a piece of brass telescope tube 6 inches long, ^ of an inch diameter, 
and ground as thin as it could be without breaking it up by emery-paper, over the 
length of 3^ inches which was left between the near sides of gutta-percha coolers, 
fitted to it in the manner represented above (see fig. 11, § 72). Streams of water 
being, as in other experiments, kept running through the coolers, and the regulating 
break (§71) being used to keep the liquids within range in the tubes of the air- 
thermometers, a small mercurial thermometer pressed against the middle of the 
brass tube, with its stem and scale projecting out through the cotton wool, indicated 
from 190° to 195° Fahr. (90°*6 Cent.). 
The regulator was not used so much as it might have been with advantage ; but, 
notwithstanding great unsteadiness in the indications of the two air-thermometers, 
the observations showed decidedly, after each reversal of the current, a cooling effect 
on the thermometer next the entering stream, in every case in which the irregulari- 
ties were not so great as to make a comparison impossible. This effect is manifest 
from the following four cases, selected merely as being those in which one of the 
