PROFESSOR THOMSON ON THE ELECTRO-DYNAMIC QUALITIES OF METALS. 701 
specimens (Pg, Pj, Copper, Iron) referred to above. Taking then the platinum 
specimens Pj, Pg, P 3 as standards, and using besides ordinary copper and iron wires, 
I commenced investigating their thermo-electric relations to as many other metals as 
I could obtain. 
86 . In experiments to determine temperatures of neutrality, the first apparatus 
which I employed for regulating the temperatures of the two junctions, consisted of 
copper vessels placed side by side in which oil could be raised by gas-burners as 
high in temperature as the mercurial thermometer can be used, that is to 340° or 
350° Cent., or somewhat above the boiling-point of mercury. To do away with 
irregularities from the flame and cold air playing unequally on the sides of these 
vessels, smaller ones were placed on wire stands within them, and were completely 
filled and surrounded with oil. In each experiment a wire or slip, about 18 inches 
long, of one of the metals to be tested, had somewhat longer wires or slips of the 
other soldered to its ends. The compound conductor thus constituted was bent into 
such a shape that the two junctions of the metals could be placed near the centres of 
the oil-baths ; it was supported in this position, carefully insulated from touching the 
Fig. 13. 
copper vessels and from all other metallic contacts; and thermometers were put 
with their bulbs in the oil as close to the junctions as possible. The gas-furnaces 
were applied below and round the sides of the large copper vessels, so that they 
could be regulated to any desired temperatures. 
87. After this apparatus had been used in several experiments, and neutral points 
between copper and iron, copper and Pj, lead wires and Pj, and brass and Pj had been 
determined, I saw reason to alter the arrangements in various respects, and had 
another apparatus constructed, according to the following description. 
88 . Two small oil-baths were made, each of an outside partly cylindrical and 
partly plane sheet of copper, and a concentric copper tube 5 inches long and i^oths of 
4 z 2 
