208 
Proceedings of Royal Society of Edinburgh. [sess. 
iron. It must be here noticed that the amount of carbonic acid 
occluded in iron is very much less than that of carbonic oxide or 
hydrogen. 
II. Thermo-Electric Properties of Hydrogenised Eon. 
For the direct measurement of the thermo-electric current, I 
divided into two a piece of iron wire of diameter mm. One of 
the pieces was put for three hours into the electrolytic bath, through 
which a current from three Daniells was passed. The thermo- 
electric current was measured, as before, by a double-coiled delicate 
galvanometer, which was gauged immediately before each experi- 
ment. The temperature of the hotter junction was measured by a 
thermometer dipped into a test-tube filled with oil, which again 
dipped into a large oil-bath with a metal cover — the junction being 
dipped into the test-tube along with the thermometer. The bath was 
heated from below by gas, the rise of temperature being controlled 
by two burners. Before the reading of the scale of the galvano- 
meter the burner was withdrawn, and the oil in the bath briskly 
stirred. After the thermometer had risen a few degrees, the thermo- 
electric circuit was closed, and after a few seconds the thermometer 
was read. Then the right and left deflections on the galvanometer 
were read, and the thermometer read again. These readings formed 
one set of observations. The maximum rise of temperature during 
one such set was 2° and the minimum rise 0°‘8. This process was 
found more convenient and accurate than the process involving the 
use of a thermometric junction. 
The thermo-electric circuit with the galvanometer contained no 
extra resistance or shunt. The galvanometer coils were joined in 
multiple arc, which reduced the galvanometer resistance to one- 
fourth. Yet the deflections were very small, ranging from four 
divisions of the scale to fourteen divisions, which correspond respec- 
tively to 8 and 28 microvolts. This smallness of the thermo-electric 
current and of the galvanometer deflections rendered the results of 
the experiments a little irregular. The following experiment is one 
of two, which were made after many preliminary experiments of 
the same kind. The temperature of colder junction was 26° C. : — 
