204 Proceedmgs of Royal Society of Edinhurgh. [sess 
copper occluding hydrogen, the effect was too small to be observed, 
probably because of the small (Quantity of the gas occluded, as well 
as the inadequacy of the process adopted. 
The iron wire I used was 0’94 mm. in diameter, and in length 
about 120 cm. After being straightened by annealing, it was put 
into an air-tight glass tube of diameter about 3 cm., the air being 
pumped off from it and replaced by carbonic oxide or carbonic 
acid gas. I then passed a strong dynamo current through the wire 
to make it yellow-hot for a few minutes, and allowed it to cool in 
these gases. At the same time, the passage between the gas-holder 
and the tube was left open, through a drying-tube. To make the 
hydrogenised iron, I employed the electrolytic process; and to 
get the distribution of hydrogen as uniform as possible, the wire 
was loosely wound round a thin circular board of wood and plunged 
into acidulated water. The platinum electrode was placed nearly 
over the centre of the board. The wire which occluded either of 
the compound gases showed a light-blue, crystalline, lustreless, 
fibrous surface, not different from that of wire which had had a 
dynamo current passed through it and had been allowed to cool in 
the air. In these cases the surface of the wire became very brittle 
and easily separated from the central portion. On the other hand, 
the iron which occluded hydrogen presented a surface of a light- 
brownish colour at first, which gradually became dark and finally 
almost black as the electrolysis continued longer, but always became 
bright when wiped with soft cloth. This colour disappeared by 
heating, and the wire recovered its ordinary iron tint. 
In the first method tried, I fixed the wire in the axis of an iron 
tube. One end of the tube was clamped to the wire and heated in a 
charcoal furnace, and five junctions were made in different parts of the 
wire, as described in a previous paper by Prof. Knott and myself.'^ 
But after v^eeks of experimenting, the difficulty of determining the 
places at which the important change takes place compelled me to 
abandon this method. Kext I made five separate triple junctions, 
each of thin iron wire, thin German-silver wire, and the iron wire 
charged with one of the gases, and put them in an oil-bath, and 
observed the five different thermo-electric currents in turn. But 
with this arrangement it was difficult to follow the change closely. 
* Journ. of the Coll, of Sc. Imp. Univ. ToMo, vol. iv. pt. ii. 
