260 
PROFESSOR H. A. WILSON ON THE EFFECT OF HYDROGEN ON THE 
cooling to the original temperature is often a little lower than before. After standing- 
cold the leak is usually higher on first heating the wire, and then falls off quickly at 
first and then more slowly. I think these variations may be ascribed to hydrogen 
dissolved in the wire like that in a new wire. The leak does not fall below a certain 
value at a given temperature, however long the wire is heated in a vacuum, even if 
the temperature is raised and lowered. Richardson’s experiments show this very 
clearly, and I have confirmed this result. Very often the leak rises slightly while the 
wire is being heated in hydrogen at a very low pressure. To explain these results on 
the theory that the large leak is due to hydrogen contained in the surface layer of 
the platinum, it is necessary to suppose that a very stable compound of hydrogen and 
platinum is formed in the surface layer which has a dissociation pressure less than 
about 0'002 millim. even at high temperatures. In addition to the combined hydrogen 
the wire may contain dissolved hydrogen, but this will be got rid of by heating in a 
vacuum. The absorption of hydrogen by platinum at high temperatures has been 
investigated by several people, and they have all found that part of the gas can be 
driven off by heating in a vacuum, while part cannot be driven oil’ in this way. It 
should be observed that the pressure of the hydrogen is never really reduced to zero 
by pumping out. It is possible that by reducing the pressure sufficiently the 
combined hydrogen could be removed, but if it has a finite dissociation pressure this 
must be very small A 
I think, therefore, that the difference between a new wire and a wire which has 
been heated in hydrogen at a high pressure is mainly due to the presence of combined 
hydrogen in the latter. 
When the combined hydrogen has been burnt out of an old wire by oxygen the wire 
gives the same leak in hydrogen at a low pressure as a new one, but the leak takes 
a long time to appear and does not fall much when the pressure is reduced. This 
seems to show that the old wire only dissolves the hydrogen very slowly. Continued 
heating in hydrogen and air appear, therefore, to produce a change in the state of the 
platinum. This change of state does not appear to affect the final value of the leak, 
and is not destroyed by heating in oxygen, which at once destroys the leak. 
This change of state may very likely be due merely to long continued heating; it 
probably consists of an alteration of the crystalline structure of the platinum. The 
formation of a compound of platinum and hydrogen in the surface layer, and its 
subsequent destruction by oxygen, might be expected to produce a change in the state 
of molecular aggregation of the platinum. Another thing which may have something 
to do with this change of state is the slow evaporation of the wire at high 
temperatures. 
In my paper on the electrical conductivity of air and salt vapours (‘ Phil. Trans., A', 
1901) I observed that the leak from hot platinum in air falls off during long continued 
[*Added March 20, 1908.—Since writing the above I have found that on heating to above 1600° C. at 
a very low pressure the large leak rapidly disappears.] 
