Diffusion of Hydrogen through Hot Platinum. 27 
We have seen that /, increases rapidly with the temperature, 
so that at high temperatures the second term would become 
the predominant one. Thus on this view the total solubility 
of hydrogen in platinum would increase rapidly with the 
temperature. According. to Roscoe and Schorlemmer’s 
‘Chemistry ’ (vol. 1. p. 139), platinum at a red heat absorbs 
3°8 times its volume of hydrogen, whereas at 100° C. it only 
absorbs.°76 times its volume. This is therefore a confirmation 
of the preceding hypothesis. 
An interesting application of these results may be made 
with the object of elucidating the mechanism of the negative 
ionization produced by hot platinum in an atmosphere of 
hydrogen, which has been investigated by H. A. Wilson *. 
Wilson showed that the negative ionization produced by hot 
platinum was increased to thousands of times its former 
value by immersion in an atmosphere of hydrogen. The 
part due to the hydrogen appeared to be proportional to 
the gas-pressure over most of the range, and varied with the 
temperature according to a formula (a6@2e—°", a and 6 constants) 
first given by one of the authors f, and shown to apply 
generally to the negative ionization from hot metals. Wilson 
also showed by experiments on the lag of the electrical leak 
behind changes in the pressure and temperature of the wire, 
that the ionization was probably not produced by the 
external hydrogen, but by the hydrogen dissolved by the 
platinum. This was in agreement with the argument, given 
by one of the authors in the paper just cited (p. 546), based 
on the magnitude of the negative leak from hot carbon, that 
the ionization in case of that substance at any rate could not 
be due to the external gas, 
We have seen that the preceding experiments indicate 
that the concentration of the dissociated hydrogen in the 
platinum is proportional to the square root of the external 
pressure. The ionization therefore cannot be a function of 
the dissolved atoms singly, but must be either due to their 
interaction (recombination or production by dissociation), or 
it may be due to some action of the dissolved molecules 
singly. If, following Wilson, we may assume that above a 
certain temperature the total solubility of hydrogen in 
platinum begins to decrease, we can explain all the results by 
supposing that the negative ions are produced by the 
collisions of hydrogen atoms in the surface layer, or in 
the body of the platinum. We see at once that the concen- 
tration N of the hydrogen atoms inside is (4,C,)?= ( “ty. 
* Phil. Trans. A. vol. ccii. pp. 248-275. 
fT Phil. Trans, A: vol. eci. pp. 497-549. 
