34 
CHEMISTRY: D. P. SMITH 
Proc. N. A. S. 
upon the condition of the occluded hydrogen at the time of observation. 
That the results were not materially affected by fluctuations of the 
measuring current is probably sufficiently attested by the regularity of 
the observations in each series. The precaution was nevertheless taken 
of substituting for the palladium wire a nearly equivalent manganin re- 
sistance, and conducting with this a similar series of measurements. The 
resistances found were in this instance constant to the fourth significant 
figure, showing that with the battery employed, which had a capacity of 
240 ampere-hours, and supplied only 4 milliamperes or less, at 2 volts, 
the variations were wholly negligible. 
Conclusion. — The fact seems to be established by the foregoing experi- 
ments that the temporary supplementary conduction exhibited by metals 
during cathodic occlusion of hydrogen, and for some time thereafter, 
is not of ordinary metallic character. Strictly regarded, this has been 
shown only in the case of palladium-hydrogen; but the great similarity 
previously shown 3 to exist in the comportment of palladium, of tantalum 
and of iron toward hydrogen, and of palladium toward oxygen, makes it 
probable that the supplementary conduction is of the same character in 
all cases. 
The outcome of the experiments is accordingly in conformity with the 
prediction based upon the first of the three explanations of the conduction 
which were enumerated above, and lends support to this conception. 
Upon either of the other two hypotheses the departure from Ohm's law 
would be difficult to understand. 
Obviously, it is to be expected that with measuring currents much higher 
than those here employed the supplementary conduction would approach 
a limiting value, owing to the inability of ordinary diffusion to maintain 
the concentration of hydrogen atoms in the regions from which these are 
driven by the current. Above this limit the dependence of resistance on 
current would not be observed. Unfortunately the likelihood of being 
able to test this further prediction experimentally seems small ; for it may 
be seen from the magnitude of the change of resistance in the range studied 
that its disappearance can be looked for only at very large values of the 
measuring current; and with such currents the heating effects would 
render doubtful any results which might be obtained. 
In concluding, acknowledgment should be made of the excellent as- 
sistance rendered in these experiments by Mr. Robert F. Mehl. 
* The assumption of the existence of molecular hydrogen within the metal seems to 
be made necessary by certain facts which are not here considered, and the present con- 
siderations will not be affected if the quantity of the molecular form is taken to be ex- 
tremely small, or even zero. 
1 Smith, D. P., and Martin, F. H., /. Amer. Chem. Soc, 38, 1916 (2577); Harding, 
E. A., and Smith, D. P., Ibid., 40, 1918 (1508). Summarized in Trans. Amer. Electro- 
chem. Soc, 34, 1918 (177). 
2 Newbery, E., /. Amer. Chem. Soc, 41, 1919 (1887). 
3 Harding and Smith, loc cit. supra. 
