1414 
within a fall of 0°.02, was again confirmed with a resistance of 
about 50 2 (calculated for liquid mereury at 0° C.). 
Concerning the threshold value of current and the potential 
differences appearing at higher currents, i.e. the phenomenon to which 
the investigations were especially directed this time (Dec. 1911) 
results were obtained which correspond pretty well with the previous 
ones (Oct. 1911) if we assume that the origin of the phenomenon is 
in the resistance itself, and at the same time make the natural assump- 
tion that the potential difference increases with the current density, and 
with conductors joined in series is equal to the sum of the potential 
differences in each of these conductors. This is shown in the table, 
in which both series of observations are combined, and holds both 
for the minimum value of the current at which the potential difference 
appears, and the value of the potential difference at a given excess 
value of current and a given temperature. 
For we must remember that the previous resistance consisted of 
7 U-shaped tubes not all precisely similar, averaging 37 2, and 
the present one of one U-shaped tube of 5) @, while the lengths 
of the tubes did not differ much. The appearance of the potential 
difference was therefore, on our supposition, to be expected in the 
last case at a slightly smaller current than in the first; on the other 
hand, the greater length which was partly compensated by a greater 
section, made it probable that in the October experiments the potential 
difference at the same temperature and current would be a few 
times larger, though not as much as seven times. 
§ 4. Questions to which the experiments give rise. There were not 
sufficient data to make out whether the resistances used really differed 
as much as was thought as regards the opportunity of receiving heat 
through heat conduction from elsewhere, in particular Joure heat. 
It would however have to be regarded as a curious coincidence that 
this conduction of heat in conjunction with other causes had ied to 
such a close correspondence in the phenomena observed. It seemed 
much more probable that the pheromena were to be accounted for 
not by disturbances from outside, but by resistance arising in the 
thread itself. 
Where such a remarkable change in the condition of the mercury 
takes place as is shown by the disappearance of the ordinary 
resistance, the appearance of a “threshold value” dependent on the 
temperature naturally gave rise to the question, if we had to do with 
a deviation from Oum’s law ') for mercury below 4°.19 K. The electron 
1) | hope to return to the new and important theory of Wien, in a further comm, 
