1419 
the value at 0° C. (in the solid state) and reduced to about half 
of that to which we could go down in the January experiments. 
TABLE III 
Potential differences at the extre- 
mities of mercury threads car- 
rying a current. /= 20 cm. 
zr* = 0,0016 mm? for Wo 
=0,004 „ » Wi 
‚Current den-| Potential i 
sity in amp.| difference | 
per mm2. in microvolts, 
5CO AP 
625 | 260 | 12.9 0 
For Ws at a strength of current of 1 amp. the current density 
which in Wizo appeared to be the threshold value was not yet 
reached. A stronger current was applied. But now a special 
disturbance arose: on raising the current to 1.5 amp. so much 
Joure heat was generated by the current in the platinum wires 
joining the mercury leg, that this reached the thin mercury thread 
and brought it up to a temperature above the vanishing point. All 
this was accompanied by a rapid boiling of the helium, while the 
ammeter showed a strong falling off of the main current correspond- 
ing to a decided rise in the resistance. From the readings it could 
be seen that the resistance of Wso had risen to that which it has 
at hydrogen temperature. This time it seemed most probable that 
the potential differences could be attributed entirely to heat introdu- 
ced from outside, so that if this could be prevented it would be 
possible to bring at these lowest temperatures the highest limit for 
the possible residual resistance still nearer to zero. 
§ 6. Experiments with an apparatus arranged so as to be sure 
that no heat penetrates to the thread from places at a. higher tempe- 
rature than that of the vanishing point. 
A mercury resistance was made, suitable for observing the potential 
changes, when a current of 3 amp. went through the same mercury 
