( 1109 ) 
metal; but moreover, such great uncertainty exists in our rough 
approximations as to confine our most probabie result in the mean- 
time to this: That within the limits of experimental error (the degree 
of purity attainable) the resistance of pure gold is already zero at 
helium temperatures. 
Let us return to platinum. The wire Pt, seems to be less pure 
than Pt; (see table V of Comm. N°. 990), and, moreover, the fixing 
of the wire on the glass may give rise to undesirable effects. 
By putting the additive resistance once more constant, to a first 
approximation, extrapolation gives for Pig Are corresponding line 
shown in the graph. But still, the resistance of Pt; may not without 
further comment be regarded as the resistance of pure platinum. 
A wire of greater diameter used by HoLBorn gave a greater relative 
decrease of the resistance from 0° to — 191°C. If we extrapolate 
these values to lower temperatures the resistance remaining at helium 
temperatures, and independent of any further change of temperature, 
would be nearer zero. One may ask if it is not possible to put the 
difference between the two wires obtained from Herragus inversely 
proportional to the thickness and in that way deduce a value for 
pure platinum unaffected by the individual treatment of each; but 
this method would lead us too far into the region of pure conjecture. 
But still, the conclusion seems to be fully established that the resis- 
tance of pure platinum is, within the limits of experimental error 
— the attainabie degree of purity — already zero at helium temperatures. 
§ 3. The change with temperature of the resistance of pure metals 
at low temperatures. 1 was formerly of the opinion that the resistance 
of pure metals reaches a minimum as the temperature is diminished, 
and then, as the temperature sinks still further, again begins to 
increase and becomes infinitely great at the absolute zero; but now 
it seems to me to be more probable that, even before the absolute 
zero is actually reached, the resistance if not zero, has become so 
extremely small that it practically vanishes, and that this remains 
the case for further lowering of the temperature. 
In view of this result, then, we must also abandon the theory 
that has served for years as a guide in our Leiden researches upon 
the resistance of metals at low temperatures, according to which it 
was imagined that the resistance would attain a minimum as the 
temperature was lowered and would become infinitely great at 7’ = 0, 
in consequence of the assumption that the electrons which are the 
actual conductors in metals would, as was expressed by me in 1904, 
begin to precipitaie on the atoms as a vapour on being cooled to 
