916 
certainly hold the more exactly as the temperatures considered are 
lower. Thus I think it highly probable that the resistance vanishes 
proportional to 7“ at the absolute zero, as — according to DrBiiE — 
c‚ vanishes in the same way as 7”. The following Table I shows 
that the resistance of gold at hydrogen-temperatures *) exhibits this 
proportionality with a fair approximation, if an additive resistance, 
caused by traces of impurities, is taken into account. 
TAB LEES 
T | Observed | 0.2687 + Observed — 
resistance Au | 5.56.10—7 74 Comp. 
| | 
| 14.18 | 0.2910 | 0.2913 —0.0003 
15.83 0.3037 0.3037 0 
| 17.30 0.3190 0.3185 + 5 
19.00 0.3412 0.3412 0 
20.35 0.3621 0.3644 —0.0023 
3. The new formula. It is easy to put down an algebraic form 
which shall be proportional to 7’ for large values of the variable 
7, but proportional to 7 for very small values. A fraction such as 
Ts 
Wa En = 
af? + 67? + cl +d 
(1) 
will do. For a large value of 7’ we can develop into descending 
powers: 
for a small 7’ into ascending powers of 7’: 
Ve de (5 ae SE. Se 
d ad’ a) ide 
From these formulae we see that the coefficient a determines the 
dependence at high temperatures, d that at low temperatures, while 
the coefficients bande determine the way in which the two regions 
are linked together. n 
In representing the platinum-resistance it would seem advisable 
1) From KaAMmERLINGH Onnes and Horsr, Leiden Comm. 142 these Proceedings 
17, 508. 
