194 
DES, A. MATTHIES SEN AND C. VOGT ON THE INFLUENCE OF 
The values in the last column were obtained as follows : take for instance that of 
gold. The mean deduced value (column 1) for its conducting-power at 0° was 83T7, 
under the supposition that the percentage decrement in its conducting-power between 0° 
and 100° was 29-307 ; the percentage decrement, however, found for pure gold was 28-30 ; 
we must therefore recalculate the deduced value to obtain a more concordant one, and 
this may be done with the help of the proportion 
First, 
then 
ence 
P:F::M 100 o:M; 00 » 
83-17x70-693 
JM-ioo 0 — 10Q 
=58-80, 
29-307:28-30 :: 58-80 : M' 100 , ; 
M' 100O = 56-77, 
and therefore the deduced value at 0° 
( 5 ) 
56-77x100 
71-7 
= 79-20. 
Reducing the others in the same manner, we are struck with the coincidence of these 
values with those really found for the annealed wires by experiment ; in fact we must 
assume that the values deduced for the conducting-power of metals are those of the 
annealed wires, even when hard-drawn ones are experimented with. What the deduced 
values for the conducting-power would be when using annealed wires of impure metals 
we are unable at present to say, for no determinations have been made in this direction. 
It must be remembered that the effect of annealing on the conducting-power of alloys 
is very small, so that the deduced values from those found for the annealed wires would 
not be very different from those deduced from the hard drawn, assuming, as we have 
done in the former part of this investigation, that the percentage decrement in the con- 
ducting-power between 0° and 100° of hard drawn and annealed is the same. 
Having thus -proved that, by using the expression 
P:F::M 100 o:M' 100O , (5) 
we may deduce the conducting-power of the pure metal from the impure one, when the 
observed values do not differ from those calculated by more than 20 to 30 per cent, 
we will now proceed to give the results of some experiments with impure metals where 
the conducting-power of the same metals in a pure state has not yet been determined. 
Tables XXIV. and XXV. contain the results. 
