236 Walter Stiles and Ingvar Jorgensen. 
In his own experiments Whetham used a Wheatstone’s bridge 
arrangement, measuring his resistance by comparison with standard 
“resistances as in Kohlrausch’s method. Whetham himself ex¬ 
perienced various difficulties, which he endeavoured to surmount. 
Thus as regards the electrodes he found bright platinum gave good 
results with extreme dilution, but a too high resistance in other 
cases. On covering the electrodes with platinum black the results 
were variable, the platinised electrodes extract salt from the solution 
and deposit it on their surface. After heating the platinised 
electrodes red hot it was found possible to measure resistances 
varying from 10 to 50,000 ohms with an accuracy of 1 per cent. 
Other possible difficulties arise from the winding of the coils in the 
resistance box, which might give rise to errors due to their capacity 
x 
Fig. 3. For explanation see Text. B, battery ; C,, C 2 , two commutators 
rotating synchronously ; G, galvanometer ; X, resistance. 
and self-induction ; and also from the thermo-electric force between 
the revolving drum and the brushes of the commutators. The 
thermo-electric force however does not give rise to any serious 
error, while any error due to winding of the resistances can of 
course be eliminated by using suitably prepared resistances. 
So far this method has only been used to compare resistances; 
it would be better to apply a method that would measure the 
resistance or conductivity absolutely than one which merely 
compares resistances. This would avoid capacity and induction 
effects in the standard resistances, and owing also to the simpli¬ 
fication of the method, should enable a continuous record to be made. 
Our own attempts in this direction with the revolving commu¬ 
tator method indicate that great difficulty arises in trying to keep 
a constant contact between the brushes of the commutator and the 
