[BARNES & JOHNSON] FALL OF POTENTIAL METHOD 137 
to flow from the mains and the temperature carefully observed. The 
temperature was maintained constant during a set of readings, but 
it varied somewhat for the different measurements on account of 
the fact that the water in the mains became warmer as the 
summer weather advanced. In correcting the results to a uniform 
temperature a temperature coefficient was obtained from a special 
set of readings. This value was found to agree with the coefficient 
given in Kohlrausch and Holborn. It is our intention in conducting 
any further experiments to pay particular attention to keeping the 
temperature of our bath constant during a series of measurements 
with different concentrations. The electric current was supplied 
from a storage cell and was passed through a standard 100-ohm coil 
connected in series with the cell. The Thomson-Varley slide poten- 
tiometer was connected in parallel with the cell and standard coil, 
and being of such high resistance, 100,000 ohms, only shunted a small 
portion of the current. To avoid the errors due to polarization we 
arranged a make and break key in the main circuit shown at K! 
and a reversing key shown at K. The galvanometer was permanently 
connected to the sliding contact on the potentiometer, and to a 
contact which could be passed successively to the mercury cups 
d, e, b, a; these connected with the two potential terminals on the 
cell and the ends of the 100-ohm resistance coil. Since the fall of 
potential through the cell and standard, when the current was flow- 
ing, was equal to that on the potentiometer, equipotential points 
could be found between d, e, b, or a successively and the potentio- 
meter. The fall of potential from d to e was then represented by 
the difference in the two readings on the potentiometer, and the fall 
of potential from b to a was likewise represented by two correspond- 
ing readings. As long as the current remained constant the resist- 
ance of the electrolyte represented by a column of liquid between 
two planes at right angles to the axis of the tube and fixed by the 
position of the potential terminals was given by the simple expression. 
R= d,—d, 
d.—d, 
In taking the readings, instead of allowing the current to flow 
and taking the position of the balance or zero current through the 
galvanometer by making and breaking the galvanometer circuit, the 
main current was made momentarily, and the deflection of the gal- 
vanometer observed before the cell had time to polarize. A second 
reading was then observed by reversing the main current at the 
reversing key. On account of rapid polarization the galvanometer 
would always show a deflection, but in case the balance was such 
Sec. III., 1902. 9. 
x 100 

