226 Mr Chree, On the effect of an electric [May 21, 
in all the operations dealing with the samples taken on one 
occasion. As the object aimed at was merely to detect differences 
between these samples, and the same plan was followed in treating 
each, this could introduce no error. 
The argentic nitrate solution usually contained about 10 gram- 
mes of salt in 500 c.c. of water. This supplied enough for a large 
number of experiments. 
No great care was exercised in keeping the solution at an 
exact strength, and though tightly corked it seemed to change 
slowly. During the treatment however of the samples taken on 
the same occasion no sensible change could occur, and as the 
titrations were always repeated and the samples taken in varying 
order even this small effect must have been pretty well elimi- 
nated. It was not attempted to determine the absolute amount 
of salt in solution but only to detect differences between the 
samples. 
In carrying out the titration a drop or two of a solution of 
yellow chromate of potassa in distilled water was used as an 
indicator, The burette from which the argentic nitrate was run 
was graduated to ;,¢.c., and could be read with fair accuracy to 
4 of a graduation. In treating the sodium and potassium chlorides 
the argentic nitrate was kept very nearly at a constant strength, 
and the average of the results obtained is included in the follow- 
ing table: 
Before current Time current During current Interval After current 
lm — run, in — elapsed, in ——— —_— 
Temp. Vol. AgNO; minutes Temp. Vol. AgNO; minutes Temp. Vol. AgNO; 
NaCl 192° 18°99 115 232° 19:02 60 224° 19-08 
KCl 182° 13:47 46 163° 13°49 28 152°) 13°55 
By “Vol. AgNO,” is meant the number of c.c. of argentic nitrate 
required to titrate 20c.c. taken from the half litre to which the 
10 c.c. taken from the saturated solution had been diluted. “Time 
current run” is the interval between starting the current and 
taking the sample during its passage. “ Interval elapsed” is the 
interval between stopping the current and taking the last sample. 
The other headings will explain themselves. In every case it is 
the average of a number of experiments that is given. 
It is obvious from the table that the effect of the current must 
be small. The accuracy of the method would scarcely justify us 
in accepting as proved any increase in the quantity of salt im 
solution during the passage of the current. On the other hand 
the increase after the current stopped seems unquestionable, though 
accompanied by a slight fall in temperature. This increase shewed 
itself in almost every experiment on the KCl. 
According to Storer’s data an increase of 1° at the temperature 
of the experiments should have increased the AgNO, equivalent of 
