28 Messrs. "Wright and Thompson on the Determination of 
with other cells. It may, however, be here noticed that, so 
long as a cell containing unsaturated zinc-sulphate solution was 
protected against concentration by evaporation, and was only 
used in connexion with a quadrant-electrometer, its indica- 
tions remained perfectly constant for many months (the tem- 
perature being the same), precisely as was found with cells 
set up with saturated zinc-sulphate solution. 
Effect of Dissolved Air on the E.M.F. of Clark's Cell. 
135. Two series of experiments were made with the object 
of finding out how far the boiling of the mercurous-sulphate 
paste (as recommended by Clark) in order to remove dissolved 
air is essential. In one series a number of cells were set up, 
using fully aerated zinc-sulphate solution and unboiled mer- 
cury (exposed to the air under a glass shade for several days 
since preparation and distillation respectively); in the other 
the paste was boiled in a Sprengel vacuum produced in the 
cell itself for some time, the cell being then hermetically 
sealed, so as to reduce the amount of residual air to a mini- 
mum. In each case the average E.M.F. of the combination icas 
sensibly identical with that of an average ordinary Clark cell 
prepared as above described and containing zinc-sulphate solu- 
tion of the same strength as that contained in the combination. 
In order to prepare these hermeticalry-sealed cells a rather 
troublesome process was employed. First a piece of glass 
tubing, about 10 or 12 millim. in bore, was blown into a 
Y- shape, and the two limbs of the Y drawn out as represented 
in fig. 3; a zinc rod was then cast so that a thin platinum wire 
was imbedded in one end ; this zinc rod was brightened and 
Fif?. 4 
sealed up on the tube so that the platinum wire projected 
(fig. 4, a). By the aid of a glass funnel with a flexible capil- 
