Chemical Affinity in terms of Electromotive Force. 27 
to the strip of platinum (and secured by a turn round the 
upper end of the narrow limb and a drop of sealing-wax) 
were bent over so as to dip into mercury-cups, a number of 
which were arranged in the arc of a circle round two others, 
like those figured in Part V. ; so that any consecutive pair 
of cups could at will be connected with the electrometer by 
the double switch. 
Either through a natural repulsion between bright zinc and 
the mercurous-sulphate paste, or through the formation of a 
faint film of grease &c. on the zinc from the file used to 
brighten it, it sometimes happened that the cell when finished 
would not work, contact not existing between the zinc and 
paste. It was found that this never occurred when the 
brightened zincs were washed successively with ether, alcohol, 
and saturated zinc-sulphate solution just before immersion in 
the paste. 
134. On comparing together a moderately large number of 
cells (upwards of fifty) thus prepared with different specimens 
of mercurous sulphate, readings being taken two or three 
times a week for some three months, the following results 
were obtained : — A slight rise in E.M.F. was often observed 
during the first few days after construction ; but at the end 
of a week at most the values became constant, and remained so 
(the temperature being constant) for long periods of time. The 
maximum variations observed between the average results of 
the series of observations for any two given cells were slightly 
less than that found to exist by Clark (whose highest and 
lowest values are respectively 1*4651 and 1*4517 volt, giving 
a difference of -0134 volt, or upwards of 0*9 per cent.). Taking 
the average of the whole set as 100*00, the maximum variation 
between two single cells did not exceed *010 volt, or 0*7 per 
cent., each cell possessing a value lying between 99*65 and 
100*35. Even amongst cells set up at the same time from 
absolutely the same materials, extreme differences of as much 
as 0*005 volt = 0*35 per cent, were sometimes observed, 
although usually the difference did not exceed *002 or "003 
volt and was frequently almost inappreciable. 
Much greater differences, however, were found to exist 
when the zinc-sulphate solution was not completely saturated 
with that salt, the variation produced being of this kind, that 
the iceaker the solution the higher the E.M.F. of the cell, the 
difference being approximately proportionate to the amount 
of dilution, and amounting to upwards of 2*0 per cent, of the 
value when considerably dilute zinc-sulphate solution was 
used. The details of these observations will be discussed in a 
future paper, along with those of similar experiments made 
