CLARK CELL AS A STANDARD OE ELECTROMOTIVE FORCE. 
589 
of a Clark cell. On May 5, it was — 12, and on May 25, ff; 0, the standard being 
an old carefully prepared small cell agreeing closely with a number of other cells. 
About a dozen comparisons were made between the end of May and July 11, and it 
was within a few parts in 10,000 of the standard always.” 
These last two cells could be immersed in a water bath, and were always so used, 
side by side. The two Rayleigh cells Nos. 1 and 2, were enclosed in two large test 
tubes, and these were immersed in water. In some of the earlier experiments they 
could not conveniently be put into the same bath as Nos. 90 and 97 ; in the later 
experiments the four cells were side by side. Even when this was done there was 
some slight uncertainty as to the temperatures of Nos. 1 and 2, on occasions when the 
temperature of the air differed by 1° or so from that of the bath ; for then the tem¬ 
perature given by a thermometer placed in the test tubes with the cells, differed 
somewhat from that of the water. In these cases the temperature of the thermometer 
in the tube was taken as the temperature of the cell. The other cells could not be 
immersed, and their temperatures were taken by a thermometer supported close to 
them. As in all cases, with one exception noted below, in which experiments were 
made, the air temperature varied very slowly, the uncertainty introduced was not 
large. In the numbers given in the Table, allowance has been made for the difference 
in temperature, if any, between the cells and the standard. The numbers are reduced 
on the assumption that the standard has the same temperature as the cell, and that 
its E.M.F decreases by 4’3 of our units per 1° C., and give the excess of the E.M.F. of 
the cell in question over our Standard No. 1 in all cases except those in which there 
is a special note. 
It will be seen from the Table that the Cells 1, 2, 90, and 97, remained very close 
together throughout, and that too, over a considerable range of temperature. An 
exception must be made for No. 97 on November 3, when it was unaccountably low. 
Finding it remain so, we removed the zinc by withdrawing the glass stopper, and 
rubbed it on a clean silk handkerchief. We also dried and rubbed with emery cloth 
the upper part of the cell and the ebonite terminals to which the wires were attached. 
After this the cell recovered its value. 
The experiments on November 19 were made with the object of testing the effect of 
a rather sudden change of temperature. The heating apparatus in the room had been 
turned off for about a week previous to November 18. After the observations of that 
day it was turned on, and the temperature of the bath rose about 5° C. in 24 hours. 
It will be seen that the cells are all wrong. The large cells 90 and 97 are too high 
(assuming the standard right). They have the E.M.F. corresponding to a lower tem¬ 
perature, lower by about 2 0, 5 C. The liquid which was saturated at 8 0, 5 has not had 
time to reach the state of saturation corresponding to the new condition. But the 
standard is itself too high, for No. 2 has fallen with respect to it bv some seven divisions. 
Now No. 2 is a smaller cell than No. 1 and contains more liquid. The zinc 
sulphate in No. 1 is very full of crystals, and it is reasonable to suppose that it takes 
