. 
TRANSACTIONS OF SECTION A. 397 
Dr. F. A. Wolff we have been glad to actively co-operate with the Bureau 
in the work it has been doing along these lines. 
I. The Reproducibility of the Cells when prepared according to the 
Specifications of Wolff and Waters.—The mean of fifteen Clark cells made 
in this way differed, a few weeks after their construction, from the mean 
of the reference cells of the Bureau of Standards by less than 14 microvolts. 
The average deviation of our cells from their own mean was not more than 
13, while the maximum deviation was only 31 microyolts. 
In the case of the Weston cells the figures were similar, but showed 
somewhat better agreement in every case; the difference between the mean 
of thirteen cells and the mean of the reference cells at the Bureau of 
Standards being in this case 4 microvolts, the average deviation of the cells 
from their own mean being 8, and the maximum deviation being 22. 
Some of the Clark and Weston cells were made in Washington by one of 
the writers and subsequently transported to Montreal. A means of direct 
comparison with the values of the cells at the Bureau of Standards was thus 
obtained. This comparison has recently been checked by a second inter- 
change of cells. 
We have also had in our possession six Weston cells made at the National 
Physical Laboratory in London. One of these was damaged in transit; the 
mean of the other five differed from the mean of our other cells by only 
5 microvolts. 
Il. The Effect of Introducing Slight Simplifications into the Preparation 
of Materials.—About twenty-five cells were constructed in order to examine 
the effects of slight simplifications in their preparation. It was found that 
the ingredient of main importance is the mercurous sulphate, which must 
always be prepared with great care. When unpurified mercurous sulphate 
was used in a cell, the average electromotive force was found to be from 
300 to 500 microvolts higher than that of the normal cell. The purification 
of the zinc and cadmium salts is not so important, and as the processes are 
somewhat tedious, especially in the case of cadmium sulphate, it is interest- 
ing to see what accuracy may be obtained with various samples of chemically 
pure commercial sulphates. It was found that no Clark cell made with this 
modification differed by more than 50 microvolts from the mean of our 
normal cells, while no Weston cell differed by more than 100 microvolts. 
Such cells are therefore sufficiently accurate for all practical work. 
The presence of basic oxide, oil, or small quantities of organic impurities 
was found to exert only a very small influence on the electromotive force. 
III. The Relative Value of Cells set up according to the Board of Trade 
Specifications and those set wp according to the Specifications of Wolff and 
Waters.—Ten Clark cells of the old ‘ test-tube crystal’ type were prepared 
from a number of different samples of chemically pure commercial materials, 
for the purpose of ascertaining how much they differed in voltage from those 
set up according to the modern specifications. The average electromotive force 
of these cells during the first seventy-five days, neglecting the first day, was 
0:31 millivolts higher than the mean of our normal cells. The average 
deviation of these cells from their own mean was considerable, about 
0:06 millivolts, which suggests that there might be a possible variation of 
three or four in the last figure of the mean for different batches of ten cells 
set up with different materials. It is therefore in good agreement with the 
value (0°00030 volts) given by Wolff and Waters. 
IV. The Ratio of the Electromotive Force of the Weston Cell to that of 
the Clark.—The determination of this ratio, which had also been obtained at 
the Bureau of Standards, was made with the object of furnishing a further 
check on the reliability of the comparison between our cells and those of the 
Bureau, and to give added assurance that no errors had been introduced by 
transportation. 
Five Clark cells were connected in series and placed in opposition to 
