CLARK CELL AS A STANDARD OF ELECTROMOTIVE FORCE. 
597 
This set then shows the tendency of Clark cells made with different materials to 
settle down to the same value in time. 
§18. Cells set up by S.S. in March and April, 1891. 
This set of cells are all in short test tubes of the Rayleigh pattern, and were 
set up by one of us (S.S.) in March and April, 1891. They contain materials which 
are indicated as above by Greek letters. Some of these are the same as w r ere used by 
Mr. Wilberforoe ; particulars of the others are given. \. This paste was prepared 
from Harrington’s specimen of mercurous sulphate, which was washed by decan¬ 
tation and on a filter paper until the washings were no longer acid. During this 
operation the paste became yellow from the conversion of the mercuric sulphate present 
into basic mercuric sulphate and sulphuric acid. The basic mercuric sulphate or 
turpeth mineral being insoluble, remained with the mercurous sulphate, whilst the 
acid passed away with the washings. After washing, pure mercury was mixed with 
the paste, and heated with it in a pan of boiling water, with the view of attacking 
any yet undecomposed mercuric sulphate. The paste was allowed to remain with the 
mercury for two or three days, and then it was dried at the room temperature. To 
prepare it for use in the cells it was rubbed with saturated zinc sulphate to the con¬ 
sistence of a cream. The paste, therefore, contained mercurous sulphate, zinc sulphate, 
mercury, and traces of basic mercuric sulphate. This was an attempt to prepare the 
paste in accordance with instructions given verbally to one of us (R.T.G.) by Dr. 
Alexander Muirhead, to whose great knowledge of the subject we are much indebted. 
Different mercury was used in some of these cells. This (p) was a sample treated 
with nitric acid for some days, then washed and afterwards distilled in vacuo. 
It will be seen from Table V. that the cells all start by being too low, but that in 
from three to four weeks of the time they were set up, the first seven, those made in 
March, have attained a steady value not far from the standard. The observation of 
52, on August 7, is abnormal. The first six cells were lent for some other experi¬ 
ments in August. 
The cells 50 and 51, made of the same materials as 25 and 26. agree with them in 
being too low. At first, the cells in which the mercury k was used, are a good deal 
too low. Nos. 48 and 49 are of the same material as 27 and 28. 
The first two tests recorded on these cells, made immediately after the cells were 
set up, were very rough, being performed by joining the cell up in opposition with 
another Clark, and passing a current from the two through a galvanometer and 
20,000 ohms. They were merely intended to see that there was no great error. They 
give us, however, some important information, for the E.M.F. sank rapidly between 
March 17 and March 21, but regained its normal value by April 12. 
The observations, given in Table IV., on Mr. Wilberforce’s cells, Nos. 25-35, 
were made two days after they were set up, and we do not possess for them in such 
