32 Messrs. Wright and Thompson on the Determination of 
checked against one another from time to time, or compared 
with Daniell cells set up as described in Part V., with amal- 
gamated pure zinc and electro-copper plates, and pure zinc- 
and copper-sulphate solutions of the same molecular strength*. 
No permanent changes in the values were observed (outside 
of the limit of the errors of observation) in the case of those 
cells which were so well sealed that neither evaporation took 
place, nor passage outwards of the fluid by capillary action 
through cracks in the sealing material. Vacuum-cells were 
thus kept unchanged for upwards of two years, as also were 
some normal Clark cells that were completely imbedded in 
paraffin-wax. In several cases, however, where the cells were 
not completely imbedded, but were only sealed up by a plug 
of paraffin-wax poured in at first round the zinc plate and the 
cork &c. supporting it, cracks formed sooner or later either in 
the paraffin-wax itself or between the glass and the wax, so 
that the fluid passed out through the cracks by capillary action 
and formed an efflorescence outside the cell. In some cases 
the action went on to such an extent as to leave the zinc 
wholly exposed, no contact finally existing between it and the 
paste : such cells were of course utterly spoilt. In other in- 
stances the zinc was only partially bared : in these cases the 
E.M.F. of the cell remained almost unaltered when saturated 
zinc-sulphate solution was employed in the first instance, but 
was lessened when unsaturated solution was originally used, 
owing to the evaporation and concentration which went on 
simultaneously with the capillary action, or subsequently to 
the commencement thereof. For example, two cells set up 
* A large number of observations on tbe E.M.F. of Daniell cells have 
shown that, wben proper precautions are taken in setting up the cells, a 
very considerable degree of constancy in value is attainable, so that such 
cells serve as good practical standards ; but that if these precautions are 
neglected, variations amounting to 5 per cent., and even more, map ensue. 
The essential precautions are : — first, that pure solutions of zinc and copper 
sulphates containing no free acid should be used, each beinir of the same 
molecular strength (Y. e. practically of the same specific gravity ; conve- 
niently the molecular strength may be near to MS0 4 , oOH.,( >) : secondly, 
that the solutions should be in separate vessels, united when required by 
an inverted U-tube, the mouths of which are covered with thin bladder 
(Raoult's form of cell ) ; thirdly, that the plates should be pure zinc amal- 
gamated with pure mercury, and copper recently electro-deposited from 
pure sulphate solution — the wires serving as electrodes, and their junctions 
with the plates being coated with gutta-percha, so that no part of the plate 
or wire is simultaneously in contact with both liuid and atmosphere ; and 
fourthly, that, if used to generate a current, the current-density must not 
exceed'some 5 microamperes per square centimetre, so that with plates 
exposing 20 square centimetres the total resistance in circuit must be at 
least 10,000 ohms, if exposing 10 square centimetres 20,000 ohms, and 
so on. 
