256 PROCEEDINGS OF SECTION A. 
eighteen divisions was obtained with this cell using a 40-cell Clark 
battery, but the deflection rapidly decreased to zero, and after that 
no conductivity could be detected. 
In order to combine the advantages of the two methods, 7.e., to 
have great insolubility and pure uncontaminated sulphur the 
following method of manufacture was finally used for the sulphur 
cell. 
Two glass rods 7:7 cm. long and :45 cm. in diameter were fixed 
parallel to one another, and with their centres 3 cm. apart by 
means of two aluminium cross bars, through which the glass rods 
passed. The length of the glass rods between the aluminium bars 
was 6:2 cm., forming a sort of Oxford frame upon which to wind 
the platinum wire. Each electrode was formed of twenty com- 
plete turns of platinum wire wound upon this frame, the total 
length of wire to each electrode being 160 cm. There were, 
therefore, altogether forty complete turns at an average distance 
of 1-5 mm. from one another. After the frame had been very 
carefully cleaned the wire was wound on, the electrodes being 
joined, one to each aluminium bar. ‘The frame was provided with 
an aluminium handle about 6 cm. long. 
The sulphur used for this cell was once distilled Chance sulphur. 
This was heated to melting, and the frame immersed in it. The 
temperature was then raised till the sulphur was so thick that on 
withdrawing the frame the latter was covered thickly all over 
with a mass of viscid sulphur. It was then suddenly immersed. 
in distilled water and left till cold. In this way the whole frame 
was covered with highly insoluble sulphur, the electrodes being 
shielded from surface action by a thick coat of sulphur. The 
conductivity of this cell was small in comparison with that given 
by some of the cells obtained in the other way, another indication 
that some at least of the cells had been spoilt by the glass-rod 
process of burying the wires, for some of them were made with 
pure sulphur. 
The cell was first tested for polarisation by the quadrant electro- 
meter. No certain positive effect could be observed, though 
owing to the residual charges in the keys necessarily used, no 
very satisfactory experiments could be made. 
For the galvanometric study of the cell it was hung by means 
of its electrodes in the thermostat. 
With forty cells the elongation upon reversing the current 
through the galvanometer was fourteen divisions. This elongation 
fell regularly in half an hour to five divisions, the temperature 
being 25°5°. 
The oil-bath was then heated, the galvanometer throw every 
few minutes being taken while the temperature was rising. 
