ELECTRICAL PROPERTIES OF PURIFIED SULPHUR. 250 
The sulphur slabs were cast in little zinc moulds. These 
were filled with sulphur, and heated till the sulphur was on 
the point of bursting into flame; they were then plunged into 
cold water. When the sulphur was sufliciently set, the dishes 
were cut off with a knife. When quite hard, they were scraped 
carefully all over to remove any surface contamination which 
might have been caused by exposure during heating or by 
immersion in the water. They were dried by warming and 
leaving some hours over phosphorus pentoxide. The weakest part 
of this method for the preparation of the cells is undoubtedly the 
melting in of the wires by means of hot glass, for it is difficult to 
ensure that the rod shall be perfectly clean, and the annealing 
produced is quite uncertain. 
At first the electrometer method was used to measure the 
resistance of the cells, but no very satisfactory evidence was 
obtained in this way. When a steady fall of electrometer 
deflection was observed on _ short-circuiting the electrodes 
through the sulphur, no constant difference in the rate of discharge 
was detected on exposing the sulphur to light. Both sunlight, 
gaslight, and magnesium light were used, but with entirely negative 
results. These results were affected by the imperfect shielding 
of-the sulphur from dusty air. 
Most of the experiments were made by the galvanometric 
method. The cells were hung by means of the ends of the 
platinum wires in jars containing phosphorus pentoxide to ensure 
dryness, the wires passing through small holes in the lids of the 
jars. The battery and galvanometer were connected in simple 
series with the cell. The battery consisted of forty Clark cells. 
It was invariably found that when first the circuit was 
completed there was great irregularity of conduction, and it was 
only after some time—a time which varied from about an hour 
with some cells to as much as a day with others—that a deflection 
could be obtained sufficiently steady for the experiments to be 
tried. 
Remarkable changes in the conductivity were observed. As 
regards the action of the light, however, the method of procedure 
was very simple. A steady deflection having been obtained in a 
darkened room, a piece of magnesium wire was lighted, so as to 
shine on the sulphurcell. Any change in the conductivity should 
have been accompanied by a change in the deflection. None was 
in any case observed. Owing to changes of zero of the galvano- 
meter, an apparent change of deflection to the extent of one or 
two divisions was sometimes noticed, but after repeated trials this 
change was found to be just as often in the direction of a decreased 
deflection as of an increased one. During numerous trials, 
however, when the galvanometer was exceptionally steady, 
perfectly steady deflections were obtained, both before and after 
