6 Journal of the Asiatic Society of Bengal. [N.S., XVI, 
the value of this powerfully hygroscopic substance as a desic- 
cant was destroyed during the time taken in spreading it onthe 
dishes and introducing it inside the case. To avoid this dbjec- 
tion, and also to test the effect of repeated doses of the desic- 
large bottle containing anhydrous phosphorus pentoxide, and 
then quickly introduced into the electrometer case through the 
ted in the interior of the case which was then immediately and 
hermetically sealed by replacing the cover in the groove round 
e opening. In this way the substance hardly came into 
contact with the external air at all. This method gave results 
Diacram VI. 
Effect of introducing fresh charge of Phosphorus Pentoxide (A) or of 
tirring the mass already inside (B). (6.12.15—23.12.15). 
2 
—— Leakage in volts/min. 
° 80 160 
— Time in hours 
of considerable interest, which are shown in Diagram V and 
Column VII of Table I. Immediately after its introduction a 
however, immediately commenced to increase, and before 
becoming after 50 hours approximately constant, reached over 
double the standard value. A second charge of the desiccant 
produced almost exactly similar effects, which were again 
reproduced by a third charge. The results of a further trial 
with this substance extending over a period of 400 hours are 
given in Diagram 
in temporarily reducing the leak. On opening the case at the 
end of the last set of measurements it was observed that there 
