concerned. In the preliminary measurements the substance 
Dracram IV. 
Top—Phosphorus Pentoxide on dishes (23.11.14—2. 12.14). 
Bottom—Strong Sulphuric acid, the best desiccant 
£ 
i 
oO. 
+ 02 
rr) 
> 
| 
So 
fs 
oC 
oO 
4 
| Sp Sine Seen ASaR eR Othe eAMLa ee Doak 
° ’ r 120 160 200 
40 80 
—— Time in hours 
was, as with sodium, bs ga out on several dishes and quickly 
introduced into the case. Diagram IV and Table I, Colu 
VI, show that the leak of the quadrants, which was initially a 
little below the permissible value, began steadily increasing on 
Dracram V. 
Effect of repeated doses of fer we Pentoxide (3 doses) 
(17.11.15—25.11.15). 
s 
= 
_ 
—~ Leakage in volts/min. 
) 40 80 120 180 200 
—— Time in hours 
sealing up the case, and reached double the maximum value in 
20 hours. No improvement was noticed even after 200 hours, 
as the leak remained between twice and three times the stan- 
dard permissible value. It might naturally be thought that 
