1920.] Desiccants in Electrostatic Measurements. 7 
was only a slight change on the surface of the phosphorus 
pentoxide. 
It seems clear that moisture could not have been the 
cause of the increase of leak which was invariably observed. 
As in the case of metallic sodium, any drying action due to 
this substance is accompanied by chemical change, which may 
similarly be accompanied by increased ionisation. It is also 
possible that the abnormal increase of leak in this case is due 
to the presence of yellow phosphorus in the sample of phospho- 
rus pentoxide which was used. The chemical action of mois- 
ture on yellow phosphorus was found by Schenck! to produce 
an increase of ionisation. 
IV. QUICKLIME. 
phuric acid that it was considered worthwhile to investigate 
tions due to chemical action might be more suitable for our 
in insulation caused by it is only temporary. This is probably 
due to the great difficulty in keeping this substance absolutely 
anhydrous, but it also appears as if its efficiency as a drying 
agent is entirely confined to the surface layer, which is rapidly 
affected by any moisture present. Diagram III shows that 
when the quicklime was first introduced into the case it 
brought down the leak to 0°0025 volt per minute, but the 
effect did not last long, and after 30 hours reached the value 
0-006, which is the usual leak during the dry season when desic- 
cants are not used. A larger amount of the substance tested 
sulphuric acid. 
| Science Abstracts, Vol. XVIII, 1915, No. 1581. 
