t Journal of the Asiatic Society of Bengal. [N.S., XVI, 
was noticed it was thought that the insulation of the quad- 
rants had been temporarily spoilt by a spider thread or some 
other conducting fibre. To test this point the case was 
bashers tes it even for the first few eee was 
obtained a dete freshly drawn sodium wire. The sodium 
press was brought very near the electrometer case, and fairly 
— lengths of wire 0°5 mm. in diameter were quickly drawn 
to dry beakers, which were then immediately introduced 
Dracram IIT. 
Top—Sodium $95 Asis 11.14—23.11.14). 
Bottom—Quick 
03 
E 
pl 
n 
— 
$ 
& 
Sa 
a 
> 
4 
. cae 40 60 80 100 
—— Time in hours 
i the case and sealed up. age operation was done so 
fected. The results are shown in Dia am III, and Table I, 
Column V, from which it may be seen that the leak never fell 
as low as the standard maximum value. 
cause of this anomalous behaviour of metallic sodium 
is still pea It may be pointed out, however, that the 
action of sodium on the amount of moisture present in the 
ate 
m 
change, ether sodium is slightly radio-active or no 
certain amount of increased ionisation may result from the 
process. 
III. Pxospuorvs Prentoxipe. 
osphorus Pentoxide i is so rapidly affected by any mois- 
Phos 
ture present in the air that it has probably not occurred to 
any physicist to use it i in order to improve electrostatic insula- 
