422 
DR. HAROLD A. ^MLSON ON THE 
The effect of cooling the inner tube by blowing a current of air through it was 
also tried. A small hole was bored at the point of the conical cap to allow the air 
to escape. The results obtained are shown in diagram No. 3. 
400 
S. 
5 
« 300 
V- 
19 
i;20C 
100 
0 
X 
■uloe 
e.) 
b6_«^ 
>(0’ 
Lr 
<=800 Votes.) 
Diagram No. 3. 
Cooling the inner tube greatly diminished the current in both, directions, and also 
caused tlie current, with the outer tube positive, to begin to show signs of attaining 
a saturation value. 
A very peculiar phenomenon was noticed in all the above experiments. On first 
a})plying the E.M.F. the current was much greater than the steady value to which 
it soon settled down. This effect was alwa 3 ^s much more mai'ked wlien no current 
had l)een passed for some time than immediately after breaking the circuit. Thus 
on first heating the tulje in the morning to about 1100°, and putting on an E.M.F. 
of 400 volts, the current would be as much as ten times its final value, to which it 
