176 Mr Richardson, On an attempt to detect radiation , etc. 
The apparatus had now assumed its most sensitive form, since 
the capacity of the charged cylinder and parts connected with it 
was only three times that of the electroscope. In a slightly 
different form of apparatus, used for the experiments at ordinary 
pressures, and insulated entirely with sulphur, the normal leak 
was reduced to about 01 scale division per hour; and in the last 
described apparatus, used for low pressures the leak was never 
greater than 1 scale division per hour. The apparatus was there- 
fore much better insulated than that with parallel plates when 
there was a normal leak of 2 divisions per hour. Experiments 
were made at ordinary pressures with aluminium and steel wires. 
The following numbers, for an aluminium wire, indicate the sort of 
readings obtained. 
Pressure 
Initial 
Reading of 
movable leaf 
Final 
Reading 
Leak in Scale 
Divisions 
Time of 
Experiment 
Leak 
per hour 
Atmospheric 
46-0 
44*2 
1-8 
18 hrs. 
o-i 
No current 
through wire 
)> 
44-2 
44-0 
0-2 
3 „ 
0-07 
Current passing 
through wire 
1 
From a large number of similar observations it was evident 
that no appreciable amount of ionisation was produced in air at 
ordinary pressures. It was still possible however that an appre- 
ciable amount might be produced at low pressures. For if a few 
ions were produced under ordinary circumstances, these, if the 
pressures were lower, would produce more by their motion in the 
electric field, i.e. when they collided with the molecules of the 
gas 1 . Thus the method might be expected to become more 
sensitive when employed at low pressures. A large number of 
observations were therefore taken (using an aluminium wire) at 
pressures varying from 3 to 0T mm. Two of the observations 
taken are subjoined : — 
1 J. S. Townsend, Phil. Mag., Feb. 1901. 
