204 ME. J. S. TOWNSEND ON THE DIFFUSION OF IONS PEODUCED IN AIR 
Methods of Conducting the Experiments with Air at Atmospheric Pressure. 
The tube B was moved into A until the disc a/ came up to a small screw that 
projected into A, and the rod F x was connected to the insulated quadrants of the 
electrometer. The tube B. : was connected to the end of by means of a piece of 
tubing the same size as A. Before any observations were made the stream of air 
was allowed to pass for one minute through the apparatus in order to blow out the 
ions that accumulate in the air in A. The quadrants (connected to F,) were then 
insulated, and the air, being ionised as it passes the tube C, carries some of the 
ions with it through the tube T x . The electrometer deflection, n 1 divisions per 
minute, was then observed. 
The positions of the two tubes B, and B> were then interchanged. The resistance 
to the passage of the air through the apparatus was unaltered, so that the stream of 
gas that passed through was the same as when the deflection n x was being observed. 
The rod F 2 was connected to the quadrants of the electrometer and the same 
observations were made, the electrometer deflections being n.-, divisions per minute. 
These experiments were repeated several times, and it was found that the numbers 
n 1 and n 2 were constant, showing that the rays emanating from the tube C did not 
vary with the time to any appreciable extent. 
When the quadrants were insulated and the air inside A at rest, a small deflection 
(usually one division per minute) was obtained arising from the imperfect insulation 
of the plug G. This deflection has to be subtracted from the deflections obtained 
when the stream of air is passing along A. The leak across G gives rise to no 
inaccuracy, as it is perfectly constant and was easily determined. In the tables of 
observations the corrected values of and n. 2 are given. 
When the tube A is charged positively the deflections n 1 and n 2 refer to positive 
ions; similar numbers for negative ions were obtained by changing the sign of the 
potential of A. 
Experiments at Lower Pressures. 
In order to make experiments with air at pressures lower than the atmospheric 
pressure, it was found necessary to make a slight change in the diffusion apparatus. 
The experiments were made with the ions produced by the radio-active substance, as 
it is a source of very constant radiation. A flat ring was soldered to the end of the 
tube A, and similar rings were soldered outside the tubes B ( and B : . The ring on 
the tube B L was 8 centims. from the disc oq', and the ring on the tube B 2 8 centims. 
from the disc a 2 . The diffusion apparatus was made sufficiently air-tight by greasing 
the ring on A, and pressing the ring on B against it. 
The arrangement of the apparatus for obtaining a stream of air at low pressure is 
shown in fig. 2. Rubber stoppers with glass tubes leading from them were fitted into 
