272 MR J. S. TOWNSEND ON THE DIFFUSION OF IONS PRODUCED IN AIR 
the holes were soldered to A in the positions shown in the figure. Ebonite plugs 
were fitted tightly into Q and R so that no air should escape by the side tubes. 
The brass rods F and G passed through holes in the plugs, and could be moved up 
and down so as to bring the metal points which were soldered to their ends to any 
desired position in the tubes. The point S at the end of F was a steel needle, and 
the point p was a short platinum wire. 
The potential of the point was raised by means of a Wimshurst machine driven at 
a constant speed by an electric motor. One of the conductors of the machine was 
connected to earth and the other to the wire V, which terminated in a point P at 
a distance of about 5 millims. from the metal disc D. The disc was connected to one 
of the rods F or G by the insulated wire W. When the Wimshurst machine was 
working no discharge took place from the point inside the diffusion apparatus while 
the earth-connected wire X rested on W. When the earth-connected wire was raised 
off W, a discharge immediately took place from the point at the end of the rod F. A 
constant discharge could therefore be obtained inside the tube for any desired time, 
independent of the initial and final variations of the machine. 
Most of the electricity discharged from the point inside A goes to earth through 
the battery, and oidy a small fraction is carried by the stream of air along the 
tube A. 
Owing to the charges carried about in the air of the room (emanating partly 
from P and partly from points of the machine), it was found that gauze screens were 
not sufficient protection for the wire leading from F to the insulated quadrants ; it 
was found necessary to cover the screens with tinfoil. When this precaution was 
taken the electrometer showed no deflection when the Wimshurst had been working 
for several minutes. 
The experiments were conducted practically in the same manner as the experi¬ 
ments on ultra-violet light. The following tables give the results of the experiments, 
the numbers in the columns n 1} n 2 , V and K having the same signification as in the 
previous tables :— 
Table XI.—Positive Ions in Dry Air. 
Expt. 
H. 
% 2 . 
y. 
K. 
1 
766 
162 
262 
324 
•0263 
2 
760 
81 
129 
334 
■0251 
3 
761 
101 
150 
378 
•0245 
4 
754 
89-4 
142 
329 
•0247 
5 
753 
180-6 
299 
324 
■0257 
6 
767 
52-4 
77-4 
342 
•0216 
I 
