452 Mr. Garrett and Dr. Willows: Chemical 
over it from left to right. The tube and electrode P are 
earthed and a field is put on at Q sufficiently large to remove 
all the ions reaching it; this number being proportional to 
the rate of deflexion of the electrometer-needle which is 
connected to Q. A field is now put on at P, and so some 
ions are removed, the reduced rate of deflexion of the electro- 
meter connected with Q measuring this amount. Then, if p 
is the proportion removed by the field at P ; A the difference 
of potential between P and the tube; b=radius of tube, 
a=radius of P; and ¢=time taken by the current of air to 
pass along P, the velocity of the ions uw is given by* 
_ pb? —a”) log. (b/a) 
2At z 
To ensure insulation, the tubes carrying P and Q were made 
several inches long, and to facilitate the fixing of the elec- 
trodes the sections of the tube A, C, could be screwed off the 
other parts. The electrodes were made sometimes of pla- 
tinum, sometimes of aluminium, the outer tube was iron. 
The velocities for the positive ions under a_ potential 
gradient of one volt per centimetre were as follows :— 
ions from ZnCl, =:0062, 
ions from Zn Bry="0059, 
ions trom Znl, =-0057; 
i. é., the velocities were the same in all cases within the limits 
of experimental error. The relative velocities are probably 
much more accurate than the absolute velocities. In order 
to be able to interpret definitely the results of the measure- 
ments, the field at should be sufficient to produce the 
saturation-current. Owing to the smallness of the velocity 
we could not fulfil this condition as well as would be desirable. 
The results for the negative ions from zinc chloride were 
very irregular. The mean of a large number of results, 
which differed considerably from each other, gave a velocity 
= ‘02 cm. per sec. under one volt per centimetre. 
During some of the experiments, we had occasion to draw 
air over the substances into a large bottle containing water. 
We usually, but not always, noticed in these circumstances 
an exceedingly dense fog, which was produced without ex- 
pansion, and which hung in the bottle for half an hour before 
dispersiont. The bottle was several yards away from the 
* ‘Conduction in Gases,’ J. J. Thomson, p. 48. 
+ Similar effects have been noticed by Townsend, Wilson, and Meissner 
(see ‘ Conduction in Gases’) for gases from electrolysis, &c. 
