[HENDERSON] DEPOSITS OF RADIUM, THORIUM AND ACTINIUM 155 
the testing vessel, the potential gradient at a point distant r from the 
axis of the vessel is 
V 
lo 
r log im 
where r1 and rz are the radii of rod and case and V is the difference 
of potential between them. 
For example in the vessel generally used by Wellisch r; = -0915 
cm., ro = 2:90 cm. If the difference of potential between rod and 
case is 1,000 volts, then it can readily be calculated that: 
The potential gradient near the rod =1890 volts per cm.. 
“ “ “ “ “ case = 60 « “ “ 
é se “ midway between rod and case = 116 volts per cm. 
The effect of ionization in the vessel would be to increase the 
potential gradients near the electrodes, particularly near the central 
electrode. Hence the potential gradient in the body of the vessel 
would be correspondingly decreased. Thus it is evident that the 
potential gradient in a large fraction of the volume of the vessel is 
very weak compared with what it would be if the field were uniform. 
The presence of insulating material in the plug holding the rod also 
tends to distort and weaken the field near it. To make matters 
still'worse, there is a considerable region in the corners formed at 
either end of the vessel between the cylindrical side and the ends, 
where the field is much weaker than in any other part of the vessel, 
thus greatly increasing the chance of recombination of the rest- 
atoms. 
Another objection to the cylindrical type of vessel is that a small 
part at least of the total activity will be deposited on the insulating 
material and remain unmeasured. Furthermore, of the rest-atoms 
to be measured part are on the rod while part are on the case, so that 
the ionizations produced are not proportional to the number of the 
rest-atoms present in the two cases. 
The parallel plate type of vessel used by Kennedy? fulfilled much 
more nearly the required conditions, but opportunity still existed 
for recombined rest-atoms to diffuse up between the plates. The 
effect of recoil was also considerable. 
It is hoped that the type of vessel finally adopted by the writer 
and used in the final experiments with radium satisfies all these re- 
quirements. A brief description is here given. The vessel is shown 
in plan and elevation in Fig. 1. The electrodes A and B were two 
