154 THE ROYAL SOCIETY OF CANADA 
Lucian,” on taking account of the active deposit on the case 
obtained results for actinium very similar to those obtained by 
Wellsch for radium. He came to the conclusion that a limiting value 
of 94-9 was reached for the percentage cathode activity of actinium 
in dry air. 
The behaviour of radium was also studied by Walmsley." 
Using a cylindrical type of vessel, he also proved that no negatively 
charged rest-atoms of radium exist, and that even at the highest 
potential gradients the activity of the central rod when anode was 
about 2:7% of that when the rod was cathode. His figures seem to 
indicate that roughly about 10% of the rest-atoms was uncharged, 
when account is taken of those collected on the case. 
The results obtained by these investigators differed from one 
another so widely that it was not clearly established whether any 
definite fraction of the rest-atoms of radium was initially uncharged, 
or if partly uncharged what the value of this fraction was. A similar 
statement would hold true for actinium and also for thorium, though 
little work had been done with the latter. 
DISCUSSION OF APPARATUS. 
In many of the experiments referred to above a considerable 
source of error is probably to be found in faulty design of the testing 
vessel. In the ideal testing vessel the following conditions should 
hold: 
1. The field should be uniform. 
2. All the rest-atoms should settle on the electrodes and none on 
other parts of the vessel, where they would remain un- 
measured. 
3. There should be no opportunity for rest-atoms, neutralized 
by recombination outside the field, or in a weak field, to 
diffuse to the electrodes. 
4. The electrodes should not be so close together as to make the 
effect of recoil too great. 
5. The active deposits collected on the electrodes should be 
measured under identical conditions. 
The field in a cylindrical testing vessel is far from uniform. A 
simple calculation will show this. Assuming no ionization within 
2 Lucian, Am. Journ. Sci., Dec., 1914. 
18 Walmsley, Phil. Mag., Oct., 1914. 
