i66 



ROYAL SOCIETY OF CANA1>a 



ments, and is satisfied that, for this purpose, it gives as accurate re- 

 sults as any of the other methods. It is also much less troublesome, 

 when it is once arranged, and is adapted to measurements for which 

 the other methods are unsuited. 



The high resistance standard itself has gone through quite an 

 evolution during the past three years. The one at present in use 

 is shown in Fig. 2. The shaded portion is ebonite, the electrodes are of 

 aluminum, and the active material consists of about a tenth of a milli- 

 gram of radium bromide. In general, the relation between the ioniza- 



0.1 Mgs. Radium 



Fig. 2. 



tion current in such a vessel and the potential difference between the 

 plates is not linear, but depends upon the nature of the active material 

 used, the kind and amount of insulation, and the distance between the 

 plates. However, it is generally possible to so adjust the distance 

 between the plates that a linear relation will exist for a considerable 

 range of potential. Nevertheless, it is necessary to get a calibration 

 curve for every such vessel. This can only be done by some form 

 of the rate method, which must ultimately be the basis of all such 

 measurements. However, in the case of a calibration curve, the rate 

 method is used under the most advantageous conditions. 



