724 *Dr. P. B. Perkins : Determination of Periods of 



were inserted and the times and deflexions for 4 beats were 

 recorded. As practically all of the emanation had dis- 

 appeared in 55 seconds, it required a certain amount of 

 practice to take and record correctly from 5 to 7 readings in 

 the time available. As the rate of production of the active 

 deposit in the chamber during the decay of the emanation 

 was so rapid, it was deemed advisable to obtain as many 

 points on the decay curve as possible since, in taking the final 

 readings with no auxiliary capacity in parallel with the leaf, 

 a large percentage of the leak was due to the active deposit 

 formed. After the last reading was taken, the vibrations of 

 the pendulum were still counted until a given second on a 

 chronometer was noted. The time of the beginning of the 

 experiment was therefore known for making the necessary 

 active deposit corrections. The growth of the active deposit 

 was then observed until it reached a maximum, about 9 min. 

 after the beginning of the experiment. 



The same general order of experiment was used for thorium 

 emanation. The rate of decay is so much slower for thorium 

 emanation that the time at which the deflexions were read 

 was taken from a chronometer. As before, deflexions were 

 observed during a given number of vibrations of the 

 pendulum, between the raising of the keys K, and K 2 . Ten 

 or twelve beats, i. e. 5 or 6 seconds, was used in general. 

 Readings extended over 12 or 13 minutes and the growth of 

 the active deposit was observed for 30 minutes after the 

 beginning of the experiment. 



A study was made of the possibility of a residual charge 

 existing in the mica condenser and thus influencing the 

 succeeding readings, especially in the case of thorium 

 emanation where charging occurred during 5 or 6 seconds. 

 No residual effect could be observed under the experimental 

 conditions used. If such an effect exists it must be very 

 small as the capacity is charged from zero to a very small 

 potential. 



Determination of the Capacity Ratios. 



The accuracy of this method depends largely on the degree 

 of exactness with which the ratios, between the different 

 capacities used, can be determined. These ratios were 

 obtained a number of times during the investigation by 

 observing the rates of leak due to a constant source of ionization 

 when the various capacities were connected in the leaf system. 

 Two methods were used : — (a) the /3 rays from a source of 

 radium emanation were fired through the window K into the 

 chamber, (b) the a ray activity of radium emanation and its 



