Weights of Radium atid 'Thorium Emanations. 59 



was again established between the air in V and the out- 

 side, and the pressure inside thus again brought to atmo- 

 spheric. After a short known interval (usually 1 minute) 

 the conductivity of the air in V was determined thus : — 

 At a definite instant the key k was removed so as to insulate 

 the rod a and its connexions; after a known time (t seconds) 

 the key 1/ was removed and the key h replaced, and the 

 deflexion of the electrometer-needle read as soon as it had 

 come to rest, and taken as a measure of the quantity of 

 emanation which had been carried into V. Immediately 

 after disconnecting the key // the tap t was again turned 

 so as to connect V to the water-pump, and a rapid current 

 of air drawn through it by opening the side tube b, which 

 had remained closed by a rubber cap during the first part 

 of the experiment. In this way the emanation was com- 

 pletely removed from V in a few minutes, after which 

 the radioactivity " induced " on the walls of the vessel 

 decayed according to the usual exponential law, becoming 

 negligible before the next experiment was performed. After 

 a known time the operations described above were repeated, 

 and the electrometer deflexion in r seconds again recorded. 

 Since it was impossible to transfer exactly the same volume 

 of air at the beginning and end of an experiment, a correction 

 had to be applied to allow for this difference ; a further 

 small correction had to be applied for any slight change of 

 -ensitiveness of the electrometer during the experiment. 



It was also necessary to allow for the natural decay of the 

 radioactivity of the emanation in the time during which the 

 diffu-ion had been proceeding. The correction to be applied 

 on this account was always small, and in the preliminary 

 experiments has been neglected ; the method of applying it 

 when necessary will be discussed when £*ivino- the final 

 observations. 



In most of the experiments, the scale of the electrometer 

 was uniform ; in one set of observations, however, it was 

 found necessary to calibrate the scale and apply suitable 

 corrections to the observations (Plug I. Table III.). 



If y is the concentration of the emanation in the diffusion- 

 vessel at any instant, then the change of concentration dy in 

 the time <h due to diffusion is given by the equation 



ty=— y//dt (1) 



in which V is the volume of the gas diffusing, and \ is ;i 

 constant depending on the nature of the porous plug. 



If y i- the value of y ;it the beginning of the experiment, 



