Weights of Radium and Thorium Emanations. 61 



[inserting this value of V in equation (1) we have 



X 4' <// 



ott 



wheilC 



ioo- •;;- = > log/i+ao. 



and since | a/ | < 1 . 



X, r a 2 £ 2 ' a 3 £ 3 



Log 



v., x r «* a " ^ V i 



,' = - V " \af- ~7 +-T7- - , + 



y «V L 2 .5 4 J 



Xf IV a/ a 2 * 2 a : 7 :; "I 



= V„L , - i + 3 -~T + ]' 



whence 



*= >■ at aV \v n t'' 



1 



:T + 3 - T + J 



V 



4:. Preliminary Experiments, 



The earlier experiments were made to compare the rate of 

 diffusion of the radium emanation with that of hydrogen 

 through the plug of plaster-of- Paris which will be referred 

 to as plug I. Although the results obtained cannot be em- 

 ployed to calculate the molecular weight of the emanation, a 

 short account of them will be given since they lead to some 

 interesting conclusions regarding the nature of the plugs 

 used, as well as throwing light on the conditions to be attended 

 ro in performing density determinations by the present 

 method. 



As a result of the first experiments, the values of X for 

 hydrogen and the emanation (mixed with air) were found to 

 be 1*77 and "138 respectively. Assuming Graham's law of 

 diffusion, these number- would lead to the conclusion that 

 the molecular weight of the emanation is 328 — a value which 

 is more than three times as great as the upper limit given 

 by Rutherford and Brooks (loc. cit.). Further experiments 

 were therefore undertaken to explain the cause of this 

 discrepancy. 



It was thought possible that the density of the gas with 

 which the emanation was mixed might influence the rate of 

 diffusion of the emanation, and therefore to test this hypo- 

 thesis experiments were made in which hydrogen was sub- 

 stituted for air. The emanation was carried into the difrusion- 

 I by a stream of hydrogen, and in order to have thi< gas 



