Difusion of Thorium and Actinium Emanations. 643 



P. 



PxD (corrected) 



76-7 cm. 



8-58 



765 „ 



8-42 



76-3 „ 



8-54 



76-0 „ 



8-58 



76-0 „ 



8-92 



75-8 „ 



8-49 



75-5 „ 



8-36 



75-0 „ 



8-11 



75-0 „ 



8-07 



42-0 „ 



9-02 



267 „ 



8-06 



16-0 „ 



8-05 



Mean . . 8-43 

 The value of D in air at 76 cm. pressure and 15° C= , lll. 



Discussion of Results. — The results given are sufficient to 

 show that apparently slight alterations in the form of the 

 apparatus may produce considerable variations in the value 

 of the diffusion coefficient as determined experimentally. 

 The minor departures from constancy of the value P x D in 

 experiments made with exactly the same apparatus and 

 amount of emanation are probably due to slight deviations 

 from parallelism of the plates. That other experimenters 

 have found as great variations can be seen from the paper 

 by Bruhat. He also gives a full list of his results, and 

 deviations from the mean value to the extent of 14 or 15 per 

 cent, may be noticed. 



The experiments show further that even without any change 

 in the apparatus it is possible to vary the value of D by 

 varying the amount of emanation in the vessel. These two 

 facts seem to indicate the reason, not only why, in the past, 

 different experimenters have found different values for D, 

 but also why there is a discrepancy between the observed 

 and the calculated values. The disturbing cause is very 

 probably the active deposit, a portion of which perhaps loses 

 its charge by recombination with the ions in the air, and 

 then behaves more or less as a gas, until it accidentally 

 strikes one of the plates. The effect was naturally more 

 pronounced where no field was employed, or only a single 

 plate. Even when a field was employed the anode showed a 

 small quantity of activity, about 5 or 6 per cent, of that on 

 the cathode. This might have been due either to electrically 

 neutral, or negatively charged particles. The portions of 

 the active deposit producing the largest effect would be the 



