in Gases at Low Pressures. 



797 



point the mobility of the ion reached a maximum value and 

 then continued to fall gradually as the pressure diminished. 

 The position of this maximum varied in different gases and 

 slightly in the same gas. It was found later that this fall in 

 mobility was due to the presence of water vapour. When 

 great care was taken to dry the gas by allowing it to stand 

 over phosphorus pentoxide for several days, and when the 

 pressure in the apparatus before the introduction of this dry 

 gas had been kept below 0*001 mm. of mercury by means of 

 a charcoal tube and liquid air for some hours, it was found 

 that the mobility continued to increase and at a greater rate 

 than the inverse pressure law requires. Fig. 5 is a curve 



Fkr. 5. 



icon 





X 













T 7 ^ 



\ 



\ 







£> 



/ 







\ 



^o 

















^ 













1 



showing the relation between the mobility and pressure for 

 air when water vapour is present. The pressure at which the 

 mobility begins to fall is in the same region as that in which 

 the product of the pressure and the mobility begins to increase 

 for the dry gas. 



At ordinary pressures water vapour has practically no 

 effect on the velocity of the positive ion, while it invariably 

 retards the negative ion. 



When the pressure was in the neighbourhood of O'l mm., 

 it was noticed that ionization by collision of the positive ions 

 took place if the maximum alternating potential was suffi- 

 cientlv great. The negative ions produced having presumably 

 much bigger mobilities reached the electroscope plate in 

 greater numbers than the positive ions, thus giving a negative 

 charge to the leaf, so that the curve OPQ (fig. 2) would 

 suddenly bend over at Q and cross the axis of potential. The 

 effect diminished and disappeared as the free path arrived at 

 dimensions comparable with the distance between the plates. 



Phil Mag. S. 6. Vol. 22. No. 131. JS r ov. 191 1. 3 G 



