E. 31. Wellisch — Mobilities of Ions in Air. 597 



of a cluster of molecules grouped round a charged nucleus, 

 that for a certain range of forces and pressures there is a tran- 

 sition stage in which the average mass diminishes from that of 

 a group of molecules to that of an electron ; and finally when 



the value of — exceeds "2 (X being measured in volts/cm and 



p in mm) the negative ions in dry air are all in the electronic 

 state. 



The present series of experiments confirms the view that in 

 dry air the tendency is for the negative ions to assume the 

 electronic state as the pressure is reduced but in other respects 

 they lead to a view of the mechanism which is radically differ- 

 ent from the foregoing. The chief point of difference lies in 

 the result that the negative carriers in general consist of two 

 distinct kinds, electrons and ions. There is accordingly no 

 justification for introducing the terms mobility or velocity as 

 applied to the negative carriers at the lower pressures where 

 the percentage of electrons is appreciable; all attempts to 

 determine the velocities under these conditions rest upon a fic- 

 titious method of averaging which appears to be unjustifiable. 



If, however, care be taken to distinguish between the elec- 

 tron and the ion, it is found that the latter has a distinct mobility 



so that its velocity is expressible in the form v = k — ; there 



appears to be no necessity for introducing any unknown 



function of — . The absence of any indication of the existence 

 p J 



of an intermediate stage between the electron and ion renders 



nugatory all attempts to determine an average velocity for the 



negative carrier. 



The mobility of the negative ion was found to be practically 

 independent of the frequency of alternation employed, in other 

 words, the applied field has little or no effect on the mobility 

 or the nature of the ion (cf. curve G, fig. 7 and curve, fig. 5). 

 Of course, it is highly probable that excessively large values of 

 the field would produce a marked effect on the mobility of the 

 ion but this statement in no way invalidates the general 

 conclusion. 



In this connection it is easy to account in general terms for 

 the results obtained by previous observers who found (as 

 Kovarik)* that the mobility depended upon the rate of alter- 

 nation of the field or (as Lattey)f that the velocity was an 

 unknown function of the field. When these observers speeded 

 up their commutating device they brought the electrons into 

 greater relative prominence ; and as a larger speed implied of 



* Kovarik. loc. cit. 



f Lattey, Proc. Eoy. Soc, Ixxxiv, p. 173, 1910-11. 



