268 - THE IONISAT[ON OF THE AIR. 



by the Ebeit was dlzvays less than lo"' ; values given b-y the 

 Gerdien Ijeing sometimes as much as 50 times greater than corre- 

 sponding values given by the Ebert. 



We conclude that each instrument measured some definite 

 (juantitx , l)nt that the meaning of the quantity differed in the 

 two cases. We proceed to an examination of the exact meaning 

 of these (juantities. 



The ions present in air at an\' time are of different kinds. 

 The simplest types are ions which have about the same mobility- 

 as the ions produced in dust-free air by ordinary ionising agents, 

 such as the rays from radio-active substances. Ions of this 

 kind carry a simple charge c, and have a mobility of i.o to I.S 

 cms. per second per volt per cm. A second type of ion is the 

 Langevin ion, very complex in structure and of low mobility, 

 varying from .0008 to .0003 cm. per second. The nature of 

 this ion is im])erfectl\' understood. Pollock has recently advanced 

 the view, based on a thermodynamic argument, that it is a com- 

 plex ion carrying absorbed moisture in the Uquid state. The 

 ion does not exist in dust-free air; so that the nucleus is probably 

 a dust particle. In addition to these two types of ions, a class 

 of ion intermediate in size and mobilitv is ncnv recognised. 

 Pollock argues that these ions are surrounded l)y an envelope 

 of water vapour. Their mobility varies between .07 and .007 

 cm. per sec. The number of these three types per c.c. varies. 

 Pollock, at Sydney, found that the number of small ions of any 

 sign varied from o to i6o; the ninnljer of intermediate ions from 

 200 to 1,000; while the number of Langevin ions varied between 

 600 and 5,500. At the Cape, as will be seen later, the number 

 of small ions probal)l\- never falls to zero ; and the total number 

 of small and intermediate ions is certainly always greater than 

 the corresponding average number given by Pollock. 



Now the conductivity iic7' of the air, as it appears in the 

 expressions of Gerdien and Ebert, involves 11, the number of 

 ions of any one sign per c.c, and z', the mobility of these ions. 

 But since z' is not a constant quantity, and since /; is made up of 

 different numbers of different ions, the conductivity could be 

 better expressed as 2 /icz' — the summation to include all types 

 of ions and all mobilities; and any instrument, if it is to fulfil a 

 useful function, should measure either this total conductivity or 

 any one of the terms iiez' which make up this total conductivity — 

 that is, the instrument should be designed to catch all ions, or 

 some definite type of ion only. 



The two forms of apparatus under discussion do neither ;, 

 and, further, the quantity actually measured by the one apparatus 

 is not the (juantity measured by the other. Ions caught by the 

 Ebert apparatus are not the same in number or in kind as are 

 those caught by the Gerdien apparatus, and, in consequence, the 

 average conductivity as determined from results obtained with 

 the one cannot agree with the corresponding quantity obtained 



