Nolan and Enright — Experiments on Large Ions in Air. 103 



The view has been put forward that the large ion is a cluster,' each ion 

 being formed by a grouping together of those of the next smaller class. On 

 this view the effect of heat is to break up the ions into smaller units, just 

 as the effect of adding certain substances is to cause the ions to form more 

 complex aggregations. 



RecomMnation of Large Jons with one another and with Small Ions. 

 Large ions disappear by recombination with one another and with small 

 ions, if tlie latter are present. Kennedy lias found^ that the nuclei which 

 form the large ions disappear at the same rate wliether they are charged or 

 uncharged. Where n is tlie number of nuclei per c.c. the rate of disappear- 

 ance obeys the law -=- = - Y?r over a considerable range. The value of the 



Ctb 



constant y determined by Kennedy is 14 x 10"'". Kennedy also measured 

 tlie rate of decay of charge. Expressing it as the rate of decay of electronic 

 charges of one sign per c.c, he determined the decay constant |3. He showed 

 that if X is the number of electronic charges per ion, fix = 2^. 



The value of (3 is variable, as the charge per ion is variable. One value 

 found for /3 was 6'3 x lO"'". This gives for the electronic charge per ion the 

 value 4-5. By direct experiments Kennedy showed that the large ion can 

 have from one to six electronic charges, the higher charges tending to appear 

 when the air is drawn away rapidly from the flame. 



The collision frequency of the ions is — -=^, where n is the number 



per c.c, o- the diameter of the ions, and u their mean velocity of agitation. It 

 has been shown' that if the ion of mobility •00033 be regarded as a compact 

 sphere of water carrying one electronic charge, its radius is 4.1 x 10"" ems. 

 Of the assumptions made here, the first is the most serious, as both from the 

 fact that de Broglie found that large ions could not be seen in the ultramicro- 

 scope,* and from our own experiments on the effects of temperature and of 

 such substances as SOj, we would prefer to regard the large ion as a loose 

 cluster. Using this value, however, as a basis for calculation, and writing 

 Mj and u^ for the mass and average velocity of the hydrogen molecule, we find 

 that the recombination coefficient 



^ = 71 ''' 4.(4-1 X 10-°)' = 28 6 X 10 . 



' J. J. Nolan, Proo. Roy. Soc, vol. xciv, p. 112, 1918. 

 ^ Kennedy, Proc. Roy. Irish Acad., vol. xxxiii, A, 1916. 

 3 j_ j_ Nolan, Proc. Roy. Soc, xciv, p. 112, 1919. 

 * de Broglie, 0, R., vol. clxviii, p, 1317, 1909. 



