Nolan and Eniiight — Experiments on Large Ions in Air. 113 



large ions. We may write the equation q = i]Nkn^, where 7i, is the 

 number of large ions and k the ratio of the total to the charged nuclei. The 

 product Niii should be constant while k is constant. We have examined 

 the results of M'Glelland and Kennedy on occasions when the concentration 

 of small ions was small, and we find that the constancy of the product holds 

 fairly well in almost all cases. We find a group of observations, taken on 

 quite different days, where N is about 70 and Hi about 20,000. Another 

 group of observations gives iV"=40 and Wi = 32,000 roughly. In this we 

 assume that the large ions carry single charges. If we take Nn^ as 1,350,000 

 and for r\ assume the value 5 x 10'" we have 



2 = 5 X 1"35 k 



= 6-75 it. 



The value of q is not likely to be as high as 13'5, so that if half the 

 nuclei are charged, that is k = 2, it would seem that we have overestimated 

 the value of ij. If our value of jj is correct, and if q is of the order 5 - 10, 

 then either k is less than 2, or the ions must carry more than one charge. 

 Simultaneous observations of the small ions, large ions, and nuclei of the 

 atmosphere are necessary for the clearing up of this and other difficulties. 



Charge and Mobility. 



As there is good agreement among all observers that the mobility of the 

 atmospheric ion is under normal conditions about '00033, there can hardly be 

 any doubt that this is the correct value for the singly charged ion. While 

 this "ion is always present, the ions of two, three, etc., times this value 

 corresponding to the different multiple charges do not show up clearly. The 

 various mobilities, '00033, '0013, "0023, 0010, etc., worked out especially by 

 McClelland and P. J. Nolan,' seem to correspond to different sizes (as the heat- 

 ing experiments show) rather than to different charges. It is possible that the 

 ion of mobility about '001, which frequently appears, corresponds to the ordinary 

 ion with three charges. But it does not ever seem to be present in sufficient 

 quantity to agree with the preponderance of trebly charged ions in flame-gas 

 which our results indicate. 



This difficulty has been discussed in a previous paper.- It was there 

 suggested that if two or three ions of the same size come together to form a 

 loose group, the resistance to motion of the group might well be increased in 

 the same proportion as the charge, so that the mobility might remain 

 unaffected. Many difficulties arise in the application of this idea to special 



' McClelland and P. J. Nolan, loc. cit. 



- J. J. Nolan, Proc. Roy Soc, vol. xciv, p. 112, 1918. 



R.I.A. PEOC, VOL. XXXVI, SECT. A. [12] 



