McClellaxd axd Kennedy — The Large Ions in the Afmosjjhere. 87 



On the other hand, the theory which Langevin has de^"eloped of the 

 formation of the large ion agrees well in its general oiitUnes with the 

 experimental results in this paper. J. J. Thomson* and Langerinr have 

 shown independently from thermodvnamical considerations that water 

 condensed in the form of in^dsihle drops may exist in the atmosphere even 

 when it is far from heing saturated ^fith water-vapour, and the probable size 

 of these drops indicated by theory when they are in stable equilibrium is in 

 good agreement with the size of the large ion as indicated by its mobilit)' . 

 Lange^^n regards these invisible water-drops as the nuclei from which the 

 large ions are formed, when they acquire a charge by collision with a small 

 ion. 



Assuming this theory of the production of the large ions, some points 

 regarding the relations between the numbers of uncharged nuclei, small ions, 

 and large ions present at any time may be further discussed. 



Langevint considers that the fraction of the liquid nuclei which is 

 charged at any time is independent of the niimber of small ions. He 

 imagines the large ions as being produced by the diffusion of the small ions 

 to the uncharged centres, the production of large ions being limited by the 

 recombination of the large ions with the small ions of the opposite sign, so 

 that if the small ions of both signs be equally numeroiis there will be 

 established a permanent state in which a fraction of the particles is 

 transformed into large ions. This fraction he considers independent of the 

 number of small ions at any time, since the formation and recombination 

 both take place to an extent proportional to the number of small ions 

 present. 



Experiments conducted by us are not in agreement with this view. 

 These experiments were made on the effect of the -y rays from radium on the 

 rate of production of large ions in deionised air stored in the gas-holder. 

 The results showed that exposure to the y rays increased the number of large 

 ions produced in any time. When the exposure lasted for somewhat more 

 than an hour corresponding to the time necessary for the maximum value of 

 c'jc (fig. 9), the value now obtained was 2 instead of -86. 



When air filtered through cotton-wool was subjected to y rays, there was 

 no appreciable increase in the number of large ions produced, a fact in 

 agreement with the theory that the failure of filtered air to reproduce large 

 ions is due to the removal of the essential nuclei by the cotton-wool. 



Theory would seem to agree with the results obtained by us, if we take 

 into account the recombination of large ions of opposite signs. Consider a 



* Conductivitr through Gases. tLe Eadium, avril, 1912. i Soc. de Physique, 19 mai, 1905. 

 R.I.A. PROC, VOL. XXX., SECT. A. [14] 



