Nolan and Enright — Experiments on Large Ions in Air. 95 



The real problem then is the nature of the original kernel. This would 

 seem to be ordinarily a particle of some hygroscopic material. In certain 

 eases it is probable that such a particle is unnecessary. Thus in the case of 

 large ions produced by spraying, splashing, and babbling of water, it is 

 probable that aggregates of water of various sizes are formed, and that thfose 

 above a certain size tend to grow towards the final nuclear size. The function 

 of the hygroscopic substance in the case of ionisation by flames, arcs, 

 phosphorus, etc., is to produce the first stage of growth of the water splierule. 

 Such an agent is unnecessary if stable or quasi-stable sizes can be arrived at 

 mechanically. These views find a general support from the experiments now 

 described. 



Effect of SO2, NH3, and NaCl on lon-prorhwtion in Hydrogen and Coal-gas 



Flames. 



A small flame of specially purified hydrogen was burnt inside a glass 

 vessel. Combustion was maintained by a supply of filtered air. The gases 

 from the flame were drawn away and fed into a stream of filtered air which 

 was drawn at a steady rate through the electrical testing vessel. This was 

 of the usual type, consisting of a cylinder with a central electrode connected 

 to an electi'ometer. The total number of ions of one sign could be estimated 

 by applying the saturation voltage to the outer cylinder. If necessary, 

 measurements of mobility could be made by plotting the current-voltage 

 graph. 



With this arrangement, no large ions due to the hydrogen flame could be 

 detected. We do not assert that none were formed ; it is possible that our 

 precautions were not rigid enough to stop the formation of ions completely. 

 We can only say that if ions were formed they were present in such numbers 

 that it was impossible to detect them by our methods of observation. When 

 SO2 was mixed with the air supply to the hydrogen flame, large ions were 

 produced in considerable numbers. The following observations were made 



on one occasion : — 



Current. 



Air passing through, 6 Scale divisions /min. 



Hydrogen flame burning, ..... 6 „ 



SO2 supplied to flame, 50 „ 



Flame extinguished, SO2 still passing, . . 6 „ 



riame relighted, SO2 still passing, . . .50 „ 



Supply of SO2 stopped. Current fell gradually to 6 „ 



SO2 passed into main air-stream, . . . 6 „ 



These tests show that SO2 supplied at the seat of combustion supplies 

 centres for the growth of the water- nuclei which when charged became large 



[10*] 



