NATURE OF THE LARGE IONS IN THE AIR. 201 



an electrical filter to remove the ions without removing 

 the unelectrified dust particles. After filtering out the ions 

 in this manner the steady state between the creation and 

 disappearance of small ions is again established, under 

 ordinary circumstances, in a few seconds, while for the 

 large ions Mr. Lusby's results (loc. cit.) indicate that the 

 new equilibrium condition occurs in about twenty-two 

 minutes. I have now measured the mobility of these 

 recreated large ions thirty minutes after the original ones 

 had been withdrawn, and find a value appropriate to the 

 humidity as shown by the relation given above. The obser- 

 vations were made in June and August of this year and are 

 included in the foregoing table. The result proves that 

 large ions are naturally produced in dusty air. 



If the air is filtered through cotton wool the incoming 

 dust is removed as well as the ions, but I have not yet 

 succeeded in showing an absence of large ions in the air 

 stream in this case, doubtless owing to the impossibility of 

 getting rid of all the dust in the tubes through which the 

 filtered air flows before reaching the testing pipe. 



The non-existence of large ions in dust free air can, 

 however, be inferred from the results of investigations on 

 the formation of clouds in closed vessels. Mr. O. T. R. 

 Wilson 1 has shown that in such experiments with natural 

 dust-free air the first visible condensation takes place on 

 small ions when the expansion ratio reaches the value 1*25; 

 a special search at this laboratory has failed to show any 

 change in the limiting supersaturation necessary for the 

 formation of visible drops with time, extending to days, 

 after the removal of the dust, so it may be concluded that 

 large ions are not produced in dust-free air. 



For experiments with fogs in dusty air an apparatus was 

 employed similar to that described by Mr. Wilson in the 



1 Wilson, Phil. Mag. June,- 1904. 



