president's address SECTIOiV A. 37 



mention of results tliey liave obtained, and throughout have been 

 iiio.st ably helped bv nw assistant, Mr. Carl Shai-pe. Owing to the 

 variable character of the natural ionisation, the work has proved 

 extremely tedious, as it is only on somewhat rare occasions that a 

 sei'ies of obsei'vations is accordant enough to give a definite measure 

 of the mobility. The ionisation is more uniform after sunset, and we 

 now observe mainly in the night time. 



All our observations have been made with apparatus constructed 

 after the pattern of that used with such success by Professor Zeleny* 

 in his determination of the mobility of the small ions. In such an 

 instrument a uniform stream of air flows through a metal tube which 

 fcrms the outer conductor of a cylindrical condenser, the ions drifting- 

 on to an inner axial electrode, due to the forces in the electric field 

 established between the tube and the axial rod. The theory of the 

 method of finding the jnobility with such an apparatus, as given by 

 Professor Zeleny, is well kno\vn ; it has been followed without modifi- 

 cation in calculating .the results of the present series of experiments, 

 (ireater uniformity in the ionisation is obtained if the air, before 

 reaching the measuring tube, is drawn through a considerable length 

 of piping. We have not notice<l any effect on the nature of the ions 

 due to the somewhat prolonged contact of the air with the metal of 

 the pipes, and in most of our experiments several metres of iron or of 

 galvanised iron pijdng have been employed. In all cases Dolezalek 

 electrometers have been used to measure the ionisation currents. 



During the investigation some definite results have been obtained, 

 of which I propose to give a general account. 



In thiukiny of M. Langevin's discovery the idea nuist have 

 occurred to many, and is, indeed, suggested by Professor Kutherford 

 in his book on Hadio-active Transformations, that the large ions may 

 be due to the piesence of water va]^()ur. My efforts to elucidate this 

 ])oint have residted in finding that there is a definite relation between 

 the mobility of the ion and the amount of moisture in the air. 



When a current of air is passed over hygroscopic substances, 

 without mechanical filtration, Mr. S. G. Lusby finds that large ions 

 are absorbed, and has noticed a loss in number amounting to 55%, 

 after the air had flowed over a tray containing phosphonis pentoxide. 

 I find, in addition, that after leaving the drying agent, those large 

 ions which still exist in the air decrease in mobility with time, and 

 that when the relative humidity changes from 80% to 4%, at a 

 temjierature of 19° C, tliey are not in equilibrium with the, new 

 Viipour ])ressure conditions mitil after the lapse of about twelve 

 minutes. Owing to the vaiiable nature of the natural ionisation, and 

 ])erha])s to other causes, the calculated mobilities exhibit considei'able 

 iiiegularitles, but show in an unmistakable manner, when the equilib- 

 rium state is established, a dependence of the mobility on the amount of 

 water vapour in the air, the reciprocal of the mobility being a linear 

 function of tlie humidity between the limits of the absolute humidity 

 represented by 0'5 and 19'0 (grms/nr^), corresponding tf) relative 

 humidities of 4% and 100%. The mean mobilities for these values of 

 the hiunidity, from results so far obtained, are 1/1280 and 1/3370 

 resfiectively. In other words, the mobility for an absolute limuidity of 



* Zeleny. Phil. Trans. A, 195, p. 103, 1900. 



