﻿638 Prof. J. A. Pollock on a 



Measurement of Mobility. 



The mobility of the ions was determined by passing a 

 stead)' stream of air through a cylindrical condenser, and 

 measuring the leak between the electrodes for various dif- 

 ferences of potential between them, as in Zeleny's* investi- 

 gation of the mobility of the small ions, and Langevin'sf 

 original determinations of the mobility of the large ones. 

 The condenser consisted of a brass tube, 164 centimetres 

 long, provided with an axinl electrode of the same length; 

 the diameter of the inside of the tube was 3*65, and that of 

 the inner rod 0*66 centimetre. The inner electrode was 

 divided into two sections, insulated from each other : the 

 portion at the mouth of the tube had, on different occasions, 

 the lengths 3*8, 7*7, and 25 centimetres, the distance between 

 the two sections being 4 millimetres in all cases. This brass 

 tube with its inner electrode will be called the testing-pipe. 



In my experience the ionization seems more uniform if 

 the air nasses through some length of tubing before being 

 used, and for the great majority of the observations given 

 here, the air, before entering the testing-pipe, travelled first 

 through 28 metres of iron piping and then through 9 metres 

 of galvanized iron pipe, the diameters of the pipes being 4*5 

 and 7*7 centimetres respectively. In all but the determi- 

 nations at humidities about 90 per cent., the air was from 

 the compressed supply of the laboratory, which is fed by a 

 Sturtevant blower worked by a motor and storage-cells. 

 The blower was open to the air of the laboratory workshop, 

 which in turn was kept open to the outer air, the measure- 

 ments being ordinarily made at night. Control experiments 

 with air drawn directly into the testing-pipe showed that the 

 piping impressed no peculiarity on the mobility determina- 

 tions, though, no doubt, diminishing the ionization. 



The estimation of mobility requires the determination of 

 a critical voltage in connexion with a series of ionization 

 currents, and although the natural ionization is very variable, 

 at times the measurements agreed among themselves suffi- 

 ciently well for the purpose of the calculation. Pig. 1 

 represents one of the best examples of the type of results 

 obtained on these occasions when the long section of the 

 inner rod was attached to the electrometer, while fig. 2, 

 drawn from similarly accordant measures, shows the form of 

 the plot of the observations when the electrometer was 

 joined to the short section of the inner electrode J. 



* Zeleny, Phil. Trans. A. cxcv. p. 193 (1900). 

 t Langevin,, Comptes Hendus, cxl. p. 232 (1905). 



X For further observational detail see Jo urn. and Proc. Eoy. Soc. 

 N. S. Wales, p. 61 (1909). 



