15 



methods suggest themselves. In the first place, ionized gas 

 may, as in Rutherford's experiment, be blown into a chamber 

 in which the ions are to be collected by an electric field. In 

 this case care must be taken to enable the ions to be passed 

 rapidly into a uniform electric field, so as not to subject them 

 at first to a weaker field. 



To pass the ionized gas through a fine gauze as Ruther- 

 ford did would appear the most convenient method to adopt 

 in such a case, although, as will be shown later, it by no 

 means, in all cases, approaches the ideal condition. 



In the second method, the ionizing agent may be allowed 

 to act upon the gas which is enclosed in a suitable ionization 

 chamber, and then either the chamber or, what is usually 

 more convenient, the ionizing agent may be removed before 

 the electromotive force is applied between the electrodes. 



Both these methods have been used in the present in- 

 vestigation. 



Since the shape of the ionization curve obtained in the 

 presence of the ionizing agent is found ^7) to depend upon the 

 effects of both general and initial recombination, it is import- 

 ant, no matter which of the above methods is used, in com- 

 paring the saturation curves with and without the presence 

 of the ionizing agent, to eliminate the effects of general re- 

 combination. Also, diffusion of the ions to conducting sur- 

 faces would produce many of the effects which one might ex- 

 pect to be shown by clusters which could be separated by a 

 sufficieutly intense field. 



§1- 



The Action of Gauzes. 

 The first experiments were carried out with an apparatus 

 as shown in fig. 1. Air was used in these experiments, and 

 after passing through a plug of cotton wool was drawn by a 

 water-pump, with suitable pressure regulator, at a speed of 

 10 cms. per second, over the surface of uranium oxide. Thence 

 it passed through a gauze and between two cylindrical 

 electrodes of 2 cms. and 4 c.m5. diameter respectively. The 

 outer electrode was connected to a set of small accumulators, 

 the other terminal of the cells being connected to earth. The 

 central electrode was 10 cms. in length, and was supported in 

 a sulphur plug, which in turn was protected by an earthed 

 ring. By suitable keys the central electrode could be con- 

 nected either to the electrometer or to earth. A Thompson 

 Quadrant Electrometer was used, giving a deflection of 75 

 scale divisions per volt. 



(7) Ijrago; and Kleeman, Trans. Roy. Soc. S.A., vol. xxix., 

 1906. 



