169 



In the first place it is necessary to consider whether any 

 disadvantages are likely to arise from the use of a sheet of 

 gauze as the lower wall of the ionisation chamber. The elec- 

 tric field must be distorted in the neighbourhood of the gauze ; 

 some very small 2>ortions of the chamber which are just over 

 the openings in the gauze must be under feeble forces, and 

 the ions made there be separated only when the potential is 

 high. It is easy, however, to show that this effect is negligible 

 bv a consideration of the ionisations due to /? ray^s. This 

 ionisation does not show initial recombination, as in the case 

 of the a rays ; a fact first demonstrated by R. Kleeman, for- 

 merly of this University. In Fig. 2 are drawn the upper 

 parts of the ionisation curves of ethyl chloride under different 

 potentials. It will be seen that in the portion which repre- 



Fig. 2. 



sents the effect-s of /3 rays only, saturation is complete when 

 50 volts are applied to the 3 mm. chamber ; but the ionisation 

 due to a rays is far from being collected completely by ten 

 times that potential. Now. if the field distortion due to the 

 gauze were appreciably effective we should find the ^ rays 

 also producing an ir^nisation which appeared to increase at 

 higher potentials ; and there is no trace of any such effect. 

 The same result shows that, although ions are very apt to be 

 drawn through a gauze by a strong field on one side, yet in 

 this case notlimg of the sort takes place. To prevent it, a 

 second gauze has been placed 3 mm. below the first, and 

 earthed, so that there are strong, equally-balanced fields on 

 both sides of the latter. 



A thin, uniform metal sheet might replace the gauze, but 

 unless it were very thin it would cut off more of the range 

 than can generally be spared : and if it were thin it would be 

 liable to flexure by the powerful electric forces, so that the 



