^6 



positively charged body of small size is brought very close 

 to a body of much larger dimensions similarly charged, the 

 repulsion which normally exists reduces in value to zero and 

 changes to an attraction when the small charged body is 

 brought within a certain critical; distance of the surface, this 

 distance depending upon the charges on the small body and 

 the surface, respectively, and on the curvature of the surfaces. 

 This seems to offer the explanation of the marked difference 

 which exists in the shapes of the ionization curves obtained 

 with broad- and with narrow-mesh gauze, as the field strength 

 is varied ; for in the case considered positive ions are blown 

 by the current of air against the lower gauze surface, which 

 is but weakly electrified when the field in the chamber above 

 is weak. As the field in the chamber is increased the fringe, 

 which passes through the gauze, is strengthened, and the 

 force which is exerted upon the positive ions by the draught 

 is not sufficient to bring them to within the critical distance of 

 the lower surface of the gauze to cause attraction. As the 

 positive ions in air are larger and more massive than the 

 negative, it is to be expected, when the lower gauze is nega- 

 tive, that for a given field in the chamber above, less negative 

 ions would be received by the electrometer gauze than in the 

 corresponding case where the bottom gauze is positive. This, 

 we have seen, does occur when the field is weak. Since, when 

 a strong enough field is applied, the same maximum value of 

 current is obtained, whether the bottom gauze be positive or 

 negative, and also for a large range of gauzes exposing very 

 different amounts of surface, the most natural conclusion is 

 that under such circumstances all the ions, of one sign which 

 reach the lower surface of the gauze are drawn through the 

 gauze without loss in number and are then collected by the 

 central electrode. 



§3. 



It has been shown by Bragg and Kleeman that the lack 

 of saturation in a mixture of air and ethyl-chloride is con- 

 siderably greater than for air. Their theory of initial re- 

 combination explains this difference, and one of the objects 

 in view at the commencement of the present experiment was 

 to decide whether such an effect persisted for a time after the 

 act of ionization by an a particle had been completed. 



As the results so far obtained show the difficulty there 

 exists in eliminating the effects of diffusion when the ions 

 are blown through gauzes, a comparative test between 

 the behaviour of such a vapour as ethyl-chloride and air 

 seemed as much as could at the time be safely derived. 



It was arranged to pass a mixture of ethyl -chloride with 

 a small proportion of air from one gasometer through the 



