﻿820 Mr. A. J. Saxton on Impact Ionization by 



producing it. Thus, since V cannot be less than 16 volts 

 and x>Y, #>16, which sets a minimum value to the total 

 accelerating field of x-\-y. 



id) The positive -ion current entering the ionization 

 chamber was measured for different values of x and y and 

 for different pressures (the retarding field being kept at 

 zero) by measuring the rate of charging up of the electrode E 

 connected to the electrometer. 



For higher pressures ('lmm. Hg) the positive-ion current 

 reaching E was only a small fraction — about one-tenth — of 

 the original positive-ion current leaving the cylinder. 



As the pressure was diminished, a greater proportion 

 reached E, indicating that at higher pressures many of the 

 ions were stopped by collision with the molecules. The 

 whole of the batteries, leads, and measuring instruments 

 were insulated on paraffin-wax blocks, and tests carried out 

 to ensure that none of the measurements of the different 

 currents were affected by spurious leaks. 



Detection of Ionization. — The first method employed to 

 detect ionization by collision by the positive ions was to take 

 fixed values of x and y and to measure the charging up of E 

 as the retarding field z was increased gradually. No pre- 

 cautions were necessary to distinguish between ionization 

 and radiation produced in G 2 E. In the case of electron 

 impact the collecting electrode is negative with respect to 

 the gauze, and so it may charge up positively by releasing 

 electrons under the influence of the radiation, or it may 

 charge up positively as it collects the positive ions. In the 

 present case, since E is positive with respect to G 2 > radiation 

 will not release electrons from E, which will collect the 

 negative ions produced on ionization. The ionization 

 currents were measured by the steady potential assumed by 

 the electrometer quadrants when connected to earth across 

 a high-resistance leak. The values of z, i. e. the P.D. 

 between G 2 and E, were corrected for this potential assumed 

 byE. 



The type of curve obtained plotting the current to E 

 against an increasing retarding field z is shown in fig. 5. 

 Great difficulty was experienced in keeping the positive 

 ionizing current sufficiently constant so as to render the 

 results comparable for different values of z. 



For small values of z the current (potential assumed by E) 

 was positive, but gradually diminished as z increased until 

 for a certain value of z it became negative, further increase 

 of z resulting in an increase of the negative current. This 



