290 Dr. Norman Campbell on the 



greater than 20. That is to say, there are practically no 

 rays entering A with speed between 20 and V 1 (which was 

 never less than 40) ; the presence o£ the rays with speeds 

 less than 20 is doubtless due to the excitation of " secondary 

 rays " in B by the electrons falling upon it. 



A further advantage of this method of varying the speed 

 of the rays may be mentioned ; it renders the effect of the 

 presence of a small quantity of gas less important. If 

 the rays travel through the chamber A with the same speed 

 as that with which they enter it, then, when the speed of the 

 rays becomes equal to the ionization potential of the gas, 

 there will be a sudden increase in the current flowing 

 from P due to ionization in the gas ; there might be some 

 uncertainty whether a " kink " in the curve at some point 

 was not due to ionization in the gas rather than to ionization 

 at the plate. But if the rays always enter the chamber with 

 the same speed, much greater than the ionization potential, 

 and always travel through the gas the same distance, while 

 only their final speed is altered, it can be readily seen that 

 there should be no sudden change in the current due to the 

 gas and no distinct kink in the curve. As a matter of fact, 

 experiments made when a little gas was intentionally intro- 

 duced, showed that the current due to ionization of the 

 gas was almost independent of Y 2 ; the presence of gas 

 reduced greatly the absolute value of R (see below), but 

 hardly changed the relative values. 



4. The apparatus as drawn is not suitable for establishing 

 the statements just made concerning the homogeneity of the 

 rays; for in order to measure the total number of rays striking 

 P we must either prevent any rays leaving P by an adverse 

 field, or must connect to the electrometer all parts on which 

 any rays leaving P can fall. Now a considerable proportion 

 of the rays leaving P fall on the tube T, and neither of these 

 methods can be adopted. The truth of these statements was 

 established by experiments in which the tube T was not 

 present, and the hole in the bottom of A was only slightly 

 larger than the fine hole (0*8 mm.) in the top of B. With 

 this arrangement practically all the rays leaving P fall on A, 

 aud the total current can be measured by connecting P and 

 A to the electrometer, or by making V 3 large and negative 

 (20 volts is sufficient), so that all electrons within A are 

 driven to P. (The results of the two methods agree.) The 

 homogeneity of the rays could then be demonstrated, and a 

 satisfactory value obtained for the ratio of the number of 

 rays falling on P to the number retained by it. The tube 

 was introduced in an attempt to cure a defect of the apparatus 



