766 Prof. F. Horton and Miss A. C. Davies on 



electrons from the side filament meanwhile not giving rise 

 to any helium radiation. This curve is similar to that already 

 given in figure 3. The curve drawn with a dotted line shows 

 the effect of the same variation in V 1 when the electrons 

 from the side filament have sufficient energy to cause a 

 considerable amount of 21*2 volts radiation to be passed 

 round into the main tube, while the lower filament is 

 supplying electrons with energies corresponding, at any 

 point on the curve, to the value of Y x marked along the 

 .i'-axis. 



From the continuous curve it may be seen that, in the 

 absence of radiation from the side filament, a small negative 

 current begins at 20*4 volts, and that at 21*2 volts a relatively 

 large amount of ionization sets in. From the dotted curve 

 it is seen that, in the presence of helium radiation from the 

 side tub?, a negative current due to photoelectric action of 

 the radiation is measured when Y 1 is less than 20'4 volts. 

 When, however, V x is increased beyond 20'4 volts, this 

 negative current begins to decrease, and a few tenths of a 

 volt later gives place to a large positive current. Thus, in 

 this case, ionization sets in when Y 1 is equal to 20*4 volts, 

 whereas, in the absence of radiation from the side tube, it 

 does not do so until Vj is equal to 21*2 volts, as shown in the 

 other curve. The only factor which is different in the main 

 tube in the two cases is that, in the circumstances of the 

 continuous curve, there is no 21 2 volts radiation present 

 until the point marked X on the curve is reached, while, in 

 the circumstances of the dotted curve, 2T2 volts radiation, 

 originating in the side tube, is present in the main tube all 

 the time. The fact that ionization sets in (or at least a 

 marked increase of ionization occurs) at a smaller value of 

 the energy of the electrons from the lower filament in the 

 dotted curve, than in the continuous curve, proves this 

 ionization to be caused by tie 21*2 volts radiation. Further, 

 the fact that the bend in the dotted curve occurs when the 

 energy of the electrons from the lower filament reaches 

 204 volts, indicates that at this point something is produced 

 which is also concerned in the production of the ionization. 



In this (xperiment the temperature of the lower filament 

 was considerably less than it was in the experiment to test 

 for the occurrence of ionization as a result of direct electron 

 impacts on abnormal atoms (the electron emission being only 

 about 0*01 of what it was in that experiment), and it can 

 be seen from the two curves that the photoelectric effect of 

 radiation produced by electrons from the lower filament is 

 yevy much smaller than the photoelectric effect of the 



