the Anode in Vacuum- Tubes. 199 



This implies that the ionization processes are constant over 

 parts 1 and 2, the change in energy being due to changes in 

 mobility of the ions and perhaps in the clustering. 



By referring to Table II. it is seen that the voltage across 

 the tube was always about 660 volts at 1*5 mm. pressure 

 when the curve changes abruptly. The voltages actually 

 registered were 660, 650, 660, 650, 652, and 640 volts. The 

 sharp change in the curve is probably occasioned by the 

 commencement of ionization in the positive column. The 

 constant voltage across the tube at this point also suggests 

 that the gradient has become large enough to effect ioniza- 

 tion by collision. This ionization reduces the anode energy 

 in two ways. In the first place ions are now travelling in 

 both directions, and collisions can reduce considerably the 

 energy of those moving towards the anode, and in the second 

 place the bombarding negative ions may combine with the 

 free positive ions and cease to have any acceleration in the 

 electric field. Additional evidence of ionization is shown by 

 the decrease in " resistance " of the tube. The potential 

 across the tube is slightly below that in the initial part of 

 section 3 (fig. 4). ( For actual values of the potential across 

 the tube see Table II.) 



Part 4 (fig. 4) commences at a pressure of 0*5 mm. At 

 this point the energy diminishes rapidly, probably owing to 

 ionization at the anode surface. Here the current is no 

 longer represented by i = ne, 



but by i = ne + n x e, 



where n x is the number of positive carriers produced per 



second at the anode ; so that the number of bombarding ions 



is reduced from 



% i , 

 n or - to n , 



e e 



which number reaches a minimum when m = /2 1? i. e. when 

 each negative ion sets free a positive ion on striking the 

 anode. It is, then, possible for the lessening number of 

 bombarding ions, aided by the motion of positive ions in the 

 opposite direction, to give up less energy to the anode in 

 spite of the increased gradient. When the minimum number 

 of negative ions is obtained, further reduction in pressure 

 produces an increase in the gradient in the tube, and hence 

 an increase in the energy given up. This part is represented 

 by the sharply rising part 5 (fig. 4). 



These hypotheses lead to a fuller explanation of the dis- 

 charge in vacuum-tubes. Suppose the gas pressure is large. 



