i)0 Mr. Richardson on the Positive Ionization produced 



addition to the cases in the previous experiments. The 

 results are shown in the fbllowinoj table : — 



Potential of First Wire 

 while Second Heated. 



Activity Induced 

 in First Wire. 



134 



27 

 137 



15 



-200vclts. 

 +200 „ 

 -200 „ 

 „ 



It will be noticed that the two numbers obtained with —200 

 volts only differ by two per cent. No importance is to be 

 attached to the difference between the values (27 and 15) 

 obtained with + 200 and volts respectively. 



The numbers on page 88 show that the activity induced by 

 heating the second wire falls off fairly rapidly wath time 

 when the first w^ire is heated. There is, however, no evidence 

 to show that the induced activity falls off with time if the 

 first wire is allowed to remain cold. A definite experiment 

 was made to test this point by allowing the first wire to 

 remain cold for 75 minutes after the second had been heated. 

 On now raising the temperature of the first wire to that at 

 which it previously gave no leak with +200 volts, 45 divi- 

 sions per minute were obtained. This current was practically 

 equal to that which would have been obtained if only two or 

 three minutes had been allowed to elapse. 



The next point that was investigated was to see if a wire 

 which had lost the power of discharging positive electrifica- 

 tion could not be revived by other methods than by charging 

 it negatively in the neighbourhood of a second hot wire. Of 

 these methods the effect of air will be considered first. 



The Adivitt/ induced hj Ah 



leak in the 

 increase by about 



To test this advantage was taken of a small 

 apparatus which caused the pressure to 

 j'^ mm. in twelve hours. The first wire was heated on the 

 afternoon of August 7th by '280 ampere, and the leak with 

 + 200 volts was found to be 1 division per minute at *0023 mm. 

 pressure. On the following morning the pressure had risen 

 to '16 mm., and the leak under the same conditions as before 

 was found to be 45 divisions per minute. The leak induced 

 by air in this way was found to fall off very rapidly with the 

 time. This is readily seen from fig. 5 (PLY.), where the induced 



