hy Hot Platinum in Air at Low Pressm'es. iSO" 



111 another case the first wire was maintained at + 20U volts 

 while the second was heated for 5 minutes, and it was 

 found that the induced activity, though definite, was -mailer 

 than when the first wire had been cliarged negatively. 



Measurements were now made to compare the activity 

 induced in these two cases. /. f., when the first wire wa- 

 positive and when it was negative with regard to the second, 

 v.hich was heated. In order that any two experiments may 

 give results which are quantitatively comparable it is 

 necessary that the state of the second (hot) wire should 

 remain the same throughout the experiments. To ensure this 

 the second wire was only heated for 120 seconds each time. 

 An experiment was first made with the first wire at —20(} 

 volts, then with the wire at + 200 volts, and so on. The fact 

 that the activity induced in the first and third experiments 

 (with the first wire negative) was approximately the same 

 shows that the state of the second wire had not greatly 

 altered. The induced acti%-ity here was small in all cases 

 owing to the short time during which the second wire was 

 heated. 



The following table gives the actual numbers which were 

 obtained. The current which was used to heat the first wire 

 was '2i^6 ampere : in all cases this current produced no leak 

 before the second wire was treated in the manner described 

 above. In each case in measuring the leak + 200 volts was 

 put on the first wire. The pressure was "Onl mm. 



Potential of First Wire 

 •when Second Heated. 



Activity Induced 

 in First AVire. 



+200 volt5. 

 -200 „ 

 .+200 „ 

 -200 ,, 



5 



11 ' 

 20 



These results lead to the conclusion that the acti\'ity 

 induced when the first wire is negative is about three times 

 as great as when the first wire is positive. A second series 

 of experiments, wdiich are more satisfactory in so far as 

 they were taken when the second wire was in a more active 

 state, and the eftect observed was therefore much greater 

 compared with incidental uncertainties, show that the dis- 

 parity between the two cases is still greater. In this case 

 the pressure was '067 mm. One observation was taken with 

 the first wire earthed when the second was beino- heated, in 



