S6 Mr. Richardson on the Po.sltice Ionization produced 



that a negative electrode does not become radioactive by 

 receiving the positive discharge from hot platinum. 



Further experiments showed that at low enough pressures, 

 even at a red heat, the positive leak could be completely 

 stopped by heating the wire continuously. It is therefore 

 quite evident that the positive leak at low pressures cannot 

 be regarded as a definite function of the temperature of the 

 wire. At pressures of 1 mm. or higher, however, results 

 w^ere obtained which indicated that there was a leak due to 

 the surrounding gas which increased with the temperature 

 and pressure and did not fall away with time. 



The fact that the positive leak decreased with time and 

 vanished at low enough pressures led to the view that it was 

 due to some foreign matter present in the wire, and that this 

 gave rise to positive ions which w^ere carried away by the 

 current. On this view the positive ions would probably 

 be carried to the negative electrode and condense there. If 

 this were so the negative electrode would acquire the power, 

 which the hot wire lost, of discharging positive electricity 

 when heated. A new apparatus was therefore constructed 

 in order to test this point. 



Experiments icith Two Hot Wires, 



The new arrangement (Pi. Y. fig. 4) consisted of two spiral 

 electrodes (AB, CD) of fine platinum wire placed parallel 

 to one another and about a centimetre apart. They were 

 supported on copper rods (AA^ kc.) which were carefully 

 insulated with sealing-wax. By maintaining a suitable 

 difference of potential across Ai Bj and Ci t)^ the spirals 

 could be kept at any desired temperature. Aj was also con- 

 nected to a battery so that its potential could be raised to 

 + 400 volts, whilst Ci Di was connected to one quadrant of 

 the electrometer, the other being earthed. The various 

 sealing-wax joints are indicated by shading in the diagram. 

 All the connexions of the electrometer were carefully pro- 

 tected from electrostatic induction by earthed conductors. 



The method of experimenting consisted in heating spiral 1 

 until it gave no sensible leak with -1-200 volts at a given 

 temperature 6. The heating current in spiral 1 was then 

 stopped whilst spiral 2 was maintained at a high temperature 

 for a definite interval of time. During this time spiral '1 

 was, of course, earthed whilst spiral 1 was kept at — 200 

 volts. The current in the second spiral was then stopped, 

 and after a short interval the temperature of the first spiral 

 was raised to 6 again, and the leak with -h 200 volts on the 

 first spiral again measured. It may be permissible to state 



