154 DR. FRANK HORTON ON THE DISCHARGE OF 



strip gave a leak of 376 x 10~ 9 ampere with a potential difference of 40 volts at a 

 temperature of 1480 C. and pressure 0'0042 millim. of mercury. On testing again, two 

 hours later, at the same temperature and pressure, the leak was 5'05 x 10"' ampere ! 



This increase of the negative leak on standing at low pressures rendered it necessary 

 to have a pressure of several millimetres when comparing the negative leak before and 

 after subliming the calcium on to the cathode. Some observations were therefore 

 made to ascertain the manner in which the leak from the hot platinum varies with 

 the gas pressure in the apparatus over the range of pressures likely to be used in the 

 subsequent experiments. It was found that with 40 volts difference of potential 

 between the electrodes the negative leak at a constant temperature was nearly 

 independent of the gas pressure between 7 '5 millims. and 3 millims. If the pressure 

 was reduced below this, the leak decreased gradually until a very low pressure was 

 reached, when it suddenly increased again to many times its former value. This 

 increase is probably due to the cause mentioned above. It did not always occur at 

 exactly the same pressure, but generally at pressures below O'l millim. Sometimes it 

 only appeared after allowing the apparatus to remain at a low pressure for several 

 hours. 



The conclusion from these experiments is that, for the purpose of comparing the 

 negative leak from platinum with that from calcium or lime, it is best to work with a 

 gas pressure of a few millimetres of mercury and to use a constant voltage of 40 volts, 

 for the current never becomes saturated at this pressure. Working with a constant 

 voltage comes to practically the same thing as measuring the saturation current in 

 each case, for the current passing under a constant electromotive force should be 

 proportional to the number of ions liberated at the surface of the cathode. 



The following table contains the values of the negative leak from the platinum strip 

 at different temperatures in helium at a pressure of 3'236 millims. with a potential 

 difference of 40 volts between the electrodes. The series of observations was repeated 

 several times during the course of two or three days, and the values given were found 

 to be practically constant. The numbers in brackets refer to the order in which the 

 measurements were made. 



TABLE I. Negative Leak from Platinum in Helium Gas at a Pressure of 



3-236 millims. 



Temperature, Negative leak per 



C. centimetre (ampere). 



1331 1- 95x10-" 



1468 3-96 x 10-" 



1542 1- 75x10-* 



1571 2-99 xlO' 7 



1610 5-91 x 10~ 7 



The values given above are slightly smaller than those given by H. A. WILSON for 

 the negative leak per square centimetre from platinum in air at a low pressure. 



