164 Prof. Richardson and Mr. Robertson : Effect of 



Mainly for these reasons the procedure adopted in all cases 

 has been, after each change in the conditions, so to regulate 

 the temperature of the wire that the saturation current was 

 maintained at approximately its original value. This could 

 be done rapidly by means of a regulating rheostat. Whilst 

 these conditions are not theoretically ideal they are at any 

 rate pretty definite, and are probably the best that could be 

 adopted for a fairly rapid preliminary examination of the 

 phenomena. In many cases a sufficiently close idea of the 

 actual temperature of the wire can be got from its dimensions 

 and from knowledge already available as to the relation 

 between emission and temperature under the given conditions. 

 It is, however, important to remember that the comparisons 

 made in this work are not between w T ires maintained at a 

 given temperature with different gas pressures, but between 

 wires at such various temperatures as will maintain the 

 thermionic current constant. These various temperatures are 

 in general a function both of the original high vacuum 

 temperature and of the pressure of the admitted gas. 



Experiments with Platinum. 



The experiments with this metal were made 

 with the tube shown in section in fig. 1. The 

 hot platinum wire was M) cm. long and 

 0*184 mm. diameter. The leads were of pla- 

 tinum 14*6 cm. long and 0*52 mm. diameter. 



The cylinder was of platinum foil and was 

 5*0 cm. long and 1'88 cm diameter. The 

 tube was cleaned by boiling nitric acid and 

 distilled water before the experiments. 



In the initial experiments the tube was only 

 exhausted cold and with the wire glowing and 

 was not baked out in the vacuum furnace. 

 This was done so as to realise the conditions 

 referred to in the old experiments on the 

 electron emission from platinum as tho»e per- ^~ 

 taining to a new wire*. Under these circum- 

 stances the characteristics were found to be 

 fairly sensitive to small quantities of hydrogen. 

 The admission of the hydrogen displaced the 

 characteristics away from the direction of 

 accelerating voltages, indicating that the hot 

 wire had become relatively electronegative with 



* Of. ' Emission of Electricity from Hot Bodies,' Chap. iv. p. 102. 



