from Hot Platinum. 



289 



sulphate was determined and no determination was more than 

 20 degrees from the true value. This agreement was held to be 

 quite good enough for the purpose. 



The following observations were made to see how the current 

 from the wire varied with the potential to which it was raised and 

 if a saturating current could be obtained. The potentials were 

 measured with a Weston voltmeter. A constant current was 

 used to heat the wire and it was sought to keep its temperature 

 constant by pumping out the gas as fast as it was given off. The 

 pressure was thus kept at about '008 mm. ; the resistance of the 

 wire is given to show the extent to which the temperature varied. 

 The maximum current was roughly 2 x 10 -7 amperes and is given 

 in scale divisions of the galvanometer. 



The slightly smaller current at the higher voltages is doubt- 

 less due to the fall in the temperature of the wire as shown by 



Volts on Circuit 



Current in 

 Scale Divisions 



Pressure of gas in 

 mm. of mercury 



Resistance of wire 



- 40 



205 



•0076 



8-752 ohms 



- 80 



318 



•0083 



8-767 „ 



-120 



306 



•0072 



8-722 „ 



-160 



315 



•0086 



8-745 „ 



-256 



275 



•0074 



8-696 „ 



-370 



268 



•0088 



8-700 „ 



the resistance column. The table clearly shows that the saturat- 

 ing potential is somewhere between 40 and 80 volts since there is 

 no increase of the current with higher potentials. No current 

 was obtained under these conditions when the wire was positive. 



In measuring the variation of the current from the wire with 

 the temperature, a potential of — 120 volts was always put on the 

 wire so as to make sure of the cm-rent being saturated. In 

 general the deflexion of the galvanometer decreased with the time, 

 but not more than might be explained by the lowering of tempera- 

 ture produced by the gas which seemed always to be given off 

 from the hot wire. At temperatures below 1400° very little gas 

 was given off and the readings of the galvanometer were quite 

 steady, while at higher temperatures this was not the case. The 

 curves for lower temperatures are therefore much better than 



