IMPARTED TO A VACUUM BY HOT CONDUCTORS. 
527 
resistance was removed. The only part of the heating circuit which remained was 
the hattery of twelve storage cells and the adjustable resistance used to regulate the 
current. The magnitude of the latter was determined by means of a small vertical 
ammeter, reading up to 4 amperes, which was inserted in the circuit. 
The first experiments were made with the apparatus shown in fig. 6, and were 
pushed to very high temperatures. In fact, the maximum cuerent from the 
FILAMENT 'J’O THE ALUMINIUM ELECTRODE REACHED THE ENORMOUS VALUE OF 
1-5 ampIire PER SQUARE CENTIMETRE OF CARBON SURFACE. These experiments Were 
made with a lamp which possessed a small air leak that had been stopped by 
embedding in paraffin in the manner already described. When the greatest currents 
were put on the lamp became hot so that the paraffin melted and the pressure inside 
the apparatus rose to 1 millim. During the course of the experiments the pressure 
was therefore not constant, but increased gradually from '006 millim. to 1 millim. 
The potential on the filament was — 250 volts, and was sufficient to saturate the 
current at all the pressures concerned. The results of these observations are shown 
graphically in the accompanying diagram (fig. 14). The values of the ordinates are 
successively multiplied by ten as we move to the left fi'om one curve to the next. 
Fig. 14. 
