284 
DR. FRANK HORTON ON THE ORIGIN OF THE 
higher temperatures are plotted to one-tenth the scale used for the lower tempera¬ 
tures. The temperatures are represented by the pyrometer readings, to which they 
are roughly proportional. I was unable to calibrate the pyrometer accurately in 
degrees centigrade up to the highest temperatures used in these experiments. The 
lowest temperature represented in the figure (pyrometer reading = 40) corresponds to 
1637° C. ; the highest temperature represented (pyrometer reading = 72) corresponds 
to about 2050° C. 
The curious difference between the values of the thermionic current at a high 
temperature for the two directions of the continuous heating current arises from the 
potential difference of the heating circuit. There was a fall of potential of some 
80 volts along the glowing portion of the filament, and a steady potential difference 
of 210 volts was maintained between one end of the filament and the anodes. The 
potential difference between the anodes and a point near the other end of the filament 
therefore changes considerably when the direction of the heating current is reversed. 
At the lower temperatures the current does not vary very rapidly with the potential 
difference when this is in the neighbourhood of 200 volts, but at the higher tempe¬ 
ratures used in these experiments the current is far from being saturated with 
200 volts, so that the total thermionic current will change considerably when the 
direction of the current in the heating circuit is reversed. If this explanation is 
correct we should expect that the difference between the emissions in the two cases 
would be much less if a considerably larger potential difference were maintained 
between the electrodes of the discharge tube. To test this, the potential difference 
was increased to 600 volts and the experiment was repeated. It was then found 
that even up to 2000° C. there was no marked difference in the value of the 
thermionic current when the direction of the continuous heating current was reversed. 
The experimental points for both alternating and direct current heating fell equally 
nearly to a single curve, showing that the electron emission from the filament under 
these conditions is the same whether an alternating or a direct heating current be 
used. 
On the theory that the electron emission depends upon the re-combination of the 
electrolytically separated constituents of the oxides of which a Nernst filament is 
made, it would be expected that a larger emission would be obtained when an 
alternating heating current is employed, for then the re-combination is complete. The 
experiment just described would be more conclusive in disproving this theory if it 
could have been performed in the absence of oxygen, but, as already stated, the 
filament could not be made to continue uniformly hot with a direct heating current 
for long enough to enable accurate observations to be made. When the experiment 
was tried in a vacuum there was the further difficulty that the liberation of oxygen 
by the continuous current altered the gas pressure in the apparatus, and ionisation 
by collisions greatly increased the measured thermionic current. Experiments were 
also made at as low temperatures as possible in pure nitrogen gas at atmospheric 
