288 
DR. FRANK HORTON ON THE ORIGIN OF THE 
temperatures were therefore taken as measuring the steady emission from the 
platinum. 
The glowing portion of the filament, the emission from which had already been 
tested, was finely powdered in an agate mortar; a little water was added, and the 
particles in suspension were transferred to the platinum tube by evaporating this 
mixture upon it, a drop at a time. In this way a thin uniform coating of Nernst 
filament material was obtained upon the platinum tube, which was heated to a bright 
red heat in the air for a few minutes before being fitted into the apparatus. After 
reducing the gas pressure to O'OOOl mm. the thermionic emission from the heated 
tube was again investigated, with, as usual, a potential difference of about 210 volts 
between the terminals of the discharge tube. A series of observations at gradually 
increasing temperatures was made, and it was found that the negative emission was 
now very much smaller than the emission from the uncoated platinum tube had been. 
The observations showed, however, that the emission was tending to increase, and 
the tube was therefore left heating for some three or four hours, after which another 
series of measurements at different temperatures was made. The emission was found 
to have increased to about 30 times the original values and was now remarkably 
constant. It was thought desirable to treat the tube exactly as the Nernst filament 
had been treated when its emission was measured, so before testing on the following 
day, air was let into the apparatus to atmospheric pressure, and the tube was raised 
to incandescence for a few minutes, after which the apparatus was evacuated with 
the platinum tube still glowing. The emissions at various temperatures were at once 
measured and were found to be about 30 per cent, smaller than those finally obtained 
on the previous day. On continuing the heating, however, the emission soon rose to 
about the same value as before. These operations were repeated on four days, and in 
every case the final steady values of the thermionic current were within 5 per cent, 
of the steady values first obtained after the initial increase had taken place. It was 
assumed that these currents represented the sum of the emission due to the platinum 
tube and that due to the material upon it, and to obtain the effect of the material 
alone the previously determined values of the emission from the clean platinum tube 
at corresponding temperatures were subtracted. In making this correction the 
current from the thermocouple welded on to the platinum tube was used as an 
indicator of its temperature, and the part of the thermionic current taken as being 
due to the platinum at each temperature was the value which had been obtained 
from the tube alone at a temperature giving the same deflection of the galvanometer 
attached to the thermocouple. 
In comparing the corrected value of the emission from the filament material heated 
on the platinum tube with the emission from the whole filament at the commencement 
of this series of experiments, the optical pyrometer readings were used for temperature 
measurements in. both cases. The comparison is best made by means of emission- 
temperature curves, and in fig. 2 three such curves are given. Of these, Curves I. 
