124 
Proceedings of the Royal Society of Edinburgh. [Sess, 
later. At this temperature a slight discharge took place from the negatively 
charged foil due to the heating alone, whether or no the foil was illuminated 
by ultra-violet light. The photo-electric discharge could not be investigated 
at higher temperatures owing to the rapid increase in this hot metal 
discharge with temperature. 
The maximum reduction in the photo-electric current below the initial 
value is thus about 80 per cent., as against 44 per cent, found by Zeleny, 
whilst the temperature at which the discharge current became a minimum 
was about 200° C. higher than that observed by Zeleny. 
After switching off the heating current the photo-electric current was 
found to have increased to 240 units, and was therefore more than double 
its previous value at the temperature of the room. In the course of a 
couple of days, however, it fell to its former value. 
This increase was always obtained when the platinum had been heated 
to temperatures above 350° C., although when heated to say 150° C., at 
temperature at which the current is less than at ordinary temperatures, it 
does not even return immediately to its full original value on taking off the 
heating current, but requires several hours to do so. 
(2) Air at a Pressure of 46 vims, of Mercury. 
The potential difference applied, 200 volts, was high enough, as pre- 
liminary experiments showed, to give saturation currents — as nearly as 
possible independent of the voltage. 
At this pressure we found that the discharge current of 58 units at 
15° C. (the unit being about four times as great as in the experiments at 
760 mms. pressure, since the sensibility of the electrometer was the same 
as before, whilst the capacity in parallel was now 0*01 mfd.) increased 
immediately to 65 units at 78° C. and to 69 units at 160° C., the heating 
current to give the last temperature being 10 amperes. An increase of 
current to 13 amperes gave a temperature of 250° C., and the photo-electric 
current decreased slowly to a final value of about 28 units in an hour. 
The actual readings obtained in one series at this pressure are given in the 
following table, which also gives the readings after breaking the heating 
circuit. It will be seen that the sensibility of the platinum slowly returns 
to practically its initial value. 
