133 
1906-7.] Photo-Electric Discharge from Platinum. 
wire, measuring the leak to a piece of gauze, an arrangement which would 
hardly give a uniform field. 
Furthermore, in Zeleny’s experiments the readings were taken at intervals 
rapidly succeeding one another, only some twenty minutes being occupied 
in taking the complete series, one reading only being taken at each tempera- 
ture, and no time being allowed for the current to take up its final value 
at any particular temperature — a factor of the highest importance. 
The results in the different gases at 46 nuns, pressure show no very 
marked difference from those at the higher pressure. In the case of 
neither carbon dioxide nor air was a point reached at which the current 
began to increase with the temperature, but this was doubtless due to the 
fact that the temperature could not be raised above 400° C. in gases at this 
pressure. 
In no case could the currents be measured accurately at temperatures 
much above 500° C. owing to the constant emission of negative corpuscles 
due to the temperature alone. This effect commences about 500° C., and 
increases very rapidly as the temperature is further raised. 
The experiments show that in air and carbon dioxide, and to a much 
less marked extent in hydrogen, time is required before the photo-electric 
current reaches its final value at any temperature, and on reducing the 
temperature to that of the atmosphere, many hours may elapse before the 
current returns to its original value. If in air or carbon dioxide the 
platinum was heated to above the temperature at which inversion took 
place (about 400° C.), a large increase in the sensibility occurred, the photo- 
electric current at 14° C. being twice as great after as before the heating, 
and requiring many hours to fall again to the original value. If heated to 
temperatures not exceeding 300° C. in these gases or in hydrogen, no such 
increase in sensibility was observed. 
Some change in the surface of the metal must therefore take place at 
these high temperatures, for if the change with temperature was due to say 
a change in the secondary ionisation in the gas, near to the surface of the 
metal of course, we should naturally expect the readings at any particular 
temperature to be the same, irrespective of the previous history of the 
metallic surface itself. The increase cannot well be accounted for by an 
absorption of gas, since it occurs in carbon dioxide, which is not known 
to be absorbed by platinum. 
The experiments carried out at very low pressures in the different gases 
show that under these conditions the behaviour of the platinum is inde- 
pendent, to a large extent at least, of the gas. In every case a comparatively 
small heating current was sufficient to increase the photo-electric sensibility 
