324 
PHYSICS: H. D. ARNOLD AND H. E. IVES Proc. N. A. S. 
no red), selected so that the light transmitted by each gave added space 
currents of the same order of magnitude. 
With the deep red glass interposed the added space currents obtained 
upon brief illumination with various heating currents through the fila- 
ment are shown in figure 1 . The effect increases regularly with tempera- 
ture, but takes a considerable time to rise and fall at the beginning and 
end of the period of illumination. The times of growth and decay depend 
upon the magnitude of the current, and the curves have the general ap- 
pearance of heating and cooling curves. We believe this red light effect 
to be chiefly due to the heating of the filament by the incident radiation. 
Support of this belief is given by figure 2, in which the same increase of 
space current is caused, first (a) by red light, and second (b) by a sudden 
brief increase of filament heating current. It will be seen that the two 
curves are similar. We have also found that the added space current 
varies with filament current in the way that it would for equal added 
filament-power increments; and that the change of resistance of the 
filament is approximately the same whether the space current increment 
is due to red light or filament current. 
With the blue glass, the energy transmission, as measured by a thermo- 
couple, was only a few per cent that of the red glass, although for a chosen 
filament temperature near the middle of the available range both trans- 
mitted radiations gave the same added space current. The heating 
effect of the blue radiation is therefore negligible, and the added space 
current may be considered as a true light effect. String galvanometer 
records of the added space current due to brief blue illumination, for 
various filament currents (temperatures) are reproduced in figure 3. 
It will be seen at once that the added space currents exhibit a behavior 
essentially different from those due to red light or heating. Whereas 
with red light the current continually increases with temperature, with 
blue light the added current at first increases and then starts to decrease. 
At low temperatures the growth and decay are exceedingly slow. As 
the temperature is raised the rates of growth and decay increase until 
at the highest temperature studied the response appears practically in- 
stantaneous. 
It is the existence of this slow response and its variation with tempera- 
ture that we believe differentiate this effect from the true photo-electric 
effect, which, as far as previously known (and in agreement with tests 
made by us on a potassium cell) is instantaneous at all temperatures. 
The true nature of the increased electronic emission under blue illumi- 
nation is not decided by our experiments, but several possibilities suggest 
themselves. It may be a light induced increase in the number of electrons 
available for thermionic emission; or it may be a light induced change 
