20 
PHYSICS: MILLIKAN AND SOUDER 
continually with time; while for lines 3126, 3650, and 5461 the sur- 
face is initially relatively insensitive and then grows in sensitiveness to a 
maximum. Now according to Pohl and Pringsheim,^ the selective effect 
in sodium shows a resonance band whose maximum is at about 340 /x/^. 
It is obvious then that if we wish to define the selective effect as the 
strong photo-sensitiveness in the neighborhood of this wave length, 
then these curves mean that the selective effect so defined is due to the 
presence of gas and grows as the gas has time to get at the surface. 
Since we did not wish to define the selective effect in any such arbitrary 
manner we interpreted our result in the following words: 
A freshly cut sodium surface is not sufficiently electro-positive to respond 
photo-electrically to light of wave-length 5461 A., but under the influence of an 
Time-Pkotpcu.rren.'t Cu-Tues for Different Wcu/e. L£>\gtKs. 
active gas, the sodium forms a new substance which is more electro-positive . 
than the freshly cut sodium and hence is photo-sensitive to longer waves. 
The photo curve taken with wave-length 5461 A., represents then merely the 
growth and decay of this substance. For sufficiently short waves, however, 
the freshly cut surface is itself so photo-electrically active that its own decay 
curve completely masks the rise and fall curve due to the growth and decay 
of the more electro-positive substance resulting from the action of gas upon 
the sodium. 
In view of these results the authors raise a question as to the correctness of 
the conclusions of Wiedmann and Hallwachs that photo currents are obtain- 
able only in the presence of gas. We suggest that if the Wiedmann and Hall- 
wachs experiments are repeated with Hght of sufficiently short wave-lengths 
instead of with visible Ught, it is not likely that the photo-sensitiveness will 
