22 
PHYSICS: MILLIKAN AND SOUDER 
(2) We have found that the angle of incidence or azimuth of polari- 
zation is not at all essential to the appearance of a strong maximum at 
the wave length at which Pohl and Pringsheim locate the frequency of 
the maximum selective effect, for we find this strong maximum when we 
work not only with unpolarized light but with normal incidence. In 
other words, the only condition for the copious emission of corpuscles 
is the coincidence of an impressed frequency with a natural frequency. 
(3) We have found that with a clean surface the curve expressing the 
relation between the frequency and the photo-current per unit of inci- 
dent energy rises steadily with increasing frequency right through the 
point assigned by Pohl and Pringsheim for the maximum of the selective 
■ — 
^ ■ 4 min. after renewing surface 
(Gas pressure 0-04 mm ) 
/ 
/ 
\ 
43 days afttr 
134 days afte 
o / 
h 
\ 
I- 
— ^ 
-\ 
d 
, 4i 
\ ■ 
J- 
■ 
^ 
/ 
> 
s 
300 3.VJ 40U 430 600 
Wave Length in 
FIG. 2 
effect and that as a surface has opportunity to be acted upon by gas a 
maximum slowly grows at the frequency indicated, and ultimately be- 
comes very large. This means that if we regard the maximum as a cri- 
terion of the selective effect, then the selective effect merges by insen- 
sible gradations into the normal so that no sharp line of demarcation can - 
be set up between them. The results stated under (1) follow from the 
photo-electric determination of Planck's 'h''^ Those stated under (2) 
and (3) follow from the data plotted in figure 2. This data was taken 
as follows: 
Light from a 220 volt quartz-mercury lamp was passed through a 
Hilger monochromatic illuminator into a vacuum chamber in which the 
