184 
PHYSICS: WEBSTER AND CLARK 
on the value assumed for Planck's h. Millikan's experiments on photo- 
electrons give 6.57 X 10"" erg sec, while the value obtained in the 
work on rhodium was 6.53. The latter value was obtained from the 
potential at the middle of the sharp upward curvature at the foot of 
each graph and is subject to any errors that may occur in determining 
this potential. This is made uncertain by the width of the slit and 
consequent lack of homogeneity of the rays. If, for example, the 
graph for really homogeneous rays turns down very sharply to the axis, 
FIG. 2. 
and the boundaries of the rays used are made indefinite by the size of 
the source, the potential to be used should be that of the steepest part 
of this curve. This hypothesis seems highly improbable since the 
maximum angle between any two rays coming through the slit is under 
0.10 in this case, giving less than J% deviation of any reflected wave 
length from the mean, and the steepest part of the curve occurs about 
1% from the lowest potential at which any rays can be detected. Con- 
sequently we shall follow the method of calculation previously used. 
The values of h obtained are given in table 1. Siegbahn and Friman's 
