22 
PROF. 0. W. RICHARDSON ON THE EMISSION OF ELECTRONS 
course, have to be done simultaneously with the measurements of the chemical emission, 
and an examination of the preceding data will show that it is hopeless to attempt to 
determine this point accurately under these conditions. For it is just as the currents 
approach the maximum value that the absolute errors become largest and the exact- 
location of the points becomes most difficult. This difficulty can be avoided if we can 
make an auxiliary determination of v 0 the threshold frequency of the light which is 
just high enough to excite any photo-electric emission at all. For if D is the point 
where the foot of a characteristic such as DA for some particular frequency v cuts the 
volt axis and DE is denoted by V, then 
eV = h(v - i' 0 ) 
where e is the charge on an ion, h is Planc!k’s constant and e and V are in the same 
absolute units. By using very large photo-electric currents, points such as D can be 
determined with great accuracy and comparative ease, whereas increasing the magnitude 
of the emission does not diminish the inaccuracies to which determinations of points 
like A are liable. 
Data to which this second method can be applied were obtained in connection with 
the Series I. x C0C1 2 of 3/8/1916 (p. 12). Observations at the end of this series showed 
that with the blue filter the photo-electric saturation current was 248 divisions per 
minute, and that with the green filter it was only 0 -6 div./min. At the same time the 
chemical saturation current was 4-0 divs./min., a capacity of 0-01 mfd. being included 
in each of these measurements. The pressure of the reaction products had risen to 
0-016 mm. at this stage. The fact that the blue light of wave-length 4347-4358 and 
