Vol. 7, 1921 
PHYSICS: MILLIKAN AND BARBER 
17 
In the following observations the plate and grid No. 1 are kept at a 
common potential and both at a fixed potential, e. g., 100 volts, above 
the emitter. Figure 2 shows the curve obtained when the plate-current is 
plotted against the potential of grid No. 2 as the potential of the latter is 
varied from 0, the potential of the filament, up to 200 volts. When 
the potential of grid No. 2 has risen above that of grid No. 1, that is, 
above 100 volts, the secondary electrons, emitted by the plate under the 
bombardment of the primary electrons having an energy of 100 volts, are 
drawn from the plate to grid No. 2 and thus reduce the current flowing 
to the plate. Further, the difference between the two constant currents 
above and below the 100 volt value of abscissae is a measure of the secon- 
dary electrons which are attracted back from the plate. The ratio of this 
difference to the initial constant plate-current gives the number of secon- 
dary electrons produced at the plate per primary electron. 
Again the difference between 100 volts and the potential at which the 
current begins to descend (see fig. 2) gives the maximum velocity of emis- 
sion of secondaries in volts. Figures 3 and 4 show similar curves taken 
with primary electrons having energies of 200 and 300 volts, respectively, 
the latter being made with largely expanded abscissae so as to permit of 
a study of the distribution of velocities of the secondary electrons. It 
will be seen from figure 4 that practically no secondaries have a velocity 
of more than 5 volts, even when the exciting primary electrons have 
velocities of 300 volts. Further, figure 4 shows, since the potential reaches 
310 volts before the current is again constant, that about 10 volts are 
required to pull out of the plate all of the secondaries generated. This 
is doubtless due to the fact that some of them are shielded by, or lost in, 
the pits and hollows in the surface of the plate. 
Figures 5 and 6 show the curves obtained with primary electrons having 
energies of 10 and 5 volts, respectively. It will be seen that the maximum 
velocity of the secondaries obtainable from figure 5 is 2 volts. The failure 
of the curve of figure 6 to drop at all as the potential of grid No. 2 passes 
through 5 volts seems to show conclusively that none of the j volt incident 
electrons are reflected, and also that the ionizing potential of the surface is 
above 5 volts; while figure 5 shows that it is below 10 volts. 
The accompanying table shows how the coefficient of secondary emis- 
sion depends upon the energy of impact of the primary electrons in the 
case of a copper surface which had been subjected to red-hot temperatures 
for many hours. The coefficients were some 30% less before such treat- 
ment. On the other hand, increase in the temperature of the plate after 
treatment appeared to decrease the coefficient. 
table 1 
Impact voltage 5 7 10 15 20 25 50 75 100 150 200 250 300 400 500 
Coef. of secondary 
emission 0 .08 .22 .30 .34 .40 .46 .58 .65 .861.00 1.10 1.26 1.25 1.25 
