Vol. 8, 1922 
PHYSICS: H. E. FARNSWORTH 
253 
Measurements of the velocity distribution of the secondary electrons 
indicate that, below about 9 volts primary velocity, a large percentage 
of the secondary electrons have velocities nearly equal to the primary 
velocity and a very small percentage have velocities less than one volt. 
For higher primary velocities the distribution changes, there being many 
secondary electrons having low velocities. The percentage of secondary 
electrons having velocities less than 5 volts, say, increases as the primary 
velocity increases. There are, however, a few percent having velocities 
nearly equal to that of the primary electrons, when the primary velocity 
is as great as 110 volts. 
The following features of the investigation show that a new factor 
enters into the production of secondary electrons from nickel at about 
9 volts: (1) the sudden change in slope of the secondary electron curve 
at about 9 volts; (2) the difference between the velocity distribution of 
secondary electrons when the primary velocity is 9 volts or less, and that 
when the primary velocity is greater than 9 volts. These results lead one 
to the interpretation that reflection of electrons from nickel occurs for all 
the primary velocities investigated, the reflection increasing between 0 
and 4 volts primary velocity and then remaining constant to 9 volts, where 
emission begins and increases with increase of primary velocity to at least 
260 volts. That there are some high velocity electrons present, when the 
primary velocity is as great as 110 volts, indicates that reflection continues 
after emission begins, so that above 9 volts there is both reflection and emis- 
sion. According to this idea the reflected electrons may be considered to 
have velocities nearly equal to the primary velocity, i.e., reflection in- 
volves a change in direction without much loss in energy, while the emitted 
electrons have a much smaller velocity. 
The above interpretation is further substantiated by a phenomenon 
which was observed when obtaining the velocity distribution of the primary 
electrons. It was found that when the primary velocity was less than 9 
volts, all of the primary electrons had velocities nearly equal to that corre- 
sponding to the applied voltage. For larger primary velocities, however, 
there were some low velocity electrons present, the number increasing with 
the primary velocity. This can be explained, in accordance with the 
previous interpretation, as being due to emission of electrons from the 
edges of the diaphragms which limit the primary beam. Thus, since below 
9 volts no emission occurs there are no low velocity electrons present in 
the primary beam. 
Since the number of secondary electrons obtained at the low velocities 
was much greater before the continued heating than after, one must con- 
clude that the reflection and emission of electrons from the atoms of gas 
on a metal surface is greater than from the metal freed from gas. 
The method used is particularly well adapted to the extension of the 
