by Ultra-Violet Light, etc. 
423 
already alluded to, but also by those of Innes and Guggenheimer 
on the velocity of the cathode particles emitted by metals exposed 
not to light but to Rontgen rays. These observers have shown that 
the corpuscles emitted when the Rontgen rays are hard, i.e. when 
the pulses are thin, move faster than those emitted when the rays 
are soft, i.e. when the pulses are thick, and they showed that as 
in the case of light the velocity of the corpuscles is independent 
of the intensity of the rays. Since the thickness of a pulse of 
Rontgen radiation is very small compared with the wave length 
of even ultra-violet light we should expect much greater energy 
in the units in the Rontgen rays (and therefore much greater 
velocities of the corpuscles liberated by their action) than in the 
case of ultra-violet light. The experiments of Lenard and Innes 
show to how great an extent this is the case ; the corpuscles 
emitted under the influence of the ultra-violet light used by 
Lenard had velocities of less than 10 s cm./sec., while the velocity 
of those emitted under the Rontgen rays used by Innes had 
velocities approaching 10 10 cm./sec., that is they had 10,000 times 
the energy of those emitted under ultra-violet light. It is probable 
that the corpuscles emitted under the action of the y rays from 
radium have a still greater velocity and that therefore the units 
in the case of the y radiation have enormously greater energy 
than those in the radiation corresponding to visible light. 
We should know at once the coarseness of the structure 
corresponding to light of any intensity if we knew the amount 
of energy in each unit of light. To get some idea of this energy 
we may suppose that it is measured by the energy in the corpuscle 
liberated when the unit collides with a molecule of a metal. 
With the ultra-violet light used by Lenard the maximum velocity 
of the corpuscles was about 10 s cm./sec., the energy of a corpuscle 
moving with this speed is about 3 x 10 -12 ergs. Light of such 
intensity that 10 -4 ergs pass one square centimetre per second 
would be very faint but would be visible if the light were in the 
visible part of the spectrum. In light of this intensity the 
number of units passing each square centimetre per second would 
be 10 -4 /3 x 10 -12 or 3 3 x 10 7 , there would therefore only on the 
average be 1 unit in 3 x 10 10 /3'3 x 10 7 or about 1000 c.c. ; thus the 
structure of the light would be of an exceedingly coarse character, 
and could perhaps best be pictured by supposing the particles 
on the old emission theory replaced by isolated transverse dis- 
turbances along the lines of force. The greater the frequency of 
the light the greater is the energy in each unit, so that if it 
requires a definite amount of energy to liberate a corpuscle from 
a molecule of a gas, light whose wave length exceeds a particular 
value, which may depend on the nature of the gas, will be unable 
28—2 
