74 
Transactio7is of the Royal Society of South Africa. 
(2) To find, if possible, a velocity distribution, from the maximum 
downwards, of the particles about the radiator. 
With regard to (1), a fairly complete summary of the work already done 
for the gases, air, and hydrogen is given in Kaye's ' X Rays,' 1917, whilst, in 
the case of (2), Richardson remarks. "The experiments hitherto made which 
bear on this point (that in general there will be other groups of electrons for 
which the maximum energy may in general have any value between hv* 
and zero) have been directed to the determination of the maximum energy 
of the whole group of electrons, and would not be expected to detect the 
simultaneous presence of groups having a less degree of maximum energy 
than that value. "f 
The experimental results of this paper were obtained during the first 
three months of 1918. 
If there exists a velocity distribution of the electrons, then it must be 
noted that the absorption coefficients of Sadler and Beatty are mean values. 
The following points are noted : 
(1) The electrons are emitted from the plate initially in every direction. 
(2) The maximum distance they can travel in air at normal pressure is 
a few millimetres, this being the length of the track of gaseous ions each 
produces by bombardment. These tracks are very irregular in shape. 
(3) The initial speed of ejection conforms approximately to Whidding- 
ton's law, Vs — kw, where is a constant and tv the atomic w^eight (more 
probably twice the atomic number) of the radiator supplying the Rontgen 
rays. 
(4) The ionisation per cm. of the path of each electron is proportional to 
its fall of kinetic energy per cm. which is governed by equation (i). 
(5) The only method of arriving at an absorption coefficient in a gas, 
if absorption coefficient in this respect has any meaning, is by measuring, 
directly or indirectly, the number of secondary gaseous ions produced at 
different distances from the plate emitting the original electrons. 
As early as 1895 Lenard showed that ^/d was practically constant for 
both gaseous and solid substances (\ is the logarithmic absorption coefficient 
and d the density of the absorber) for cathode particles emerging from a 
window of aluminium in a discharge tube. He showed that the absorption 
coefficient was logarithmic, and it measured the change in energy of the 
cathode particles in their passage through matter. FrimanJ has shown 
that when the error for diffusion has been allowed for, the absorption 
coefficient of the fast-moving /3-rays from uranium-X in oxygen, carbon 
dioxide, and the vapour of acetone is approximately proportional to the 
* hv = the quantum of characteristic Rontgen radiation which falls on the plate 
emitting the electrons. 
t ' Proc. Roy. Soc./ a, xciv, p. 272 (1918). 
t 'Ann. der Physik,' xlix, 4, p. 373 (1916). 
