76 
Transactions of the Royal Society of South Africa. 
density of these substances. In more complex substances there is a marked 
deviation from proportionality. 
The method employed in this work for the determination of the absorption 
coefficients of the corpuscular rays in various gases was the " pressure 
variation " method used by Beatty. The gases used were air, O3, COo, SOg, 
which were passed slowly through a shallow ionisation chamber, a section 
of which is given in Fig. 1. The gas pressure could be varied from approxi- 
mately zero up to atmospheric pressure, and the ionisation measured with 
an ordinary Wilson electroscope. The quantity of radiation falling into the 
chamber was standardised in the usual way with a standardising electro- 
scope, and as silver X rays w^ere used throughout, a screen of aluminium 
•075 cm. thick was placed in the path of the beam of X rays before it entered 
the chamber in order to eliminate the very easily absorbed L-radiation from 
the silver plate. Two complete sets of experiments were performed, giving 
the two sets of curves shown in Fig. 2. In the upper four, using the four 
gases successively at various pressures, the silver X rays fell on a sheet of 
gold leaf stuck on a sheet of waxed filter-paper which })acked the chamber. 
In the lower four the conditions were precisely the same, except that the 
gold leaf was now absent. The ordinary ''reflection " method was employed 
for obtaining a homogeneous beam. 
The ionisation chamber possessed two unique features. The first was the 
electrode, which was a grid of about 2 cm. mesh of new, very fine carbon 
filament used in electric lamps. This was charged to a potential of about 
240 volts. The second was the thinness of the gold screen. Any attempt 
to detect a true velocity distribution of the cathode particles characteristic 
either of the incident X rays or of the absorbing screen would require an 
infinitely thin screen so as to eliminate the effects of particles emerging from 
deeper layers in it. Hence it was that only one gold leaf was employed at 
the back of the chamber. Its thickness was calculated to be '000008 cm. 
Using formula (i) we have, according to Whiddington, a (for gold) = 2*54 
X 10*^, X = '000008 cm., Vo = 108 x 10^. On calculating Vx it is seen that 
the fall in speed of the fastest cathode particles, or of those having an 
initial velocity of, say, -^Vo, will be inappreciable compared with their maximum 
initial velocities if they were generated at the back of the gold leaf and had 
to penetrate it in order to emerge into the ionisation chamber. 
Prejparation of the Gases. 
Air. — Atmospheric air roughly dried by bubbling through strong H^SO^. 
Oxygen. — From the electrolysis of dilute H^SO^j. The ozone was re- 
moved by passing the gas through a glass tube immersed in a sand-bath kept 
at about 260° C. and then dried by passing through strong HgSOj,. 
Carhon dioxide. — From commercial HCl and marble and dried with 
H2SO4. 
