PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 
361 
It was evident then that the sudden change was due to an increase in the 
ionization in air, and was such as would be given if nitrogen gave an X-ray 
spectral line of absorbability in aluminium (\/p) M equal to about 2‘5, or wave-length 
about 0'56 x 10~ 8 cm. 
On continuing the S0 2 ionization experiments with X-rays of shorter wave-length 
still, the ratio of ionizations rose again. This effect indicated that sulphur emits 
a characteristic X-radiation of shorter wave-length than that from nitrogen and 
oxygen (as might be expected from analogy with the K and L series of radiations), 
and as the primary radiation became of shorter wave-length than the radiation 
characteristic of sulphur, this characteristic radiation was excited, and was accom¬ 
panied by the corresponding corpuscular radiation from that substance producing 
an increased ionization in sulphur dioxide. The rise and fall of the ratio 
ionization in air ,, • , , • 
; — ; --- :——with decreasing wave-length or increasing penetrating power is 
ionization m fe<J 2 
shown in fig. 6, curve 3. 
ri 
o 
• fH 
-p 
<6 
ts 
H 
5 
o 
• i—i 
4-3 
Ionization in Air Ionization in Air lonizationin Air 
: - ~ - A-- 
IonizationinCopper Ionization in Aluminium Ionization in S0 2 
Absorbability in aluminium 
Fig. 6. 
Corpuscular Radiation Method. 
To obtain further confirmation of this, it was decided to compare the ionization in 
aluminium with that in air. As the ionization can only be directly measured when 
the substance is in the gaseous state, the method adopted was of necessity an indirect 
one. Ionization is usually produced by the high speed electrons from a gas passing 
through the same gas. Instead of this the electrons from aluminium were allowed to 
