PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 
353 
aluminium. This means that the special additional corpuscular radiation is excited 
in sulphur only by primary radiation slightly more penetrating than that necessary 
to excite the corresponding corpuscular radiation in aluminium. 
In order to obtain the ionization in a substance of atomic weight differing 
considerably from that of nitrogen and oxygen, and thus to get rid of the effect 
of the neighbouring spectral line in one of the ionized substances, the ionization in 
copper was compared with that in air by the same method as in the case of aluminium 
and air. 
rpi v , • ionization in air , . . 
ine uncorrected results tor the ratio ■;— ; - ; - ; - are shown graphically m 
ionization m copper 
tig. 7. The effect of the varying thickness of metal from which the corpuscles 
emerge is seen in the downward slope of the curve with decreasing wave-length 
of the primary radiation; but for this effect the line should be approximately 
horizontal. 
Producing the relative ionization line PQ towards the left we get the relative 
ionization if there had been no additional corpuscular radiation associated with the 
spectral lines of nitrogen and oxygen. Thus the ordinate at C should be multiplied 
by the ratio to bring it to the true relative value of the ionization for a primary 
radiation of absorbability OM. 
Treating the values in this way we get the curve 1, fig. 6, giving the relative 
ionizations in air and copper for various radia'tions. 
The shape of the curve shows that the explanation given in the above cases is the 
correct one; for beyond the wave-length 0'56xl0~ 8 (\/p in A1 = 2‘5) the value of 
