PROF. C. G. BARKLA ON X-RAYS AND THE THEORY OF RADIATION. 
327 
transformed diminishes also, the relation being almost a linear one within the range 
of these experiments. 
[Originally the writer determined the wave-length of a primary radiation from the 
speed of the electrons it ejects, i.e., from the equation ^mv 2 = hn, where h is Planck’s 
constants, — v being obtained from Whiddington’s law. These results are plotted 
in fig. 4a as they are somewhat more regular than those in fig. 4.] 
0-5 
0-4 
io-3 
0-2 
i ,Br Zn Cu b 
fi Co Fe, Cr | 
5 io 
15 
20 
25 
Wave-length o 
t primary radiation (K d/ ) 
xio 9 cin. 
Fig. 4. Showing the fraction of the energy of primary radiation absorbed—-K absorption—which is 
transformed into fluorescent (characteristic) X-radiation of series K, in elements Cr, Fe, Co, Ni, Cu, Zn. 
It will be noticed that by producing these lines backward so as to get the trans¬ 
formation fraction when the primary has a wave-length just shorter than the fluo¬ 
rescent (characteristic) radiation, this fraction approaches 0‘5 for the heavier elements; 
there is evidence, too, that from the heaviest substances the maximum transformation 
coefficient does not exceed about 0'5. This is seen by plotting the maximum trans¬ 
formation coefficient for various elements against their atomic weight or by continuing 
the curve in fig. 4 to the left; the value found by extrapolation is evidently very near 
to 0'5. 
In arriving at these results the energies of primary and fluorescent radiations have 
been compared by their total ionizing powers. Now the only possible error arising 
