558 Mr. C. G. Barkla on nergy of 
of the primary, the same method was employed as has been 
described for air. 
A sheet of paper weighing °373 gr. was placed in the 
primary beam with its plane inclined at an angle of 45° with 
the direction of propagation of the primary rays and with the 
line joining corresponding points of the two apertures R 
and 8. The paper was so placed that the primary beam 
entering electroscope N was not intercepted by it. The 
absorption of the secondary radiation by plates of aluminium 
was, within 2 or 3 per cent., the same as that of the primary 
radiation, showing that the difference in character between 
these radiations was exceedingly small. The energy of 
secondary radiation was then compared with that proceeding 
from air. Correction was made for the absorption of pri- 
mary and secondary radiations in the paper. The absorption 
produced by a sheet of paper of the same material and thick- 
ness as that used in the experiment amounted to about 
7 per cent. of the energy of the radiation normally incident 
upon it. There was also a small correction made for the 
absorption of part of the secondary radiation from air which 
passed through the paper. 
The results are given below :— 
Ratio of Secondary Elec. Deflexion 
Secondary Radiator. to Deflex. of Primary Electroscope. 
BIS OY. PAPEL) hs ewe ce we cage nd 38°'9 to 100 
194 c.c. of air at 74 ems. press. and oe 
17° ©. (-239 er.) eee ip ls: ee 
Correcting for absorption in the paper :— 
Enerey of sec. radiation from °373 gr. paper _ 42'8 
Ener oy of sec. radiation from ‘239 or, air D3) 

This gives as the ratio of the energies of secondary radiation 
from equal masses of paper and air 113 to 110. These are 
as nearly equal as the possible errors in estimating the volume 
of air productive of secondary radiation and in the electro- 
scope deflexions admit. 
We thus arrive at the conclusion that the energy of radiation 
from gases and light solids, the radiation from which differs 
little from that of the primary producing it, is for thin layers 
of equal thickness proportional to the density of the sub- 
stance; that is, for radiation of given intensity the energy 
lost in secondary radiation is proportional to the quantity of 
matter passed through. 
Recent experiments on secondary radiation proceeding 
from thin sheets of metal have also led me to believe—though 
accurate measurements have not yet been made—that from 
these also, though the character of the secondary radiation in 
some cases differs enormously from that of the primary, the 

ee ee ee 
