
Secondary Réntgen Radiation. 557 
that calculated on the electronic theory of matter from the 
determined values of the number of molecules per cubic 
centimetre. 
The number of ions (here denoting a molecule + or — an 
electron) necessary to produce the intensity of radiation expe- 
rimentally determined would be between 10°? and 10%! per 
cubic centimetre. This is approximately 10" times the actual 
number of molecules. 
Quantitative measurements thus support the theory that the 
constituents of the molecules are the radiators. 
The laws governing the secondary radiation from gases are 
thus in perfect agreement with the theory that this radiation 
proceeds from the negative corpuscles or electrons during the 
acceleration of their motion by the intense electric fields in 
the Réntgen pulses passing through the medium containing 
them. 
No quantitative measurements of value have previously 
been made on the radiation from solids. 
An experimental determination of the energy of secondary 
radiation proceeding from a brass plate which totally absorbed 
a primary beam was made by H. 8. Allen*. He found that 
the number of ions produced in eee hydrogen by 
the secondary rays trom brass was about 3,45 part Cat the 
number which would have been produced by the primary 
beam if it had been totally absorbed by the gas. This result 
has no special significance regarding the amount of trans- 
formation into secondary radiation, as an unknown fraction 
—but a very large one—of the total energy of secondary 
radiation was absorbed by the metal itself and transformed 
into heat. Only by using very thin sheets can an approxi- 
mation be made to the fraction of the energy used up in the 
production of secondary radiation, or even to the character 
of the radiation proceeding from the metal. Sheets of metal 
of greater thickness give secondary beams of greater average 
penetrative power, for the deeper layers are only reached by 
the penetrating rays, and the most penetrating secondary 
rays penetrate to the surface in greatest proportion. Conse- 
quently the addition of more layers of metal results in the 
superposition of simply the most penetrating secondary rays, 
and the composition of the radiation proceeding from a 
thick plate is thus entirely different from that which is set up 
in each layer. 
To measure the energy of secondary radiation from a solid, 
the radiation from w hich differed little in character from re 
* Phil. Mag. [6] iii. p. 126 (1902), 
