[EvE} IONIZATION WITHIN CLOSED VESSELS 77 
Discussion. 
The problem appears to be too complex for reduction to a simple 
formula. It is, however, possible to give a satisfactory general explana- 
tion of the results, and to show that the ionization effects obtained are 
not out of harmony with the view that y and X rays are similar in type. 
The explanation is derived from Bragg’s theory of the nature of 
the ionization produced by Réntgen and ¥ rays, and from the remark- 
able experimental verification obtained by C. T. R. Wilson by photo- 
graphs of ionization effects in an expansion chamber. 
In the case of Réntgen rays ionization appears to be due to the ex- 
pulsion of electrons from air or other substances, and these electrons, 
having paths about a millimetre in length, produce groups of electrons 
along those paths. 
In the present experiments the testing vessels cut down the in- 
tensity of the Réntgen rays to an extent mainly dependent on the thick- 
ness and density of the walls. The electrons are projected from the 
air and from the walls, their velocities are of the order 10° em. sec.!, 
and their irregular paths about a millimetre in length. Hence there is 
least ionization with thick or heavy walls which cut off the primary 
and secondary Rôntgen rays and prevent them from causing volume 
ionization of the air within. 
But with y rays the ionization is mainly due to the electrons with 
much higher velocity expelled by or from the y rays, chiefly on meet- 
ing the walls of the vessel. These electrons have velocities of the 
order of those of the # rays from radium C, 10% em. sec.-', with an 
average path of one or two metres. Moreover many electrons, before 
absorption, will be reflected again and again from the sides of the 
vessel, ionizing the air along their paths. 
The extent of ionization due to multiple reflection may be gauged 
from the results of Kovarik!, who experimented with / rays striking 
reflectors above and below his testing vessels. 
Tonization Measurements. 





Kovarik 3 rays with Testing vessels 
multiple reflectors. and y rays. 
Eben ay à | 100 100 
TA OR Me Res Soe OG Pie ee 70-5 69 
Or sare RP Senin pcos ee 69-5 69 
Beye ie ater Can 65-0 | 68 
AD ER D PO D TROT 51 | 58 
Cardboardir Enr. 42 49 



1 Phil. Mag. Nov. 1910. 
