176 Mr. D. C. H. Florance : A Study of the Ionization 



Consider the shape of the ionization-chamber and the con- 

 sequent paths of the ft rays. Each plate is 7 cm. in diameter, 

 the distance apart is 1 cm., and the active material is spread 

 over a surface 3 cm. in diameter. If the ft rays pass out 

 equally in all directions, not more than 20 per cent, of the 

 total number of ft rays emitted will traverse a distance less 

 than 1*3 cm. ; and about 30 per cent, will have a path of 

 3*5 cm. in air. Hence, instead of the effective distance 

 between the plates being 1 cm., the average path between 

 the plates, allowing for the obliquity of the rays, is roughly 

 2*2 cm. Allowing for this correction, the value of the 

 absorption coefficient should be *014, instead of the calcu- 

 lated value *0064. The experimental value is, however, 

 *04 cm. -1 . 



No account has been taken of scattering and multiple 

 reflexion, and it is obvious that these factors are of great 

 importance in the present case. Since the upper plate is 

 Al foil, the ft rays reflected from the walls and top of the 

 vessel will have a certain ionizing effect on the gas between 

 the plates. It is difficult to evaluate this effect. The 

 number of ft rays striking the walls decrease as the pressure 

 in the vessel increases ; but the ionization due to these 

 reflected ft rays will rise to a maximum and then diminish, 

 The ft rays turned back by the air above the Al foil will 

 first increase to a maximum with pressure, and then decrease 

 on account of the absorption of the ft rays between the 

 plates. The ionization due to these ft rays will increase to 

 a maximum and finally decrease as the pressure is raised. 

 There will also be a small effect due to the ft rays finally 

 reflected back from the plate carrying the active material. 

 A certain number of the primary ft rays that penetrated the 

 bottom plate will be scattered back into the gas. Wilson 

 (Proc. Roy. Soc. March 1912) has recently shown that the 

 absorption coefficient in aluminium of reflected ft particles 

 may have average values ranging from 14 to 40. 



Hence, taking into consideration scattering and multiple 

 reflexion and the consequent variation of ionization due to 

 increase of pressure, the discrepancy between the experi- 

 mental and the calculated value of the absorption coefficient 

 can be explained. The curve (fig. 3) can be resolved into 

 its various components in order to allow for the separate 

 /3-ray effects, but the data at present available are too 

 incomplete to give definite values for all the factors 

 involved. 



An experiment was made to show definitely that the high 

 value obtained for the absorption coefficient was due mainly 



