114 BIOLOGICAL EFFECTS OF RADIATION 



increases also, but more gradually. Since in the cubic centimeter of air 

 under scrutiny there are also ions which have been produced by the beta 

 particles, we may extend the comparison to them. The number present 

 per primary photon increases also with the distance in much the same 

 way as that of the beta particles, except that at the greater distances the 

 increase is still less marked. This is because on the average the beta 

 particles there have less energy, since some of them have been produced 

 by secondary photons. In the case of the gamma rays of radium this 

 problem has been investigated experimentally in the writer's laboratory 

 with the following results: Starting with radium filtered by 2 mm. of 

 brass and 4 mm. of bakelite, the number of ions produced per second per 

 centimeter of atmospheric air increases rapidly with the distance from 

 the source up to about 100 cm. At this point it is approximately 60 per 

 cent higher than very near the source. Beyond the 100-cm. distance 

 the increase is less rapid. Nevertheless, at a distance of 900 cm. (the 

 maximum used in the experiments) the value is more than twice that 

 obtained very near the source, and the curve indicates a further increase 

 beyond this distance. Since there is no very sharp break in the curve, 

 it is impossible to say just what thickness of air is necessary to bring about 

 equilibrium between the primary photons and the secondary beta rays. 

 For practical purposes 300 cm. probably suffice. The above results 

 indicate that most of the secondary electrons produced by gamma rays 

 are unable to emerge beyond a 100-cm. layer of atmospheric air around 

 the source. However, the actual distance covered by the beta particle 

 in its zigzag path may be much greater then this. An appreciable pro- 

 portion of the secondary electrons have enough energy to carry them 

 beyond the 2-, and even 3-meter layers of air. 



In view of the foregoing considerations it is evident that the ionization 

 produced in a small volume of air at a given point in a beam of (hard) 

 X-rays is due partly to the secondary electrons which are liberated there, 

 and partly to electrons which have been liberated elsewhere and subse- 

 quently reach the volume. If this ^'olume with imaginary boundaries 

 is now surrounded by walls of dense matter, such as the solid material 

 of which ionization chambers are made, it is to be expected that in general 

 the ionization in it will not be the same as before. Most, if not all, of the 

 secondary electrons liberated in the surrounding air, which before reached 

 the volume under consideration, are now stopped by the chamber walls. 

 On the other hand, other secondary electrons set free in the material of 

 the chamber, contribute to the ionization of the air within. Under 

 certain conditions the loss and gain through the presence of the walls 

 may balance. In this case, the chamber measures the actual ionization 

 which would be produced in the same volume of free air. Such a chamber 

 may be said to have no "wall effect." But, it is very important to note 



