172 BELL SYSTEM TECHNICAL JOURNAL 



of them starting out with values of kinetic energy which are con- 

 siderable fractions of the energy of the original photon. ^^ 



It is this escort of fast electrons which should be the immediate cause 

 of the ionization due to the short-wave gamma-rays. One easily sees 

 how this result is destined to complicate the question of the nature of 

 the cosmic rays. If it can be proved that the ionization attributed to 

 these is caused by fast electrons, this may mean that the electrons are 

 coming into the lower atmosphere from above, in which case most 

 physicists would say "the cosmic rays are fast electrons." On the 

 other hand it may mean that the electrons are produced in the lower 

 atmosphere by photons coming from above, in which case most 

 physicists would say "the cosmic rays are of the nature of gamma- 

 rays." Yet in both cases the immediate agent would be the same; 

 the term "cosmic ray " would be applied in the one case to it, and in the 

 other to the ultimate cause of the immediate agent. This fine dis- 

 tinction, partly physical and partly verbal, is the source of a lot of 

 confusion. 



Consider now the evidence about the nature of this immediate agent. 

 About three years ago the physicist Skobelzyn was engaged in studying 

 the Compton scattering of gamma-rays from a certain radioactive 

 substance, using the "expansion" or "cloud-chamber" method. By 

 this method (it has often been described elsewhere, i" I will therefore 

 omit the details) the ions formed in a glass-walled container of gas 

 are in effect rendered visible, each becoming the centre of a droplet of 

 water, which appears on a photograph taken when the gas is illumin- 

 ated. If during the experiment the chamber is traversed by a fast 

 electron, its path is marked out by an unmistakable train of droplets. 

 If a strong magnetic field is applied to the gas meanwhile, the trails 

 of the fast electrons are visibly curved, and their speeds may be com- 

 puted from the curvatures and the field strengths. Skobelzyn applied 

 a field of 15000 gauss, in which the paths of the electrons liberated from 

 atoms by the gamma-rays were curled up into beautiful spirals. But 

 on examining the six hundred and thirteen photographs which he took, 

 he found twenty-five trains of droplets resembling the curled ones 

 due to the electrons of known cause, but not perceptibly curved at all! 

 and two others of which the curvatures were perceptible, but so slight 



" See for instance the article of H. Kulenkampff {Phys. ZS. 30, 561-567; 1929) 

 where he plots distribution-curves deduced by the quantum-mechanical theory of 

 Klein and Nishina for electrons ejected by light of wave-lengths 2.4- 10"^^ ^nd 

 24-10"". He also considers the influence of the photons in continuing to eject 

 new electrons after their first, second and later impacts; it appears that one photon 

 is likely to start off several electrons at various points of the beam, instead of disap- 

 pearing (so far as ionizing-power is concerned) after its first impact. 



1- As for instance in my Introduction, pp. 45-46, or the first article of this series. 



