The nature of the penetrating component of cosmic rays

The measurements by Neddermeyer and Anderson (1937) of the absorp­tion of cosmic-ray particles of low energy by metal plates differ in certain respects from those by Blackett and Wilson (1937). The former results showed that, in the energy range 1∙2 x 108 to 5 x 108 e-volts, two types of particles exist, an absorbable group assumed to behave as theory predicts of electrons and a much more penetrating group, attributed provisionally to heavier particles. On the other hand, we found that all the rays with energy under 2 x 108 e-volts were absorbed like electrons, while for rays of greater energy the average energy loss was very much less. Though a very few energetic particles were found to have a high energy loss, insufficient evidence was then available to justify classifying them as of a nature distinct from the less absorbable rays. Thus we obtained definite experimental evidence that the energy loss of the great majority of the rays varies rapidly with their energy. We concluded, therefore, that the energy loss of a normal electron varies with its energy. We now believe this to be probably false, since the success of the cascade theory of showers, in explaining the transition curve in the atmosphere, and a large part, at any rate, of the phenomena of the transition curves of showers and bursts, has provided fairly strong evidence that there must be a very few energetic rays at sea-level, which have the full radiation loss of electrons, even in heavy elements. It follows that the great majority of the rays, for which the energy loss certainly varies rapidly with energy, are probably not normal electrons. We therefore agree with the view of Neddermeyer and Anderson that it is likely that there are two types of particles present, though the difference in behaviour only exists for energies over 2 x 108 e-volts.