596 DISCUSSION ON THE 
ordinarily rash. If my extrapolation is not significant, theirs must 
represent that lack of significance raised to a very considerable power. 
Let me not, however, overstress the precision of this quantitative 
comparison. It as scarcely necessary to do so, since there is no other energy- 
releasing act of this order of magnitude that can produce this soft component 
of the cosmic rays. There are, to be sure, other atom-building acts which 
might produce as soft or softer rays, and some of these may well be taking 
place, such, for example, as the formation of carbon out of helium, but a 
consideration of the abundance of the elements involved shows that such 
rays would in general be of negligible intensity. The fact that there are 
but some five abundant elements is here a powerful guide. 
So far I have dealt only with the formation of helium out of hydrogen. 
Cameron and I have shown that our complete cosmic ray curve is con- 
sistent with the existence of three other bands corresponding to the three 
other abundant groups of elements which we have called the oxygen 
- group, the silicon group, and the iron group, but the evidence was here 
necessarily qualitative in view of two recently demonstrated cosmic ray 
facts. The first is that referred to above, namely, the evidence that the 
rays come into the earth’s atmosphere as practically pure photons. For 
this means that we may not assume that the harder rays have reached a 
condition of equilibrium with their secondaries at the points at which we 
measure them. ' 
The second is that interesting new fact discovered by Bothe and 
Kolhérster™ a couple of years ago: that the penetrating power of the 
secondary beta rays may apparently, with increasing energy, rise up to, 
or possibly surpass, the penetrating power of the original photons. This 
seems to me to furnish the most simple explanation of the fact brought 
to light by Regener, as well as by Cameron and myself in our work at 
great depths in water, namely, that there is a very small component of the 
cosmic rays of intensity of the order of 1/500 of that of the whole cosmic 
ray beam as it enters the atmosphere, which has a penetrating power 
thirty-five or forty times that of the main cosmic ray band of which we 
have been speaking. The hardest rays of appreciable intensity that should 
be formed by the atomic building process are those of iron, and their 
energy should be about seventeen times that of the helium rays. There 
are, indeed, more penetrating atom-building rays that might be formed, 
but the abundance of the elements corresponding to them is so small that 
these rays should be exceedingly feeble—probably too feeble to make the 
formation of these heavy elements out of hydrogen a good theory as to 
the origin of the hardest component of the cosmic rays. 
There is, however, already some little evidence, which we hope soon to 
complete and extend, that the penetrating power of these very hard rays 
increases more rapidly than does their energy, and in the present state of 
our knowledge this is the best working hypothesis to account for this one 
point which might raise doubts about the completeness of the atom- 
building explanation of the cosmic rays. If, however, it might be assumed 
that proton-annihilation does not take place in the stars, but does take 
place, with atom-building, in interstellar space, then proton-building might 
12 Bothe and Kolhorster, Zeit. fiir Physik, 56, 751; 1929. 
