————— ee ore 
EVOLUTION OF THE UNIVERSE. 593 
very rapidly discharged our electroscope when it got within 80 em. of 
water of the top of the atmosphere. In other words, there is a definite 
and strong cosmic ray band of penetrating power about five times (energy 
ten times) that of the hardest gamma rays of thorium. The analysis of 
this band on its other side, that is, on its short wave-length side, into much 
lower intensity bands but of energies of the order of 100,000,000 volts and 
180,000,000 volts respectively, has been made by Cameron and me," 
though these results are less certain. The existence, however, of a very 
strong cosmic ray band at about 25,000,000 volts, fairly sharply limited on 
both the long wave-length side and the short wave-length side, is altogether 
definitely established by our experiments. This band carries of the order 
of 90 per cent. of the total cosmic ray energy entering the atmosphere. 
Now, the experimental fact of the existence of such a band is only 
interpretable, so far as I can see, by the assumption of some kind of an 
energy-emitting atomic transformation of definite energy-releasing value 
taking place continuously throughout interstellar space. The other 
suggestions which have been made to account for a band having such 
energy can, as it seems to me, be quite definitely eliminated. There are 
but two of them. First, some kind of a cosmical electrical field has been 
postulated of sufficient total potential difference to impart the necessary 
25,000,000 volts to electrons falling through it. There is indeed known 
to be a symmetrical field surrounding the earth, but its value is definitely 
known to be about one million volts and its direction is such as to drive electrons 
out, not in ; in other words, it has one twenty-fifth the requisite strength and 
is of the wrong sign. No potential difference of 25,000,000 volts, sym- 
metrical with respect to the earth and positive at its surface, can possibly 
be assumed without, as I think, colliding head-on with other well- 
established facts ; and even if such a field existed, it would produce a more 
or less continuous spectrum, like the general X-ray spectrum, instead of 
a well-marked band. I cannot see any way of making this suggestion 
reproduce even remotely the experimental situation. 
The second effort has been to find a way by which rays due to the 
annihilation of protons (which are at least thirty-five times too energetic) 
could be softened down to the observed values. The most interesting 
of these suggestions has been made by Dr. Zwicky, who seizes upon the 
nebular red shift recently discovered by Hubble to suggest that if at the 
creation, which he estimates from the maximum possible amount of 
matter in interstellar space would have been about 10" years ago, annihila- 
tion rays began suddenly to be formed, and if these same rays are still 
hurtling about in a spherical universe and are subject to the red shift, 
these rays, with all that have been created since, would now constitute a 
spectrum having a definite limit to its frequency on the long wave-length 
side, but falling off very slowly in intensity on the short wave-length side, 
where the limit would be 25,000,000 volts. In my opinion, this is the only 
conceivable mechanism by which annihilation rays could become reduced 
in energy to that observed on the long wave-length side of the band which 
carries nine-tenths of the energy of the cosmic rays; but the shape of 
the spectrum on the other side is completely wrong, since it represents a 
1 Millikan and Cameron, Phys. Rev., 37, 235; 1931. 
1931 QQ 
