608 DISCUSSION ON THE 
I think that a possible test of the theory is that, if I am right, cosmic 
rays cannot be formed uniquely of photons, but must contain, like the 
radioactive rays, fast beta rays and alpha particles, and even new rays of 
greater masses and charges. I have shown that the momenta of such 
rays must be reduced by the expansion in about the same ratio as that 
of the photons. 
The mass of a star should be determined by the weight of its original 
atom, and it is conceivable that stars might be born with different masses. 
If the mass of the original atom is too big, the star must be finally broken 
up by radiation pressure, and the original atom must give birth to a cluster 
of stars, chiefly formed by stars of maximum mass. If the star comes 
from an atom, both masses and luminosity are determined by the weight 
of the original atom. Thus thistheory accounts fora mass-luminosity relation. 
The actual theory does not completely bring order into the chaotic 
state of cosmogony imposed by the fact of the expansion of space. 
Explanation of the approximate equipartition of energy between the 
stars, or evolution with loss of mass along the Russell diagram, might be 
dismissed with regret. But I do not see any way to retain these processes 
of evolution, because they are altogether too slow. 
If I had to ask a question of the infallible oracle alluded to by Sir 
James Jeans, I think I should choose this: ‘Has the universe ever been 
at rest, or did the expansion start from the beginning?’ But, I think, 
I would ask the oracle not to give the answer, in order that a subsequent 
generation would not be deprived of the pleasure of searching for and of 
finding the solution. 
If the total time of evolution did not exceed, say, ten times the age of 
the earth, it is quite possible to have a variation of the radius of the 
universe going on, expanding from zero to the actual value. I would 
picture the evolution as follows: At the origin, all the mass of the universe 
would exist in the form of a unique atom; the radius of the universe, 
although not strictly zero, being relatively very small. The whole 
universe would be produced by the disintegration of this primeval atom. 
It can be shown that the radius of space must increase. Some fragments 
retain their products of disintegration and form clusters of stars or 
individual stars of any mass. When the stars are formed, the process of 
formation of the extra-galactic nebulz out of a gaseous material, proposed 
by Sir James Jeans, could be retained for the star-gas filling the space. 
The numerical test works out equally well for this case. 
Whether this is wild imagination or physical hypothesis cannot be said _ 
at present, but we may hope that the question will not wait too long to be 
solved. 
We want two things. First, a theory of nuclear structure sufficient 
to be applied to atoms of extreme weights. For these atoms, the problem 
cannot be separated into a nuclear problem and a problem of surrounding 
electrons ; because it is easily seen that the K ring would merge into the 
nucleus. We must wait, but we may trust the physicists that we do not 
have to wait too long. The second thing we want is a better knowledge 
of the nature of the cosmic rays. The correlation of the theory of nuclear 
structure with further observations on the cosmic rays must answer yes 
or no to our question ; and we shall prefer this answer, however incomplete 
it may be, to any answer of any infallible oracle. 
ee ee 
