578 DISCUSSION ON THE 
atomic changes are spontaneous, like radio-active change, the difficulty as 
to stability also disappears. But these milder changes cannot provide 
enough energy for the long lives—millions of millions of years—through 
which the stars have to all appearances existed. The equipartition of 
energy in the motion of the stars, as well as in the orbits of binaries, and 
also the small masses of what appear to be the oldest stars, all point to 
extremely long stellar lives. To provide adequate total radiation through- 
out these, we need annihilation of matter, and nothing less will serve. If 
we cannot have this, we must conclude that the universe of stars is still 
quite young, in spite of looking so old ; its appearances of great age must 
all be deceptive. 
Recent developments of the theory of relativity certainly give much 
support to this possibility. These suggest, somewhat strongly, that the 
whole universe may be expanding, while recent astronomical observations, 
if they have been rightly interpreted, indicate that it actually is expand- 
ing, and this at so rapid a rate that it becomes a mere transitory and 
ephemeral structure compared with what we recently thought ; the spectra 
of the great extragalactic nebule seem to indicate that these bodies are 
running away from one another so fast that they cannot have been 
running for long. This reduces the whole life of the universe to a matter 
of hundreds of thousands of millions of years at most, and, incidentally 
in so doing, brings almost complete chaos into the already chaotic problem 
of stellar evolution. 
Another interpretation of the observations—nearly but not quite 
identical with the foregoing—is that the universe retains its size while we 
and all material bodies shrink uniformly. The red shift of the spectra of 
the nebula is then due to the fact that the atoms which emitted the light 
millions of years ago were larger then than the present-day atoms with 
which we measure the light—the shift is, of course, proportional to distance. 
The final end here is a universe in which all matter has shrunk to nothing. 
On the other hand, if the universe is expanding, the stars are merely 
pouring ovt their radiation into a bottomless pit, since the space to be 
filled with radiation is for ever increasing in amount. The total energy © 
of the universe is for ever decreasing in amount, because radiation does 
work in pressing out the boundaries of the universe—just as a gas loses 
energy and so cools when it expands, and presses back the boundaries of 
its ‘universe.’ Thus, the mass of the stars is continually changing into 
energy, while this energy in turn changes into mere additions to the size of 
the universe. There is conservation neither of mass nor of energy. Nor, 
if the evidence of the cosmic radiation is to be trusted, is there any con- 
servation of matter. Matter turns into energy and energy into mere 
bigness of space. 
a ee ee 
Such a universe can never attain a state of maximum entropy; such — 
states abound in our map, but no railway line on which we can possibly : 
be leads to them. The universe can never come to rest; it is destined to 
go on changing for ever—continually swelling in size, and continually 
dissolving into nothing but size, yet never attaining either complete dis- 
solution or truly infinite size. It seems likely that after a sufficient time, 
the different galaxies or star-systems will be scattering away from one 
another with speeds greater than that of light, so that radiation will be 
