146 | ANNUAL REPORT SMITHSONIAN INSTITUTION, 1946 
satellites are as we find them and move as they do because they were 
born that way; and that may be the correct answer. We have some 
evidence that the stars may have been catastrophically born; and, if 
they were, how can we escape the likelihood that minor bodies also 
appeared at that time and came under the gravitational control of the 
nearby stars? Meteors, meteorites, comets, asteroids, and the small 
satellites are almost certainly the products of fragmentation. Why 
not date at least some of them, if not all, from the Crowded Days of 
Chaos, and not struggle to make them the regularized offspring of 
the orderly dynamical processes that are observed, at this late time, 
smoothly operating in our isolated solar system ? 
ON T. AND THE OVERCROWDING 
The isolation of the sun with respect to its neighboring stars is 
average for this part of the Milky Way. The stars in the Pleiades, on 
the other hand, are relatively much closer together, and in the center 
of a globular star cluster, like Messier 13 in Hercules, the separation 
of one star from another must be less than one-hundredth our distance 
of 4.3 light-years from Alpha Centauri, the nearest known neighbor. 
Nevertheless, in these densest clusters the stars are still well separated. 
We see in them no evidence of frequent collisions. 
Similarly, our galaxy of stars is pretty well isolated in metagalactic 
space. The Clouds of Magellan (much smaller galaxies than our 
own) are perhaps within the spherical star haze that surrounds our 
own galaxy; but the nearest great spirals (the Andromeda Nebula 
and Messier 33) are nearly a million light-years away. Nevertheless, 
the galaxies, large and small, in this part of the universe are perhaps 
10 to 50 times as numerous per cubic megaparsec 7? as in most of the 
metagalaxy that is now under observation. The typical galaxies 
average a million or so light-years apart, and their random speeds of 
a few hundred miles a second are not sufficient to provide numerous 
collisions and encounters. 
The dominant motion of the galaxies, as far as we can ascertain it, 
is the spectroscopically measured radial motion. Five hundred ob- 
jects have been measured with the large reflecting telescopes. After 
V.M. Slipher’s pioneer work on the speeds of these external galaxies, 
Milton Humason of Mount Wilson has contributed most of the ob- 
served values. With this material on motion, Hubble has derived the 
well-known linear relation between the red shift and distance. If we 
interpret the red shift in the spectra of the galaxies as the result of 
recession, then the linear relation is between velocity and distance. 
For many good reasons such an interpretation of the red shift is gen- 
122A megaparsec is 3,260,000 light-years. 
