Supplement to “ Nature,” May 12, 1923 
vii 

to make these calculations for globes of gas of various 
dimensions. Let him start with a globe containing 
Io grams, then 100 grams, 1000 grams, and so on, so 
that his mth globe contains ro" grams. They mount up 
in size rather rapidly. No. 1 is about the weight of a 
letter ; No. 5, a man; No. 8, an airship; No. ro, an 
ocean liner ; after that comparisons are difficult to 
find. The following table gives part of his results : 
No. of Globe. Xtherial Pressure. Material Pressure. 
30 000000016 0-99999984 
31 0000016 0-999984 
32 0-0016 0:9984 
33 0-106 0-894 
34 0°570 0°430 
35 0-850 0-150 
36 0-951 0-049 
37 0-984 0-016 
38 0-9951 0:0049 
39 0-9984 0-0016 
40 0:99951 0:00049 
It is obvious why I omit the rest of the table; it con- 
sists of long strings of o’s and 9’s. But for the 33rd, 
34th, and 35th globes the table becomes interesting ; 
and then lapses back into 9’s and o’s again. Regarded 
as a tussle between ether and matter to control the 
situation, the contest is too one-sided to be interesting, 
except just from Nos. 33 to 35, where something more 
exciting may be expected. 
Now let us draw aside the veil of cloud behind 
which our physicist has been working and let him 
look up into the skies. He will find there a thousand 
million globes of gas all of mass between the 33rd and 
35th globes, The lightest known star comes just 
below the 33rd globe; the heaviest known star is 
just beyond the 35th globe. The vast majority are 
between Nos. 33 and 34, just where the etherial 
pressure begins to be an important factor in the 
situation. 
It is a remarkable fact that the matter of the universe 
has aggregated primarily into units of nearly constant 
mass. The stars differ from one another in brightness, 
density, temperature, etc., very widely; but they 
all contain, roughly, the same amount of material. 
With a few exceptions they range from } to 5 times 
the mass of the sun. I think we can no longer be in 
serious doubt as to the general cause of this, although 
the details of the explanation may be difficult. Gravita- 
tion is the force which condenses matter ; it would if 
unresisted draw more and more matter together, 
building globes of enormous size. Against this, 
etherial pressure is the main disruptive force (doubt- 
less assisted by the centrifugal force of the star’s 
rotation) ; its function is to prevent the accumulation 
of large masses. But this resistance, as we see, only 
»begins to be serious when the mass has already nearly 
reached the 33rd globe ; and if indeed it is efficacious, 
it will stop the accumulation before the 35th globe 
is reached, because by then it has practically com- 
pletely ousted its more passive partner (material 
pressure). We do not need to know exactly how 
strong the resistance must be in order to prevent the 
accumulation, because, when once the resistance begins 
to be appreciable, it increases very rapidly and will 
very soon reach whatever value is required. All over 
the universe the masses of the stars bear witness that 
the gravitational aggregation proceeded just to the 
point at which the opposing force was called into play 
and became too strong for it. 
ASCENDING AND DESCENDING TEMPERATURE 
STAGES. 
It was shown by Homer Lane in 1870 that as a’ 
gaseous star contracts its temperature will rise. 
Betelgeuse is typical of the first stage when the tem- 
perature has risen just far enough for the star to be 
luminous. It will go on contracting and becoming 
hotter, its light changing from red to yellow and then 
to white. But evidently this cannot go on indefinitely. 
When the condensation has proceeded far enough the 
material will be too dense to follow the laws of a 
perfect gas. A different law then begins to take 
control. The rise of temperature becomes less rapid, 
is checked, and finally the temperature falls. We can 
calculate that the greatest temperature is reached at 
a density of about } to $ that of water. The sun is 
denser than water, so that it has passed the summit 
and is in the stage of falling temperature. So long 
as the temperature is rising the brightness of the star 
scarcely changes. It is becoming hotter, but smaller. 
Calculation shows that the increased output of light 
and heat per square metre of surface, and the decreased 
area of the surface, very nearly counteract one another, 
so that the total output remains fairly steady. But 
on the downward path the falling temperature and 
diminishing surface both reduce the light, which falls 
off rapidly between the successive stages or types 
which we recognise. That is entirely in accordance 
with what is observed to happen. 
Taking any level of temperature, a star will pas: 
through it twice, once ascending and once descending. 
In the main we have been in the habit of classifying 
stars according to their surface temperature, because 
it is on this that the spectral characteristics of the 
light, its colour, and the chemical elements revealed, 
chiefly depend. But that classification mixes together 
stars from an early ascending stage and from a later 
descending stage. For example, a star like Betelgeuse 
just beginning its career is put in the same class with 
a dense red star which has run its course and reached 
its second childhood. They are both red stars of low 
