108 
Gaseous Stars 
[ch. Ill 
values of N*/A given in table (134) were calculated for the value 2*5. 
To pass to the value /a = 26, they must (§ 94) be multiplied by about 
(2’6/2’5) 8 ‘‘ or T40. When this is done we obtain the values: 
Sun 620, Sirius A 490, a Centauri B 1050, Capella A 350, etc. 
For uranium N*/A is 355, so that for stellar atoms the actual values will 
be of the order of 37, and the values just calculated are about 16 times too 
large. The values of 6 such that reduction by the requisite factor d 7 6 reduces 
them to the uniform value of 37 are : 
Sun 0’69, Sirius 071, a Centauri B 0'64, Capella A 074. 
For the average typical star we may suppose that the necessary reduction 
in temperature is one of 30 per cent, of the temperature in the gaseous state. 
At first sight this may appear too slight a change to produce any profound 
changes in our views of stellar mechanism. But the pressure of radiation 
$aT* is reduced by a factor of (0’70) 4 or 0 - 24, so that it becomes quite small, 
relative to the total pressure, even in quite massive stars, and quite insignificant 
in all others. 
The free electrons and atomic nuclei are so small that after the tempera 
ture has been reduced they continue to exert pressure in accordance with 
Boyle’s law. Their pressure is only 70 per cent, of what it originally was, 
so that for the star to remain in equilibrium, the atoms must make up 
the deficiency of 30 per cent, in the gas-pressure as well as the deficiency of 
76 per cent, in the pressure of radiation. With an atomic number of 95, 
there are about 95 times as many free electrons as atoms, so that each 
atom must exert about x 95, or 40, times the pressure exerted by a 
free electron, and so about 40 times the pressure it would exert if Boyle’s 
law were accurately obeyed. 
To exert a pressure of this magnitude the atoms must be so closely 
packed as to be almost in contact, or rather so closely packed that their 
effective volumes occupy almost the whole of the available space. Such a 
condition may be properly described as a liquid, or semi-liquid, state. 
Our main conclusion, then, is that if Boyle’s law is assumed to be obeyed 
throughout the interior of a star, the observed capacity of the atoms for 
stopping radiation demands an impossibly high atomic weight. We can 
reconcile the observed opacity with reasonable values for the atomic weights 
by supposing the density to be so great that Boyle’s law is not obeyed, but 
the deviations from Boyle’s law have to be so great that the matter must be 
supposed to be in a liquid or semi-liquid state. This refers of course only to 
the central regions of the star ; the outer layers must in any event be gaseous. 
We leave the discussion at this point for the moment, because an entirely 
independent chain of evidence, to be presented in the next chapter, points 
still more conclusively to the liquid or semi-liquid state, as does also a third 
line of evidence to be put forward in Chapter X.