CONSTITUTION OF THE STARS— EDDINGTON 135 



atoms and electrons to the passage of X-rays, and so find how many- 

 photons leak out into space per second. We can compare this result 

 with observation — that is to say, we can see whether Professor 

 Shapley catches as many of them with his telescope as (according to 

 our calculation) he ought to catch— in short, whether the star is 

 actually as bright as our calculation makes it. 



In the last few years we have found a complication in the calculation 

 which I must now explain. At an earlier stage we had to ask the 

 physicist to supply a formula giving the temperature of a gas when the 

 pressure and density are known. Not unreasonably he will object: 

 "You have not given me enough information. What is the gas? — 

 oxygen? iron vapor? mercury vapor? or what?" We cannot say. 



But on second thoughts he withdraws the objection. "Never 

 mind. Ordinarily it would make a big difference, but at the high 

 temperatures we are concerned with it makes practically no difference 

 what element we take. The atoms will be almost completely ionized; 

 that is to say, their satelUte electrons will be moving as free particles. 

 We only want to know the average weight per free particle. The 

 number of satellite electrons in an atom is roughly half the atomic 

 weight — so that we shall have roughly 2 units of weight per particle; 

 for example, oxygen of atomic weight IG breaks up into 9 particles, 

 namely 8 electrons and a nucleus; iron of atomic weight 50 breaks up 

 into 27 particles." Owing to this remarkable property it has been 

 possible to make considerable progress with the theory of the interior 

 of a star without knowing what chemical elements it is composed of. 



Those are the physicist's second thoughts. But on third thoughts 

 he exclaims, "Bother! There's hydrogen." The rule that there are 

 two units of weight per particle does not work for hydrogen. It has 

 atomic weight 1 and splits up into a proton and electron, so that the 

 average weight per particle is % — instead of 2. That makes a vast 

 difference. 



A year or two ago the physicist had some alarming fourth thoughts 

 about neutrons; but neutrons are absorbed very easily by atomic 

 nuclei, and I think they will have only a transitory existence on the 

 sun, as on the earth, and never form an appreciable part of the 

 population. So we won't worry about fourth thoughts. The crux of 

 the matter is that, for the purposes of these investigations, there are 

 just two kinds of matter, namely, hydrogen and not-hydrogen. Hydro- 

 gen gives a much lower temperature than not-hydrogeii, and therefore 

 lower brightness for a star of the same mass and radius. Our compari- 

 son of theory and observation can therefore be used in two ways. We 

 can calculate the brightness of a star, assuming the material to be 

 not-hydrogen, compare it with observation, congratulate ourselves on 

 the partial agreement we find, and ponder over the possible sources of 

 the discrepancies which remain — one possible source of discrepancy 



