SURVEY OF THE PROBLEM 
7 
/ 
1 more directly 
nfirmation that 
neter measure- 
’hese show that 
enormous bulk, 
them. We are 
f stars beyond 
un. 
Lty rather than 
der 250 million 
ut the mass, or 
and 100 times 
s. It is rather 
has something 
ion km. radius 
ly, the force of 
to escape from 
rth. Secondly, 
} the spectrum 
aid produce so 
close up round 
it gives a more 
0-001; for even 
too great to be 
e conservative 
e found in the 
whose diameter 
sun, would not, 
ive at us; it is 
se may, through 
n the accepted 
it and accord- 
g temperature 
cists in general 
,rs as the latest 
hat any of the 
o Dr H. Jeffreys 
> still commonly 
ire types (K and 
ordinary types of spectrum could be produced in bodies diffuse enough 
to behave as a perfect gas, and supposed that Lane’s theory, if it had any 
astronomical significance, must refer to some pre-stellar stage of develop 
ment. 
About 1913 a revolution of ideas occurred and the “Giant and Dwarf 
Theory” of E. Hertzsprung and H. N. Russell soon gained general 
acceptance. Setting aside certain misgivings which have arisen since 1924, 
we shall summarise the main points of the theory. In principle it was a 
revival of the ideas of Lane and Lockyer; the novel point was the adapta 
tion of these ideas to the observational data, so that each star could be 
assigned its particular place in the scheme. The stars start to be visible 
as cool red stars of type M with low density and enormous bulk. They 
contract and in obedience to Lane’s condition rise in temperature*, passing 
up the spectral series K, G, F to A and B —i.e. the reverse of the pre 
viously accepted order. At some stage of the contraction the density 
becomes too great for the perfect gas laws to apply, the rise of temperature 
is checked, and ultimately the star cools down again as a solid or liquid 
would do; in this last stage it returns down the spectral series to type M 
and ends in extinction. On this theory the stars which had been classed 
together indiscriminately as type G, for example, must be divided into 
two groups, the one making the ascent, the other on the descent, the one 
a nearly perfect gas, the other a very imperfect gas behaving similarly to 
a liquid. The surface conditions being similar, as evidenced by the spectral 
type, the outstanding distinction is that the ascending series or giants 
have much greater volume than the descending series or dwarfs. The greater 
volume and surface of the giant stars gives them greater luminosity, and 
when the absolute magnitudes are studied the division into two groups is 
easily seen. The separation is shown in the types M, K, G and F ; it is 
not to be expected in type A, which marks the turning-point for most 
stars. Naturally it is most striking in type M, where the stars in the most 
diffuse and most concentrated state are brought into contrast; the one 
group clusters about absolute magnitude + l m -5, the other about + 10 m -5, 
aryl there is a clear gap of about 6 m in which no M star has yet been 
detected. 
According to the statistics there is little or no change of absolute 
brightness with type along the giant series; this would be expected since 
the rising temperature and decreasing surface area will keep the total 
light about the same. In descending the dwarf series the decreasing 
temperature and decreasing surface combine to give a rapid falling off of 
brightness. 
I ,. 
* The theory applies to internal temperature, and it was generally taken for 
granted that the observed photospheric temperature would keep step; but this is 
by no means inevitable.