THE MASS-LUMINOSITY RELATION 
179 
/ 
end of the series; for example, Alcyone although the brightest of the 
Pleiades is by no means the hottest. Exceptions of this kind become rarer 
as we go down the series and apparently the band narrows. Perhaps the 
effect of diverse initial conditions of the stars wears out after they have 
been a long time on the main series. The luminosity-spectrum relation is 
primarily an empirical one, whereas the luminosity-mass relation is based 
on a definite theory; there is no reason to anticipate that the former will 
be as exact as the latter for individual stars. 
It is difficult to know what to make of this constancy of central 
temperature. It may be an accidental relation due to the exhaustion of 
the sources of sub-atomic energy balancing the decrease in demand as the 
mass diminishes; but this seems an unlikely adjustment. Taken at face 
value it suggests that whether a supply of 680 ergs per gram is needed 
(V Puppis) or whether a supply of 0-08 ergs per gram (Krueger 60) the 
star has to rise to 40,000,000° to get it. At this temperature it taps an 
unlimited supply. 
The obvious difficulty is that temperature is a statistical attribute, and 
if we turn to the individual processes, anything characteristic of 40 million 
degrees is still moderately abundant at 20 million degrees. It is true that 
at low temperatures we have experience of critical transitions, but the 
analogy seems inapplicable. Does energy issue freely from matter at 
40,000,000° as steam issues from water at 100°? 
12-2