THE SOURCE OF STELLAR ENERGY 
313 
of many giant stars in the system complicates the problem, since it becomes 
impossible to treat it as a single coeval cluster. 
Jeans considers that his cosmogonic theories require, or at least render 
desirable, a greater concentration of the stars in the early history of the stellar 
system; he had indeed postulated this in some of his investigations before 
the foregoing explanation of the expansion of the system was proposed. 
In particular, he welcomes the very long time-scale and the closer concen 
tration of the stars in the past as rendering more probable the event which 
he is forced to postulate as the origin of the solar system, viz. the close 
approach of another star to the sun. This provoked the subtle reply that 
since we know that this event occurred within the last 10 10 years, there 
is not much gain in rendering it more probable in a distant past when it 
did not occur. To this it may fairly be answered that the 10 10 years is the 
interval between the event itself and a direct consequence of the event 
(viz. the evolution of beings capable of speculating on it) and does not in 
any way serve to date it in the evolution of the universe. But the argu 
ment is double-edged; as the time relation between the event and its 
consequence does not date it, so the place relation between the event and 
its consequence does not locate it; and, if we locate neither time nor place 
—if we consider the probability of the event happening not to the sun 
10 10 years ago, but somewhere in the stellar system at some time—it 
already becomes highly probable without an extension of time limit by 
the aid of the hypothesis of radiation of mass. 
220. We attempt to sum up the arguments for and against the hypo 
thesis of radiation of mass, i.e. radiation of a considerable fraction of a 
star’s mass during its life-time. 
1. Unless we admit the hypothesis the time-scale is somewhat cramped 
though it cannot be definitely shown to be inadequate. 
2. It appears to account satisfactorily for the relative numbers of 
stars between different limits of brightness. 
3. It appears that very few stars are born with masses below 2-5, so 
that most dwarf stars must have lost a great part of their original mass. 
The radiation of mass is a natural explanation of this loss though perhaps 
not the only explanation. 
4. On this hypothesis the age of dwarf stars of types G to M is vastly 
greater than the maximum possible age of a giant star. The presence of 
giants and dwarfs together in clusters of coeval stars is a most serious 
objection to the theory. On the other hand dwarfs of small mass although 
present are not so overwhelmingly abundant as in ordinary regions of the 
stellar system. 
5. It has certain consequences of interest in the evolution of double 
% 
Ì