158 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1926 
In 1914 Professor H. N. Russell, of Princeton, propounded a 
scheme of stellar evolution whose main features at least have won 
general acceptance. According to this scheme, all the stars are 
moving down the same evolutionary ladder. Some start at the top, 
some perhaps join in part of the way down, but all pursue the same 
course and all end in the same way. At the top of the ladder are 
stars of the very highest luminosity, radiating perhaps ten thousand 
times as much light and heat as our sun. Moving down the ladder, 
the luminosity of the stars decreases, we pass stars like Sirius 
radiating some forty times as much as our sun, then, well down on 
the ladder, our sun and stars of similar luminosity; finally, on still 
lower rungs, are stars so faint as to be almost invisible. No doubt 
there are even lower rungs occupied by stars which have become 
perfectly invisible, but these need not concern us here. 
Since the appearance of Russell’s theory, it has gradually emerged 
that the stars on the highest rungs are of far greater mass than 
those on the lowest rungs. Not only so, but all the stars on any 
one rung—i. e., all stars having the same luminosity—are of approxi- 
mately equal mass, and there is a gradual diminution of mass as 
we pass down the ladder. If, then, as there is no serious reason 
for doubting, the stars are all moving down the ladder as their 
evolution progresses, it follows that they must all the time be di- 
minishing in mass. Having reached this conclusion, it becomes 
natural to conjecture that the diminution of mass precisely repre- 
sents the output of radiation. The hypothesis becomes something 
more than a conjecture when it is found that it satisfies every quan- 
titative test which can be applied to it. 
Since the rate of radiation of the stars on each rung of the ladder 
is known, it becomes an easy matter to calculate the rate at which 
they would be moving down the ladder on the hypothesis that their 
diminution of mass is the exact equivalent of their radiation. A 
simple addition then gives the time which a star would take, on 
the same hypothesis, to pass from any one rung of the ladder to any 
other. . It is found, for instance, that the time from Sirius to our 
sun is about 6,400,000 years; from the brightest of known stars to 
the faintest is of the order of two hundred million million years, 
while from the brightest to our sun is rather over seven million mil- 
lion years. It is significant that these hypothetical ages for different 
types of stars fit in well with estimates that can be made from certain 
purely astronomical evidence which is wholly independent of any 
hypotheses as to the source of steller radiation. Unfortunately the 
evidence is all too technical for discussion here, but it leaves little 
room for doubt that the long-standing problem of the origin of stellar 
radiation has been solved, and that the solution is the amazingly sim- 
