152 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1926 
stars are much too far away for their distance to be measured in this 
manner, and in no event could the method tell us the distances of the 
most remote stars in the universe, for it can not succeed unless the 
star under observation is seen against a background of even more 
distant stars. It is only quite recently that other methods have 
provided a measure for sounding the furthest depths of the universe. 
The most fruitful of these methods depends on the special proper- 
ties of a certain class of stars called “Cepheid variables,” after their 
prototype, the star 6 Cephei. These stars do not shine with a steady 
light; at intervals which are always perfectly regular, but may range 
for different stars from a few hours to several days, they flash out 
to two or three times their original brightness. Just as the mariner 
recognizes a highthouse from amongst a crowd of other lights by the 
regular succession of its flashes and the nature of these flashes when 
they come, so the astronomer recognizes a Cepheid variable by the 
regularity, period, and nature of its light variations. In 1912 Miss 
Leavitt, of Harvard Observatory, discovered a simple relation be- 
tween the periods and the luminosities of the Cepheids which occur 
in the Smaller Magellanic Cloud; the slower the light variation of 
the Cepheid the more luminous it is—broadly speaking, its lumi- 
nosity varies inversely as a definite power of its period. More re- 
cently Doctor Shapley, the present director of Harvard Observa- 
tory, has shown that this relation, now generally known as the 
“ period-luminosity law,” is true of Cepheid variables in general. 
Whenever the astronomer detects a Cepheid variable and can meas- 
ure the length of its period, he can deduce the amount of light it 
emits. By comparing this with its apparent brightness, as observed 
through a terrestrial telescope, it is easy to determine its distance 
from us. The method is simply that of a mariner who estimates his 
distance from land by identifying a lighthouse, looking up its candle- 
power in a book of reference, and noticing its apparent brightness 
at the spot where he happens to be. The analogy to the parallactic 
method would of course be if the mariner, knowing the speed of his 
ship, should try to estimate his distance from land by noticing the 
rate at which a church spire or chimney on the coast appeared to 
move against a background of distant hills. This method does not 
demand the existence of a lighthouse of known candlepower, but 
would obviously be useless for a mariner far out at sea, and, as we 
have already noticed, it could in no case give the distance of the 
most remote objects visible. 
The discovery of the “ period-luminosity law” opened up a new 
world as regards exact survey of astronomical distances. It was 
first used by Doctor Shapley himself to determine the distances of 
the remarkable objects known as “globular star clusters.” These, 
