RECENT PROGRESS IN ASTROPHYSICS. 13 
found to vary from —8.6 to —14.6 kilometers, a difference 
far too large to be ascribed to errors of observation. Mr. 
Campbell found that these apparent irregularities can be ex¬ 
plained by assuming a period of 3 days and 23 hours. Now, 
if we apply Doppler’s principle, that the spectrum of a star 
receding from us would be shifted toward the red end of a 
normal spectrum, and that of one approaching us toward 
the violet end, we have the hypothesis that Polaris is a bi¬ 
nary system, that its center of gravity approaches our sun 
with a velocity of 12 kilometers, and that the Pole star re¬ 
volves around this center once in about four days, thus ap¬ 
proaching us in a cycloidal curve. But this is not all. The 
mean velocity of the system was about 8 kilometers larger 
three years previously. From this Mr. Campbell draws the 
conclusion that Polaris is a triple system. 
The periodic doubling or widening of the spectral lines 
of C Centauri is remarkable for the fact that it takes place 
at intervals of 2 and 6 days. A physical explanation is not 
far to seek. If we attribute the doubling of the spectrum 
to two different sources of light, and again apply Doppler’s 
principle, we have the hypothesis of a binary system in 
which the one component describes around the other a rela¬ 
tive eccentric orbit whose major axis is directed toward our 
solar system. The time of revolution would be 8 days, and 
the eccentricity such that the passage around the periastron 
between two quadratures would last 2 days and the rest 6 
days. 
A very important advance in this line, however, is a 
caution, which has been given by Humphrey, Mohler, and 
Wilsing, against pronouncing these hypotheses as certainties. 
They have proved by experiment that a change of pressure 
in the source of light produces a change in the normal 
spectrum. When we consider that a change in pressure 
may be produced by interior chemical or gravitational 
forces as well as by exterior disturbing bodies, we find the 
field for theories rather wide and uncertain. 
Before passing to the third part of this report the writer 
