264 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1911. 
distance by parallax measurements, that we can obtain them, and 
that we can hope to increase our knowledge of stellar systems. 
Another interesting variable is u Herculis, whose orbital elements 
were determined by Schlesinger at Allegheny from radial velocity 
measurements with the spectroscope. He finds that the brighter 
star is about 5,000,000 miles in diameter—six times our sun—is 7.5 
times as massive but only one twenty-seventh as dense as the sun. 
The fainter star is 2.9 times as massive but only one-seventieth as 
dense. The parallax of this star is not known, but if it is as luminous 
as Algol the brighter star must give out about 8,000 times as much 
light as the sun. 
There has been a very marked advance in recent years in stellar 
spectroscopy, particularly in the line of the determination of the 
radial velocities of the brighter stars, and several observatories are 
now engaged in this work. Accurate radial velocity measures were 
first obtained by Prof. Campbell at the Lick Observatory in 1896 or 
1897, and for many years he was practically the only one doing that 
work. Campbell’s work has been the determination of the radial 
velocity of all stars in the sky, containing spectra with well measur- 
able lines, which are brighter than the fifth visual magnitude. This 
work is now practically completed, and a preliminary value of the 
direction and magnitude of the sun’s motion in space, with numerous 
other interesting and valuable deductions, are just being published. 
In his work and that of Frost, of the Yerkes Observatory, who 
is measuring the radial velocities of Orion type stars, many spec- 
troscopic binaries—stars whose radial motion varies, and which are 
hence accompanied by invisible companions, as distinguished from 
visual binaries where both stars are seen—have been discovered, 
and it is believed that not fewer than one in three of all stars must 
have a companion of approximately the same size, thus eliminating 
in these cases all possibility of a planetary system like our own. 
Great advances have been made in determining the orbits, the char- 
acter of the motion around one another, of these binaries, and the 
two institutions most active in this line of work are the Allegheny 
and the Dominion Observatories. Of the 70 spectroscopic binary 
orbits determined, our observatory has obtained 16, which, con-_ 
sidering that the aperture of its telescope is only half or less that 
of others engaged in the work and that it has been established only 
a comparatively short time, is a creditable showing. The great 
strides made in the determination of spectroscopic binary orbits 
has led to no less than three summaries of the results, containing 
deductions of important conclusions from them, by Campbell, 
Schlesinger, and Ludendorff. I have not time to enter into the 
results deduced from these discussions except to say that it was 
shown that most binary systems probably originate from a revolving 
