May 17, 1912] 
the magnitude disclosed by the work of the Galilee 
and of the Carnegie. 
The Carnegie is at present making a circum- 
navigation eruise and is expected back in New 
York towards the end of 1913, having left the 
same port in June, 1910. Up to February 1, 1912, 
this vessel has already covered about fifty thousand 
miles. She left Manila on March 23, in command 
of W. J. Peters, bound for the Fiji Islands. 
Owing to the non-magnetic construction of the 
Carnegie and the absence in consequence of any 
deviation corrections, it is possible to obtain and 
communicate results expeditiously. The data are 
promptly transmitted to the chief hydrographic 
establishments issuing magnetic charts in order to 
enable them to make the necessary corrections from 
time to time. 
SATURDAY, APRIL 20—AFTERNOON SESSION 
Vice-president Edward C. Pickering in the chair 
Symposium on Stellar Spectroscopy—Radial Ve- 
locity (Illustrated) : WiLLIAM W. CAMPBELL. 
Our knowledge of motions of the planets around 
the sun and of the satellites around the planets 
has been reduced to a system, of remarkable accu- 
racy. Knowledge concerning the motions of the 
stars is growing rapidly: shall we be able to 
develop this knowledge into a system? 
Stellar motions are studied through their com- 
ponents known as proper motions and radial veloct- 
ties. Proper motion studies of the stars have been 
fruitful, but limited in value by our ignorance of 
stellar distances. Radial velocities have the ad- 
vantage that they are independent of stellar dis- 
tances. Z 
All observed stellar motions contain components 
due to the motions of the observer. The first step 
in studies of stellar motions is to determine the 
elements of the solar motion and to eliminate its 
effects from the observed motions of the stars, 
thus leaving the motions with reference to the 
stellar system. 
The direction of the solar motion has long been 
fairly well known: the solar system is approaching 
a point 10° or 15° southwest of Vega. The speed 
determined from 1,200 radial velocities is 193 kilo- 
meters (12 miles) per second. 
The velocities of the stars are functions of their 
spectral classes; 7. e., of their effective ages. The 
young stars are traveling slowly—12 kilometers per 
second, on the average; the middle-aged stars more 
rapidly—28 + km. per second; and the old stars 
the most rapidly—34 + km. per second. Our sun, 
SCIENCE 
793 
as a middle-aged star, is traveling with a speed, 
193 km. per second, far below the average of its 
class. We do not know why stars increase their 
speeds as they grow older. 
Amongst the brighter and nearer stars those 
resembling our sun in effective age predominate; 
and they partake somewhat of the solar motion. 
Neglecting these brighter middle-aged stars, the 
remaining stars form a fairly homogeneous mix- 
ture of stars of all ages. 
Radial velocity data increase our estimate of the 
scale of the universe about 50 per cent. above 
proper motion estimates. 
Observers with great telescopes have found that 
one star in eighteen, down to the 9th magnitude, is 
a double star, and that one in eleven or twelve 
brighter than the 5th magnitude is double. The 
spectrograph has shown that at least one star in 
four, down to the 5th magnitude, is a double star; 
with components so close together that they have 
never been seen separately in our most powerful 
telescopes. For young stars the observed propor- 
tion is greatest, the components are closer together, 
and the orbits are rounder. For the old stars the 
proportion of observed doubles is smallest, the 
orbits are larger and more elongated. The propor- 
tion of discovered doubles, especially amongst the 
old stars, is rapidly increasing. Conclusion: 
Double stars have been joined and developed from 
parent gaseous masses, substantially as mathemat- 
ical astronomers had predicted. There is the 
utmost need for cooperation amongst astronomers 
in observing the radial velocities of stars between 
the fifth and seventh magnitudes. 
Objective Prism Spectra: EDWARD C. PICKERING. 
On the Prospect of Obtaining Radial Velocities by 
Means of the Objective Prism: FRANK SCHLEs- 
INGER. 
Relations between the Spectra and other Charac- 
teristics of the Stars (Illustrated): Henry N. 
RUSSELL. 
Among the stars whose distances can be meas- 
ured with some approach to accuracy, and whose 
teal brightness can thus be determined, there 
exists, with few exceptions, a very marked relation 
between the actual brightness and the class of 
spectrum. The stars of spectrum A (resembling 
Sirius) are on the average about 50 times as bright 
as the sun; those of spectrum F'5 (like Procyon) 
about five times as bright as the sun; those with 
spectra like the sun’s (class G) are nearly equal 
to the sun in brightness; while the orange stars of 
spectrum K average only one sixth as bright, and 
