MOUNT WILSON OBSERVATORY. 203 



The conclusion, therefore, may be drawn that, statistically considered, the 

 mass of a binary system is a function of its absolute magnitude and not of its 

 spectral type. 



The results indicate that the spectroscopic absolute magnitudes of the 

 early F-type stars as derived by the method used for the A stars are more 

 nearly correct than the previous values derived mainly by extrapolation from 

 stars of later type. The correction to the latter system is found to be +0.42 

 magnitude, in agreement with the value +0.5 found by Adams and Joy. 



The geometrical mean masses for the different types of stars agree with the 

 values derived by Scares, except for the B and A stars. The difference in 

 these two cases almost certainly is due to the data for the double stars used 

 in the two investigations. 



The Space-Velocities of the A-Type Stars. 



Stromberg has completed his study of the motions in space of the A-type 

 stars to which reference was made in the last annual report. The chief result 

 of this investigation is the proof of the existence of three groups of stars which 

 differ from one another in their systematic motions. Two of these are to be 

 identified with the well-known Taurus and Ursa Major groups, but the 

 majority of the stars belong to a central group with an ellipsoidal distribution 

 of motions. This central group has the same group-motion but a smaller 

 dispersion in velocities than the ellipsoidal group to which stars of later types 

 belong. The low value of the solar motion as derived from the A-type stars is 

 due to the large proportion of stars which belong to the Ursa Major group. 



The Asymmetry in the Distribution of Cosmic Velocities. 



The lack of symmetry in the distribution of stellar motions in space has 

 been made the subject of a special investigation by Stromberg. He now 

 finds that this phenomenon, which is especially marked among stars of high 

 velocity, can be traced to all classes of stars and to the globular star-clusters 

 and spiral nebulae. It shows itself as an increase in the sun's velocity with an 

 increase in the internal velocities of the group of objects to which the solar 

 motion is referred. No star with a space-velocity exceeding 100 km. has 

 yet been found which is moving toward the portion of the sky lying between 

 the limits of galactic longitude —10° and +150? 



The phenomenon appears to be a perfectly general one everywhere in space. 

 As a purely tentative hypothesis and one calculated to promote discussion, 

 Stromberg has suggested the following explanation: If a certain reference 

 system is assumed to exist in space referred to which excessive velocities 

 for some reason do not occur, it can be shown that this "velocity-restriction," 

 in combination with the internal motions of the objects in the different 

 groups, will produce an effect similar to that which is observed. The sun's 

 velocity in this frame of reference will probably be the same as that which 

 is found when it is referred to the spiral nebulae. This is about 400 km. 

 towards the apex a = 304°, 8= +81°. The hypothesis is worthy of consider- 

 ation as an attempt to explain an extraordinarily far-reaching phenomenon. 



The Spectrum of o Ceti. 



The spectrum of this important variable star has been the subject of a 

 further investigation by Joy. The radial velocity over seven cycles of the 

 star's variation in light shows changes which doubtless have a vital bearing 

 on the interpretation of the behavior of this class of variables. The velocity 



