DEPARTMENT OF MERIDIAN ASTROMETRY. 237 



pointed out whose motion is very nearly directed toward the antapex 

 of solar motion. The absolute agreement between the observed mean 

 proper-motion of groupings arranged according to distance from their 

 convergent, and the value of the proper-motion computed on the basis 

 of true parallelism of motion of the group, pointed to a physical con- 

 nection between its members. The radial velocities of six stars furnish 

 a further criterion for the reality of the group. They were found to 

 satisfy the conditions of the hypothesis, if we assign a velocity to the 

 group of 21 kilometers per second relative to the sun. 



The location of the convergent point of the group, taken in conjunction 

 with the group velocity, indicates that we are dealing with a group of 

 stars practically at rest in space. Observed parallaxes were only 

 available for two stars, but in both cases the agreement with the com- 

 puted value is well within the probable error of determination. From 

 the computed parallaxes it appears that the stars of the group are 

 relatively near the solar system. 



The solution for solar motion by zones of galactic latitude furnished 

 a very discordant direction for solar motion in the case of the treatment 

 of the south galactic polar region. Analysis of the proper-motions in 

 this region disclosed a startling dissimilarity between the distribution 

 of the motions of early and late type stars. For stars whose proper- 

 motion does not exceed 20" per century the ratio of the cross-motions 

 -(- to — is 5.8 for type A, 0.2 for type G, and 0.5 for type K. For 

 large proper-motions the ratio is 1.4 for type F, 0.6 for type G, and 0.3 

 for type K. Radial velocities for early-type stars in the region of the 

 south galactic pole are scarce, but what evidence there is points toward 

 a relative drift between early and late type stars. The discussion has 

 not been entirely finished. 



Following is a summary of a paper by Dr. Albrecht, who has been 

 continuing his study of standards of wave-length: 



Stellar radial velocities are known to contain systematic errors. 

 These may be separated into two divisions : (a) systematic differences 

 between the results for the same stars as observed at different observa- 

 tories, and (6) systematic differences dependent upon the stellar spectral 

 type, which were found by Campbell and designated by the letter K. 

 The former must clearly be due to causes within the control of the 

 observer, i. e., the instruments and the methods of measurement and 

 the reduction of the spectrograms. The latter are due in part to these 

 same causes, though they may — and probably do — depend also upon 

 causes within the stars themselves. 



Owing to the great perfection of modern stellar spectrographs, errors 

 from instrumental causes must be nearly negligible. Dr. Albrecht has 

 previously shown: (1) that a large percentage of the spectrum lines 

 vary in w^ave-length progressively as a function of the stellar spectral 

 type ; (2) that for many of the lines — and these must be used for the 

 determination of the stellar radial velocities — the laboratory values of 



