236 CARNEGIE INSTITUTION OF WASHINGTON. 



galactic poles. If such were the case, the ratio of axes in the projected 

 ellipse should be least in the galactic region, which is contrary to the 

 testimony of this investigation. The evidence, such as it is, tends to 

 the conclusion that the proper-motions within and those without the 

 galaxy are differently distributed, the non-galactic ellipsoid being for 

 earlj' types more prolate. This favors the idea that the the non-galactic 

 stars are outside the galaxy as the result of a selective process. 



5. The well-known increase of velocities with advancing type is 

 distinctly shown. If account be taken of the value of solar velocity 

 as derived from the separate types, the increase in the semi-axes of the 

 velocity-figure, and consequently in the pecuHar motions, is still more 

 marked, and is progressive through type F. 



AMiat may prove to be a vertex of preferential motion for small 

 proper-motion stars within the galaxy has been found. The directions of 

 motion of four groups of stars, two of which are the Pleiades and Prae- 

 sepe groups, converge sharply at a point in R. A. =4*^ 45'", Decl. = — 15°. 

 Each group, besides agreeing as regards amount and direction of proper- 

 motion, is largely composed of early-type stars, which for the most part 

 lie in or near the galaxy. As it has been shown that the large proper- 

 motion stars lie in great part outside the galaxy and that the vertex 

 of preferential motion determined from them is quite distinct from the 

 vertex derived from treatment of stars of average proper-motion, both 

 within and without the galaxy, it is probable that the slow-moving ga- 

 lactic stars will furnish still another vertex of preferential motion. 



In order to make a complete investigation of the phenomenon all 

 parts of the sky were treated for a possible preferential motion toward 

 the given vertex. By plotting the regions where such effects were 

 shown, it soon became manifest that a belt was described coincident 

 with that in which lie the majority of the bright B-type stars and the 

 thickly clustered A-type stars. Consistently with the theory of prefer- 

 ential motion, there was no manifestation of the phenomenon in those 

 portions of the belt more nearly approaching the vertex or antivertex. 



In order to further test the supposition that the preferential motion 

 is mainly confined to the bright belt, all regions 90° from the convergent 

 point were examined. The preferential motion should be a maximum 

 in such a zone. Only 1 1 per cent of the zone showed a marked evidence 

 of the phenomenon sought, and only 26 per cent showed some evidence. 

 Rejecting the region common to the two belts, only 8 per cent showed 

 strongly the effects of preferential motion and only 14 per cent some 

 evidence. On the other hand, considering that portion of the bright 

 belt where preferential motion might be expected to be evidenced, 54 

 per cent shows the effect strongly and 77 per cent shows some effect. 

 It may therefore be concluded, with a reasonable degree of certainty, 

 that a portion of the early-type stars in the galaxy exhibit a preferential 

 motion toward a vertex at R. A. =4'^ 45"*, Decl. = —15°. 



In the Astronomical Journal, Nos. 635-636, a group of stars was 



