4fi 



Sven Wickseil 



of q or & 1 be used. But it will be affected if we correct for a $ 1 varying with 

 galactic latitude. In the main we characterize the velocity ellipsoid as follows: The 

 velocity ellipsoid is an ellipsoid with three unequal axes. The shortest axis is directed 

 toward the galactic pole and the longest one near the galactic point a — 273° 3 = — 15°. 

 The excentricités of the ellipsoid in sections transverse to the three axes are e 12 — 0.76, 

 e is - 0.84, e 23 = 0.B7. If 'O-j varies with galactic latitude those excentricités shall be 

 slightly corrected, else they are highly independent of the assumptions regarding the 

 density- and luminosity-laws. 



The most interesting feature of the ellipsoid as found here from the proper 

 motions is that it has not the same form as the ellipsoid of the radial motions *. 

 Indeed, the radial motions produce an ellipsoid having the medium axes directed 

 toward the pole of the Galaxy. Gyllenberg has found that for all stars brighter 

 than the magnitude 4.9 we have the axes ** 



3j - 19.7 km. 



a 2 = 12.6 » 



a, = 1 6 9 » 



and this form is qualitatively the same for all spectral classes save the B- and 

 ilf-type, of which there is procentually very few among the stars brigher than 6.0. 



Taking the mean of the square axes for the types A, F and G Gyllenberg's 

 values will give: 



o 1 = 21.2 km. 

 a 2 = 13.3 » 



o g = 17.5 » 



The transgalactic axis a 3 lies within 6° of the pole of the Milky Way and the 

 longest galactic axis points towards a = 273° 8 — — 12°. 



Now, of course, those values are not directly comparable with our values for 

 different q as the direction of the axes also varies with q . The more careful 

 comparison shall be saved till a following chapter, where the characteristics referred 

 to a fixed galactic system of coordinates can be discussed. Already here we can, 

 however, note some circumstances. It will be seen, for instance, that the ratio of 

 the two galactic axes is the same in the ellipsoids of proper motion and radial motion. 

 Further that the galactic axes will be of the same magnitude in the two ellipsoids 

 if q is taken about 0.45, but that the transgalactic axis requires a q of the order 

 of magnitude of 0.80. The question is now which of the two sorts of axes shall 

 be compared? Against using the galactic axes speaks the exceedingly low value 

 of q for which the vertex positions do not agree. Against the other axis, however, 

 speaks the fact that it is determined from galactic stars when using the proper 

 motion, but from non-galactic stars when employing the radial velocities. But, any- 

 way, the last objection seems to be the one of least importance. 



* This fact was first noted by Gyllenbkkg. Compare Medd. Nr. 59. 



* Those figures are as yet unpublished. They have been kindly placed at my disposal. 



