KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 60. NtO 8. 23 



With regard to the great uncertainty in the unknown quantities in the former 

 case it seems not to be justified to introduce K into the equations of condition and 

 it will therefore be appropriate to consider the latter solntion as the definitive one. 



It is evident that the Apex-value thus found has nothing to do with the solar 

 Apex commonly adopted. This circumstance gets its natural explanation when we 

 adopt Shapley's opinion, that we are surrounded by an ellipsoidal stel] ar duster, 

 whose plane of symmetry inclines only 12° towards the galactic plane. This local 

 duster contains the majority of the B and A stars known, and will possibly be 

 identified with Charlier's »Galaxy» and Gould's belt of bright stars. Thus it 

 becomes possible also to understand the origin of Charlier's great parallaxes for 

 clusters. From his supposition, that the Helium-stars are phenomena characteristic 

 of the Galaxy, he has been led to quite too small dimensions for our stellar systems. 

 The .B-stars are evidently phenomena appearing in cluster formations. In the new 

 Draper-Catalogue we thus find 26 B — B 3 stars in and close up to the galactic cluster 

 h & x Persei. Hertzsprung's and the author's measurements of effective wave- 

 lengths in N. G. C. 1647 show that the majority of the stars belonging to this cluster 

 must be considered as A and F stars, but that few, if any, B stars are to be found 

 in this group. Measurements by Bergstrand not yet completely finished över 

 effective wave-lengths in the står cloud in Cygnus also show many A stars but 

 probably very few B stars. 



With the adoption of the local cluster theory 198199 it also follows that the 

 solar Apex should be dependent on the average distance to the objects employed in 

 its determination. Should bright stars of earlier types be employed, we obtain a 

 solar Apex with respect to stars within the cluster. Again, should we employ remote 

 stars we ought to obtain an Apex for the solar motion with respect to the galactic 

 system. Our Apex calculated above should consequently best correspond to the 

 motion and velocity of the local cluster with respect to the globular cluster system. 



It is interesting to see how the declination for the solar Apex increases, the 

 fainter the stars employed are. Thus Dyson and Thackery 58 found: 



245° 

 268 

 278 

 280 



272 



Comstock 42 found: 



280 

 288 



D 



Mag. 



+ 16° 



m m 

 1.0 — 4.9 



+ 27 



5.o —5.9 



+ 33 



6.0 — 6.9 



+ 38,6 



7.0-7.9 



+ 43 



8.0-8.9 



+ 58 



8.o 



+ 71 



ll.i 



The concordance between the Apex-value derived from cluster motions and 

 that of Comstock is striking, and perhaps more than a contingency. 



