stellar clusters 



31 



M{x') = £ [— 0.8 ±1.3], 

 M{y') = B[— 18.0+ 2.2], 

 M{z') = s[— 1.7 ±1.0], 



o{x) = s [10.7 +1.0], 

 a{y) = s[ll.8± 1.6], 

 a(^') £ [ 8.0 ± 0.7]. 



These numbers confirm our general conclusions from the diagrams. Especially 

 we note that the dispersions in the x- and in the ^'-coordinate are of the same 

 order of magnitude and only half the dispersion in the ^/-coordinate. The values 

 of M(x'), M(t/'} and give us the coordinates of the centre of the system of 



the globular clusters. 



17. It remains to determine the scale s. This determination is much more 

 difficult here than regarding the ordinary clusters. The relative distribution of the 

 latter was the same as that of the B-stars and it was very plausible to assume 

 both systems to be grouped around the same centre, all the more as we got nearly 

 the same value of the scale from the centre-coordinates as from the dispersion (in 

 the s coordinate). Regarding the globular clusters the things stand otherwise. There 

 can be no question but that the system of the globular clusters and the system 

 of the jB-stars (and that of the ordinary clusters) have different centra. We conclude 

 that either the globular clusters form a system being a part of our Galaxy though 

 not uniformly distributed around the centre of the Galaxy or they form a system 

 separated from the galaxy. In the latter case we are eventually led to the con- 

 clusion that our Galaxy might itself be a globular cluster forming part of the whole 

 system of globular clusters. Such a conclusion, improbable as it may seem to be, 

 has indeed been earnestly advocated from competent side. There are also observa- 

 tional results speaking for an extra-galactic situation of the globular clusters, though, 

 for ray part, I do not consider them conclusive. These results will be considered 

 in the next paragraph. 



On the other side there are direct reasons for considering the globular clusters 

 related to the Galaxy. First their apparent distribution which is nearly symmetrical 

 to the Milky Way or more precisely to a circle somewhat south of the path of the 

 Milky Way, as it would be if symmetrically placed to the XF-plane of the Galaxy. 

 If the Galaxy would be a member of a system of globular clusters, it is a singular 

 chance that the XY-plane (the plane of symmetry) of the Galaxy would pass through 

 the centre of this system, whereas such a coincidence needs no explanation if the 

 globular clusters form a part of the Galaxy. 



I find another support for the galactic nature of the globular clusters in the 

 distribution of the planetary nebulae. It seems not to have been noted before that 

 the apparent distribution of these objects is nearly of the same nature as that of 

 the globular clusters. I give in figure b of plate IV the apparent distribution of the 

 planetary nebulae in the same projection as that used for the globulars, the square 

 öC^^ being placed at the midst of the diagram at the side of GCy^. The similarity 

 with the distribution of the globular clusters is obvious. The planetaries seem to 



