CHAPTER I. 



Ordinary clusters. 



1. Among the stellar clusters we distinguish according to William Herscuel 

 two different kinds of objects: globular clusters and ordinary clusters. The latter 

 are also shortly called only clusters. This distinction is originally only made in 

 regard to the form of the objects, the globular clusters being round and regular in 

 form whereas the other clusters are characterized by a more or less irregular distri- 

 bution of the stars belonging to them. Judging from appearance alone one may 

 well consider it natural that an ordinary irregular cluster may in the course of 

 time be able to develop into a globular cluster and that on the other side the glo- 

 bulars may only be considered as a station on the way to new diffused clusters. 



Neither of these possibilities seems, however, to be in accordance with nature. 

 The two kinds of stellar clusters seem, indeed, to be distinguished from another 

 not only in regard to form but also in more essential respects, so that we have 

 good reason to consider them as two essentially different representatives for com- 

 pound stellar systems which have, genetically, nothing at all with another to do. 

 The most weighty reasons for this genetic distinction between the two kinds of 

 clusters are obtained from a discussion of their distribution on the celestial sphere 

 as well as of their distribution in space We will discuss here these distributions 

 separately for each one of the two kinds of clusters beginning with the ordinary 

 clusters. 



In the three great catalogues of Dreyer we find in all 657 clusters registered. 

 The apparent distribution of these clusters on the celestial sphere is shown ou 

 Plate I of my memoir Preliminary Statistics of Nebulae and Clusters». Meddel 

 N:o 56. The adherence of the clusters to the Milky Way is strongly pronounced, 

 and would be even more accentuated if a galactic system of coordinates were used, 

 as will be found below. 



The use of all this material for a discussion of the distribution of the clusters 

 in jpace is unfortunately scarcely possible, owing to the summary character of the 

 description given by the observers. Another difficulty is the heterogeneity ^of the 

 material caused by the different apprehension of the observers of the Heeschelian 

 scheme used at the description of clusters and nebulae. The relative scale of each 



