16 



C. V. L. Chai-lier 



surrounding stars. Even with the most regular clusters — the globular clusters — it is 

 not possible to determine a definite limit. The statements regarding the diameters 

 are therefore to a great deal dependent on an arbitrary estimate of the observer. 



For abolishing this arbitrariness it is preferable to base tlie computation not 

 on the diameters themselves but on the dispersion in the radial distribution of the 

 stars of a cluster. I have observed some of the clusters on the Franklin- Adams 

 Charts for determining this dispersion and give here an instance. 



Using a common pocket lens and a glass plate ruled into squares having a 

 side of 1.2,5 mm (=5' on the F. A. C) I counted the stars in the cluster and in 

 its vicinity and obtained the number given in the following table. Each count 

 gives the number of stars shown on the F. A. C. within a square of the dimensions 

 5' X 5'. The direction against the pole is to the left and increasing right ascension 

 is downwards. 



Count of stars around N. G. C. 1912 = M 38. 







3 



0 



2 



1 



2 



0 











8 



3 



4 



2 



6 



1 







2 



4 



1 



2 



7 



8 



8 



6 



5 



3 



S 





2 



7 



13 



(5 



8 



G 



5 



5 



3 



3 



9 



16 



23 



14 



ii 



4 



7 



2 



5 



1 



8 



11 



1« 



11 



3 



G 



3 



2 



1 



3 



8 



4 



13 



8 



4 



5 



2 



3 



4 



3 



6 



5 



4 



5 



5 



5 



1 



3 







5 



3 



4 



5 



3 



3 











4 



3 



4 



2 



2 



4 







Tiie counts show that the normal number ol stars in this region, independent 

 of the clustering, amounts to approximately, four. Subtracting them we get a cluster 

 of the following constitution 





3 



4 





3 



9 



2 



4 



5 12 



19 



10 



2 



4 7 



12 



7 







9 



4 





The total number of stars is 116. The distribution is somewhat dissymme- 

 trical and 1 compute the dispersion in the direction of the circle of declination as 

 well as perpendicular to it. I get og^ô'.o, a,^ = 5'.6 and in the mean a = 5'. 3. 



The diameter may generally be estimated at 5 or 6 times the dispersion, 

 hence for this cluster at 25' à 30', whereas Melotte from the F. A. P. gives the 

 value 20', a notably smaller value. 



