218-220] The Evolution of Star -Clusters 221 



219. Some features of similarity between spiral nebulae and star-clusters 

 have already been noted ; these fall in naturally with the hypothesis we are 

 now considering. We have already mentioned (5) that the velocities in space 

 of the star-clusters are approximately of the same order of magnitude as those 

 of the spiral nebulae. The masses again are of the same ordef~ofmagnitude. 

 At a reasonable estimate the mass of the galactic universe is that of 

 1500 million stars each of mass equal to 1*7 times that of our sun (2 x 10 33 gms.), 

 giving a total mass of 5 x 10 42 gms., which is of the same order of magnitude 

 as our estimate of the mass of the Andromeda nebula M. 31 (p. 217). This 

 nebula is perhaps the biggest of known nebulae, just as our galactic universe 

 is the biggest of known star-clusters. Shapley inclines to an estimate of 

 about 100,000* or possibly moref for the mean number of stars in more typical 

 clusters. This may correspond to a mass of the order of 10 38 which is com- 

 parable with the mass of the nebula M. 101 conjecturally determined on 

 p. 217. Finally according to Pease and Shapley ( 6) many of the so-called 

 globular clusters are in reality of a flattened shape, suggesting that the plane 

 of the spiral nebula persists as the plane of symmetry of the resulting star- 

 cluster, this being of course the galactic plane in our own universe. 



On the other hand before the hypothesis can be finally accepted some 

 obvious features of dissimilarity between the spiral nebulae and star-clusters 

 will demand explanation. The known star- clusters are very few in number 

 compared with the spiral nebulae, and their observed distribution in space is 

 different ( 5, 6). Also the star-clusters are on the whole probably more 

 distant than the spiral nebulae. Curtis J gives 0*033" as the average annual 

 proper motion of 66 large spiral nebulae, whence, the order of magnitude 

 of their linear velocities being known, Curtis suggests an average parallax 

 of the order of O'OOOS". On the other hand Shapley ( 6) has estimated the 

 nearest star-clusters to have a parallax of only about 0'00012; a later study 

 of the distances of 69 globular clusters has led him to the conclusion that the 

 nearest clusters of all, &> Centauri and 47 Toucanae, are distant just under 

 7000 parsecs (-or = 0'00014), the furthest, N.G.C. 7006, is distant 67,000 parsecs 

 (-or = 0-000015), while the mean distance of 69 is 23,000 parsecs O = 0*000044). 

 These facts suggest problems which have to be solved rather than fatal diffi- 

 culties. For the present we may confine ourselves to discussing the general 

 theoretical problem of the possibility of nebulae evolving into star-clusters. 

 As we have found it impossible to progress in a forward direction from nebula 

 to star-cluster, we must attempt to pass backwards from star-cluster to nebula. 



220. Of all the star-clusters known to us our own universe is naturally 

 the best known. Let us try to reconstruct the nebula out of which, on our 

 present hypothesis, it must have formed. 



* The Observatory, 39 (1916), p. 453. f Astrophys. Journ. 48 (1918), p. 181. 



J Astron. Soc. Pacific, 173 (1918). Astrophys. Journ. 48 (1918), p. 154. 



