150 SEE— DYNAMICAL THEORY [April 19, 



systems made up of a small number of bodies, but the visual double 

 stars, as the brighter and more easily recognized systems, appear to 

 have components more nearly equal than the much greater number 

 of systems,* which remain invisible at the distance of the fixed stars. 



The association of thousands of equal stars in a cluster must 

 therefore depend on something besides a chance distribution, or par- 

 tition of the primordial nebulosity. For although the clusters are 

 very far away, and the double stars in a cluster would thus appear 

 single from the perspective effect of distance alone, yet the distance 

 would not prevent fainter single stars from appearing on the back- 

 ground of the cluster if they were present. Perrine points out in 

 Lick Observatory Bulletin No. 155 that in cluster there is rarely 

 a difference of more than two magnitudes among the stars com- 

 posing it. This difference probably depends on difference in the 

 spectral types, rather than on difference in mass. The conclusion 

 that the great equality in luster depends on the essential equality in 

 the redistribution of dust within the system therefore seems unavoid- 

 able, as a necessary result of known laws of nature actually proved 

 to be in operation. If therefore this argument regarding the origin 

 of clusters, based on the equality of the stars, is admissible, the 

 explanation may as confidently be depended on as the law of gravi- 

 tation itself. For the testimony of the sidereal universe to its truth 

 seems to be absolutely overwhelming. There are in all over one 

 hundred globular clusters, and they include millions of stars; so that 

 the observed order of nature obviously rests on a fundamental cause. 



Accordingly, if we admit the truth of this theory of clusters, 

 which now seems to be well established, through the evidence pre- 

 sented by hundreds of globular clusters, and by the analogous evi- 

 dence offered by thousands of nebulae, we have at the same time an 

 equally satisfactory proof of the universality of the operation of 

 repulsive forces in nature. With his usual penetration Herschel saw 

 in the accumulation of density and brightness towards the centers of 

 these masses an incontestible proof of the existence of a clustering 

 power operating throughout the sidereal universe. 



Now by exactly reversing his argument we have an equally valid 

 proof of the operation of repulsive forces, to give the original distri- 



* Resembling planetary systems. 



