i9i-'.l OF THE GLOBULAR CLUSTERS. 147 



only possible average effect of their action on the visitor will be to 

 exert a drag on its motion. Some of the quiescent stars may be 

 slightly disturbed by the passing body, but as the effect of one appulse 

 is likely to be comparatively small, the stars in the shell will readjust 

 the relations among themselves easily, while the visitor will suffer a 

 considerable retardation of its oscillation. And after many appulses 

 the visitor will have its motion restricted to the shell like the motion 

 of the multitude of stars composing it. 



. This explains in a simple manner the capture process by which 

 clusters are built up, and given such accumulation of density towards 

 their centers. For the clusters are made up of a series of shells, and 

 if the effect of one shell is of this type, the effect of all the shells 

 will be an integration of these damping effects. It is no wonder 

 therefore that all the clusters show such pronounced accumulation 

 of density towards their centers. It is the inevitable outcome of this 

 capturing of foreign bodies in the course of immeasurable time. 



In section III. we have admitted the possibility that defects in 

 our photographs will account for the central density in clusters ex- 

 ceeding that of the atoms in a globe of monatomic gas in convective 

 equilibrium; but in view of this capture process, it seems much more 

 likely that the stars are accumulating in these centers beyond the 

 normal density for a mass of monatomic gas. Thus have the clus- 

 ters been built up to such extraordinary accumulation that they justly 

 excited the wonder of Sir William Herschel. 



XI. The Globular Clusters can be Explained Only by the 



Capture Theory. 



The figures of the clusters, nebulse and other sidereal systems 

 impressed Herschel with the view that there is a clustering power in 

 nature, everywhere gathering the stars into globular swarms, and 

 moulding the nebulosity into figures of greater and greater sym- 

 metry {PJiil. Trans., 1789, pp. 217-219). This is the earliest outline 

 of the modern capture theory as applied to clusters and nebulse of 

 symmetrical figure. It is evident that this process gives a good ex- 

 planation of the origin of the clusters, and that they can be explained 

 in no other way. 



