DISSOLUTION. 547 



able. For the diffused matter produced by such conflicts 

 must form a resisting medium, occupying that central re- 

 gion of the cluster through which its members from time 

 to time pass in describing their orbits — a resisting medium 

 which they cannot move through without having their ve- 

 locities diminished. Every additional collision, by augment- 

 ing this resisting medium, and making the losses of velocity 

 greater, must aid in preventing the establishment of that 

 equilibrium which would else arise; and so must conspire 

 to produce more frequent collisions. And the nebulous 

 matter thus formed, presently enveloping the whole cluster, 

 must, by continuing to shorten the gyrations of the moving 

 masses, entail an increasingly active integration and re- 

 active disintegration of them; until they are all dis- 

 sipated. Whether this process completes itself inde- 

 pendently in different parts of our Sidereal System; or 

 whether it completes itself only by aggregating the whole 

 matter of our Sidereal System; or whether, as seems not 

 unlikely, local integrations and disintegrations run their 

 courses while the general integration is going on; are ques- 

 tions that need not be discussed. In any case the conclu- 

 sion to be drawn is, that the integration must continue until 

 the conditions which bring about disintegration are reached ; 

 and that there must then ensue a diffusion that undoes the 

 preceding concentration. This, indeed, is the con- 

 clusion which presents itself as a deduction from the persist- 

 ence of force. If stars concentrating to a common centre of 

 gravity, eventually reach it, then the quantities of motion 

 they have acquired must suffice to carry them away again to 

 those remote regions whence they started. And since, by 

 the conditions of the case, they cannot return to these remote 

 regions in the shape of concrete masses, they must return 

 in the shape of diffused masses. Action and reaction being 

 equal and opposite, the momentum producing dispersion, 

 must be as great as the momentum acquired by aggregation; 

 and being spread over the same quantity of matter, must 



