224 SCIENCE PROGRESS. 



that he is working out a complete theory of a dense 

 medium of elastic spheres. 



40. We might perhaps carry the doctrine of " streams " 

 farther. Such a stream as above supposed is always dis- 

 solving, the spheres comprising it escaping by diffusion 

 into the surrounding medium, just as we are told the par- 

 ticles of water which form a visible cloud in the sky are 

 not the same from instant to instant. And the state of 

 equilibrium would be when streams form as fast as they are 

 dissolved. But the energy of relative motion of the spheres 

 forming a stream may be a little greater or less. The less it 

 is, the less is the tendency to diffuse, and the longer the 

 stream survives. So that in the case of equilibrium we 

 should expect to find that those parts of the system which 

 have stream motion in the highest degree would on the 

 average have less energy of relative motion, and therefore 

 greater density. 



41. The following system would not indeed be in com- 

 plete equilibrium, but might not require any great force to 

 maintain it, namely, a number of dense masses moving in a 

 comparatively rare medium, the spheres of the rare medium 

 having considerably greater mean kinetic energy than that 

 of relative motion of the spheres within one of the dense 

 masses. So that if one of the dense masses were sur- 

 rounded by an elastic envelope, the external pressure on 

 that envelope due to spheres of the rare medium with 

 higher mean velocities should be equal to the internal pres- 

 sure due to the dense medium within. If there were an 

 elastic envelope equilibrium would be complete. In the 

 absence of an envelope the spheres of the dense masses 

 would escape, but only by the comparatively slow process of 

 diffusion at constant pressure. 



42. Now I have assumed the molecules to be merely 

 elastic spheres. If, as Professor Tait supposes, they also 

 attract each other with finite force, that molecular attrac- 

 tion might overcome the tendency to diffuse, and the sys- 

 tem described in the last paragraph might be permanent. 



S. H. Burbury. 



