Corpuscules ofLe Sage. 329 



their mutual collisions, and by collisions with mundane atoms, the 

 whole stock of gra vine energy is being gradually reduced, andthere- 

 fore the intensity of gravity gradually diminishing from age to age. 



Or, 2nd, suppose mundane matter to be spread through all 

 space, but to be much denser within each of an infinitely great 

 number of finite volumes (such as the volume of the earth) than 

 elsewhere. On this supposition, even were there no collisions 

 between the corpuscules themselves, there would be a gradual di- 

 minution in their gravific energy through the repeated collisions 

 with mundane matter which each one must in the course of time 

 suffer. The secular diminution of gravity would be more rapid 

 according to this supposition than according to the former, but 

 still might be made as slow as we please by pushing far enough 

 the fundamental assumptions of very small diameters for the 

 cage-bars of the mundane atoms, very great density for their 

 substance, and very small volume and mass, and very great 

 velocity for the ultramundane corpuscules. 



The object of the present note is to remark that (even although 

 we were to admit a gradual fading away of gravity, if slowenough), 

 we are forbidden by the modern physical theory of the conser- 

 vation of energy to assume inelasticity, or any thing short of 

 perfect elasticity, in the ultimate molecules, whether of ultra- 

 mundane or of mundane matter — and, at the same time, to point 

 out that the assumption of diminished exit-velocity of ultramun- 

 dane corpuscules, essential to Le Sage's theory, may be explained 

 for perfectly elastic atoms, consistently both with modern ther- 

 modynamics, and with perennial gravity. 



If the gravific corpuscules leave the earth or Jupiter with less 

 energy than they had before collision, their effect must be to 

 continually elevate the temperature throughout the whole mass. 

 The energy which must be attributed to the gravific corpuscules is 

 so enormously great, that this elevation of temperature would be 

 sufficient to melt and evaporate any solid, great or small, in a 

 fraction of a second of time. Hence, though outward-bound 

 corpuscules must travel with less velocity, they must carry away 

 the same energy with them as they brought. Suppose, now, the 

 whole energy of the corpuscules approaching a planet to consist 

 of translatory motion : a portion of the energy of each corpuscule 

 which has suffered collision must be supposed to be converted 

 by the collision into vibrations, or vibrations and rotations. To 

 simplify ideas, suppose for a moment the particles to be perfectly 

 smooth elastic globules. Then collision could not generate any 

 rotatory motion ; but if the cage-atoms constituting mundane 

 matter be each of them, as we must suppose it to be, of enor- 

 mously great mass in comparison with one of the ultramundane 

 globules, and if the substance of the latter, though perfectly 



