4 On the Constitution of Matter, 



posed to be mere points or centres of force, having inertia, but 

 no magnitude ; then the potential energy due to the relative 

 position of the atoms of any solid body is infinite. These 

 difficulties disappear if we accept the fact which stares us in the 

 face, that the final action between two atoms, when the distance 

 is diminished, is repulsive ; and that for any two atoms of a solid 

 body, there is a relative position of stable equilibrium short of 

 actual contact or coincidence. 



Certain considerations in connection with the Dynamical 

 Theory of Gases, require us to suppose that the particles of a 

 gas are not atoms, but molecules, or collections of atoms. The 

 facts revealed by the spectroscope indicate also that the particles 

 of vapours, even in their most attenuated condition, have a com- 

 plicated constitution, and that they probably consist of a con- 

 siderable number of atoms. Presuming that the mutual action 

 of atoms is the cause of, and of the same kind as, that which 

 takes place between molecules ; that action may be rather rudely 

 represented by the annexed diagram. The abcissa represents 

 the distance between two atoms, and the ordinate the force, the 

 positive ordinate representing attraction. Supposing one atom 

 fixed at 0, A is a position of stable equilibrium for the other, 

 and corresponds to the solid state at the absolute zero of tem- 

 perature. B is a position of unstable equilibrium, and corres- 

 ponds to the liquid state. Distances greater than OB correspond 

 to the condition of gas or vapour. It is necessary to suppose 



that in all cases, AB is small compared with OA, and that t- 



is large at the point A, but generally small at the point B. I do 

 not know that we have any means of forming an opinion as to the 

 relative magnitude of OC, but such a point as C must exist if 

 we assume the transition from Newton's Law to that of 

 gaseous repulsion to be gi'adual. This necessitates the existence 

 of a maximum negative ordinate DQ. It may be remarked that 

 the distance OD would correspond to the saturation point at the 

 absolute zero of temperature ; for supposing a number of equidis- 

 tant atoms in a confined space, if the distances be intermediate in 

 magnitude between OB and OD, the equilibrium is unstable, but 

 for distances greater than OD, it is stable. For refractory solids 

 EP must be large, and DQ comparatively very small. Por gases, 

 AB, or EP, or both, are probably very small, and DQ must be 

 very large, compared with its magnitude for non volatile sub- 

 stances. 



If a small velocity be impressed upon the moveable atom in 

 the direction Ox, it wiU oscillate about A. If the initial velocity 

 be increased to a certain value, that is under the action of a 

 certain amount of heat, the atom wiU just reach B, and there 



