the Foundations of Dynamics. 23 



to the gas molecules — and, further, that the molecules which 

 are in motion are not the same as those which are under strain. 

 The energy, that is the portion of energy concerned with the 

 particular amount of activity contemplated, is continuously 

 leaving the molecules recovering from strain and transferring 

 itself to those in accelerated motion, and as continuously 

 leaving those in retarded motion and transferring itself to 

 those acquiring strain*. 



Further to avoid misconception it is necessary to admit 

 distinctly that in speaking of the pressure-energy of a mass 

 of gas we are subjecting ourselves to an unnecessary though 

 convenient limitation. It is chosen as an example just because 

 so much is known about it ; it would have been easy to choose 

 elastic solid examples, where the elastic stress is not so readily 

 analysable, and where the idea of potential energy and the 

 statistical grouping of a mass of molecules was not optional 

 but necessary, in our present state of ignorance. I know well 

 enough that the pressure of gas in a reservoir is really due to 

 the motion of the particles, and I am willing freely to con- 

 template the case of Davy's " repulsive motion," no real 

 contact or elastic impact between particles ever occurring, 

 but only a rapid swing or asymptotic orbit round their 

 common centre of gravity ; but this rapid centripetal swing 

 of particles getting within each other's molecular range is at 

 present essentially an unknown process, and has to be rele- 

 gated to the sphere of potential energy quite as much as in 

 the not perhaps really dissimilar cases of elastic impact and 

 gravitation. 



Whenever an atom after collision retains its energy intact, 

 there is no need to say that there has been either transference 

 or transformation ; mere retention or storage is sufficient for 

 practical purposes (though if we attend to the details of a 

 collision we find there really is always a double transference 

 and a double transformation t) ; and if this is typical of what 



* See, however, the footnote on preceding page. 



t E.g., suppose a quick molecule A strikes or otherwise collides with a 

 slow but similar molecule B. The mechanism of the collision, being un- 

 known, is unimportant, and is most conveniently thought of in terms of 

 elasticity. A and B undergo distortion as they approach, and a certain por- 

 tion of A's energy is communicated to whatever medium it is which keeps 

 them asunder. This medium by its recoil then drives them apart, sending 

 off Bwith the energy which A originally had, and transferring to A the small 

 portion received from B. [This matter is discussed at greater length in 

 Appendix 2.] Any part of the energy which was not transformed is also 

 not transferred. Everything transferred from A to B necessarily under- 

 went rapidly a double transformation. When other occurrences are 

 analysed it will be found that the above is typical of what constantly 

 happens. Yet it is not always necessary to attend to all details, and in 

 the text above I speak as if A and B passed through each other, each 

 retaining its original energy. 



