214 SCIENCE PROGRESS. 



Then -j- = - Kf, where K is a constant depending on the 



constitution of the body. And the actual constant value of 



/will be given by the equation f= y ^- And /"may be 



considerable if K be small. But where are we to look for 

 the supposed disturbing cause ? Clearly it cannot be looked 

 for in any mutual actions of the gas molecules ; for that is 

 always effacing the disturbance already existing. It can 

 be looked for (if at all) only in the action between the gas 

 and external bodies. 



20. Now we have appealed to the spectroscope as 

 proving that our molecules have many degrees of freedom, 

 because they, or their constituent atoms, communicate 

 vibrations of many kinds to the luminiferous ether, and are 

 capable of receiving from the ether vibrations of the same 

 kinds. The vibration energy or internal energy of our 

 molecules is thus capable of transmission by what we call 

 radiation from these molecules to external bodies. It is 

 our duty also to believe that some of it is scattered into 

 infinite space and so lost to science, as much as if it ceased 

 to exist, quia de non apparentibus et 11011 existentibus eadem 

 est ratio. On the other hand energy of translation is, so far 

 as we know, incapable, while in that form, of being communi- 

 cated to the ether. 



21. If this be true, it follows that some action is always 

 taking place between the molecules of our gas and external 

 bodies. It may be said that this action will not be sensible 

 except at high temperatures, i.e., when the vibrations are 

 very intense. But according to our theory there is even 

 at low temperatures a minority of molecules in the same 

 state of intense vibration which the majority have at 

 high temperatures. And it may be questioned whether we. 

 can, as now the fashion is, treat minorities as non-existent. 

 It is improbable indeed that this interaction between gas 

 molecules and external bodies should be sufficiently im- 

 portant to affect the determination of y. But one conse- 

 quence must, I think, be admitted, viz., that the law 

 £ -Mx + T) can never express with complete accuracy the state 

 of our system. For in our proof of that law we assumed in 



