SOME VIEWS ON LORD KELVIN'S WORK 435 



takes place by finite steps. If we consider the radiation from a 

 system of atoms and electrons, such as is presented by the 

 kinetic theory of gases, enclosed in a perfectly reflecting en- 

 closure at a fixed temperature, the view expressed by Lord 

 Kelvin in the passage quoted earlier is that the emitted energy 

 may consist largely of pulses due to collisions. This view may 

 still be regarded as in harmony with modern requirements of 

 Planck's theory, as it would naturally involve the emission of 

 energy in discrete quanta from a molecule, when an electron was 

 expelled from it or detached by the influence of other molecules. 

 The sudden expulsion would in this way constitute a pulse, of 

 some definite form depending on the constitution of the atom, 

 which would carry with it a definite quantit}' of energy into the 

 enclosure. As is pointed out above, there is no need to suppose 

 that this energy takes the form of actual wave motion. The energy 

 per wave-length is the same in the unresolved form of the pulse, 

 as when resolved mathematically by Fourier's theorem or experi- 

 mentally by any form of resolving apparatus ; and would restore 

 the radiated energy to the system equally effectively in this form 

 in the process of absorption as in the form of regular wave-trains. 

 These suggestions are similar to those put forward by Prof. 

 Sir J. J. Thomson — namely, that the quanta of energy which 

 have been proved to exist, do not indicate a molecular structure 

 for radiant energy, but merely that emission occurs when some 

 system within the atom is ruptured and that the change involves 

 a definite quantity of energy. It is clear that if we adopt the 

 idea that the quanta of energy introduced by Planck are to be 

 identified with pulses all of some definite form, considering the 

 agreement of Planck's formula with experimental facts, we must 

 regard these pulses as constituting practically the whole of the 

 energy emitted by the radiating body. In this line of specula- 

 tion the steadiness of the emission of such a sequence of pulses 

 at any given temperature is to be accounted for by the existence 

 of some instability or weak connection in the atomic constitu- 

 tion, leading readily to expulsion of an electron or rearrange- 

 ment of the atomic system in some new equilibrium configuration; 

 and in the statistical steadiness of the conditions within the 

 enclosure this instability may belong to all or, though less likely, 

 only to one large group of the colliding molecules, all of this 

 group being in a similar state of motion or atomic constitution 

 which is subject to variation with increase of temperature. Some 



