254 THE ORIGIN OF LIFE 



vibrations into higher and higher nodal sub-divisions. 

 And this, it has been calculated, should take place 

 very rapidly. But we know that if it takes place 

 at all it must take place to a very small extent, 

 the energy from radio-active bodies being the 

 nearest approach we have to any such continuous 

 loss of energy. The answer, which therefore seems 

 obvious, is that the assumption — that by repeated 

 collisions the frequencies are altered without an ac- 

 companying change in the structure of the molecule 

 — is unsound, and that the increase in the energy 

 of higher vibrations as the temperature is raised 

 must be sought for in the increased number of 

 molecules so changed as to radiate in this par- 

 ticular manner. Thus no amount of heating will 

 cause a gas to give out a continuous spectrum, 

 and new frequencies must be accompanied by new 

 atoms or molecular connections. 



In the case of solids and liquids all sorts of 

 molecular connections may be produced, from the 

 casual juxtaposition to the closest chemical union ; 

 and therefore a continuous spectrum may be ex- 

 pected. But the continuous transformation of 

 vibrations from high to higher, as in the case 

 of billiard balls, is not to be expected, and, in 

 fact, such a transformation would lead to wrong 

 expectations, by a false analogy, with the processes 

 which actually take place. 



It may therefore be said with safety that if a 

 vibration becomes intense in any part of the 

 spectrum without any change of a molecular or 



