42 Larmok, Physical Aspect of tJie Atomic TJieory. 



compacted that the thermal collisions do not induce 

 sensible internal commotion. 



But it is time to conclude this discursive survey. We 

 have recognised that the Daltonian molecular theory is 

 still the indispensable guide, if we wish to continue con- 

 structive efforts in the physical elucidation of nature, and 

 are not content to take down our scaffoldings for the 

 sake of logical symmetry, and, in the future, make the 

 most of the edifice as it now stands. While we are 

 certain with Dalton that molecules are very definite, 

 identical, structures, it has been seen that, when we 

 inquire into the detail of their constitution, though many 

 guiding principles mainly of electric and spectroscopic 

 types have been made secure, yet we have not much more 

 than their distant analogy with familiar dynamical systems 

 to aid us. But for many branches of the science know- 

 ledge of detailed molecular structure is not required. The 

 pioneering example of this kind was the kinetic theory 

 of gases. The domain of electrodynamics is now 

 securely founded on the displacements and movements of 

 electrons, each of which may be considered merely as a 

 point at which the unitary electric charge is concentrated, 

 so small are the unknown nuclei of the electrons com- 

 pared with their distances apart. In the same way the 

 wide domain including the course and equilibrium of 

 reactions in dilute systems can be studied by pure 

 numerical statistics in the manner of Guldberg and 

 Waage, or by the more generalised but fundamentally 

 equivalent thermodynamic methods associated mainly 

 with Willard Gibbs. But the aim of structural chemistry 

 must go much deeper ; and we have found it difficult, 

 on the physical evidence, to gainsay the conclusion that 

 the molecular architecture represented by stereochemical 



