PHILOSOPHY OF PHYSICS — MILNE 22J 



existence of combininir ratios and so on — but ^vas, nevertheless, a 

 •irand tool of investigation before individual atoms were observed. 

 The dynamical theory of gases led to relations between observable 

 quantities, like Maxwell's prediction of the independence of gaseous 

 viscosity of pressure, subsequently verified. In each case the start- 

 ing point was a superb effort of the imagination. Rutherford's dis- 

 covery of the atomic nucleus led to Bohr's magnificent conception of 

 atomic structure as involving discrete electronic orbits about a center, 

 and Bohr's theory opened up the possibility of unraveling the spectra 

 of the elements and predicting their chemical properties. 



Electronic orbits were a hypothesis. Was this a necessary hypoth- 

 esis? In 1925 Heisenberg answered this question by passing to the 

 other method of theoretical physics, the method of introducing noth- 

 ing but '* observables ", to use Dirac's word. Abandoning the men- 

 tion of orbital sizes and frequencies, Heisenberg showed that the 

 transition i-elations between the observed energy levels in atoms 

 could be obtained by starting from these levels alone. 



This great revolution in thought led to the foundation of the 

 quantum mechanics. But it was not a revolution in method. In 

 1905 Einstein had applied the same method to the notion of simul- 

 taneity of events. He showed that when we examine simply the 

 observations which it is possible to make by which we habitually 

 assign epochs to events, then two events which appear to be simul- 

 taneous to a given observer are not in general simultaneous to a 

 second observer in uniform motion with respect to the first observer. 

 Einstein swept out of tenability the concept of an absolute simul- 

 taneity, previously uncritically accepted as intuitive. Einstein's 

 method was to tie an observer down to stating his tests for what 

 he was disposed to call simultaneity; unless he came prepared to 

 state tests for simultaneity, a judgment as to the simultaneity or 

 otherwise of two distant events was valueless. It was this necessity 

 for concentrating on evidence, which Einstein forced physics to take 

 into account, that has had so profound an effect on the development 

 of physics. It is true that the statement made above, concerning an 

 observer in uniform motion with regard to a second observer, in- 

 volves a conventional assignment of epoch to an event and involves, 

 further, a definition of uniform velocity, which, in turn, implies a 

 definition of distance and a definition of uniform time. The defini- 

 tion of distance employed by Einstein involved the introduction of 

 the concept of the rigid body, which cannot be defined, so that Ein- 

 stein's treatment was not free from conceptual taint. We shall 

 later endeavor to avoid such departures from the ])urely observa- 

 tional method. Nevertheless, we may with Jeans si:)eak appropriately 

 of the " Einstein-Heisenberg policy " of concentrating only on 



