20 THE LIMITATIONS OF SCIENCE 



waves are not mechanical but electro-magnetic, similar 

 to those we now use in wireless telegraphy. But the 

 difficulties are still pursuing us. We may assume that 

 such waves can pass through space, but we cannot con- 

 struct a mechanical model of an atom or ether which 

 will produce or maintain these vibrations, nor have we 

 any evidence that electro-magnetic disturbances can 

 affect the optic nerve and produce the sensation of 

 light. The prediction, which I made when this essay 

 was first published, that we shall return to a modified 

 form of the corpuscular theory with the electrified par- 

 ticle, the constituent of the atom, as an agent, has been 

 verified. The principle of relativity, recently an- 

 nounced by Professor Einstein, absolutely requires it. 



If a general atomistic theory, which seems to be the 

 only practicable hypothesis, involves these inherent dif- 

 ficulties, and if it presents a real peril to correct scien- 

 tific thinking, the question arises, whether some general 

 mechanical explanation of all physical phenomena is 

 possible which is not so limited. 



Rankine, in the same essay, proposes a method which 

 he calls the science of energetics. As we have been 

 able to frame with some success a theory of physics by 

 using a hypothetical method, we should have even more 

 success in combining all the branches of the science 

 into one general theory if the abstractive method were 

 extended and applied for the purpose. Instead of sup- 

 posing the various physical phenomena to be con- 



