152 Physical Interpretation of MicJielson-Morley Experiment. 



massive bodies and grows small as the masses grow smaller 

 or the distances increase. 



There is at present no method by which we can test the 

 accuracy of the above formula. The Michelson and Morley 

 experiment will naturally not serve as a test, since the 

 correction-term has been actually chosen so as to cope with 

 their results. The delicacy of our optical instruments in 

 other experiments is insufficient to detect a term such as has 

 been suggested above. Thus, theoretically, there will be an 

 alteration in the accepted law of aberration for different 

 directions, and there will also be a slight deviation from 

 SnelFs law of sines. But the former of these changes will 

 involve less than one-millionth of a second of arc in our 

 telescopic observations ; while the latter would amount to 

 less than one-thousandth of a second of arc in our measure- 

 ments on the spectroscope. Both these quantities are at 

 present utterly beyond our powers of observation. 



The above formula is not opposed to the axiom of 

 Relativity that the velocity of light in free pether is a 

 constant. For we must mean by " free aether " any portion 

 of aether freed from the influence of ordinary matter. Now 

 by the above formula, when the moving matter has been 

 taken to a sufficient distance so that its influence may be 

 disregarded, the velocity of light becomes equal to a 

 constant V. And all theories of light are bound to take 

 note of the Fizeau effect, where it is definitely proved by 

 experiment that the motion of matter may influence the 

 velocity of light. A distinction is sometimes made between 

 the case of a ray passing through moving matter, and one 

 passing near moving matter. The distinction, however, if 

 pressed too rigorously, would appear to disregard the prin- 

 ciple of Huyghens and the atomic structure of matter. 

 When a ray of light passes from the atom A to the atom B 

 we are not concerned in Physical Optics, merely with what 

 happens in the cylinder through A and B ; strictly speaking 

 the whole of the aether must be taken into account, though, 

 of course, the predominant part in the propagation of the ray 

 is taken by the aether within the cylinder. The flowing- 

 water in the Fizeau experiment consists of separate moving 

 molecules, each of which may have its effect, and com- 

 paratively few of which lie directly between A and B. 



The really important distinction between the influence of 

 motion in the two cases is that in the Fizeau experiment 

 we can give, owing to the work of Lorentz, a reasonable 

 description of the machinery whereby the influence is 



