Interpretation of the Michelson-Morley Experiment. 151 



double-refraction effect, as well as Trouton & Banking's 

 attempt by means of a change in electrical resistance, 

 having both failed. (See Lord Rayleigh, Phil. Mag. 6th 

 series, vol. iv. 1902 ; Rankine, Science Progress, July 

 1908.) 



We may add to the failure of these tests the difficulty of 

 believing that all solid substances, whatever their elasticity, 

 should experience the same contraction when travelling with 

 the same velocity. 



The failure to confirm the Fitz Gerald contraction has led 

 to the propounding of the theory of Relativity, in which, 

 with a good deal of metaphysical subtlety, all our notions of 

 time and space are revised. It seems to have been over- 

 looked that a third solution is possible, viz., that, even 

 though there be no dragging effect, such as Lodge searched 

 for, a change in the constants of the aether may perhaps take 

 place, owing to the motion of a large mass in the neigh- 

 bourhood. Thus at the surface of the earth, the aether may 

 be supposed to alter slightly in its properties, owing to the 

 orbital motion. We may imagine that the velocity of an 

 aethereal disturbance parallel to the motion of the earth 

 is unaltered, but that at right angles to the orbital motion 

 the velocity of a disturbance is 



i v - w ■ 



where V is the velocity of light in free aether, and u is the 

 velocity of the earth in its orbit. Any other simple function 

 of Y and u agreeing with the above as far as squares of small 

 quantities would, of course, answer the purpose equally well 

 in the present state of our experimental knowledge. 



The correction-term suggested above is applicable to 

 measurements at the surface of the earth ; at considerable 

 distances from the earth the effect of its motion ought clearly 

 to be taken as smaller. Thus it is absurd to suppose that our 

 optical determinations are sensibly affected by the orbital 

 motion of Jupiter or Saturn. We may make allowance for 

 the falling-off in the effect by supposing that the correction- 

 term is more nearly represented by an expression of the 

 form 



where G is a function of the mass and the distance, akin to 

 gravitation, in so far as it is large in the neighbourhood of 



