Density of the Mther. 499 



had made me sanguine of seeing if it existed (see page 776, 

 Phil. Trans, vol. 184). Could certainly have detected ^th of! 

 a band, probably ^ th ; but nothing was seen. 



23. The experiment was so good that it was decided to try 

 a higher voltage ; so, putting the town mains in series with a 

 storage-battery, 230 volts were applied to each coil, giving a 

 current of 2S amperes on the meter, or 7 amperes to each 

 coil. The bands were again watched while the current was 

 being put on and off and reversed, sometimes slowly and 

 sometimes quickly, but I never saw the slightest motion of any 

 kind. The current could not be left on very long, because of 

 possible heat disturbance penetrating from the windings of 

 wire : for the coils themselves naturally became hot, since 

 9 horsepower was being expended upon them, but the water- 

 jacketing succeeded in keeping the cores quite cool. 



24. Hence I conclude, with security, that if there is any 

 a?ther-flow at all along magnetic lines of force, it bears to the 

 velocity of light a ratio less than the ratio of ^ th of a wave- 

 length to 32 times the length f a bobbin. The number 32 

 being justified by the fact that there are four bobbins in 

 optical series, and that the light is sent four times through 

 each, in opposite directions simultaneously ; so that the whole 

 virtual length of path along which the slightest acceleration 

 or retardation would have caused a shift of the bands, is 

 32 X 44 — 1408 centimetres. The ratio of a fiftieth of a wave- 

 length to this distance is less than 10~ 9 , so I conclude that 

 any hypothetical aether drift, along a field of intensity 1400 

 CG.s., is less than 30 centimetres, or one foot, per second; in 

 other words, the aether flow is not as much as -^-th of a milli- 

 metre per second, or say 16 inches an hour, for each c.G.s. 

 unit of field intensity. 



25. The total drop of magnetic potential applied was 

 equivalent to 32 X 60,000 = nearly two million c.G.s., and this 

 certainly did not hurry the light by so much as 55th of a 

 wave-length. As a matter of fact ^th shift could certainly 

 have been seen, under the conditions ; but I estimate ^th, to 

 keep on the safe side. 



So the deduction from this experiment was, either that^the 

 theory about longitudinal magnetic drift is erroneous, or else 

 that the aether of space is at least eight times as dense as 

 platinum. 



2Q. That was the experimental conclusion thirteen years ago. 

 I have now brought forward theoretical arguments which 

 suggest that there is a sense in which what we call the 



2 M2 



