Density of the JEiher. 495 



being 10~ 6 x lO" 3 x 10- ]2 =10" 21 , at the exhaustion of a 

 millionth of an atmosphere, even supposing the whole contents 

 of the tube were electrons, — but it is the optical density; and 

 that would appear to be something more like 10~ 6 (n 2 — l)/n 2 y 

 where n is the refractive index for common air. The optical 

 fraction being ^qVoj this amounts to \ . 10 _s ; and at a speed 

 of 10 9 cm. per sec. it would seem to give something almost 

 appreciable in the way of drift. But the torrent of electrons 

 must be only a small fraction of the whole residual gas ; and 

 some positive charges are sailing up in the other direction, in 

 numbers probably sufficient to compensate for their smaller 

 speed. So the outlook for experiment on this kind of influence 

 of electricitv on light is not encouraging. — whether the 

 question be examined from the point of view of magnetic flux, 

 or from that of electric current. 



The estimates of this paper are: — That the aether has a 

 magnetic inertia-coefficient which can be represented by 

 likening it to a mechanical medium with a density of 10 12 

 grammes per cubic centimetre; that the intrinsic aetherial 

 kinetic energy is 10 33 ergs per cubic centimetre ; and that 

 any streaming along magnetic lines of force is, in centimetres 

 per second, one four-millionth of the number of C.G.s. units 

 of intensity in the magnetic field. 



Part II. 



Experimental Attempt to observe the Hypothetical 

 Magnetic jEther Drift. 



16. During the years 1892-93, being urged thereto as- 

 certain mathematical theories, which are referred to and partly 

 quoted in an Appendix in inverted order of date, I made a 

 careful examination of the question of an aether flow along- 

 magnetic lines of force. This experiment has not been pub- 

 lished, and in view of the above estimate its details ought to be 

 recorded. As the result was negative I shall make the record as 

 brief as possible. In July 1892 I constructed four bobbins 

 or frames of brass, each 44 centimetres long, with a core of 

 elliptical section of great eccentricity, so that the major axis of 

 its aperture was about ten centimetres and the minor axis about 

 one centimetre. Each bobbin was provided with carefully 

 glazed ends of optically worked plate-glass, that could be 

 clamped on in a water-tight fashion, and with a means for 

 supplying any desired liquid into the interior of the core. The 

 surrounding brass-sheet walls of the core w T ere double, so that 

 between the core and the winding there was an annular space 

 for water- jacketing, in which a current of cold water could be 



