concerning the Constitution of Matter. 199 



In speaking of the medium I have more particularly in 

 mind the hypothesis of a turbulently moving liquid, as pro- 

 pounded by Sir W. Thomson* and modified by Prof. Fitz- 

 gerald f . Without entering on the subject mathematically 

 (which I am unable to do), it may be noticed that, according 

 to this view, the sether is made up of interlacing vortex- 

 filaments, the interspaces between which are filled with 

 quiescent or irrotationally moving liquid. The network of 

 filaments is equivalent to a sponge-like compressible solid, 

 whose pores, however, are completely filled with incompressible 

 fluid. The medium, taken as a luhole, is equivalent to an 

 incompressible solid ; but as the effectively liquid (i. e. irrota- 

 tionally moving) portion of the medium cannot suffer any 

 strain which calls up an opposing stress, it appears that in 

 considering the strain-figure we should have to deal with the 

 sponge-like compressible solid, made up of vortex-filaments. 



9. Gravitation. — Disregarding this last speculation, let 

 us return to the more general question of strain, and for sim- 

 plicity take the case when the strain-figure is symmetrical 

 about a point 0. So far as this one strain- 

 figure is concerned, we may, roughly speaking, -p. ~ 

 divide the medium into two portions, in one 

 of which the strain tends to increase, while 

 in the other it tends to decrease. If these 

 portions are separated by a single surface, then, 

 in the present case of symmetry, this surface 

 (A) will be spherical. On one side of A the 

 medium will be so strained that a small additional 

 strain of the same type would correspond to increased potential 

 energy, while on the other side of A a small increase of strain 

 would correspond to decreased potential energy. If the strain 

 outside A corresponds to increased potential energy, two distant 

 strain-figures will repel one another ; but if the strain outside 

 A corresponds to decreased potential energy, two distant 

 strain-figures will attract one another. If gravitation is to be 

 explained in accordance with the theory of this paper, the 

 above would seem to indicate the nature of the explanation. 

 There must be some type of strain which would of itself be 

 produced in the medium, were it not that in some other 

 portion of the medium there would be an accompanying strain 

 of a type corresponding to stability. Perhaps we may also 

 vaguely infer why the gravitative attraction exerted by a 

 body is proportional to its mass, for we have supposed both 



* B. A. Keport, 1887, p. 486; Phil. Mag., October 1867, p. 342. 

 t t Nature/ May 9, 1889. 



