Thirdly, the ether has been conceived to be the ordinary elastic gases or atmospheres freely expanded into space. But these have no co-efficient of elasticity sufficient to give them such expansion, and they would be liable to condensation and compression by their own gravity about the planets, which would cause a rise of temperature and dissipation of energy which would rob the ether of its permanent character. Besides, Maxwell has shown that our atmosphere expanded into space would be far too rare in the interplanetary spaces to satisfy the required conditions; nor is there any molecular velocity at all adequate to the propagation of wave-energy with the velocities observed, as will be shown further on.
This brings us to the fourth concept, which is that of a pure primordial gaseous plenum, of sufficiently high tension, and in the condition assumed by gases in a rarefied receiver, where the mean path is so long in proportion to the mean distance that a symmetrical movement arranges itself, according to the law first pointed out by Maxwell as a corollary from the equilibrium of pressure observed in confined gases, and the performance of gases in a rarefied space first observed by Crookes, that particles in free collision in space tend constantly to rearrange their motions automatically so as to move uniformly in all directions in radial lines from every point. With the gases experimented with the mean path at normal pressure and density is very short (only about 260000 of an inch when the molecules have a mean distance from each other of 7000000 of an inch), but, at the extreme of rarefaction which we are able to effect (about 1000000 of an atmosphere, when the distance of the molecules is still only 70000 of an inch), the mean path rises to about four inches; and C. T. Preston has shown ("Nature," vol. xxiii, p. 463), that could we carry the exhaustion to the third power of that obtainable, so that the distance of the molecules apart should be so much as one seventh of an inch, the mean path would be raised to 60,000,000 miles, since it increases in the triplicate ratio with the distance. But with the ether no such rarity need be postulated, since mean free path is but a question of size of molecule, and in comparison with the hydrogen-molecule the size of the particle can only be infinitesimal. It is clear, however, that, with the enormous velocity due to the particle, all the effect of continuity would be produced, so far as vision is concerned, by a mean distance apart, not merely of one seventh of an inch, but of many miles.
We may therefore assign to the ether any required free path, and any necessary density, tension, and velocity, all of these latter being imperceptible to molecular structures which float in and are permeated by it. The motions also of molecules in a gas so constituted would be practically as unaffected as in free space, since it is demonstrable that the resistance to motion offered by a medium such as the hypothesis calls for would be in the ratio of the motion of the moving mass to that of the particle of the medium, which in the case supposed would