Hence we have but two ways of accounting for gravitation:—either it is due to differences of pressure in a substance continuously filling all space, except where matter displaces it (?), or it is due to impacts, in some respects analogous to those of the particles of a gas which have been found to be capable of accounting for gaseous pressure. Now, all attempts as yet made to connect it with the luminiferous ether, or the medium required to explain electric and magnetic distance-action, have completely failed; so that we are apparently driven to the impact theory as the only tenable one. 141. To this theory Le Sage of Geneva devoted a singularly acute mind during the whole of his exceptionally long life; but, for all that, his posthumous tract on the subject is but little in advance of the results he had arrived at in his eighteenth year. He assumes the existence of ultra-mundane corpuscles; in infinite numbers, even compared with those of the particles of matter; of dimensions excessively small, but flying about in all directions with velocities enormously great. Portions of gross matter virtually screen one another to a certain extent from the pressure due to this perpetual rain of corpuscles; but only on the sides turned towards one another. Hence a lone body would be equally battered on all sides; but the introduction of a second mass interferes with this arrangement, and diminishes the pressure on the side next it. It is easy to show that the amount of this diminution, for _given small masses_, is inversely as the square of their relative distance. But when larger masses are taken account of, this diminution of pressure will not be (as gravity is) directly as the quantities of matter present, unless the further assumption is made that matter, whether by the great distance between its particles, or by the cage-like form of these particles, is almost perfectly permeable to the corpuscles; so that, practically, the corpuscles rain upon each of the interior particles of a mass as freely as if it had been alone in space. Some of the postulates of this theory are hard to grant, and there is additional difficulty as to the mode in which the supply of energy of the corpuscles is to be kept up. To enter into details on this subject is not in accordance with our plan. We therefore refer the reader to Sir W. Thomson’s account of Le Sage’s theory (Proc. R.S.E., 1871), and his suggestions for its improvement, based upon his theory of vortex-atoms.[47]