It is only when a planet attains a mass at least a quarter that of the earth that it is capable of retaining water-vapour, one of the most essential of the gases; but with so small a mass as this, its whole atmosphere would probably be so limited in amount and so rare at the planet's surface that it would be quite unable to fulfil the various purposes for which an atmosphere is required in order to support life. For their adequate fulfilment the mass of a planet cannot be much less than that of the earth. Here we come to one of those nice adjustments of which so many have been already pointed out. Dr. Johnstone Stoney arrives at the conclusion that hydrogen escapes from the earth. It is continually produced in small quantities by submarine volcanoes, by fissures in volcanic regions, from decaying vegetation, and from some other sources; yet, though sometimes found in minute quantities, it forms no regular constituent of our atmosphere.[18] The quantity of hydrogen combined with oxygen to form the mass of water in our vast and deep oceans is enormous. Yet if it had been only one-tenth more than it actually is, the present land-surface would have been almost all submerged. How the adjustments occurred so that there was exactly enough hydrogen to fill the vast ocean basins with water to such a depth as to leave enough land-surface for the ample development of vegetable and animal life, and yet not so much as to be injurious to climate, it is difficult to imagine. Yet the adjustment stares us in the face. First, we have a satellite unique in size as compared with its primary, and apparently in lateness of origin; then we have a mode of origin for that satellite said to be certainly unique in the solar system; as a consequence of this origin, it is believed, we have enormously deep ocean basins symmetrically placed with regard to the equator--an arrangement which is very important for ocean circulation; then we must have had the right quantity of hydrogen, obtained in some unknown way, which formed water enough to fill these chasms, so as to leave an ample area of dry land, but which one-tenth more water would have ingulfed; and, lastly, we have oxygen enough left to form an atmosphere of sufficient density for all the requirements of life. It could not be that the surplus hydrogen escaped when the water had been produced, because it escapes very slowly, and it combines so easily with free oxygen by means of even a spark, as to make it certain that _all_ the available hydrogen was used up in the oceanic waters, and that the supply from the earth's interior has been since comparatively small in amount.