The bulk of the moon is about one-fiftieth part that of the earth, and an easy calculation shows us that, taking the area of the Pacific, Atlantic, and Indian Oceans combined as about two-thirds that of the globe, it would require a thickness (or depth) of about forty miles to furnish the material for the moon. We must, of course, assume that there were some inequalities in the thickness of the crust and in its comparative rigidity, so that when the critical moment came and the earth could no longer retain its equatorial protuberance against the centrifugal force due to rotation combined with the tidal undulations caused by the sun, instead of a continuous ring slowly detaching itself, the crust gave way in two or more great masses where it was weakest, and as the tidal wave passed under it and a quantity of the liquid substratum rose with it, the whole would break up and collect into a sub-globular mass a short distance from the earth, and continue revolving with it for some time at about the same rate as the surface had rotated. But as tidal action is always equal on opposite sides of a globe, there would be a similar disruption there, forming, it may be supposed, the Atlantic basin, which, as may be seen on a small globe, is almost exactly opposite a part of the Central Pacific. So soon as these two great masses had separated from the earth, the latter would gradually settle down into a state of equilibrium, and the molten matter of the interior, which would now fill the great oceanic basins up to a level of a few mile below the general surface would soon cool enough to form a thin crust. The larger portion of the nascent moon would gradually attract to itself the one or more smaller portions and form our satellite; and from that time tidal friction by both moon and sun would begin to operate and would gradually lengthen our day and, more rapidly, our month in the way explained in Sir Robert Ball's volume.