tracting upon the nucleus — a process which would leave no cavernous spaces between the crust and the nucleus for the waters to retreat to. From time to time the contracting crust would give way, exactly as a non-contracting crust would give way under the pressure of an expanding nucleus. The scene of such a catastrophe would be marked thereafter by a great crater at the place where the crust first gave way, and a series of radiating streaks marking the places where the crust was split open all around that spot. The signs of events such as these in the moon's earlier history are very manifest. There is the great lunar crater Tycho, which is clearly visible to the naked eye, near the lower part of the disc of the moon; and from this as a centre radiations extend in all directions, some of which run right across the visible lunar hemisphere, and probably extend right round the moon. These also can be seen with the naked eye; and they are so well marked in photographs of the moon that some supposed the earlier photographs by Draper and Rutherford in America, and by De la Rue in this country, were in reality only photographs of a peeled orange, the crater Tycho representing one end of the core, and the radiations corresponding to divisions between the sections of the orange. Besides this most remarkable case, there are six others, centres of radiating streaks on the moon's visible hemisphere, and doubtless others upon the unseen hemisphere. We have here clear evidence of the tremendous nature of the forces which were at work throughout the moon's frame in the earlier stages of her history, the disturbance in particular by which the radiations from Tycho were made having apparently wracked the whole frame of the moon. Directly, indeed, these considerations do not affect the theory we are considering, because no large portion of the lunar seas can by any possibility have retreated beneath the surface during this stage of her existence. But as showing the enormous store of heat which existed at that time (by far the larger part of which must have remained unexhausted when the next stage began) the consideration of these amazing evidences of disturbance has an important though indirect bearing on our subject.
After the crust had parted with the greater portion of the heat which it had possessed when first formed, it would cool and therefore would contract but slowly. The nucleus, on the other hand, which had before contracted more slowly than the crust, would now contract more rapidly, leaving spaces between itself and the crust. And then two things would happen. One would be the manifestation of vulcanian energy in consequence of the heat generated by the crust as it crushed its way downwards upon the retreating nucleus. The other would be the influx of water wherever it found access to the cavernous spaces between the crust and the nucleus. It is probable that before this vulcanian era of the moon's history was completed a considerable portion of the lunar waters had taken its place permanently beneath the crust. It should be noticed that this era corresponds with a part of the earth's existence which is as yet far from being completed, even if it can be regarded as much more than begun. It is far from unlikely that the era during which a planet's crust is thus kept in constant activity by the retreating motion of the nucleus synchronizes with the period during which life exists on the planet's surface. During all this period, which may have lasted tens of millions of years, not only were portions of the waters of the moon gradually taking up their place in cavernous spaces between the crust and the retreating nucleus, but another process must have been at work to exhaust the lunar seas. When water falls upon a large land-surface in the form of rain, so that the surface is thoroughly drenched, a portion probably disappears permanently from the water-circulation of the globe. Of course, the greater portion is conveyed into the sea in the form of running water. Then, again, the drying of the surface means that the water which had moistened it is taken into the air again in the form of aqueous vapor. And this eventually assumes the form of visible cloud, and after sundry changes (during which it may many times in turn appear as cloud or disappear as vapor) it falls again in rain, and may be either restored in this way directly to the sea from which it came, or so fall on land-surface as to run into some stream communicating by brook, rivulet, river, and estuary with the ocean. And some portion of the water which falls on land-surfaces, passing below the surface, feeds internal streams, and eventually appears again in the form of spring-water. But it cannot be doubted that a portion of the water which falls on dry land soaks its way downwards, very slowly, perhaps, but steadily and continuously, thus removing itself from sight, and pro tanto diminishing the planet's surface-waters.
How much of the water would be re-
