RING-PLAINS.—­These are by far the most numerous of the ramparted enclosures of the moon, and though it is occasionally difficult to decide in which class, walled-plain or ring-plain, some objects should be placed, yet, as a rule, the difference between the structural character of the two is abundantly obvious.  The ring-plains vary in diameter from sixty to less than ten miles, and are far more regular in outline than the walled-plains.  Their ramparts, often very massive, are more continuous, and fall with a steep declivity to a floor almost always greatly depressed below the outside region.  The inner slopes generally display subordinate heights, called terraces, arranged more or less concentrically, and often extending in successive stages nearly down to the interior foot of the wall.  With the intervening valleys, these features are very striking objects when viewed under good conditions with high powers.  In some cases they may possibly represent the effects of the slipping of the upper portions of the wall, from a want of cohesiveness in the material of which it is composed; but this hardly explains why the highest terrace often stands nearly as high as the rampart.  Nasmyth, in his eruption hypothesis, suggests that in such a case there may have been two eruptions from the same vent; one powerful, which formed the exterior circle, and a second, rather less powerful, which has formed the interior circle.  Ultimately, however, coming to the conclusion that terraces, as a rule, are not due to any such freaks of the eruption, he ascribes them to landslips.  In any case, we can hardly imagine that material standing at such a high angle of inclination as that forming the summit ridge of many of the ring-plains would not frequently slide down in great masses, and thus form irregular plateaus on the lower and flatter portions of the slope; but this fails to explain the symmetrical arrangement of the concentric terraces and intermediate valleys.  The inner declivity of the north-eastern wall of Plato exhibits what to all appearance is an undoubted landslip, as does also that of Hercules on the northern side, and numerous other cases might be adduced; but in all of them the appearance is very different from that of the true terrace.

The glacis, or outer slope of a ring-plain, is invariably of a much gentler inclination than that which characterises the inner declivity:  while the latter very frequently descends at an angle varying from 60 deg. to 50 deg. at the crest of the wall, to from 10 deg. to 2 deg. at the bottom, where it meets the floor; the former extends for a great distance at a very flat gradient before it sinks to the general level of the surrounding country.  It differs likewise from the inner descent, in the fact that, though often traversed by valleys, intersected by deep gullies and irregular depressions, and covered with humpy excrescences and craters, it is only rarely that any features comparable to the terraces, usually present on the inner escarpment, can be traced upon it.