When a glacier diminishes in size it leaves its lateral moraines stranded on the flanks of the valleys. Successive shrinkings may thus occur, and _have_ occurred at intervals of centuries; and a succession of old lateral moraines, such as many glacier-valleys exhibit, is the consequence. The Mer de Glace, for example, has its old lateral moraines, which run parallel with its present ones. The glacier may also diminish _in length_ at distant intervals; the result being a succession of more or less concentric terminal moraines. In front of the Rhone-glacier we have six or seven such moraines, and the Mer de Glace also possesses a series of them. Let us now consider the effect produced by a block of stone upon the surface of a glacier. The ice around it receives the direct rays of the sun, and is acted on by the warm air; it is therefore constantly melting. The stone also receives the solar beams, is warmed, and transmits its heat, by conduction, to the ice beneath it. If the heat thus transmitted to the ice through the stone be less than an equal space of the surrounding ice receives, it is manifest that the ice around the stone will waste more quickly than that beneath it, and the consequence is, that, as the surface sinks, it leaves behind it a pillar of ice, on which the block is elevated. If the stone be wide and flat, it may rise to a considerable height, and in this position it constitutes what is called a glacier-_table_. (See Fig. 6.) [Sidenote: GLACIER TABLES ACCOUNTED FOR.] Almost all glaciers present examples of such tables; but no glacier with which I am acquainted exhibits them in greater number and perfection than the Unteraar glacier, near the Grimsel. Vast masses of granite are thus poised aloft on icy pedestals; but a limit is placed to their exaltation by the following circumstance. The sun plays obliquely upon the table all day; its southern extremity receives more heat than its northern, and the consequence is, that it _dips_ towards the south. Strictly speaking, the plane of the dip rotates a little during the day, being a little inclined towards the east in the morning, north and south a little after noon, and inclined towards the west in the evening; so that, theoretically speaking, the block is a sun-dial, showing by its position the hour of the day. This rotation is, however, too small to be sensible, and hence _the dip of the stones upon a glacier sufficiently exposed to the sunlight, enables us at any time to draw the meridian line along its surface_. The inclination finally becomes so great that the block slips off its pedestal, and begins to form another, while the one which it originally occupied speedily disappears, under the influence of sun and air. Fig. 20 represents a typical section of a glacier-table, the sun's rays being supposed to fall in the direction of the shading lines. [Sidenote: TYPE "TABLE."] [Illustration: Fig. 20. Typical section of a glacier Table.]