In the first place, we have to consider the singular abrasion of the surfaces over which the glacier has moved, quite unlike that produced by the action of water. We have seen that such surfaces, wherever the glacier-marks have not been erased by some subsequent action, have several unfailing characteristics: they are highly polished, and they are also marked with scratches or fine _striae_, with grooves and deeper furrows. Where best preserved, the smooth surfaces are shining; they have a lustre like stone or marble artificially polished by the combined friction and pressure of some harder material than itself until all its inequalities have been completely levelled and its surface has become glossy. Any marble mantel-piece may serve as an example of this kind of glacier-worn surface. The levelling and abrading action of water on rock has an entirely different character. Tides or currents driven powerfully and constantly against a rocky shore, and bringing with them hard materials, may produce blunt, smooth surfaces, such as the repeated blows of a hammer on stone would cause; but they never bring it to a high polish, because, the grinding materials not being held steadily down, in firm, permanent contact with the rocky surfaces against which they move, as is the case with the glacier, but, on the contrary, dashed to and fro, they strike and rebound, making a succession of blows, but never a continuous, uninterrupted pressure and friction. The same is true of all the marks made on rocky shores against which loose materials are driven by water-currents. They are separate, disconnected, fragmentary; whereas the lines drawn by the hard materials set in the glacier, whether light and fine or strong and deep, are continuous, often unbroken for long distances, and rectilinear. Indeed, we have seen[A] that we have beneath every glacier a complete apparatus adapted to all the results described above. In the softer fragments ground to the finest powder under the incumbent mass we have a polishing paste; in the hard materials set in that paste, whether pebbles, or angular rocky fragments of different sizes, or grains of sand, we have the various graving instruments by which the finer or coarser lines are drawn. Not only are these lines frequently uninterrupted for a distance of many yards, but they are also parallel, except when some change takes place in the thickness of the ice, which may slightly modify the trend of the mass, or where lines in a variety of directions are produced by the intermittent action of separate glaciers running successively at different angles over the same surfaces. The deeper grooves sometimes present a succession of short staccato touches, just as when one presses the finger vertically along some surface where the resistance is sufficient to interrupt the action without actually stopping it,--a kind of grating motion, showing how firmly the instrument which produced it must have been held in the moving mass.