The mode of aggregation of water-worn materials is very different. Examine the shingle along our beaches: we find it so distributed as to show that the fading tide-wave has carried the lighter materials farther than the heavier ones, and the successive deposits exhibit an imperfect cross-stratification resulting from changes in the height of the tide and the direction of the wind. Moreover, in any materials collected under water we find the heavier ones at the bottom, the lighter on the top. It is true that large angular boulders may occasionally be found resting upon beach-shingle, but their presence in such a connection is easily explained. They may have been dropped there by floating icebergs, or have fallen from crumbling drift-cliffs.
I should add, in speaking of drift-materials, that, while we find the large angular boulders resting above them, we occasionally find boulders of unusual size mingled with them; but, when this is the case, such massive fragments are more or less rounded, polished, and marked in the same way as the smaller pebbles, or as the surfaces over which the glacier has passed. This is important to remember, because, when we examine the drift in countries where the ice, during the glacier-period, overtopped nearly all the mountains, so that few fragments could fall from them upon its surface, we find scarcely any angular boulders, while the drift is interspersed with larger fragments of this character, carried under the ice, instead of on its back. Another distinction between water-worn deposits and drift consists in the fact that the former are washed clean, while the latter always retains the mud gathered during its journey and spread throughout its mass.
In summing up the glacial evidences, I must not omit the moraines, though I have described them so fully in a previous article that I need not do more than allude to them here; but any argument for the glacial theory which did not include these characteristic walls erected by glaciers would be most imperfect. We need hardly discuss the theory of currents with reference to the formation of terminal moraines, extending across the valleys from side to side. Any current powerful enough to bring the boulders and _débris_ of all sorts of which these walls are composed to the places where they are found would certainly not build them up with such regularity, but would sweep them away or scatter them along the bottom of the valley. That this is actually the case is seen in the lower course of the valley of the Rhone, where there are no transverse moraines, while they are frequent and undisturbed in the upper part of the valley. This is no doubt owing to the fact, that, when the main glacier had already retreated considerably up the valley, the lateral glaciers from the chains of the Combin and the Diablerets still reached the valley of the Rhone at a lower point, and barred the outlet of the waters from the glaciers above. A lake was thus formed, which, when the lower glaciers retreated up the lateral valleys, swept away all the lower transverse moraines, and formed the flat bottom of Martigny. In this case, the moraines were totally obliterated; but there are many other instances in which the materials have been only broken up and scattered over a wider surface by currents. In such remodelled moraines, the glacier-mud has, of course, been more or less washed away. We have here a blending of the action of water with that of the glacier; and, indeed, how could it be otherwise, when the colossal glaciers of past ages gradually disappeared or retreated to the mountain-heights? The wasting ice must have occasioned immense freshets, the action of which we shall trace hereafter, when examining the formation of our drift-ponds, of our river-beds and estuaries, as well as the river-terraces standing far above the present water-level.