If we examine the matter fully we shall be led to conclude that the _transformation of a land-surface into a sea-bottom_ will probably completely obliterate every trace of glaciation which that land-surface may once have presented. We cannot, for example, expect to meet with polished and striated stones belonging to a former land glaciation; for such stones are not carried down bodily and unchanged by our rivers and deposited in the sea. They become broken up by subaërial agencies into gravel, sand, and clay, and in this condition are transported seawards. Nor even if we supposed it possible that the stones and boulders derived from a mass of till could be carried down to sea by river-action, could we at the same time fail to admit that such stones would be deprived of all their ice-markings, and become water-worn and rounded on the way.[137] Nor can we expect to find boulder clay among the stratified rocks, for boulder clay is not carried down as such and deposited in the sea, but under the influence of the denuding agents becomes broken up into soft mud, clay, sand, and gravel, as it is gradually peeled off the land and swept seawards. Patches of boulder clay may have been now and again forced into the sea by ice and eventually become covered up; but such cases are wholly exceptional, and their absence in any formation cannot fairly be adduced as a proof that that formation does not belong to a glacial period. The only evidence of the existence of land-ice during former periods which we can reasonably expect to meet with in the stratified rocks, consists of erratic blocks which may have been transported by icebergs and dropped into the sea. But unless the glaciers of such epochs reached the sea, we could not possibly possess even this evidence. Traces in the stratified rocks of the effects of land-ice during former epochs must, in the very nature of things, be rare indeed. The only sort of evidence which, as a general rule, we may expect to detect, is the presence of large erratic blocks imbedded in strata which from their constitution have evidently been formed in still water. But this is quite enough; for it proves the existence of ice at the time the strata were being deposited as conclusively as though we saw the ice floating with the blocks upon it. This sort of evidence, when found in low latitudes, ought to be received as conclusive of the existence of former glacial epochs; and, no doubt, would have been so received had it not been for the idea that, if these blocks had been transported by ice, there ought in addition to have been found striated stones, boulder clay, and other indications of the agency of land-ice.