the gravel removed. We shall see hereafter how these boulders, derived from the land-drift and scattered along the coast, may be distinguished from those cast ashore by icebergs. Notwithstanding the number of facts thus far collected respecting glacial phenomena in America, certainly forming in their combination a very strong chain of evidence, the scientific world has, nevertheless, been slow to admit the possibility of the former existence of glaciers over such a wide, unbroken expanse of level land. This backwardness is, no doubt, partly due to the fact, that, as glaciers have hitherto been studied in mountainous countries, their presence has been supposed to imply the presence of mountains, this impression being strengthened by the downward and onward movement of existing glaciers, so long supposed to be exclusively due to the slopes along which all modern glaciers advance. Were it true that glaciers move solely or mainly on account of the sloping bottom on which they rest, and that they can advance only on an inclined plane, all the phenomena concerning drift, polished and furrowed surfaces, boulders, etc., in America, would hardly justify us in assuming a moving sheet of ice as their cause. But we have seen that the phenomena of glaciers, like those of currents, are in great part meteorological. The Gulf-Stream does not flow toward the English shore because the ocean-bottom slopes eastward; nor does the cold current of Baffin's Bay run down-hill when it pours its icy waters southward upon our northeast coast. Their course is determined by laws of temperature, and so have we also seen that the motion of glaciers is mainly determined by conditions of temperature, although, in this case, an internal mechanical action is combined with external influences; and while it is true that glaciers, as they now exist, are dependent upon the shape of the valleys in lofty mountain-chains, yet under different geographical conditions the same phenomena may be produced over level, open countries.