of the glaciers still remaining in the high mountain-ranges dividing the eastern part of the continent from California, still less respecting any indications of their former extension. There can be little doubt that such traces exist, and as soon as the so-called parks between Pike's Peak and Long Peak are explored, we may hope for information on this point. Indeed, the investigation may be spoken of as already undertaken; for among the exploring parties now on their way to that region are some intelligent observers who will not fail to make this point a subject of special study. But it is well known that the usual characteristic marks of glaciers extend over the whole surface of the land in the eastern half of the continent, from the Atlantic shores to the States west of the Mississippi, and from the Arctic Sea to the latitude of the Ohio, in its middle course, while within the range of the Alleghanies they stretch as far south as Georgia and Alabama. In no other region where these traces have been observed do they extend over such wide tracts of country in unbroken continuity, this being of course owing to the level character of the land itself.
The continent of North America, east of the Rocky Mountains, is, indeed, an immense uniform plain, intersected from east to west only by the ranges of low hills running in the direction of the St. Lawrence and the Canadian lakes, and from northeast to southwest by the Alleghany range stretching from Alabama to New England, where it trends towards the Canadian hills in the ridges known as the Green and White Mountains. This coast-range has a short slope towards the Atlantic, and a long one in the direction of the great Mississippi valley. With the exception of some higher points of the Alleghany range, the surface of this whole plain is glacier-worn from the Arctic regions to about the fortieth degree of northern latitude, the glacier-marks trending from north to south, with occasional slight inclinations to the east or west, according to the minor inequalities of the surface. There is, however, no decided modification of their general trend in consequence of the range of hills intersecting them at right angles for nearly the whole width of the continent between latitudes forty-six and fifty; indeed, the Canadian, or, as they are sometimes called, the Laurentian Hills, formed a no more powerful barrier to the onward progress of the immense fields of ice covering the continent than did the small hummocks, or roches moutonnées, in the Swiss valleys to the advance of the Alpine glaciers. In fact, these low hills may be considered as a succession of roches moutonnées, trending in a continuous ridge from east to west, over which the masses of northern ice have moved unimpeded to the latitude of the Ohio.
Owing to the absence of high mountain-ranges over this vast expanse of land, the glacial phenomena of America are not grouped about special centres of dispersion, or radiating from them, as in Europe. During the greatest extension of the ice-fields, there were but few prominent peaks rising above them, and dropping here and there huge boulders on their surface, to be transported to great distances without losing their rough angular character. And when the temperature under which these vast frozen masses had been formed rose again, the wasting ice must have yielded first on its southern boundary, gradually and uniformly retreating to the Arctic regions, without breaking up into distinct glacial regions, separated from one another, each with its local distribution of erratic boulders and glacier-marks radiating from circumscribed areas on higher levels, as they occur everywhere in Europe. It is true that there are a few localities within the Alleghany range, on the Green and White Mountains, and in parts of Maine, where it is evident that local glaciers have had a temporary existence; but even throughout this eastern coast-range the elevation of the mountains is so slight, and their trend so uniform in a northeasterly and southwesterly direction through twenty degrees of latitude, that the localization of the phenomena is less marked
