At this time the Cordilleran region was nearly a thousand feet (possibly even more) higher than at present, but the eastern plains were rather lower. The eastern margin of the Cordilleran ice-sheet, at the period of its maximum extension, came in places near to the western one of that from the Laurentide region; but whether these periods were synchronous or the two areas actually joined is open to question.

1

As the ice retreated, it often left moraines to mark its halts, and the land generally is covered with sheets of stony glacial deposits of variable thickness, which sometimes filled valleys and blocked up glens. Erratics can be found which in some cases are nearly six hundred miles away from the parent rock, and are perched not seldom some hundreds of feet above it. The same diversity of opinion as to the history of these clays with boulders exists in America as in England. Some authorities consider them to be almost wholly the product of land-ice-the ground moraine of the retreating glacier, augmented by material either incorporated into the ice itself, or carried upon its surface. Others think that, though mainly produced by the action of ice, they have generally been deposited under water. Both parties agree that in some cases streams running in the direction opposite to that in which the ice was advancing might be dammed up till their valleys were

1 There is a singular driftless area in Wisconsin and parts of the adjacent States. See Wright, "Man in the Glacial Period," p. 101.

converted into lakes; but the second school of geologists asserts that whatever greater elevation the land may have had at the outset, this was followed by a subsidence sufficient to bring a considerable part of the lowland region beneath the level of the

sea.

Scattered erratics, often of great size, occur in this American region, and sometimes form trains as in England. Kames in some districts, as in Maine and the south-eastern part of New Hampshire,1 seem to be no less wonderful than in Ireland. So far as I can judge from the evidence published, and from the few sections which I have seen, the glacial phenomena of North America often resemble those of Great Britain, only they are generally on a more gigantic scale. Some of the glacial clays containing erratics, like the tills of Switzerland and of Britain, must be in some way or other the produce of land-ice. Here and there also intercalated beds of peat or of other vegetable matter indicate a retreat, and a subsequent return, of the ice-sheets. But in other places the boulder clays either show signs of stratification or pass up into stratified silts, sands, and gravel. These in some cases may have been deposited in extramorainic lakes, but in others they must be of marine origin. Of the latter there are many proofs, but

1 "Man and the Glacial Period," pp. 77-81.

2 Especially in the reports of the U.S. Geol. Survey, which contain numerous admirable reproductions of photographs representing the various phenomena.

FIG. 22.-Map showing the alleged effect of the Glacial Dam at Cincinnati. (Claypole.) (Trans. Edinb. Geol. Soc.)

On

it may be sufficient to quote a single instance. the flanks of Mount Royal, at a height of nearly five hundred feet above the sea, the surface of the Ordovician limestone, when uncovered in the process of quarrying, is often found to be smoothed, polished, and striated. This is no local or accidental phenomenon, so that it can hardly indicate anything but the passage of an ice-sheet. That surface is covered by boulder clay containing blocks of similar limestone, of the Laurentian gneisses, and of gabbro from the so-called Norian System, above which comes, apparently without any marked break, another clay containing Yoldia (Leda) arctica, which is overlain by the Saxicava Sand,1 a stratum full of marine mollusca. In this position, far away from the present estuary of the St. Lawrence, it is impossible to appeal to any scooping up and transference of material from the Atlantic sea-bed; the moulding also of the rock surface and the direction of the striations show that the ice came from the Laurentide hills, and not up the Gulf of St. Lawrence. In other parts of North America, even in the neighbourhood of the Cordilleran ice-sheet, beach terraces occur. Some can be proved

1 A list of the fossils from the Leda Clay and Saxicava Sand is given by Sir. W. Dawson, "The Canadian Ice Age," pp. 211-2€8. Bones of Phoca Grænlandica, Beluga catodon, and Megaptera longimana have been found in these or similar deposits at Montreal or far inland; the last at an elevation of 440 feet above sea-level. The marine mollusca occur frequently on Mount Royal up to 470 feet, and they have been found in one place at 500 feet. Sir W. Dawson, op. cit., p. 199.

by the presence of molluscs to be of marine origin, but others may possibly be lacustrine, and may mark the shores of those inland sheets of water which were produced by interruption of the drainage system of the country (Fig. 22). There can be little doubt that in Northern America such lakes have been more important factors in the production of stratified drifts, especially of the finer clays, than in our own country; and the fact of considerable oscillation of level, even since the. Glacial Epoch, appears to be proved by the observations on the heights of the old shore-terraces in the region of the great lakes.1

Outlying glacier systems of course existed in favourable situations in the great mountain regions far to the south of these two main centres of ice dispersion, but it is needless to do more than mention them. There can be no doubt that for a considerable time the temperature of a large portion of North America was materially lowered, while the amount of precipitation apparently was not diminished, so that ice-sheets, originating from the two districts. already mentioned, trespassed upon the lowlands. But though the evidence of their presence here is indubitable, that of marine action in some regions seems to be no less certain. Accordingly, it often becomes difficult to assign the glacial deposits to

1 There is an important paper on this subject by W. J. McGee, "Pleistocene History of North-Eastern Iowa," in the Eleventh Annual Report of the U.S. Geological Survey, part i. See also Professor J. W. Spencer, Quart. Jour. Geol. Soc., xlvi. (1890), p. 523.