present St. Clair and Detroit rivers, by which the earliest outHow of the old glacial lake probably passed southward and thence ran east as a glacial River Erie, at first tributary to Lake Lundy. As soon as that very briefly existing glacial lake was drained away, the river followed the lowest part of the shallow bed of the present Lake Erie along all its extent, which then had an eastward descent of probably 200 feet, allowing no lake or only a very small one to exist in the deepest depression of the basin; and north of Buffalo it coincided with the course of the Niagara river.

Gilbert, Wright, and Spencer, have thought that for a long time the outflow of the three great lakes above Lake Erie passed by the way of Lake Nipissing to the Mattawa and Ottawa rivers. It seems to me far more probable, however, that the epeirogenic uplift of the Nipissing region, which had elevated it already about 400 feet during the existence of Lake Warren, continued so fast that both the Trent and NipissingMattawa passes were raised the additional 50 feet needed to place them above the level of Lake Algonquin before the glacial retreat uncovered the country east of them so that outlets could be obtained there.

With the continuance of the uplift of the Lake Superior basin after the formation of the Algonquin beach, the mouth of Lake Superior and the Sault Ste. Marie came into existence; and this movement allowed the lake level at Duluth to fall probably 40 or 50 feet beneath the Algonquin and present shore line. Subsequent differential elevation of the eastern and northern parts of the basin, as compared with Duluth, has again brought the west end of the lake up to the Algonquin shore, but not until the St. Louis river, while the water surface stood considerably lower than now, had deeply eroded its broad channel through the very gently sloping expanse of till from Fond du Lac to the harbor of Duluth and Superior.

The differential uplift of the Algonquin beach, as compared with Chicago and the previous mouth of Lake Warren, has been about 60 feet near the mouth of Lake Huron and at Duluth; 110 feet at the mouth of Lake Superior; 200 feet at Lake Nipissing; and 240 to 290 feet at Barrie, Lorneville, and Orillia, on Lake Simcoe. A broad lobe of the waning ice sheet, terminating on the highland area between the south end of Georgian bay and the west end of Lake Ontario, appears to have delayed the elevation of that district, so that subsequent to the formation of the Algonquin beach more uplifting took place there than at the north side of Georgian bay and about Lake Nipissing. The ascent of the Algonquin beach in nearly 200 miles from the mouth of Lake Huron northeasterly to Lake Simcoe averages about a foot per mile; and thence in

about 135 miles north to Lake Nipissing it descends at an average rate of about eight inches per mile.

While the eastern part of the Lake Algonquin area was being much uplifted, with the formation of other beaches below the first, probably the southern part of the Lake Michigan basin remained with a very slight change of attitude or none, having previously risen to approximately its present height, which it has since held with little or no change. But the northeastward elevation raising the country where Lake Algonquin and now Lake Huron have outflowed, gradually caused the water level at Chicago to rise some 40 feet above its old Algonquin level, which is shown by a sublacustrine terrace formed by the Algonquin wave erosion and beach accumulation.

On the Saugeen river, Ontario, and near the south end of Georgian bay, fresh-water shells are found in beds belonging to stages of Lake Algonquin respectively about 40 and 100 feet below the main and earliest Algonquin beach, or 90 and 78 feet above the present lake and bay.

Lake Lundy. From the Forest beach at Crittenden, Erie county, N. Y., marking the latest level of Lake Warren, there is a descent of 125 feet between 860 and 735 feet above the sea to the earliest strand of the glacial Lake Lundy, which for a time occupied the northeastern three-fourths of the Lake Erie basin. A more conspicuous principal Lundy beach, 30 feet lower, on which is the "ridge road" named Lundy lane, near Niagara Falls, has an eastward ascent of 30 feet in about 40 miles from Font-hill, Ont., to Akron, N. Y., five miles north of Crittenden. Lake Lundy opened through a strait about 30 miles wide into the Lake Ontario basin. Its outflow passed eastward, across the country close north of the Finger lakes, to the Mohawk and Hudson valleys, still partly filled by the receding ice-sheet and permitting a series of mouths of Lake Lundy to be found at successively lower levels, until as the ice-border withdrew the water soon sank to the lowest point of the Ontario-Mohawk watershed at Rome, N. Y., where its level long remained, forming the Iroquois beach. One of the stages of the sinking Lake Lundy or incipient Lake Iroquois, probably nearly midway in altitude between the Lundy and Iroquois beaches, I find to be indicated by my studies of eskers in Rochester and Pittsford, N. Y.+

* J. W. Spencer, "Deformation of the Lundy Beach and Birth of Lake Erie," this Journal, III, vol. xlvii, pp. 207-211, with map, March, 1894.

1893.

Proceedings of the Rochester Academy of Science, vol. ii, pp. 196-198, Jan.,

Lake Iroquois.*This glacial lake, outflowing at Rome to the Mohawk and Hudson, occupied less area in the west part. of the Lake Ontario basin during its earliest stage than during the later and probably longer enduring lake stage by which the high Iroquois beach in that region was formed. Previous to the date of the western development of the Iroquois beach, the early water level stood at one time only a little higher than the present Lake Ontario at Toronto and Scarboro Heights, 6 to 15 miles east of Toronto, as compared with the altitude, doubtless absolutely lower than now with regard to the sea, which the land then held in that part of the lake basin. This is shown by the occurrence of fossil fresh-water mollusks of fourteen species, and wood of ash, oak, and American yew, in beds at Toronto, described by Coleman, which now are 33 to 51 feet above Lake Ontario, or 280 to 298 feet above the sea. All the mollusk species are now living; but four are restricted, so far as known, to waters tributary to the Mississippi. A boulder-bearing surface deposit above these beds proves that the front of the ice-sheet was not far distant; but the climatic conditions of that time, clearly indicated by the fauna and flora, were as mild as now. There next ensued, probably, a gradual rise of the lake, due to an uplifting of the country about its outlet at Rome, until it stood at the level of the well defined Iroquois beach, which has a height at Toronto of about 200 feet above Lake Ontario. Thick fossiliferous delta deposits had been, meanwhile, brought into the north edge of the lake at Toronto and several miles eastward along the lakecliff section of Scarboro Heights, described by Hinde; and repeated re-advances of the ice-front, one during, and another after, the delta accumulation, formed, at the locality last noted, two deposits of till or boulder-clay.

In a limited sense the Toronto and Scarboro fossils may be called Interglacial, since they lie between deposits of glacial drift; but they seem better referred to moderate oscillations. of the ice boundary than to the distinct glacial epochs which Coleman and Hinde infer from them. Both these beds and

* J. W. Spencer, "The Deformation of the Iroquois Beach and Birth of Lake Ontario," this Journal, III. vol. xl. pp. 443-451, with map, Dec., 1890; and papers previously cited. Thomas Roy (in paper by Sir Charles Lyell), Proceedings Geol. Soc., London, vol. ii. 1837, pp. 537, 538. Sir Charles Lyell, Travels in N. A., in 1841-42, vol. ii, chapter xx. E. J. Chapman, Canadian Journal, new series, vol. vi, 1861, pp. 221-229, and 497. 498. Sandford Fleming, Can. Jour., same vol. vi, pp. 247-253. George J. Hinde, Can. Jour., vol. xv, 1877. pp. 388413. A. P. Coleman, Am. Geologist. vol. xiii, pp. 85-95, Feb. 1894. Geol. Survey of Canada, Report of Progress to 1863, pp. 912, 913. James Hall, Geology of New York, Part iv, 1843, pp. 348-351. Baron Gerard de Geer, On Pleistocene Changes of Level in eastern North America." Proc. Boston Soc. Nat. Hist., vol. xxv, 1892, pp. 454-477, with map; also (excepting the map) in Am. Geologist, vol. xi, pp. 22-44, Jan., 1893. G. K. Gilbert, F. B. Taylor, E. W. Claypole, G. F. Wright, and Warren Upham, as cited for Lakes Warren and Algonquin.

the richly fossiliferous Leda clays, which last overlie the latest glacial drift in the St. Lawrence, Ottawa, and Champlain valleys, may be referred to the closing stage or Champlain epoch of the Ice age; and they both testify, like the partially forest-covered Malaspina ice-sheet in Alaska, of the close sequence of a warm climate, with luxuriant plant and animal life, during and immediately after the recession of the icesheet. The transition from the Glacial to the Champlain climate seems readily explained by the epeirogenic depression which ended the Glacial period.*

The height of Lake Ontario is 247 feet; and that of the old Iroquois outlet crossing the water-shed at Rome is 440 feet, above the sea level. Thence the Iroquois beach in its course northward adjacent to the eastern end of Lake Ontario has a gradual ascent of about five feet per mile along a distance of 55 miles to the latitude of Watertown, where the highest beach is 730 feet above the sea, showing that a differential uplift of about 290 feet has taken place, in comparison with the Rome outlet. From Rome westward to Rochester, the beach has nearly the same height with the outlet; but farther westward it descends to 385 feet above the sea at Lewiston and 363 feet at Hamilton, at the western end of Lake Ontario. Continuing along the beach north of the lake, the same elevation as the Rome outlet is reached near Toronto, and thence eastnortheastward an uplift is found, similar to that before described east of the lake, its amount near Trenton and Belleville.above Rome being about 240 feet. It is to be added that northward from Rome the Iroquois beach becomes divided into a series of distinct beaches, marking stages in the northeastward rise of the land and having near Watertown a vertical range of 80 feet below the highest and oldest, which was before noted; and that westward a similar series of strand lines also lies below the highest, likewise before noted, which there, however, contrary to the order northeastward, was the newest. The highest beach near Watertown was probably contemporaneous with the fossiliferous beds of Toronto; some of the intermediate northeastern beaches corresponded to the delta deposits of Scarboro; and the lowest northeastward lake level was continuous with the highest at Toronto, Hamilton, Lewiston, and east to Rome.

Between Lakes Warren and Lundy the old water level near the west end of Lake Ontario fell 125 feet, minus some amount to be subtracted for the progressing northeastward elevation of the land. The two Lundy shores are 30 feet apart verti

* J. D. Dana. Trans. Conn. Acad. of Arts and Sciences, vol. ii, 1870, p. 67; this Journal, III, vol. x. pp. 168-183, Sept., 1875. Warren Upham, Glacialists' Magazine, vol. i, pp. 236-240, June, 1894.

cally. From the lower and main Lundy beach the water fell about 480 feet to the earliest stage of Lake Iroquois when the Toronto fossil shells lived in the edge of that lake, excepting that here again some undetermined amount must be subtracted to compensate the concurrent rise of the land. Adding these vertical intervals together, we have 635 feet, which probably may be reduced 100 feet, more or less, for the effects of the accompanying epeirogenic uplift. We have left some 500 or 550 feet, to be subtracted from the altitude of the old Chicago outlet of Lake Warren, believed to have been then approximately as now, 590 feet above the sea, to give the earliest altitude of the Rome outlet. It thus appears, as I concluded from a similar computation four years ago, that the Rome outlet was at first only 50 or 100 feet above the sea level.* It was gradually uplifted, participating in the differential rise of the whole Ontario basin, to about 300 feet above the sea while the outflow continued here, and to probably 350 feet or more, lacking less than 100 feet of its present height, by the time when the much farther retreat of the ice permitted the extension of the sea to Ogdensburgh and Brockville, on the St. Lawrence river near the mouth of Lake Ontario. Intermediate between Lake Iroquois and the Champlain incursion of the sea, the glacial Lake St. Lawrence, into which Lake Iroquois was merged by the retreat of the ice-sheet from the northern side of the Adirondacks, filled the Lake Ontario basin for a considerable time at levels below the Iroquois beaches.

As the area of Lake Warren was being differentially much elevated during the earlier existence of that lake, and as the area of Lake Algonquin was similarly uplifted in part or wholly contemporaneously with the Iroquois basin, so this region was being rapidly raised and tilted upward to the north. and east while the lake level, held constantly without important downward cutting at the Rome outlet, inscribed many shore lines on the slowly moving land. All the movement throughout the whole region probably was upward; but the position of Rome, and its greater rise than western parts of the basin during the existence of Lake Iroquois, caused the old beaches westward to have now declining gradients.

Lake Hudson-Champlain.†―The absence of marine fossils

* Bulletin Geol. Soc. Am., vol, ii, pp. 260-262.

Warren Upham, Bulletin Geol. Soc. Am., vol. i, p. 566; vol. ii, p. 265; vol. iii, pp. 484-487 (first using this name). C. H. Hitchcock, Geology of Vermont, 1861, vol. i, pp. 93-167, with map. J. S. Newberry, Pop. Sci. Monthly, vol. xiii, 1879, pp. 641-660. F. J. H. Merrill, this Journal, III, vol. xli, pp. 460-466, June, 1891. W. M. Davis, Proc. Boston Soc. Nat. Hist., vol. xxv, 1891, pp. 318-334. S. Prentiss Baldwin, "Pleistocene History of the Champlain Valley," Am. Geologist, vol. xiii, pp. 170-184, with map, March, 1894. Baron de Geer, as cited for Lake Iroquois.