Ce | Ghee ete er ras .F Se al mr ya ; : { 4 " t od ee re -* es, rhe ly E Seniors | vit | OF THE 2) eee * = ee 4 ; “ i WITH : i Notes on the Origin of our Lower Great Lakes. a> - , es BY | | ; . % W. SPENCER, B.A. Sc, Ph.D, F.G.S, : oan King’s College, Windsor, N. S. > oan ‘ . a Spencer.} 300 {March 18, Discovery of the Preglacial Outlet of the Basin of Lake Erie into that of Lake Ontario; with Notes on the Origin of our Lower Great Lakes. By J. W. Spencer, B. A. Sc., Ph.D., F.G.8., King’s Oollege, Windsor, NV. 8. (Read before the American Philosophical Society, March 18, 1881.) SUMMARY. The object of this paper is to bring before the scientific world the follow- ing observations, bearing on the Preglacial Drainage and origin of our Great Lake Basins : 1. The Niagara escarpment, after skirting the southern shores of Lake Ontario, bends at nearly right angles in the neighborhood of Hamilton, at the western end of the lake ; thence the trend is northward to Lake Huron. At the extreme western end of the lake this escarpment (at a height of about 500 feet) encloses a valley gradually narrowing to four miles, at the meridian of the western part of the city of Hamilton, where it suddenly closes to a width of a little more than two miles, to form the eastern end of the Dundas valley (proper). This valley has its two sides nearly paral- lel, and is bounded by vertical escarpments, which are capped with a great _ thickness of Niagara limestone, but having the lower beds of thes pes composed of Medina shales. On its northern side the escarpment extends for six miles to Copetown ; but westward of this village it is covered with — = 1881.] 301 [Spencer, adrift, but it is not absent. On its southern side the steep slopes extend for less than four miles to Ancaster, where they abruptly end in a great de- posit of drift, which there fills the valley to near its summit, but which is partly re-excavated by the modern streams, forming gorges from two to three hundred feet deep. To the north-eastward of Ancaster these gorges are cut down through the drift to nearly the present lake level. Westward of Ancaster, a basin occupying a hundred square miles, where the drift is found tc a great depth, forms the western extension of the Dundas valley. With the north-western and western portions of this drift- filled area the upper portion of the Grand river and Neith’s creek were formerly connected. The Grand river, from Brantford to Seneca, runs near the southern boundary of this basin, then it enters its old valley, which extends from Seneca to Cayuga, with a breadth of two miles, and a depth, in modern times, of seventy-five feet, having its bed but a few feet above the surface of Lake Erie. Near Cayuga, the deepest portion of the river- bed is below the level of Lake Erie. 2. The Dundas valley and the country westward form a portion of a great river valley, filled with drift. Along and near its present southern margin this drift has been penetrated to 227 feet below the surface of Lake Ontario, thus producing a cation with a lateral depth of 743. feet, but with a computed depth, in the middle of its course, of about 1000 feet. _ 8. The Grand river, at four miles south of Galt, has, since the Ice Age, left its ancient bed, which formerly connected with that of the Dundas valley, as did also Neith’s creek, at Paris. 4. Lake Erie emptied by a buried channel a few miles westward of the present mouth of the Grand river, and flowed for half a dozen miles to near Cayuga, where it entered the present valley, and continued this channel (reversed) to a place at a short distance westward of Seneca, whence it turned into the basin referred to above, receiving the upper waters of the Grand river and Neith’s creek as tributaries, and then emptied into Lake Ontario by the Dundas valley. This channel was also deep enough to drain Lake Huron. 5. Throughout nearly the whole length of Lake Ontario, and at no great distance from its southern shore, there is a submerged escarpment (of the Hudson River Formation) which, in magnitude, is comparable with the Niagara escarpment itself, now skirting the lake shore. It was along the foot of this escarpment that the river from the Dundas valley flowed (giving it the present form) to eastward of or near to Oswego, receiving many streams along its course. 5. The western portion of the Lake Eric basin, the south-western coun- ties of Ontario, and the southern portion of the basin of Lake Huron formed one Preglacial plane, which is now covered with drift or water (or with both) to a depth varying from fifty to one hundred feet, excepting in channels where the filling by drift is very great. A deep channel draining _ Lake Huron extended through this region, leaving the present lake near the Au Sable river, and entering the Erie basin between Port Stanley and Spencer.} 302 [March 18, Vienna, at a depth near its known margin of 200 feet, but ata penne depth in the centre sufficiently great to drain Lake Huron. 6. The Preglacial valleys (now buried) of Ohio and Pennsylvania+for example ; the Cuyahoga, Mahoning (reversed), and Allegheny (deflected), formed tributaries to the great river flowing through the Erie basin and the Dundas valley. 7. The bays and inlets north of Lake Huron are true fiords in chaiiakee, and are of aqueous origin. 8. The Great Lakes owe their existence to sub-aérial and fluviatile egen- cies, being old valleys of erosion of great age, but with their outlets closed by drift. Glaciers did not excavate the lakes and had no important action in bringing about the present topography of the basins. 9. The old outlet of the Niagara river, by the valley of St. David’s, was probably an interglacial channel. I. INTRODUCTION. Whilst residing in Hamilton, Ontario (1877-80), a portion of my time was devoted to studying the geology of the neighborhood. At first it began in connection with Lieut. Col. Grant, H. P., Sixteenth Regiment,/and some other gentlemen, in making collections of fossils ; as this locality is one of the best for obtaining Niagara Fossils (and also those of the Hud- son River Formation from the drift pebbles in the beaches) in Canada. In 1874, the present writer published in the Canadian Naturalist a sketch of the local geology. In 1878, he laid the plan of collecting the information necessary for preparing an exhaustive paper on the Geology of the region about the Western End of Lake Ontario. When systematic work was commenced, the information gained required so much time for its study that it has long delayed the publication. A large number of new species of Niagara fossils (twenty-nine of the Graptolite family alone) were ob- tained. The present state of the work is, that a paper on the Palsozoic Geology, and another on the Palxontology, containing descriptions of many new fossil species, are ready for publication. A third portion, on the Surface Geology, is under way ; and the investigations on this subject have, step by step, carried the writer outside of his original field,—having as- sumed an importance never anticipated ; and have resulted in this advance notice of a few of the most striking facts concerning the origin of our great lakes. The completion of the work will be further delayed until oppor- tunity will have been afforded to study some questionable points, espe- — cially such as relate to the drift deposits of the region, and others hav-. iug a broader bearing on the physical geography of the lake regions in Pre- glacial times. . In the present paper, all discussion relating to the vexed glacial fy pou: sis is scrupulously avoided, except those questions bearing on a true ws al nation of the origin of our great lakes. . In the study of the surface geology, the first great question that pre sented itself was, ‘‘ What is the origin of my native valley, Dundas?’ The 1881.] 303 (Spencer, possibility of Lake Erie flowing down through the Dundas valley (though it suggested itself) did not seem probable, owing to the high lands between the two great lakes. However, in the Canadian Naturalist, 1874, I re- ferred to it as having been produced by a ‘‘mighty river.’’ This was like one of those gratuitous hypotheses that are common, now-a-days, for attributing to a continental ice sheet most of the causes of the present phys- ical features of the continent, which do not readily explain themselves. Sub- sequently, Mr. George J. Hinde refers to it as having been scooped out by a glacier. This assertion will be found in the sequel to be.a perfectly un- tenable hypothesis. Certainly, the origin of the valley was obscure, yet it showed that the excavation of a cafion of such magnitude required a pro- portionately great agent; and no present stream would account for even a _ small portion of the excavation. However, in this paper it will be seen that its existence was unquestionably occasioned by the action of a mighty river, as originaily suggested. This outlet of Lake Erie also perfectly ac- cords with, and accounts for the preglacial drainage of Pennsylvania, as made known at the close of last year by Mr. Carll, of the Geological Sur- vey of that State. TI. TopoeraPuy OF THE REGION ABOUT THE WESTERN END OF LAKE . ONTARIO, _ ‘The Niagara Escarpment.—This range of hills commences its course in Central New York, and extends westward, at no great distance south of Lake Ontario. It enters Canada at Queenston Heights, and thence its trend is to the western end of the lake, where, near Hamilton, it turns northward. _ and extends to Cabot’s head and Maintoulin island. Everywhere in Canada, south of Lake Ontario, it has an abrupt fall looking towards the northward ; but at Thorold and other places to the eastward its bow is more broken than at Grimsby, and westward. At Hamilton, the brow of the escarpment varies from 388 to 396 feet above Lake Ontario.* About five miles east of Ham- ilton, the escarpment makes an abrupt bend enclosing a triangular valley, down which Rosseaux creek, and other streams now flow. This valley is about two miles wide at its mouth, and has a length of about the same distance. About five miles westward of Hamilton, the Niagara escarpment be- comes covered with the drift deposits of a broken country, or rather ends abruptly in the drift of the region. Above the range, the country gradually rises to the divide between Lake Ontario and the Grand river, or Lake Erie, without any conspicuous features. South-eastward of Hamilton, at a point about five miles from the brow of the escarpment, where the Ham- * Prof. Dana places the mean level of Lake Ontario at 233.5 feet above ocean- level; the Canadian Geological Survey, at 232 feet; the New York Central Railroad, at 249.84; the Geological Survey of Pennsylvania, takes 5/3 feet as the mean of the results of determin!ng the level of Lake Erie; tie Welland canal levels show Lake Erie as being 326.75 feet higher than Lake Ontario; and the Hamilton and North Western Railway a difference of 328 feet, both of these last routes being short lines with direct courses. Therefore the height of Lake Ontario should be about 245 feet above the sea, Spencer. } . 304 [March 18, ilton and North Western Railway reaches the summit, the altitude above Lake Ontario is 493 feet. At Carpenter’s quarry, two miles southward of the ‘‘mountain’’ brow, at the head of James street, the altitude reaches 485 feet; and near Ancaster the summit is 510 feet above Lake Ontario. From eastward of Grimsby (for twenty miles) to near Ancaster, the es- carpment presents an abrupt face from 150 to 250 feet below the summit (having a moderate amount of talus at the base), thence it- extends by a more or less steep series of slopes to the plane, which gradually — (sometimes by a succession of terraces), to the lake margin. On the northern side of the town of Dundas, the abrupt face of the es- carpment looks southward, and extends four or five miles westward, until the exposure becomes covered by the drift deposits near Copetown station, similar to the termination at Ancaster on the south side of the Dundas valley, but not by an abrupt ending as at the latter locality. About two miles east of the G. W. Railway station, at Dundas, the trend of the range bends more to the northward, and from this point there is a marked differ- ence in the configuration of the country below the summit. The range, after extending beyond Waterdown, turns still more to the northward and passes near Milton, and Limehouse station (on the G. T. pom adh and thence extends to Georgian bay.’ The height of Copetown aboy lake is 502 feet. On the west side of Glen Ppencer it is 409 feet, and east- 7 coming within four feet of the surface). At Waterdown the altitude i is over 500 feet (?) and at Limehouse the brow of the range (though only the lower beds of the Niagara limestones occur) is 810 feet. The features of the surface of the country above the highlands north of Dundas are much more varied than south of the Dundas valley. As the trend of the es- carpment turns northward around the end of the lake, the face of the slope looks towards the eastward. But the country does not present the steep declivities as exhibited along the southern side of Lake Ontario; for the vertical face is usually less than 100 feet, and the country between it and the water has a more uniform pitch. Basin of Lake Ontario. As is well known, Lake Ontario consists of a broad; shallow (considering its size) basin, excavated on the southern margin out of the Medina shales, and having its southern shores from one to several miles from the foot of the Niagara escarpment. The Medina shales form the western margin (where not covered with drift) toa point near Oakville. From this town toa point some distance eastward of Toronto, the hard rocks are made up of the different beds of Hudson River Epoch ; while the soft Utica shales occupies the middle portion, and the Trenton limestones the portion of the Province towards the eastern end of the lake. The country at the western end of the lake consists of slopes gently rising © to the foot of the Niagara escarpment, noticed before. Sometimes this ele- vation is by terraces, and again by inclinesso gentle, as between the foot of the escarpment at Limehouse (on the G. T. Railway) and the lake, where 1881.] 305 [Spencer, the difference of altitude above the water is more than 700 feet, without any very conspicuous features. At the western end of the lake, the two shores converge at an acute angle. At about five miles from the apex of this angle is the low Burling- ton beach, thrown across the waters in a slightly curved line; which forms the western end of the open lake. Burlington lake, thus formed, is connected with the open lake by a canal of the same name, made where there was a former shallow opening between the waters within and without the beach. This beach is made up of sand and pebbles (mostly of Hudson River Age), and is more than four miles long, but nowhere is it half a mile wide. No mean depth of Lake Ontario can be fairly stated. For geological pur- poses it has no mean depth, because it is simply a long channel with the adjacent low lands covered by back-water. West of the meridian of the Niagara river the lake is evidently filled with more silt then eastward, as we find that the bottom slopes more gradually towards the centre, where the mean depth (increasing from the westward) of the channel may be fairly placed at 400 feet below the pres- ent surface of the waters. In this section of the lake, the average slope from both shores may be stated at 30 feetin a mile. At a short distance east of the 78th meridian, the character of the late bottom changes in a most conspicuous manner. Here we find a deeper channel which extends for more than ninety miles, having an average depth of about 90 fathoms or 540 feet, with, in some places, a trough of about 600 feet depth, gener- ally near the southern margin of the 90-fathom channel. Here and there is a deeper sounding—the deepest being 123 fathoms or 738 feet. The long channel, surrounded by the 90-fathom contour line, is situated at a mean distance of not less than twenty miles from the Canadian shore, whilst its southern side approaches in some places to within six miles of the Amer- iean shore, with which it is parallel. This 90-fathom channel varies from three to twelve miles in width. Its broadest and deepest portion is south of the Canadian peninsula of Prince Edwards’ County. The mean slope of the lake bottom, from the Canadian shore to this deep channel just pointed out, may be placed at less than twenty-five feet in a mile, with variations from twenty to thirty feet in that distance. The mean slope from the New York shore line to the 90-fathom channel may be placed at sixty feet ina mile, but varying generally from fifty to ninety feet. On examination we find that the greater portion of this slope belongs to a belt which descends much more rapidly than the off-shore depression. That the southern side of Lake Ontario has a submerged series of escarp- ments or one moderately steep and of great. dimensions, is manifest when we come to study the soundings. In fact, if the bed of Lake Ontario were lifted out of the water, this submerged escarpment would be more con- spicuous than the greater portion of the present one, known by the name ofthe Niagara. In many places the descent from the table-land above the Niagara escarpment is no more precipitous than the slopes of the sub- eperiedr) 306 sae [March 18, merged Cambro-Silurian (Hudson River, in part, if not throughout the entire length) rocks, with its sloping summit, in part crowned by a gently sloping surface of Medina shales. Nearly north of the mouth of the Gen- esee river we find that within a single mile the soundings vary from forty- three to seventy-eight fathoms (between contour lines). This gives a sud- den descent in one mile of 210 feet. As the soundings are not taken con- tinuously to show to the contrary, most of the change of levels msy be within a few hundred yards. In the region of these soundings the deepest water outside of the 78. fathom line is 84 fathoms, while from the shore to the 48-fathom sounding the least distance is four and a half miles, thus giving the greatest mear slope of the lake bottom at sixty feet in a mile, before the escarpment is reached. : An excellent series of soundings can be studied in a line nearly north- ward from Pultneyville, N. Y.: . Distance from Pultneyville| Depth of Sounding. scr sounals lala 0.5 miles. 42 feet. ce 1.0 e 5 60 feet per mile. 1 Bans 126. ** tae e 4.125 ‘ 246 “ 50. ** “ 5.0 «* \ Face of the Bie. ** \ 144 “ be 6.0 «* J escarpment. 582“ 210 *“ 6 7.0 ‘e 624 “é 49 ‘é - 66 10.0 s 642 *“* 6\ m 12.0 < 738 ** 48 <* fs Fig. 1. From this table it will be seen that in a distance of less than two miles the slope of the escarpment is the difference between 582 and 246 feet, or 336 feet as actually recorded. At Hamilton, the Niagara escarpment is only 388 feet above the lake, which is two miles distant, whilst the present slope at Thorold is spread over rearly twice that distance. That this escarp- ment is not local is easily seen. For a distance of over forty miles, from near Oswego westward, it plunges down 300 feet or more in a breadth varying from less than two to three miles. Eastward and westward of this portion of the lake this submerged escarpment can be traced for nearly 1881.] 307 [Spencer. one hundred miles, but with the portion deeper than the 70-fathom con- tour having more gradual soundings, as the base of the hills either origi- nally hada more gradual slope, or the lake in its western extension has subsequently been filled with more silt. Although we have not soundings made very close together, yet the admirable work of the United States Lake Survey is more than sufficient to prove the existing of a continuous escarpment that has an important bear- ing on the Preglacial geography of the region, and on te explanation of the origin of the Great Lakes themselves. The soundings do not show a conspicuous escarpment after passing west- ward of the meridian of Niagara river, partly on account of the sediments filling this portion of the lake, and partly because the lake in all proba- bility never had its channel excavated to so great a depth as farther east- ward. Attention must be called to the fact that the depth of the Niagara river is 12 fathoms near its mouth, but that the lake around the outlet of the river has not a depth exceeding four fathoms with a rocky bottom. Another escarpment at the level of Lake Ontario, now buried, was dis- covered by the-engineers of the enlargement of the Welland canal, accord- ing to Prof. Claypole (Can. Nat. Vol. ix, No. 4). When constructing, No. 1 lock, at Port Dalhousie, it was found that at its northern end, there was an absence of hard rock which formed the foundation of its southern end. Rods more than 40 feet long were pushed into the slimy earth without meeting any hard rock bottom. This discovery will be noticed in the sequel. * Basin of Lake Erie. The exceedingly shallow basin of Lake Erie has its bottom as near a level plane as any terrestriai tract could be. Its mean depth, or even maxima and minima depths from its western end for more than 150 miles, scarcely varies from 12 or 18 fathoms for the greater por- tion of its width. The eastern 20 miles has also a ked no deeper than the western portion. Between these two portions of the lake, the hydrography shows an area with twice this depth (the deepest sounding being 35 fathoms). This deepest portion skirts Long Point (the extremity, a modern peninsula of lacustrine origin), and has a some- what transverse course. An area of less than 40 miles long has a depth of more than 20 fathoms. The deeper channel seems to turn around Long Point, and take a course towards Haldemand ceunty, in our Canadian Province, somewhere west of Maitland. The outlet of the lake, in the di- rection of the Niagara river, has a rocky bottom (Coniferous limestone). The study of this lake at first appears less practicable than that of Ontario, but, when its former outlet and its tributary rivers are described, the writer trusts that he will have made some observations, that may help to clear the darkness that hangs about the history of our interesting lake region, before the advent of the Ice Age. The Dundas Valley and adjacent Cations, We may consider that the *See Report of Chief Engineer of Canadian Canals, 1880. PROC. AMER, PHILOS, soc. xx. 108. 2M. PRINTED MARCH 30, 1881. Spencer. } , 308 {March 18, Dundas valley begins at the ‘‘ bluff’ east of the Hamilton reservoir, and extends westward, including the location of the city of Hamilton and the Burlington bay, at least its western portion. With this definition, the width at the Burlington heights (an old lake terrace 108 feet above present level of the water) would be less than five miles. At a mile and half westward of the heights, the valley suddenly becomes narrowed (equally on both sides of its axis of direction, by the Niagara escarpment making two equal concave bends, on each side of the valley, whence the straight upper portion extends, the whole resembling the outline of a thistle and its stem), from which place it extends six miles westward to Copetewn, on the northern side ; and three and a half to Ancaster, on its southern side. The breadth between the limestone walis. of this valley varies somewhat from two to two and a half miles. The summit angles of the limestone walls on both sides are decidedly sharp. . ' Dundas town is situated in this valley, its centre having a height of about 70 feet abeve Lake Ontario, but its sides rise in terraces or abrupt hills; and on ascending the valley, we find that between the escarpments are great ranges of parallel hills separated by deep gorges or glens, excavated in the drift by modern streams. This rugged character continues until thé summit of the Post Pliocene ridges have a height equal to that of the es- carpment. As the gorges ascend towards the westward, they become smaller, until at some distance south-west ef Copetown and Ancaster, the divide of the present system of drainage isreached. Some of these streams have cut through the drift, so that they have only an altitude above the lake (which is seven miles distant) of 240 feet, while the tops of the ridges im mediately in the neighborhood are not much less than 400 feet high, though they themselves have been removed to a depth of about another hundred feet, for the drift has filled the upper portion of the valley to the height of 500 feet above Lake Ontario. Even to the very sources of the streams, the country resembles the rivers of our great North Western Territories (or those of the Western States), cutting their way through a deep drift at high altitudes, which is not underlaid by harder rocks, showing deep valleys rapidly increasing in size and depth, as they are cleaning out the soft material, and hurrying down to lower levels—a strong contrast to the fea- tures in most other portions of our Province. On the southside of the Dundas valley, a few unimportant streams, mostly dry in summer, have worn back the limestone escarpment, over which they flow, to distances varying from a few yards to a few hundred, making glens at whose head in spring time some picturesque cascades can be seen. — At Mount Albion, six miles east of Hamilton, there are two of these larger gorges, whose waters, after passing over picturesque falls, 70 feet high, and through glens several hundred yards in length, empty into the triangular valley noticed before. On the north side of the Dundas valley, besides small gorges with their streams comparable to those on the south side, there are several of much larger dimensions ; for example, that at Waterdown, six miles north of Hamilton. Stilllarger is Glen Spencer which has a cavon half 1881. | 309 {Spencer, a mile long; 300 feet deep and between 200 and 300 yards wide at its mouth, At the head of this is Spencer Falls, 135 feet high, and joining it laterally there is another cafion, with a considerable stream flowing from Webster’s Falls, which, however, is of less height than the other. The waters feed- ing their streams come from northward of the escarpment, and belong to a system of drainage different from those streams which flow down through the drift of the Dundas valley, and are of much greater length. At the foot of Spencer Falls, the waters strike the upper portion of the Clinton shaly beds. The Falls now are two feet deeper than twenty years ago. Yet the stream is small, and makes a pond below in the soft shales. But this difference in height does not represent the rate of wearing or recession of the precipice. That the stream is mnch smaller than formerly is plainly to be seen, for at present it has cut a narrow channel, from ten to fifteen yards in width, above the falls, and from four to six feet deep, on one side of the more ancient valley, which is about 50 sh wide and 30 feet deep, excavated in the Niagara dolomites. The surfaces of the sakeociciak in both sides of Glens Si dccar and. Webster present a peculiar aspect. That on the north-eastern side has a maximum height of 520 feet above the lake. On the same side, a section made longitudinally shows several broad shallow glens nearly a hundred feet deep crossing it and entering Glen Spencer. The surface of the rocks is glaciated, but not parallel with the direction of the channels. On the south-western side of the same cavon, we find that a portion of the thin beds.of Upper Niagara limestone have been removed. This ab- sence is not.general, for it soon regains its average height of about 500 feet. Dundas Marsh. The eastern end of the Dundas valley contains a large swamp, nearly three miles long, with a breadth of about three-fourths of a mile, known in the early settlement of the country by the name of Coote’s Paradise. This marsh was formerly connected. by a small rivulet with Burlington bay, but this was subsequently closed by the G. W. Railway, when the cutting of Desjardin’s canal through Burlington heights was completed. Into this marsh all-the drainage of the Dundas valley is deposited, causing it to fill up at the rate of one-tenth of a foot per annum. Burlington Heights. Across the eastern end of the Dundas swamp and some of its branches,are the Burlington Heights, varying from a few hundred yards to nearly a quarter ofa mile in width, and over 100 feet in height, which have been an old beach, at a time when the lake level was at the same eleva- tion, for we find that a lake beach extends along the flanks of the escarp- ment, both eastward and northward for a considerable distance at the same level. This is mentioned here as forming a most, conspicuous terrace, and as changing the physical character of the western extremity of Bur- lington bay, and the outlet of the Dundas valley. Various terraces and beaches are found, both at lower levels, and also fragments at higher alti- tudes, or along the side of the ‘‘mountain,’’ until some attain a height of 500 feet above Lake Ontario. Spencer. ] 310 [March 18, - The Grand River Valley. The Grand river of Ontario rises in the County of Gray, not more than twenty-five miles from Georgian bay. Thence it flows southward, and at Elora the river assumes a conspicuous feature. Here it cuts through the Guelph Dolomites to a depth of about feet and forms a cavon about 100 feet in width with vertical walls. At this place it is joined by a rivulet from the west, which has formed a tributary canon similar to that of the Grand river itself. The country in this region is so flat that it appears as a level plane. Farther southward the river winds over a broader bed, and at Galt the present river valley occupies a portion of a broad depression in a country indicating a former and much more extensive valley. In fact, the old river valley existed in Preglacial times, for the present stream has re-excavated only a part of its old bed at Galt, leaving on the flanks of one of its banks (both of which are) composed of Guelph Dolomites, a deposit of Post Ter- tiary drift, in the form of a bed of large rounded boulders mostly of Lau- rentian gneisses. The country for four miles south of Galt is of similar character, forming a broad valley, in which the present river flows. At this distance from Galt the river takes a turn to the sovth-westward ; but at the same place, the old valley appears to pass ina nearly direct line with the course of the present bed (before the modern turn is made to the west. ward). As this portion of the valley now entered, has not to any extent been cleaned out by modern streams, it forms a broad shallow depression in the country extending for a few miles in width. Yet, it is often occu- pied with hills composed of stratified coarse gravel belonging to that belt, which extends from Owen Sound to the County of Brent, and called by the Canadian Geological Survey ‘‘ Artemesia gravel.” It is through a portion of this valley that the Fairchild’s creek flows. Many streams derive their supplies of water from the Beverly swamps, which also feed the Lindsay creek, that empties over Webster falls and flows down Glen Spencer through the Dundas valley to Lake Ontario. The G. W. Railway, at four miles south of Galt, enters this valley and continues in it or its branches as far as Harrisburg, though the deeper de- pression is near St. George (a short distance west of Harrisburg). After leaving what I consider its more ancient bed, south of Galt (unless the country between the present bed and Fairchild’s creek was an island), the Grand river flows southward to Paris and Brantford, having a deep, bread valley. At the latter place the valley may fairly be placed at a few miles in width, while further to the eastward the river winds in an old course, which had formerly a width of over four miles (see map). In the region - of Brantford the valley is bounded by a somewhat elevated plateau. At Paris, Neith’s creek enters the Grand river from the west, and has a val- ley almost comparable in size with that of the latter at this town. At Paris, the Grand river cuts through the plaster-bearing Onondaga forma- tion. Similar rocks appear at various places along the river, at places where the river has cleaned out:a portion of one side of its ancient valley. At the Great Western Railway crossing, east of Paris, the bed of the 1881.] 311 [Spencer. river has an altitude of 495 feet above Lake Ontario, while at Brantford it is 410 feet (this elevation may not be perfectly accurate) above the same datum. From Brantford the river winds through a broad valley, with a general easterly direction, to Seneca, where the immediate bed is about quarter of a mile wide, flowing at the southern side of a valley, more than two miles wide, and 75 feet below its boundaries, which are 440 feet above Lake Ontario (see profile on subsequent page). At Seneca the bed of the present river-course is 365 feet above Lake Ontario, or only 37 feet above Lake Erie. (The H. & N. W. Railway levels give Lake Erie as 328 feet above Lake Ontario, whilst the Report of the Chief Engineer of the Wel- land Canal states that the difference of level is 326? feet. As these two levels agree so nearly, and as the other figures refer to the railway levels, I have followed them here.) Eastward from Seneca the river continues to have its broad valley as far as Cayuga. To near this town the waters of the Welland canal feeder reach, at a height of about 9 (2) feet above Lake Erie. From Seneca to Cayuga the direction of the valley is nearly south, but at the latter place it abruptly turns nearly to the eastward, and in a short distance it passes to a flatter country and flows over Coniferous limestone. After a sluggish flow, it enters Lake Erie (passing through a marshy country) at Port Maitland, more than fifteen miles in a direct line from Cayuga. It must be remembered that, from Seneca to Cayuga, the valley is broad and conspicuous. At only a short distance south of the river, at Seneca, the summit of the country is occupied by a gravel ridge.* Returning to the valley of Fairchild’s creek, we find the stream princi- pally flowing in the former bed of the Grand river, abandoned a few miles below Galt since the Ice Age. This creek crosses the Great Western Rail- way at a level of fifteen feet below the crossing of the Grand river, ata few miles to the westward. Again, the Fairchild’s creek crosses the Brantford and Harrisburg railway at an altitude of 407 feet above Lake Ontario, or a little below that of the Grand river at Brantford, although it empties into it a few miles east of the city just named. Fairchild’s creek is now of moderate size meandering through the drift for a width of two miles. This drift is in part stratified clay. The Grand * The General Manager and Chief Engineer of the air line of the G. W. Rail- way have recently kindly furnished me with a profile of the railway crossing over the Grandriver. A similar favor has been kindly conceded by the Chief Engineer of the Canada Southern Railway. From both of these lines of levels (about a mile apart) we find that the hard rock appears in the drift at a few feet below the bed of the river. but at a level below that of the surface of Lake Erie. The stream, at these places, occupies the eastern portion of tbe valley about two miles from the nerthresn or aorth-western boundary of the valley, marked: by the contour line of 440 feet above Lake Ontario, noticed south of Seneca, but which also occurs westward of Cayuga, near the general bend in course of the river. On both of these profiles, at about half a mile to the westward of the present site of the river, adepression in the drift occurs to a depth but little inferior to that of the present river-bed. This appears to mark the place where the ancient channel leaves what is now the modern direction of the river fora nearly direct line to the Erie basin, Spencer. ] 31 2 [March 18, river from Brantford eastward, is generally excavated from the drift deposits, although occasionally one side of the valley shows rocks of Onondaga formation, exposed by the removal of the drift in modern times. It:is also desirable to call attention to the fact that in the region of Brantford, much of the Onondaga Formation is shaly and forms the surface country rock,’ covering a broad belt, whilst from Seneca eastward, the surface of the country is more generally covered with Coniferous limestone. ' Country between the Grand River and Dundas Valleys. The watershed between these two present drainage systems is at only a short distance south- west of Copetown, and the distance in a direction from the Fairchild’s to the Dundas side of this divide is less than seven miles, with an average altitude of less than 480 feet (the same as that of the Fairchild’s creek as it - crosses the Great Western Railway). The highest point that I have leveled is 492 feet above Lake Ontario. On receding westward from the divide, the country gradually descends to the Fairchild’s creek, which, as it crosses the Brantford and Harrisburg Railway, is 407 above the lake. It is considerably lower where it enters the Grand river. The region between the divide and the Grand river is traversed from north-west to south-east by a considerable number of streams, all with relatively large valleys, cut in the drift, since the present system of drainage was inaugurated in Post Glacial times. The country from Jerseyville (about 465 feet above lake) slopes gradually to the Grand river, from six to eight miles distant to the southward. ! On examination, it may be seen that the country is too high to permit the Fairchild’s creek or Grand river, as they are at present situated, to flow over the height of lanc into the upper portion of the Dundas valley. As referred to before, the Niagara limestone forming the summit of the es- carpment at Ancaster and eastward has a height of about 500 feet. These beds dip at only about 25 feet in a mile (to about 20 degrees west of south) and are not generally covered by a great thickness of drift, but in many places are exposed on or near the surface. Westward of Ancaster these limestones are nowhere to be found, but the country is only covered with drift. At a short distance west of this village, we find streams flowing north-easterly and easterly with very deep valleys in the drift, indicating the absence of the floor of limestone to a depth of over 250 feet below the surface of the escarpment. But on going westward we find that the streams have not cut to an equal depth, but still running deeply through drift. Eventually we reach the divide, after which we find that other systems of streams also cut deeply in the drift running in a south-easterly direction to join the Grand river ; but the Niagara limestone ‘is absent from a consider- able extent of country. On the northern side of the Dundas valley the escarpment after reach- ing Copetown is buried by the drift. Although the line of buried cliffs recedes scmewhat to the northward of the Great Western Railway, yet there are occasional exposures, as at Troy and other places in Beverly and Flamboro, where the underlying limestones come to the surface. At 1881.) 313 [Spencer. Harrisburg the limestones are known to be absent for a depth of more than 72 feet, as shown in a deep well in the drift. In the town of Paris one well came upon hard rock at 10 feet below the surface, whilst another at 100 feet in depth, reached no further than boulder clay. This last well must have been ina buried channel of Neith’s creek, as outcrops of gypsum-bearing beds of the Onondaga Formation frequently occur near the summit of the hills. From what has just been written, it is easily seen that the Niagara limestones are absent from a more or less horizontal floor (which is over 500 feet above the lake, on both the northern and southern sides of the Dundas valley) which continues from Dundas westward to near Harrisburg, where it meets a portion of the Grand River valley. But almost immediately west of Ancaster we find streams running northward at right angles to the escarpment, and cut- ting through drift to the depth of almost hundreds of feet. In fact, if we draw a line from Dundas to northward of Harrisburg (a mile or two), and another from Ancaster southward to the Grand river, we have two limits of a region where the limestone floor has been cut away from an otherwise generally level region. The southern side of this area is the south- ern margin of the Grand River valley, between Seneca and Brantford, and the western boundary is composed of Onondaga rocks east of Paris (which perhaps forms an island of rocks buried more or less in drift). _ Additional proofs may be cited. About a mile south of Copetown a well. was sunk to the depth of 100 feet before water was obtained. At two miles south east of the same village there is small pond only 240 feet above Laxe Ontario, or more than 260 feet below the neighboring escarpment. This is in drift. Again, at a mile north of Jerseyville, the country has a height of 465 feet, with a well in the surface soil to a depth of 40 feet. A small rivulet flows in a valley a few hundred yards south of the last, named well which has a bed 485 feet above the lake. At about a mile west of Jerseyville, the altitude is 468 feet with a well 52 feet deep. Again, at about two miles west of the same village, near the county line, the altitude is 460 feet, with a well 57 feet deep (the bottom being lower than the Fair- child’s creek more than three miles to the westward). About a mile north of the last named station is a ravine 436 feet with the adjacent hills forty. feet higher, and rising in a mile or two to. about 500 feet. All these wells are in the drift. From exposures near Ancaster, it appears that the un- stratified drift has not an altitude of much more than 400 feet. And as we know that some of these superficial beds are stratified clay, and over most of the country just described not a boulder is to be seen, neither on the sur- face nor in the material taken from the greater portions of the wells, it is probable that the water is only obtained on reaching the more porous boulder clay below. It has also been noticed that two wells, at least, are 100 feet deep before reaching water, therefore we may fairly place this as about the inferior limit of stratified superficial clays. _By reference to the - accompanying map, it will be seen that westward of the meridian of Ancas- ter there is an area of over 100 square miles, where the Niagara floor is Spencer. } : 314 {March 18, known to be removed everywhere to a depth of 100 feet, and in its eastern portion to more than 260 feet, and still nearer Lake Ontario to a measured depth of more than 200 feet below its waters. III. TopoGRAPHY AND HyDROGRAPHY OF LAKES SUPERIOR, mance HURON AND St. Chater. As the origin of all our great lakes is so closely related, it will not be out of place to describe briefly some of the features of the upper lakes that ap- pear most striking. In the present paper it is only intended to call atten- tion to some of the existing physical features of these great basins of water that appear to show a relation which existed in a more or less common origin of all our lakes. Though I have frequently visited many localities on these lakes, for this portion of the present paper I am particularly in- debted to General Comstock, Superintendent of the U. 8. Lake Survey, ' who kindly furnished me with the lake charts. Lake Superior. This lake may be described as a large basin with a level or gently undulating bottom and steep margins. The mean depth may be placed at 800 feet below its present surface. Very few soundings exceed 900 feet. Of these, one near the centre of the basin is 954 feet, and another, not far from Duluth, is 1026 feet—the maximum depth of the lake, as shown by the hydrography. The depth of the lake at three or four miles from the shore is generally as great as in its centre; in fact, it is often deeper near the shore on its north-western side. However, about the Apostle islands, between the Pic- tured rocks and St. Mary’s river, and in some of the bays, the waters are shallower than in the open lake, with their floors more or less gradually sloping as they recede fromthe land. As is well known, the lake is gener- ally surrounded by crystalline or metamorphic rocks, which rise from sev- eral hundred to even twelve or thirteen hundred feet above its surface. In short, the near shore hydrography simply shows that the present sub- merged margins of the lake are composed of the bases of the same rugged hills that surround it above the water. The margin of this mountain- bound basin forms a strong contrast with its floor, which, at most, is only — a slightly undulating plane, extending for nearly its whole length and often for little less than its breadth, excepting in its south-eastern portion and some other places referred to above. In fact, the lake bottom is quite as level as most extensive planes which are now subjected to SE SEE action. That this great plane is not covered with any great depth of drift-deposit (excepting locally) appears evident on examining the character of the bottom of the lake in the soundings just off Keweenaw point, and these to the northward. In the various localities hard and rocky bottoms are alike found in both places, at the same depths, so frequently that they cannot be regarded as only rocky eminences protruding through the silt covered bottom which is generally observed. The general direction of the deepest channel, for more than 200 miles along the north-western margin of the lake, appears to point to a river 1881.) 315 (Spencer. course in the region of its south-western extremity, and the few deeper holes to have been produced by some receding cascade from the adjacent shore to which there appears to be a transverse deep channel south of the mouth of Gooseberry river. Again, Prof. N. H. Winchell calls attention to the depression in the low country between the Chocolate river (east of Marquette), and Train bay (near the Pictured rocks), as the only place where there could have been connection between the basins of Lakes Superior and Michigan. It may be remarked that some of the deeper soundings put in towards this portion of the coast, whilst, tothe westward and eastward, the present lake bottom slopes more gradually. The soundings, however, that are near the shore, show a rocky bottom, excepting north of Laughing Fish point (Sable river), and along a narrow channel north of the mouth of Chocolate river. The Jake is very shallow for some distance westward of the St. Mary’s river. ; Lake Michigan. This lake may be said to consist-of a broad long plane, the northern half having a mean depth of about 600 feet, whilst the soundings in the southern half are not much more than half that measure- ment. The deepest sounding recorded is 870 feet, in the latitude of the southern end of Green bay. Throughout the whole length, the lake ap- pears to be traversed by a deep channel, and in the northern end by more than one. Although the pitch of the bottom from the shore line is more or less gradual—generally less than 49 feet in a mile—yet, along the eastern side there is a precipitous escarpment extending for a considerable distance, which in one place suddenly descends, in a horizontal line of little over a mile, from 17 to 93 fathoms, or 456 feet, and increases 60 feet more in the distance of another mile. The conspicuous channels in the submerged plane extend far northward to near the end of the lake. An interesting sounding east of the meuth of the Manistique river shows a depth of 448 feet, at a distance of two miles from the shore, whilst all the adjacent depths do not exceed 11 fathoms. This appears to be a continuation of the deep soundings, ten miles to the southward, but the surrounding lake bottom is covered with drift to a great depth, wherever the Niagara limestones have been removed. It is more than probable that this great depth is in a rock-bound channel of an ancient water course, which elsewhere has been filled with drift. It seems proba_ ble that it was a portion of a buried channel extending through the valley of the Manistique lakes to the depression in the country south of Lake Su. perior, alluded to above, and formed a Preglacial connection between the valleys of Lakes Superior and Michigan. Prof. Winchell regards the val- ley between the two lakes along the Chocolate and White Fish rivers (the latter emptying into Little Baie De Noc), as indicating the ancient con- nection. This route seems less favorable, as both Little Baie de Noc and Green bay are shallow compared with Lake Michigan, for the greatest depth, which is near an outlet through Portes des Mortes, is only 32 fathoms, whilst generally the bay does not exceed 100 feet. PROC. AMER. PHILOS. SOC. xIx. 108. 2N. PRINTED MARCH 30, 1881. Spencer.] 316 {March 18, Green bay is separated from: Lake Michigan by a Niagara escarpment facing the westward, and rising two or three hundred feet above the waters. There appears not to have been any closer connections between these two basins at any previous time than at present, excepting when the waters were at a higher level. We are told that from Green bay for 400 miles to the Mississippi river, a broad, low depression occurs in the coun- try and may have been a former outlet for Lake Superior. This valley is filled with drift even if it ever had a sufficient depth.* Grand Traverse bay has a considerable depth in both of its branches, especially in the eastern. Here we find depths to 612 feet, whilst its north-, ern mouth is now filled, so that it does not exceed 120 feet. The north eastern portion of the basin of Lake Michigan has a general depth of less than 100 feet, but with deeper channels running through it. Many of the soundings about the Straits of Mackinac show a rocky bottom at no great depths. The channel between the 10-fathom contour margins is not much more than a mile and a half wide, and thongh generally shal- lower, contains a hole 252 feet deep. In proceeding outward, the deepest channel passes northward of Mackinac island, having a depth not exceed- ing 216 feet, and a width of less than a mile. Again, a depression of the country extends from near Chicago, on Lake Michigan, towards the Mississippi river, which, in some places, is known - to be filled with drift to a depth of more than 200 feet, according to Dr, Newberry. This is along the Illinois river, whose valley is from two to ten miles wide ; whose mouth is 200 feet lower than Lake Michigan ; and. whose upper streams, near Chicago, are only a few feet higher than the neighboring lake. Lakes Huron and St. Clair. Of these water basins we can make four, divisions. The first section may be made to include the shallow basin south of a line drawn from Thunder bay, or Presqu’ Ile, to Kincardine, in Canada, and Lake St. Clair. The second basin comprises the deep chan- nels of Lake Huron, and extends northward to the Manitoulin islands and the Indian peninsula; the third, the north channel between the Manitoulin islands and the Huronian hills, to the northward; the fourth, Georgian bay proper. ; The first of these divisions is represented by shallow water, seldom thirty- * Since writing the above, I have fortunately been.able to see General War- ren’s Report on the Transportation Route from the Mississippi river to Green — bay (via the Wisconsin and Fox rivers). In this report we find that the bottom of the valley alluded toin the text hasa maximum height of 208.8 feet above Green bay, and also that Lake Winnebago (on Fox river) is 169 feet above the same water. This small lake discharges by the Fox river, which flows over hard limestones down a series of rapids. Therefore Green bay never discharged its waters into the Mississippi river, and this depression in the country between the Great river and Lake Michigan (the Green bay portion) was not a former outlet of Lake Superior, since it was within about 200 feet of the present level. This fact strengthens the probable correctness of the suggestion that Lake Su- perior emptied into the northern end of Lake Michigan directly. Also, Green bay has evidently the character of a fiord. The outlet of Lake Michigan could only have been by the low country along the Illinois river. 1881.] 317 (Spencer. five fathoms deep, but with a channel of about fifty fathoms depth running through it, towards the direction of the north angle of the Au Sable river, near Brewster’s mills. Saginaw bay, belonging to this section, is like Green bay, shallow even at its mouth, where it is less than 100 feet deep. _ Lake St. Oluir is a flat plane, with its bed varying from 18 to 21 feet below its surface, and is altogether modern. At Detroit the drift is 180 feet deep. The three south-western coun- ties of Ontario are low and flat, and covered with drift varying gen- erally from 50 to 100 feet in thickness below the level of Lake Erie. In_ places it is known to be absent to a depth of 200 feet below the same lake in portions of a buried channel to be noticed below. In fact, all the evi- dence appears to show that the southern end of Lake Huron and the west- ern end of Lake Erie, with the intervening region, constituted oné plane underlaid by a considerable depth of Erian shales, reposing on the thick development of Coniferous limestone, and traversed by deep channels running through it. _ The section of Lake Huron under consideration is mostly excavated out of Upper Erian shales in a direction at right angles to the trend of the formations. The denuding action was lessened when the waters in the deeper northern part of the lake subsided to a level having a southern mar- gin bounded by hard Coniferous limestone, covered to no very great depth with Upper Erian shales subjected to only sub-aérial action—the whole traversed with water courses in deep channels. The second division into which, for convenience, I have made of Lake Huron, is that portion between the line drawn from Presqu’ He to Kincar- dine, and the Manitoulin islands to the northward. This is the deepest portion of the lake and extends in a direction running from north-west to south-east. It consists of a broad plane traversed by several deep channels. The average depth of this plane below the surface of the lake does not ex- ceed 75 fathoms, although there are channels much deeper, one of which is represented by a depth of 117 fathoms. There is also one isolated sounding, which reaches 125 fathoms or 750 feet, this being the deepest spot known. The deeper channels appear to lead from the northern portions of the lake, and unite as they proceed southward, being separated by elevations indicating peninsulas or islands. Two of the principal channels appear to proceed from Mississagua strait (between Manitoulin and Cockburn islands), and from south of Manitoulin island, eastward of the Duck islands. However, the channels in the marginal portions of the lake are generally more obscured by drift or silt than towards the central waters. The channel, if such you can call it, proceeding from the Mackinaw straits is of inferior depth to those leading from the more northern end of the lake. This portion of the lake is excavated out of the rocks of the various for- mations from the Niagara to the Coniferous limestones. But most largely out of the more or less soft rocks of the Onondaga group, along the strike of these formations, thus giving the eroding agencies the power of remov- ing the softer basal rocks, and of producing an escarpment of the Conifer- Spencer.] | 318 [March 18, ous limestone looking to the northward, until it was finally undermined, and worn back to its present position, submerged beneath the shallower waters of the southern portion of the lake, or buried in drift deposits. On the northern side, the lake has not made so much encroachment, as it is bounded by the hard Niagara limestones of the Manitoulin islands, and the Indian peninsula of Canada, the strata dipping down beneath the lake. Yet it must be noticed that these rocky shores are indented by nu- merous deep bays transverse to their directions, The North Channel. This is generally a shallow water, the greatest depth being only 204 feet. To the northward, we have the Huronian rocks forming the boundary. The islands, especially towards the eastward, and near the whole north shore, are generally composed of Trenton limestone. The southern margin of the channel, bounded by Manitoulin and the other islands, is often composed of Hudson river, more or less, shaly rocks, overlaid by the Niagara limestones (where not removed by denudation), constituting an escarpment facing the northward. In fact, the whole of the north channel is principally scooped out of the Hudson River Formation, which attains a considerable thickness in this region. Dr. Robert Bell states that he has observed fifteen anticlinal folds trav- ersing the group of the Manitoulin islands ; and it is in these that we find the great indentations, and lakes in the islands, as well as the straits which separate them. Doubtless many of the southern ends of the Manitoulin lakes and channels are filled with drift. For example, the mouth of South bay is only 33 feet deep, whilst the upper portion is generally deep, one place giving a sounding of 156 feet. In fact, the north channel may be considered as a broad continuation of the Spanish river westward. The Mississagua river points directly to Mississagua straits, which are 204 feet deep, as deep as any part of the channel itself. Thessalon river has a direction towards False Bay De Tour, which is 186 feet deep. Vermilion river flows amongst the islands west of Cloche mountains and probably had a connection to Lake Huron through some of the buried channels across Manitoulin islands, as between Manitou- aning and South bays. The narrow channel between the peninsula of Cloche mountains and Manitoulin island is less than 60 feet deep and appears to be a modern con- nection with Georgian bay. Georgian Bay. The eastern and northern margins of this bay are com- posed of crystalline rocks; the south-eastern, of Trenton limestone ; whilst the western is made up of the Hudson River shales capped with the Niagara limestone, on the Indian peninsula and Manitoulin island. This basin is principally excavated out of the Utica shale, and the somewhat harder rocks of the Hudson River Formation. It lies along the junction of the various formations, and thus on the removal of the lower soft layers of rock, an escarpment was produced which has subsequently and slowly continued to be undermined. In the channels connecting this bay with Lake Huron, there are many 1881.] 319 [Spencer. small islands separated by shallow water. Yet in one place there are two deep soundings reaching 51 and 42 fathoms. The bay itself is much deeper than these passages, for there is an escarpment submerged to a depth of 498 feet immediately off the Indian peninsula, at Cabot’s Head, which is itself 324 feet above the bay. This peninsula is deeply indented with bays or fiords. There is a depression from the southern end of the bay, through the val- ley occupied by Simcoe, Balsom, Rice and other lakes, to the Trent river, emptying into the Bay of Quinté, an arm of Lake Ontario. This will be alluded to again. . IV. Tar Burrep Rrver CHANNEL IN THE DUNDAS VALLEY AND ITS Ex- TENSIONS. That the Dundas valley is that of an ancient river valley now buried to a great depth with the débris produced in the Ice Age, becomes apparent on a careful study of the region. However, untila key was discovered the mystery of its origin was found to be very obscure. My own labors at studying this region may fairly be stated as the first systematic attempts at the solution of the present configuration of the western end of Lake Ontario and the adjacent valley. Assertions have been made that it ‘was scooped out by a glacier, but this wild hypothesis was only a statement made without any regard to facts. ; From the description of the topography, given in Section II, of this paper, it will be seen that the apparent length of the rock-bound valley is six miles, with a width of over two miles; then it widens suddenly to four miles (with concave curves on both sides), after which it gradually increases in width as it opens into Lake Ontario. The direction of the axis of the . valley is about N. 70° E. The summit edges of the rock-walls on both sides are sharply angular and not rounded@dor truncated. This angularity is not due to frost action since the Ice Age, to any extent, as is shown by the character of the talus. The rocks of the summit are frequently covered with ice markings, but I am not aware of any locality where they have been observed as being parallel with the true direction of the valley, but on all sides one can observe them (sometimes at only small angles of Jess than 30 degrees) making conspicuous angles with its axis. One exception may be made to this statement. On a-projecting ledge of Clinton limestone, at Russel’s quarry, near Hamilton, at a height of 254 feet above the lake, and 134 feet below the summit of the ‘“‘ mountain,”’ after the removal of some talus, I observed that the surface was polished, but with scratches so faint that they could scarcely be compared with those of fine sandpaper on wood ; and the direction, if determinable, was parallel with the overhanging es- carpment. There are many tributary cuzons, which are evidently of greater antiquity than the Ice Age, which could not have excavated by the present streams, and are at all sorts of directions compared with the striated surface of the country. The topography of the lower lake regions precludes the idea of a glacier flowing down the valley to the north-eastward. Again, as the direc- Spencer. ] 320 {March 18; tion of the ice was towards the south-west, the waters from the melting glaciers could scarcely flow up an escarpment many hundreds of feet in height. Even if the Niagara escarpment did not exist elsewhere, the non- parallelism of the strie, and edges of the escarpment with their angular summits, is sufficient to prove the non-glacial origin of the valley in the hard limestone rocks. Moreover, at the eastern end of the narrower por- tion of the valley, there are two concave curves facing the lake, which of necessity would have been removed if such a gigantic grinding agen had been moving up the valley. This glacier-origin of the valley being an absolutely untenable hypothesis, I sought for some fluviatile agent capable of effecting the present configura- tion of the region. At the time, no idea occurred that even the great valley of the present is only a miserable remnant of one of gigantic proportions obscured by hundreds of feet of drift. The question arose, could Lake Erie have ever emptied by this valley? This suggestion did not hold its ground for any length of time, because the present levels are all too high. Near Galt, the traces of the true origin first presented themselves. A branch of the Great Western Railway extends from Galt southward for about four miles in the valley of the Grand river, after which, without making any im- portant ascent, it passes into the broad older valley, described above as that in which Fairchild’s creek now flows. After a careful examination of the ™ region, and of the railway levels, I came to the conclusion that this was an old buried valley. It then became apparent that if the Grand river had occupied the site of the Fairchild’s creek, that the latter probably flowed down the Dundas valley, and that the Grand river, being one of the largest of the rivers of Ontario, might have been a sufficient cause for the great excavation at the western end of Lake Ontario. Having procured all the levels that bore on the subject which were available, it became necessary to connect several places myself by instrumental measuremeuts, which work was accomplished last July, with the aid of Prof. Wilkins. As the whole floor of Niagara limestones is absent, as has previously been shown, the proof that the ancient Grand river flowed down the Dundas valley was completed, and of this discovery there was published a local notice last August. Significant and interesting as this fact was, relative to the change of systems in our Canadian drainage, a still more important issue was involved. When taking the levels between the Dundas valley (modern) and the Grand river, it was found that the whole calcareous floor was removed from a basin several miles in width, and that all the wells were sunk to a considerable depth in the drift before water could be obtained. On glancing at the map it will be seen that the Grand river from Brantford to Seneca meanders through a broad course, which in its ancient basin is several miles in width, but that from Seneca the valley is narrower, and the course of the stream more direct, as far as Cayuga. At Seneca the valley is two miles wide, and seventy-five feet deep. Also the bed of the Grand river at Seneca is in drift which is only 37 feet above the lake into which it now empties. As has been pointed out in the section 1881.] 321 (Spencer. on the topography, this broad valley continues to Cayuga within a few miles of the lake, whence its former probable course was by a nearly direct line to Lake Erie, now filled with drift, near the present bend in the river towards the eastward. At Cayuga, the rock beneath the drift-bed: of the river is below the lake level, on the margin of its ancient valley. (See note, Section II.) " ' Having observed the connection between the Dundas valley, Grand river and Lake Erie, it dawned on me that I had established the knowledge ofa channel having a very important bearing on the surface geology of the lake region. It now became apparent that Lake Erie had flowed by the Grand river reversed to a point west or north-west of Seneca, and thence by the Dundas valley into Lake Ontario; also that the upper waters of the Grand river, previously discovered as passing down the Dundas valley, were really tributary to the outlet of Lake Erie, and joined it somewhere south of Harrisburg ; and that the basin between the Brantford (and the Grand river of to-day) and the Great Western Railway, at Cope- town, formed an expanded lakelet along the course of the ancient outlet of Lake Erie, scooped out of the softer rocks of the Onondaga Formation be- fore noticed. As the waters excavated a bed in a deeper channel, ef course this lakelet would become an expanded and depressed valley, such as we often see amongst the hills of drift, at a short distance westward of Dun- das. Possibly the Grand river divided and flowed around an island, the western side of which is occupied now by the town of Paris. At any rate, Neith’s creek, at that town formed a large tributary to the river then flowing down to Lake Ontario. Along the course from Cayuga to Lake Ontario all obstacles to the outlet of Lake Erie appear to be removed. But along the present course of the Grand river, eastward of Cayuga, the waters flow over Coniferous lime- stone. But this difficulty is removed on observing that the river, filled with drift, approaches Lake Erie to within a direct distance of about six miles, but that at this place it leaves its southward course and also its con- spicuous valley and flows eastward, in the same manner as the Niagara river, above the Whirlpool, left its old choked-up outlet by the valley of St. David, and cleaned out a new channel for itself through several miles, in hard rock, from Queenston southward. _ We have recently seen by a note in the second section of this paper, that the Grand River bed is near the eastern margin of its ancient valley at Cayuga. From northward of this town, at about half a mile to the west- ward of the river, a deep depression in the drift indicates the deeper por- tion of the ancient river as it left the modern channel direct for the Lake Erie basin. Also along this route the hard rock is known to be absent to a depth below the surface of Lake Erie. - In Ohio; the Geological Survey considers that Maumee river emptied into the Wabash. If the waters of Lake Erie ever passed by this route into the Mississippi river when they were at no higher level than at present, then there must be a channel buried to a depth reaching at least 170 feet above Spencer. ] J22) the lake, as that is the elevation of the divide between the mibee waters of these two rivers. The outlet of Lake Erie, indicated i in this paper, is s known enamel along its route to have no rock-bed for a distance below the surface of the higher lake, and to a probable depth shown below ig Re pee to empty Lake Huron. ues Again, Mr. Carll has shown that the Allegheny draidele passed skies re Dunkirk into the Erie basin at a place just opposite to 0 its Ones as a nt by the present writer. - . mo Much of the Dundas valley is underlaid by stratified Erie clay, which is known to extend toa depth of 60 feet below the surface of Lake Ontario, _ In the upper part of the valley, eens according to Dr. Rob have exposed some deposits of unstratified clay filled with angular shingle, 3 4 ' 1 — ~—— : i et. AStaly St Sits ear) Be , os TA ee see “Ne Gar ; - By - et tear ion Meee cal Lee ee % “ Pe es 4 Fi oA > : a uemroaGee, oH: aE We = = cA r - rage Se Se my - . . - r rn ‘ . ; > nas ie be > ee oR ST ee eres en on A Pc a! As henna Saale = + ‘tie, Dba eS “7 . “ EP ER LETTE PO Ns SM | he aw knee ioe x a, De Oe en, , . . ~ t 4 ] . 4 a | : | P , J “ : - F a ) : : bie ao i s, : : | ¥ é us tert as rs - ation os = | _— ‘ - . r ; x ay | : : - S _ a o