A- ISV. T O -1 — UJ CO o I- d o cr g P 2 THE Duration of Niagara Falls AND THE HISTORY OF THE GRE^T LAKES. By J. W. SPENCER, A. M., Ph. D., F. G. S. Author of " Reconstri'ction of the Antillean Continent, " Etc. 1881-1894. SBCOND EDITION. NEW YORK : The Humboldt Publishing Co., 64 Fifth Avenue. COPYRIGHT By J. W. SPENCER. 1895. a^s/ PREFACE. To the encouragement of Prof. J. P. Lesley, State Geologist of Pennsylvania, is largely due the author's interest in the investi- gations into the history of Niagara Falls and the Great Lakes, chapters of which have appeared from time to time in the lead- ing scientific journals, the first of which was printed in the pro- ceedings of the American Philosophical Society of Philadelphia, in 1881. Since that time more complete chapters have ^ appeared. These have been compiled into their natural (although not appearing in their chronolgical) sequence owing to the pro- found interest taken in research by Andrew H. Green, Esq., the president of the Commissioners of the State Reservation at Niagara, and are here published in order to give a scientific istory of the great cataract of the world, so that the informa- tion may extend beyond the limit of a few specialists. The full history of the lakes has not yet been told, but enough is known <}^ to write a somewhat complete story of the Falls. To Mr. F. B. "; Taylor is due a special acknowledgment, among other contribu- "^ tions, for having recently given us details of the drainage of -^ the Huron basin by way of the Nipissing Straits and the Ottawa J^ valley before the waters of the upper lakes changed their ^ drainage to the Niagara river, and thus accelerated the recession ^ of the FaUs. ^ THE AUTHOR. WASHiNGTOif, D. C, December 1, 1894. (3) /\ nrM2o^^ C O N T E N T SI PAGE. Preface 1 Chapter I. The Evidence of High Continental Elevation during the Formations of the Valleys of the Great Lakes 7 II. Origin of the Basins of the Great Lakes 14 III. Ancient Shores, Boulder Pavements and High Level Gravel Deposits in the Region of the Great Lakes 27 IV. Deformation of the Iroquois Beach and Birth of Lake Ontario. Appendix — The Iroquois Beach North of the Adirondacks 44 V. Deformation of the Lundy Beach and Birth of Lake Erie. 58 VI. Deformation of the Algonquin Beach and Birth of Lake Huron 64 VII. High Level Shores of Warren Water (Gulf) and their Deformations 74 VIII. Pleistocene Subsidence versus Glacial Dams. Appendix — Channel Over Divides Not Evidence per se of Glacial Dams 85 IX. The History and Duration of Niagara Falls 99 (5) ILLUSTRATIONS. PAGE. Plate I. Camera obscure drawing of the falls in 1832 Frontispiece II. (9) Boulder pavements on Georgian bay 37 (10) Ancient boulder pavements of Algonquin beach 37 III. Niagara Falls from the Canadian side 99 IV. Niagara gorge below the Whirlpool 101 V. Whirlpool rapids 109 Figure 1. Map of the Giilf of St. Lawrence 10 2. Map of the ancient Laurentian river and its tributaries.. 15 3. Section of Cut-Terrace and beach 28 4. Section of Construction Terrace 28 5. Section of Cut-Terrace with boulder pavement 29 6. Plan of Burlington beach 29 7. Section of Cut-Terrace, with beach concealed 30 8. Plan of Barrier beach in front of the lagoon 32 9. In Plate II 37 10. In Plate II 37 11. Section of gravel ridges 43 12. Map of the Iroquois gulf 47 13. Map of the Lundy gulf 59 14. Map of a portion of the Algonquin gulf 66 15. Deserted strands of Warren gulf ... 75 16. Section across the St. Clair valley 79 17. Map of Niagara cafion 103 18. Map of Whirlpool 104 19. Section across Whirlpool ravine 105 20. Map of recession of falls 106 21. Map of Foster's flats 109 22. Section across Foster's flats 109 23. Section across end of gorge 110 24. Longitudinal section of Niagara gorge Ill 25. Section across narrows of gorges 112 26. Section across gorge at Johnson's ridge 113 27. Section of gorge near Horseshoe Falls 113 (6) CHAPTER I. The Evidence of Continental Elevation during the For- mation of the Valleys of the Great Lakes.* If, in the growth of the American continent, the moulding of the land features had not largely depended upon its projection above the sea, favoring or retarding the action of rains and rivers in sculpturing its surface, there would be li tie interest as to what was its relative height, before the commencement of the Pleistocene period. But we find valleys vastly greater than the meteoric agents could have pro- duced under existing conditions Thus, there are not only deep canons, but also vast depressions, descending to levels far below the sea, which are now filled with the earlier drift accumulations, or form channels submerged beneath ocean waves, or constitute basins occupied by lakes. Hence, in the study of the drift itself, in the investigation of the lake history, or in the research upon the growth of modern rivers, we necessarily inquire what was the altitude of the continent that would permit of the mouldings and channelings of the original rock surfaces. Following the period of high continental elevation, the geologist sees in the valleys and old channels, still below the level of the sea, and in the high level beaches, extensive submergence, succeeded by a re-elevation, but not to the original height, when the continent was being chiseled out by the ancient rivers. That this re-elevation is still going on is shown by the northward tilting of the comparatively recent marine accumulations along the St. Lawrence valley and Gulf coast, and the raised beaches in the lake region, as well as by the shoaling of the waters of Hudson's Bay during the present period of observation. As general statements do not satisfy investigation, it becomes necessary to search for definite measurements of the former height of the continent among the archives of the geological past. Let us first seek for the testimony recorded by the Mississippi river. •Reprinted from Bcll. Gkol. 8oc. Am. vol. I, pp. 65-70, 1889, where the text appears under Htle of "The High Continkntai. Elevation Precedino the Pleibtocbmb Period." (T) 8 BiJKiED Channel of the Mississippi. For a distance of 1,100 miles, measured in a direct line, above the mouth of the " Father of Waters," the modern valley is merely- maintaining its own size, or more generally is being slowly filled by the deposition of river alluvium upon its floor. There are only two exceptions, of a few miles each, where the river is scouring out the rocky floor, and these are over barriers recently exposed there during the changes of the Pleistocene period. To such an extent has the ancient valley or canon been filled, first with drift, and this covered with river alluvium, that its original rocky floor is now buried to a depth of 170 feet, even at La Crosse, a thousand miles from the Gulf of Mexico.* Farther south the depth of these loose deposits increases, until at New Orleans a boring of 630f feet below sea level does not penetrate the southern drift, nor even reach to its lowest members. The lower 500 miles of the ancient Mississippi were excavated out of Eocene or Cretaceous deposits, while the valley above the mouth of the Ohio has the form of a caiion, excavated out of Paleozoic rocks, varying in width from ten to two or three miles, and having a depth (exclusive of the portion now filled) of from 150 to 550 feet, according to the late General G. K. Warren. From this inspection of the river, it is easily seen that no natural rainfall could so increase the volume of the discharge as to remove all the deposits which now fill the old valley, much less excavate the original and immense canon. A vastly greater elevation of the conti- nent was necessary. Even were the whole continent uniformly elevated 630 feet, together with the remainder of the unknown depth of the ancient Mississippi river, at New Orleans, the canon of the upper part of the river would require a still greater relative elevation of the northern country in order to give suflicient channeling power to the flowing waters; but the slope of the floor of the partially buried valley is much less than that of the modern, as was formerly shown by the author. I Here, again, is the proof that the country drained by the upper waters of the Mississippi once stood, relatively to that in the region of its mouth, much higher than at present. Of the amount, which was at least many hundreds of feet, we have no absolute measurement; nor can we ascertain it by calculation, for there is no register of the excess of the amount of rainfall during the epoch of the greatest sculpturing over that of the present day. *Geol. wis., Vol. I, 1883, p. 253. tE. W. Hllgard, Am. Jour. Sc, 2nd Ser., Vol. XLVIII, 1869, p. 333. $ Am. Nat. Vol. XXI, 1887, pp. 168-71. Peofotjnd Submergence of Yalleys. 9 While these records of the Miseissippi, which have been only partially deciphered, do not furnish all of the desired information, yet as far as they go they are invaluable. Passing from the buried channel of the Mississippi to its continua- tion, now submerged beneath the waves of the Gulf of Mexico, we find evidence indicating such a stupendous continental elevation as to be almost incredible, were it not supported by collateral evidence, upon both the Pacific and Atlantic coasts. The soundings off the coast of the delta of the Mississippi indicate the outer margin of the continental plateau as submerged to a depth of 3,600 feet, indented by an embay- ment of another hundred fathoms in depth, at the head of which there is a valley a few miles wide, bounded by -a plateau from 900 to 1,200 feet above its floor. This valley is now submerged to a depth of 3,000 feet, and is the representative of the channel of the ancient Mississippi river, towards which it heads.* On the Pacific coast, in the region of Cape Mendocino, Prof. George Davidson has identified three valleys now submerged to from 2,400 to 3,120 feet, and several of inferior depth. These measurements are those of the valleys where they break through the marginal plateaus of the continent, at about six miles from the present shore, where it is submerged to the depth of 100 fathoms. f The soundings along the Atlantic coast reveal similar deep fjords. The long-since known extension of the Hudson river, beneath the Atlantic waters, is traceable to the margin of the continental plateau, acquiring a depth of 2,844 feet, in front of which the soundings show a bar, covered with mud, which, however, is now submerged to the depth of only 1,230 feet. The unpublished soundings off the mouth of the Delaware river bring to light another valley, the floor of which is now covered by ocean waves to nearly 1,200 feet — its continuation seaward not having been ascertained. (Lindenkohl.)J Were the continent elevated only 600 feet, the Gulf of Maine would be replaced by a terrestrial plain, in some places 200 miles wide, but traversed by rivers, one of which, towards its mouth, would be 2,064 feet deep — that is to say, the bottom of the fjord is now submerged 2,664 feet. Even this great depth may not be its maximum, for along the line between the opposite banks, at the mouth, now beneath 100 fathoms of water (which is approximately the depth to which the real margin of the continent is submerged), we find that the sea is nearly * J. W. Spencer, "The Mississippi River During the Great River Age," New Haven, 1884, p. 2. tGeo. Davidson, Bull. Cal. Acad. Sc , vol. II, 1887, p. 265. t Appendix 13, Rep. U. 8. Coast and Geodetic Survey for 1887 (1889), pp. 270^73. 10 SUBMEEGED LaURENTIAN YaLLET. 5,000 feet deep. Whether this represents an erabayment of the ocean setting towards the valley or a continuation of the fjord is not determined. The St. Lawrence river and gulf bear the same testimony of the existence of deep fjords extending from the rivers through the now submerged plateau forming the margin of the continent; and the lower part of Saguenay river flows between stupendous walls and constitutes Peofound Depth of the Great Lakes. 11 a fjord whose waters reach a depth of 840 feet. In the St. Lawrence river, a little below the mouth of the Saguenay, there is a channel 1,134 feet below the surface. This increases in depth in passing sea- ward. In the region of the centre of the modern gulf, the floor of the old channel is now submerged 1,878 feet, and the adjacent valley 1,230 feet; thus showing the caiion as being over 600 feet deeper. As at the mouth of the channel through the Gulf of Maine, so at the mouth of that of the St. Lawrence, there is a deep chasm; for inclosed between the banks, 100 fathoms below the surface, there is now the depth of 3,666 feet, with water 2,t00 feet deeper just seaward of it. Although this ancient valley is over 60 miles wide at its mouth and was a narrow channel, yet it is not as broad as some portions of the modern so-called river. The breadth of the submerged valley through- out its windings for a length of 800 miles or more, is remarkably regular, only gradually increasing its magnitude in passing seaward. Other and smaller channels are visible in the soundings: thus, south of the Straits of Canso, between Nova Scotia and Cape Breton island, there is one 1,200 feet deep, according to the British admiralty charts, while adjacent soundings show less than 600 feet of water. Hudson's Bay rarely exceeds a depth of 600 feet, yet at the outlet the channel is 1,200 feet deep. This depth increases in passing down the straits, where the scanty soundings show 2,040 feet before reaching the mouth. Here, in Hudson's Straits, the old valley is a chasm across a mountain system, whose peaks, upon the southern side rise to 6,000 feet above tide. The canon of the St, Lawrence also crosses the trend of two mountain systems, but these are of no great height. The same is true for any of the other submarine valleys described. The record of a former high continental elevation is again inscribed in the depths of the Great Lakes — Ontario reaching to 491 feet below ocean level, Superior to nearly as much, Michigan to 800, and Huron to 150 feet. The lake basins are merely closed up portions of the ancient St. Lawrence valley and its tributaries. Their distance from the sea would necessitate not merely a general elevation of the conti- nent, but also a greater amount of elevation towards the head-waters of the system, as has been shown with regard to the excavation of the upper portion of the ancient Mississippi caiion. The lake basins are all excavated out of Paleozoic rocks, except a part of that of Lake Superior. The soundings do not afford all the information that we desire, yet they demonstrate the presence of submarine valleys reaching upon all our coasts to depths of 3,000 feet or more. Again^ the soundings 12 Origin of the Ancient Yallets. show that within comparatively short distances from their mouths the depth of the valleys, below the surface of the seas, sometimes did not exceed from 1,200 to 1,800 feet, but that beyond, there was a greater increase in depth, within the last few leagues. While depressions in the earth's surface are made and modified by terrestrial crust movements, yet the leaving open of great, yawning chasms is not ol sufficiently well known occurrence to attribute all of the submerged valleys upon the American coast to such an origin, especially when we consider the great length of the submerged channel of the St. Lawrence river (800 miles), its various windings and its uni- formly increasing size, until it passes into the great chasm, just before it reaches the margin of the continent. The idea of the excavation of these submerged valleys by glaciers — some of which are outside of glacial regions even of the past — is too untenable for a moment of serious consideration. Irrespective of the causes which have deter- mined the location of the channels here described, it appears that they have been made, one and all, by the excavating power of rivers and lateral streams pouring down the hillsides. These, together with the other meteoric agents, have also, to a greater or less extent, removed the Paleozoic, and also the Triassic rocks, from the depressions now occu- pied by the Gulfs of St. Lawrence and Maine, which have, however, been more or less affected by terrestrial movements. The length of time required to excavate the channels of these great rivers commenced as far back as the Paleozoic days. However, the culmination of that of the Mississippi was not until the later Tertiary, or even the Pleistocene period. As the St. Lawrence valley, now sub- merged to a depth of over 1,200 feet for a distance of 800 miles, is mostly cut out of rocks of the Paleozoic group, except a belt of the Triassic (across the lower portion, more or less involved in mountain uplifts), its antiquity must be very great. The culmination was also probably in the later Tertiary era, like that of the Mississippi and the channels on the California coast, for there are submerged Tertiary rocks off the coast of Massachusetts and Newfoundland, at elevations much higher than the beds of the old channel. Although the excavating forces took so many periods to form the valleys, and required a high continental elevation, yet the extreme altitude of over 2,000 feet appears to have been of comparatively short duration, for otherwise the deep chasms in which the sub- merged channels terminate would have extended farther inland than we find them, and would have been headed by more gentle slopes, in place of precipitous cliffs, over which the waters of the former rivers Resemblances to the Fjobds of Nokwat. 13 were pecipitated in great cascades. In the fjords of Norway, merging into rapidly contracting valleys, or headed by great vertical walls, hun- dreds of feet in height, having the structure named cirques, may be seen to-day the counterpart of the coast of the American continent, when its marginal plateaus stood 3,000 feet higher than at present; yet Nor- way stood once much higher than now, but was afterwards submerged, from which depression it has only recently been re-elevated so that its plateaus, close upon the sea, rise to a height of three or four thou- sand feet, and its mountains still higher. The old hydrography is more or less distorted by warpings of the earth's crust, which, however, do not obscure the valleys, although rendering the features somewhat more complex. The amount of distortion has yet to be determined. CHAPTER II. Origin of the Basin of the Great Lakes of America.* Intkoduction. EvBN as recent as a decade ago very little was known as to the origin of the Great Lakes of America. While we find such o-eneralized statements as "most lakes are due to terrestrial crust- movements," yet such crust-movement had not been tested in the American lake-region. Again, from the time of early geological inves- tigations in America, statements are found that the basins were the result of erosion ; but the methods of erosion were not explained, and this was the more necessary as most of the basins have rock-bound outlets. Later, in some geological literature, the method of excavation was hypothetically attributed to glaciers. Such was the unsatisfactory condition of our knowledge of the problem when the writer first com- menced the study, in attempting to solve the origin of Dundas Valley, at the western end of Lake Ontario, more than a dozen years ago. This investigation has developed results bearing not only upon the origin of the lake-basins, but also upon the physical history of the lakes, and broader questions of the building and sculpturing of the continent. The methods of investigation have been the studying — (1) of the hydrography of the modern lake-basins and submerged channels upon the coast of America ; (2) of the deep wells bored into, or through, the drift deposits, by which buried channels, and their relation to or con- trast with the modern valleys, have been discovered ; (3) of the elevation of the continent ; (4) of the direction of the glaciation in the lake- region ; and (5) of the now high-level beaches, in which are recorded continental uplifts, together with the deformation of the old surfaces, owino- to unequal terrestrial movements or warpings of the earth's crust. t The lakes which have been the basin of the more careful investigation are Ontario, Erie, Huron, and Michigan, with the respective altitudes of 247, 573, and, of the last two, 582 feet above the sea (see the Map, p. 15). * From the Quarterly Journal of the Gk u^gical Society of London for November, 1890, Vol. XLVI, pp. 523-533. + In the field-work I here acknowledge the assistance of Professors D. F. H. WilMns, W. W. Clendenin, and W. J. Spillman. (14) Ancient Laubentian Kivee. 15 16 The Ontario and Ebib Basins. Features of the Ontario Ba^sin. Lake Ontario, as was shown in an earlier publication,* is a basin bounded on its southern side by escarpments, often precipitous, of which some of the steps are now submerged. At the foot of the sub- merged escarpments a valley like that of an ancient river may be recognized from the western part of the lake to near the eastern end, but there it disappears, for reasons to be noted later. The deepest part of this valley is 738 feet beneath the surface of the lake. From this trough the floor of the lake rises gradually, or with occasional low steps, to the northern shore. In short, the basin was once an old land- valley traversed by a river. At the western end of the lake borings have revealed an old channel, having a lateral depth of 292 feet. This is the continuation of the canon of the Dundas valley, which is about two and a half miles wide, bounded by rocky walls nearly 500 feet high, capped with Niagara limestone. Down this valley the waters of the ancient Erie basin once flowed, f If the waters of Lake Ontario were withdrawn, its present basin would be a broad valley, continuous with that of the St. Lawrence val- ley, having a breadth of 30 or 40 miles. Into this plain, at a point about 20 miles east of Toronto, there is a channel, approaching the shore, whose bed is 4Y4 feet below the surface of the lake, \ but with boundaries submerged to only 200 feet. This depression trends south- ward and joins that at the foot of the submerged escarpment before mentioned. § features of the Erie Basin. The floor of Lake Erie is a broad flat plain, now rarely submerged to a depth of more than 84 feet, and usually less. Only a small area, situated directly south of the western end of Lake Ontario, is of greater depth, and there the greatest sounding is 210 feet. || But from this region the Erie valley was drained by the Grand river and Dundas Valleys into the western end of Lake Ontario, as was shown in 1881; for the Niagara river did not then exist. Numerous tributaries of the modern shallow lake flow over deeply buried channels, the deepest of those discovered being 228 feet below the lake surface, as described by Dr. Newberry, ^ although the floor of that portion of the lake is nowhere over 84 feet below the surface of the water. * "Discovery of the Preglacial Outlet of the Basin of Lake Erie into that of Laiie Ontario/ by J. W. Spencer ; Proc. Am. Phil. Sec, Philad. 1881. + See " Discovery of the Preglacial Outlet of Lake Erie," etc. X See "British Admiralty Chart of Lake Ontario." § See U.S. Lake-Survey Charts of Lake Ontario. I See U.S. Lake-Survey Charts of Lake Erie. ^ IT Geology of Ohio. Hdbon and Michigan Basins. 17 Similar channels, buried to depths below the floor of the eastern end of Lake Erie, near Buffalo, have been described by Dr. Julius Pohl- mann. * The borings into many others in the region of the western end of the lake have been recorded by Prof. T. Sterry Hunt,f and prove the existence of similar buried channels. The original recognition I of the vallej^-like character of the basins of Ontario and Erie was based upon the above-mentioned characters^ and upon others now supplemented by a more perfect collection of facts; but the greatest difficulty in the way of explanation was in the occurrence of the rock-bound outlet of Lake Ontario, — a difficulty which observations have at last dispelled, as will be seen later on. Features of the Huron Jiasi?K The southern half of Lake Huron is a plain traversed by valleys and submerged to form only a shallow lake. Northward of this shallow basin, and extending obliquely across the lake for 90 miles, there is a submerged escarpment rising to a height of from 300 to 450 feet, facing northeastward. The deeper part of the lake then trends north- ward in the direction of Georgian Bay. At one point the extreme depth of the submerged valley reaches 750 feet. The absolute depth of the rock in the deepest channel between Lake Huron proper and Georgian Bay is not known, but soundings show 306 feet; and as there is a deep channel upon the western side of Georgian Bay it becomes highly probable that a deeper and connecting channel is filled with drift, like those known to occur elsewhere, beneath the lakes.. From the straits, between the islands, the narrow channel in Georgian Bay, just referred to, extends southeastward and is submerged to a depth of 51(» feet. This is at the foot of the Niagara escarpment,, which extends, as a strong topographic feature, from the head of Lake- Ontario, and, rising in places to 1,700 feet above the sea, into the pen- insiila between Georgian Bay and Lake Huron proper. The channels at the foot of escarpments, submerged or otherwise, in Lake Huron and Georgian Bay are fragmentary records of the history of the lake- valleys. 11 Features of Lake Michigan. This lake is divided into two basins. The more northern and larger basin has a maximum depth of 864 feet. It is, in part, bounded by * Papsr read before the A.mer. Assoc. Advance. Science, 1883. + See Reports Qeol. Canada, 1863-66. % See " Discovery of the Preglacial Outlet of Lalce Erie," etc. II See U. S. Lake-Survey Chart of Lake Huron, and the Canadian Chart of Georgian Bay.. 3 18 BcTRiKD Valleys South of Geoegian Bay. vertical submerged escarpments, one of which, upon the eastern side, has a height of 500 feet. While the deepest sounding at the modern outlet of the lake is only 252 feet, there are adjacent channels buried to unknown depths. But these have been imperfectly explored. Into this shallower portion of the lake, however, the fjord of Grand Traverse Bay has a northernly trend, — it is 612 feet deep. This and the smaller fjords indicate the existence somewhere of a deep channel connecting with the Huron basin, as much as the river-valleys buried beneath the drift materials of the modern floor of Lake Erie prove deep channels throughout that basin, although not shown by the sound- ings; for the Lake-Michigan valley is carved out of undisturbed and almost horizontal Paleozoic rocks, the newest of which are Coal Measures. The southern basin of Lake Michigan is separated from the northern by a plateau submerged to a depth of from 300 to 342 feet; whilst the southern basin itself is now 576 feet deep. The area of this portion of the basin is now much smaller than that of the Prepleistocene valley, as its margins have been filled with drift, and now forms broad plains bounding the lake. Beneath these deposits is a deeply buried channel, leading to the valley of Lake Huron, and to be noted further on. Buried Valleys Bevealed by Borings. The deep wells revealed the existence of the buried channel down, which the waters of the Erie Valley orginally drained, and thus estab- lished the relationship of the Erie with the Ontario Basin. But the most important series of borings were those between Georgian Bay and Lake Ontario, for here we have the connecting link between the valleys of the upper lakes and that of Lake Ontario, and indeed the key to the origin of the valleys of the lakes. Between Georgian Bay and Lake Ontario, a distance of about 95 miles, a portion of the country is comparatively flat composed of a series of rising plains; but there are also high transverse ridges of drift, having a general trend of east and west. It is upon the northern side of the drift ridges that Lake Simcoe, with a diameter of about 20 miles is situated. But upon the northern side of Lake Simcoe there is another series of drift ridges trending towards the northeast. Both of these series of ridges rise to bet wee q 200 and 550 feet above Lake Huron, these measurements being the extreme variation in their height. From Georgian Bay to near Lake Simcoe, for a distance of 30 miles, the country is low and flat, with a known absence of rock to far below Buried La.ukentian Yalley. 19 the level of the bay. Lake Simcoe is 140 feet above Georgian Bay, but upon its northern side, at Barrie, a well has been sunk in the Drift, without penetrating it, to a depth of 280 feet below its surface. Thirty miles further inland, south of Lake Simcoe, at Newmarket, a well was in the process of being bored. It had reached a level below Georgian Bay and was yet in d rift deposits when visited. In another well several miles to the westward, near the side of the ancient buried valley at Beeton, rock was reached at 50 feet below the surface of Georgian Bay. Between Newmarket and Richmond Hill there are several deep wells on the heavy drift ridges which cross the country. But at Richmond Hill, at a height of 217 feet above Georgian Bay, there is a well 400 feet deep without penetrating the drift. This proves the thickness of the Drift of the higher ridges crossing the old valley north of the well to be not less than VOO feet in the old channel. Southward of Richmond Hill the country falls away in a series of more or less rolling steppes to Lake Ontario, but these plains show the absence of rock along deeply-cut valleys to far below the level of the upper lakes. Upon the western side of this chain of borings, but a few miles distant, there is the Niagara escarpment. Upon the eastern side of Lake Simcoe the country is covered with flat limestones, rising to 150 feet above that lake. From the known absence of rocks along the line of borings and stream excavations, between a high mountainous escarpment upon one side and a rocky floor upon the other, and from these borings reaching to 200 feet or more below the upper lakes, without penetrating the drift but stopping in quicksand, there has been discovered the existence of the only channel of antiquity which could now draw off- the waters of the upper lakes, if the drift were removed. Although none of the borings have reached the original rocky floor, yet the depth of the buried valley is suggested by the channel close upon the northern side of Lake Ontario, now submerged to 474 feet, which is deep enough to drain the last drop of water out of Lake Huron. We have now found the ancient Laurentian channel from Lake Mich- igan through Lake Huron and Georgian Bay, and thence buried beneath drift deposits until it is again recognizable throughout nearly the whole length of Lake Ontario, being joined at the western portion by an ancient outlet of the Erie valley (the ancient Erigan River). But the relative maximum depression of the channels, as far as explored, is disturbed by terrestrial warpings, to be described hereafter. Across the southern part of the peninsula of Michigan, between hills risirg upon either side to heights of sometimes 800 or 1,000 20 BOEIED HUKONIAN RlVER ChANNEL. feet above Lake Huron or Lake Michigan, there is a valley whose western portion is occupied by tlie Grand river, and the eastern by a small river emptying into Saginaw Bay. At the divide between these rivers the land does not exceed 100 feet above the lakes. The tojDOgraphic features of the valley show its original opening as having been into the Huron Valley by Saginaw Bay; but a consider- able proportion of the modern drainage is in a direction opposite to that of the valley, or flowing towards Lake Michigan — that is, the drainage has been reversed. The maximum depth of the western portion of this buried valley is not known, but there is an absence of rock, as shown in several borings, to between 100 and 200 feet below the lake level. But farther east in this trough there are several deep wells, in one of which the drift is 500 feet below the floor of the side of the valley, or 350 feet below the surface of Lake Huron.* Hence we have established the great depth of the buried valley between the southern part of Lake Michigan and Lake Huron, whose ancient river I name the Huronian. Other buried valleys and channels "submerged could be given, but they all indicate the origin of the basins of the lakes as the valleys of a great river and its tributaries — a river of such high antiquity that the rains and rills had already ground off the surrounding hills to broaden the valleys. But for all this evidtnce, there are now rocky barriers forming an apparent obstacle in the way of a complete solu- tion of the problem. The Glaciation of the Region. At the present stage in the investigation this subject can be quickly dismissed. The question whether glaciers can erode great lake basins is hardly pertinent, for nowhere about the lakes is the glaciation parallel to the shores or vertical escarpments which are associated with the lakes. Indeed, the direction of the strife is often at high angles, even to 90°, to the trend of the vertical walls of rock bounding or crossing the lakes. Nor are the faces of these great walls of lime- stone polished by an agent moving along their faces. That there are no striae parallel to some local inlet or valley would be perhaps rash to assert; but, if so, it is a mere coincidence, with no bearing upon the origin or moulding of the Great Lake valleys. Hence we are forced back upon a conclusion that the lakes were subaerial valleys in spite of the barriers, and the fact that the floors of most of the basins are below sea-level — that of Ontario being nearly 5(i0 feet. *Thi8 is at the Sanitarian Well at Alma, Mich., the record being furnished by Prof. Charles A. Davis. Deformation of Raised Shobes. 21 Ihe Former High Continental Elevation of North America. If the lakes and valleys originated from atmospheric and river erosion, then the continent stood at much greater elevation than at present, as shown by the depth of the lakes themselves. But there is much collateral evidence that in the later Tertiary days, probably dur- ing the Pliocene, the continent was very high. This is shown by the submerged valleys of the St. Lawrence Gulf, of the Gulf of Maine, off New York, at the mouth of the Mississippi river, upon the Pacific coast, and in Hudson Strait. These indicate that eastern America stood for longages at between 1,200 and 1,800 feet above its present altitude; and the whole continent in more recent times, but for a brief period, at upwards of 3,000 feet.* Hence the former continental elevation was sufficient to satisfy all demmds for the erosions of the lake- valleys; but the rocky barriers still demand explanation, both on account of the present obstructions not having impeded the erosion of the valleys, and on account of their subsequent closing the valleys, in part, into lake basins — the necessary observations for the explanation having long eluded investigation. Deformation of Raised Shores and Beaches. At the close of the episode of the newest till, the region of the Great Lakes was submerged to a depth of at least 1,700 feet, as is recorded in the beaches which overlie the till. These high beaches only remain as fragments about ancient islands; but if we descend to beaches of lower levels we find them well developed and containing all the necessary evidence for explaining the rock-barriers at the out- lets of the lakes. Gen. G. K, Warren, Corps of Engineers, U. S. A. was the first to suggest the closing of the lakes by warpings of the earth's crust.f Portions of the high-level beaches about the lakes have long been noted. But it was Mr. G. K Gilbert who first connected the beaches upon the southern and eastern sides of Lake Ontario, and measured their great rise towards the northeast; but, as he did not apply his discovery to the explanations of the lake basins, it was first applied by the present writer.^ The results of Mr. Gilben's investiga- tion of beaches in New York and Ohio, and of the writer's researches in Canada, Michigan, New York, and elsewhere, are sufficient to form a chapter by themselves, and are still only in part published, but I will *"High Continental Elevation preceding the Pleistocene Period," by J. W. Spencer, Bull. Geol. Soc. Am. vol. 1, 1889; and Gen. Mag., May, 1890. t Appendix 13, Report of Chief of Engineers, U. 8. A , 1875. t See " Notes on the Warping of the Earth's Crust in its Relations to the Origin of the Basins Of the Great Lalies," Amer. Nat. Feb , 1887, pp. 168-71. 22 Deformation of the Ircquois Shoee. draw upon them only to the extent of explaining the barriers across the outlets of the old valleys. The most important raised beach of the Ontario basin is the IroquoU* At the western end of the lake it now rests at 363 feet above the sea, but rises slightly to the east and still more toward the north, until at four miles east of Watertown it is 730 feet above the sea. Still further northeastward, near Fine, on the borders of the Adirondack wilderness, it reaches an elevation of 972 feet above the sea, beyond which recent measurements carry it to 1,500 feet above the sea. At the western end of the lake the uplift is scarcely two feet in a mile in the direction of N. 28° E. At and beyond the northeastern end of the lake the uplift is found to have increased to five feet in a mil*', and in the region of farther observation to seven feet in a northeast- ward direction. Thus in the deformed water level I have already measured a barrier of about 600 feet raised up at the outlet of the lake. Of this, about 530 feet is confined to the region of and beyond the eastern end of the lake, where the later Pleistocene barrier across the ancient Laurentian valley has appeared. Wl ile we know what are the maximum soundings in the river, yet the old channels are 80 filled with drift that their depths are not revealed. Still, we know that in one portion of the channel cut out of limestone and more or less filled with drift, the sounding is 120 feet. A short distance beyond, the channel across the Laurentian gneisses shows soundings of 240 feet. The maximum depth of the lake-basins is 738 feet. The deformation recorded in the beaches is more recent than the episode of the upper till. Consequently, if the continent were at a high level, with the warping, known to have occured since the drift was deposited, removed, as shown by the above figures, there would be not only no barrier, but a sufficient slope in the Laurentian valley for the drainage of what is now the Ontario Basin. Furthermore, the presence of the rocky barriers of the Rapids of the St. Lawrence, further east, are wholly accounted for by the terrestrial warpings of the region. Hence I have demonstrated, after a decade of study, that no barrier existed across the Ontario valley when it was being carved out by the ancient St. Lawrence, and that this barrier is of quite modern origin. Southeast of Georgian Bay the average measured warping is four feet per mile, in mean direction of N. 20° E. This will account for a portion of the barrier closing the Georgian outlet of Lake Huron. The *" Iroquois Beach; a Chapter in the Geological History of Lake Ontario," Proc. Roy. Soc. Canada, 1889. Origin of the La.ke Basins. 23 more elevited beaches in the region of Lake Huron record a still greater change of level. At the outlet of Lake Erie, Mr. Gilbert and myself find a differential uplift of about two feet per mile, and this is sufficient to account for any rock basin to the recently formed basin of Lake Erie. The warping affecting the Michigan Basin has been that towards the north and east; and even in the hurried channels south of Lake Michigan there is no evidence of an ancient drainage to the south, as their beds were too high compared with those of the northern, although the latter have been elevated recently'' by warping. Conclusions from the Observations. The valleys of the great lakes here studied are the result of the erosion of the land-surfaces by the ancient St. Lawrence (named by the writer Laurentiaa) river and iis tributaries, during a long period of continental elevation, until the streams had reached their base planes of erosion, and the meteoric agents had broadened the valleys. This condition was at the mix mum just before the Pleistocene period. The clo-ing of portions of the old Laurentian valley into water basins occurred during and particularly at the close of the Pleistocene period, owing, in part, to drift filling some portions of the original valley, but more especially to terrestrial warpings of the earth's crust, which, to a sufficient degree, are measureable. Discussion.* The Chairman noted that there were one or two Fellows who had a local knowledge of the area, but the question of the origin of lake-basins in general was raised in the paper. Dr. Hinde did not think that Dr. Spencer's explanation of the origin of the American lake basins was the true one. The submerged deep channels of the alleged ancient rivers, to Avhich the erosion was said to be due, were traced towards the east end of Lake Ontario, where tbey ceased ; and their discontinuity through the barrier formed by the hard gneissoid region of the Thousand Isles was attributed to differential elevation, or so-called earth-warping. '1 his assumed warping where barriers existed could always be brought in to account for them. He (the speaker) asked where the beaches existed near Kingston, at the east end of Lake Ontario, on the difference of level of which the supposed warping was based. From his own observations on the region, he doubted the existence of the alleged buried channel between Lakes Huron and Ontario, and he did not think that the acknowledged * Before the Geological Society of London. 24 Discussion. great thickness of Drift now covering the elevated area between these lakes should be regarded as proof of the presence of a former channel directly connecting them. With reference to the supposed old channel between Like Erie and Lake Ontario, by way of the Grand river and the Dundas valley, the water of Lake Erie was supposed to have run up the valley by which the Grand river now came dovan to the lake. All these lakes and the elevated regions between them had been covered by glaciers, and their movement had been in a contrary direc- tion to that of the present water drainage ; and judging by the amount of drift material transported by the ice from lake-basins over the adjoining land surface, he believed that the glaciers had been important factors in their excavation. If, on the author's views, there had been a recent submergence to the extent of 1,700 feet, where were there traces of marine remains over the lake region west of the meridian of Kingston, though such were not uncommon in the clays of St. Lawrence and Ottawa rivers? Also on the author's hypothesis of a former great lake who^e surface would be at a considerable elevalion above the sea, what barrier was there at the south end of Lake Michigan, near Chicago, to keep such a lake from draining into the Mississippi valley ? Prof. Bonney thought that Dr. Hinde's criticism was not a valid one, as he had not understood that the author denied the occupation of the lakes by iae, though he did not uphold their glacial origin. He could not understand the formation of Georgian Bay by ice and the preservation of Manatoulin Island. He was struck with the similarity of the author's sections and those of the Lake of Como, published by the la'e Mr. J. Ball, which he had previously shown to be adverse to the glacial theory of the origin of the lake-basins. It was not safe to argue from the absence of remains of marine organisms; for elsewhere they were commonly wanting in deposits formed under circumstances similar to these, yet undoubtedly marine. He again could not follow Dr. Hinde in his objections to differential movements of the earth's surface, and insisted on the great movements of recent times, as evi- denced along the Frazer river and in Norway. Only last autumn he had seen distinct evidence of comparatively modero depression along the Dalmatian coast. He suspected some changes even in historic times. The buried river channels described by the author were par- alleled in Switzerland. He did not deny the efficiency of ice to pro- duce such effect, but it did not bring about what had been attributed to it by some geologists. Discussion. 25 Dr. Irving congratulated the author on the results he had placed before the society. He thought Dr. Hinde had not followed Dr. Spencer's arguments throughout, an, for instance, in the case of the connection between Huron and Ontario. He was glad to find the main points of his own theoretical conclusions as to the inability of ice to excavate confirmed by the author's observations in Norway and America. He saw nothing startling in the " warping" hypothesis. Mr. Clement Reid had no objection to Dr. Spencer's views of " warping," He thought all turned on accuracy of observation in tracing the terraces, and he wished to know whether it was absolutely certain that the same terrace was traceable throughout the whole distance. Rev. E. Hill called attention to the fact that tracts of Lake Superior were now below sea level, and yet no marine deposits are forming there. He called attention to the advantage of the Hydrographic Survey, which the author had utilized, and which we had in vain asked for in England. The depth of the Saguenay valley would be also accounted for by the author's explanations Prof. Seeley was prepared to accept the ancient drainage of the Laurentian river as now set forth. But he did not think it followed that the ancient valleys hid been excavated by the river any more than they were the work of ice. The general course of the Laurentian lakes followed the outcrop of the strata sufficiently to suggest that the lakes were originated by earth- movements. The main work of excavation seemed to him attributable to marine denudation in times when the level of the land was lower. And as tidal waters retired from the valley which they had cut out, the river drainage necessarily occupied these inlets after the land was elevated. Mr. Whitaker asked why objection was raised by Dr Hinde to deductions from borings in America when in England they were accepted. No other evidence of buried channels was to be had, some- times. He, would like to have some idea of the number of borings on which the author relied. The author, in reply, answered Dr, Hinde and Mr. Whitaker that he had only written a condensed account of the origin of the basins, not of the lakes themselves. Th» re were no escarpments in the place where Dr. Hinde had asserted their existence. There were scores of deep wells sunk in the drift between Lake Simcoe and Georgian Bay, where deep drift was shown. Similar sections were shown at the southeast end of the lake. He gave fuller details of the extension of 4 26 Reply by Author. these borings to the S, E. He cited instances of modern buried channels of a similar nature to those which he had described, and which evidenced a high continental elevation. To Professor Seeley he replied that he had no objection to the assistance of sea-waves, in part, enlarging the valleys in some pre-Pleistocene times. The old Erie- Ontario channel has been warped two feet per mile, which would account for the obstruction of the ancient valleys. Mr. Gilbert and he had traced one particular beach continuously round Lake Ontario. The elevations he had deduced from observations were founded on accurate instrumental measurements along this line, and similar obser- vations had been made by him m other areas. There was no evidence of barriers in the Erie-Ontario valley other than such as were due to differential elevation or partial filling with drift. The pre- Pleistocene drainage of the Lake Michigan basin was not to the south; hence no ban ier greater than at present was net ded, as^ explained in the paper. There were no beaches about Kingston, on account of the low alti- tude, but he had traced beaches in other parts of the region. If we were to follow the differential eleva' ion we should find that there were no Canadian highlands at the close of the episd^e of the upper till, but he could not now enter into the ice-hypothesis. He gave instances of the absence of marine organisms in undoubted marine beaches, and instanced the discovery of a whale in beach deposits upon which the evidence of warping was partly founded. CHAPTER III. Ancient Shores, Boulder Pavements and High-Level Gravel Deposits in the Region of the Great Lakes.* I. Characteristics of Ancient Shore-lines in the Region of THE Great Lakes. The land features throughout the lake region drained by the St» Lawrence river owe their formation largely to the action of waves sculpturing rocky or modeling earthy shores. That the waves have not always been coniined to the margins of the modern lakes is seen in the gea- cliffs and beaches, from which the waters have long since receded. These features, still remaining, are sometimes in the form of bold relief, and sometimes in the form of narrow sand or gravel ridges, delicately trace! over a flat country. In some places these ridges approach near to the lakes; in other localities they are miles away, and at varying altitudes up to hundreds of feet above their present waters. The raised shore-lines are no longer water levels, for terrestrial movements, since the lakes have receded from them, have commonly lifted them up to unequal altitudes. Whilst some of these old shores represent former lake boundaries, there seems to be little reason to doubt that the higher sea-cliffs and beaches formed the coast of brackish water inlets or arms of the sea. Besides the deformation arising from the unequal terresti'ial move- ments, the shores have been in many places defaced by the action of rains, rills, rivers and land^lides, until their broken continuity renders them somewhat difllicult to follow over long distances. The object of this chapter is to describe the characters of the old raised and deformed water-margins, by which they can be identified. The ancient coast- lines differ in no respect from the modern, but they are often easier to follow, as there are no waters to restrict one's footsteps. Were the lakes to be suddenly drained, but a few years would elapse before the deserted margins would be as difficult to mark out with precision as any of those from which the waters have long since receded. * Reprinted from Bull. Geol. Soo., Ah. Vol. I., pp. 71-86, 18S9. 28 Characteristics of Deserted Shores. With notable exceptions, the lakes are generally bounded by banks of clay or sand, stratified or unstratified. The waves have in places cut into these deposits, leaving high clay bluffs; in other localities the coast rises gently from the water-line. In front of these shores, whether high or low, beaches often occur. The typical beach forms a ridge of stratified sand and gravel, rising from three to five feet, or even more, above the surface of the water. The ridge may vary from a few yards to as many scores, or even hundreds in width. In the more perfect form, there is a slight depression behind the ridge which is some- times occupied as a bay, lagoon or swamp (fig. 3). Whilst the beach may Fig. S — Section showing the Floor of a Cut Terrace on which rests a Beach, b and c = Beaches broken into rldgelets . d = A frontal sand bar . W= Old water- level . form a frontal barrier, in shallow water, distant from the shore, it may rest directly against the coast, forming a terrace (a, fig. 4), behind which there is no depression. In this case the s irface of the terrace is apt to be defaced by landslides or washes; but the beach, whether in Fig. i. — Section showing the Floor of a Terrace of Construction. a= Terrace of constructi jn resting on cut terrace. P= Frontal pavement of boulders. W=i Old water-level. the form of a terrace or off-shore barrier, is very often wanting when the currents are cutting into and washing away the coast (fig. 5). Under such a condition, if a beach be formed, it is narrow and tempo- rary, as it is liable to be washed away or covered by landsli les. The eastern and southeastern coast of Lake Huron commonly illustrate the absence of true beach structure. Another excellent example may be seen at Scarboro heights, a few miles east of Toronto, on Lake Ontario, where the clay banks rise to the height of more than 200 feet and Charaotekistics of Deserted Shores, 29 extend for a distance of nine miles. Here the cliffs are being eroded. The waves are not forming a permanent beach, but the currents are drifting the materials several miles to the west to build up the barrier- beach in front of Toronto harbor. Fia. 5.— Sectio7i showing the Floor of a Cut Ttrrace without Bench but with Boulder Pavement. P = - Boulder pavement. W=. Old water-level. In the formation of beaches there is a tendency to straighten crooked coast-lines by the construction of bars in front of inlets, which are thus converted into bays or lagoons. Burlington bay, at the western end of Lake Ontario, is an illustration. Here, a narrow beach (c, fig. 6) cuts off a bay five miles long, whose depth is considerable, reaching to 78 feet. This is particularly a well-chosen example, for at the head of the bay there is a spit— named Burlington heights (A, fig. 6), rising to 108-1 Its feet above the lake — cutting off an older bay, now represented by the Fia. C— Map of the Western End of Lake Ontario. 6=Burlingtoii beach, separating Burlington bay from the Lake. /i= Burlington heights, an ancient beach 108-116 feet high, separating Dundas marsh from Burlington bay. Dundas marsh. This spit, when the waters were at its level, formed a portion of an ancient shore (to be described in a future chapter) in the same manner as Burlington beach forms a portion of the modern lake- shore. In places, where the waves break upon the more exposed coast, the beaches are apt to be piled up a few feet higher than their mean level. The opposite result is seen where the ridges are fashioned as spits and 30 Chara.cteri8tics of Deserted Shores, pass below the surface of the water in the form of submerged bars. The increase in the depths of the water in front of the beaches is usually very gradual. The study of the modern and ancient shores is reciprocal. By the former, still washed by waven, we can identify the latter; and by the examination of the floor in front of the raised beaches, we can more fully understand the action of waves upon the modern coasts, than where the subaqueous deposits cannot be seen. The muds, derived from the encroachment < f tlTe waves upon the land, are assorted; the coarser materials being those which form the beaches, and the finer clay, that which constitutes the off-shore silt deposits, leveling up the inequalities of the lake bottom and forming very flat submerged plains, which are rendered apparent upon the withdrawal of the waters. In the examination of old shores, the occurence of flat or very gently inclining plains, abutting at constant levels against rising hills, is as certain an indication of old coast lines as if beaches were found there; but the exact height of the water-line cannot be recognized, as the water may have been five or it may have been twenty feet deep. When this condition obtains, there may remain here and there a fragment of a temporary beach (c, fig. 7), covered by a landslide (s, fig. 1), but exposed by a stream or artificial cutting into the hillside, or there may be a barrier in front of an ancient bay or lagoon {h, fig. 6). WiQ.7.— Section showinp a Cut Terrace with a fragmentjof Old Beach partly concealed by a Landslide. 6 = Boulder pavement. c = Fragment of old beach. d = Drift. s = Landslide. t<; = 01d water-level. While the greater proportion of the lake coast is composed of drift deposits, there are places where the water-margins are bounded by rocks. Here the structure is similar, although not so well developed, and the banks may assume the form of vertical cliffs. Generally speak- ing, the beaches in front of these rocks are not so well developed as where there have been shore deposits of boulder clay to supply the waves with pebbles. However, some of the higher and older coast- markings remain in the form of such "sea-cliffs," in front of which there are comparatively flat plains. ChakA-oteristics of Deserted Shores. 31 Another structure, when present, is very characteristic of many portions of the ancient shores, or, indeed, is occasionally seen in front of the modern beaches. This is a pavement of boulders (derived from adjacent shores of boulder clay), occupying a given zone (P, figs. 4 and 5). This zone is in front of and a few feet lower than the level of the true beach; the boulders having been left just below the water-level as the waves made encroachments upon the coast. Again, the boulders have been more or less pushed up to this line by the waves forcing up the coast- line to which these boulders have been frozen. When these deposits occur adjacent to the modern beach, they may be seen rising out of the water, but they are also found outward in the lake to the depth of several feet (figures 9 and 10, p. 3 8). In front of an elevated shore, the boulders may be arranged in the form of a zone, even a few hundred yards in width, throughout a vertical range of a few feet, which may be increased to 30 or 40 feet where there is a succession of beachlets close together, marking the gradual recession of the waters. But the upper level of these zones never quite reaches that of the beaches. In traveling along a flat country these pavements of boulders are as certain indications of shore- lines as are any other forms of the beaches (fig. 9, p. 38), Boulders left on the hillsides by the action of rains, washing out the finer materials of the drift clay, are not arranged in belts of symetrical level. The boulder pavements do not usually occur where the adjacent coast is not composed of boulder clay, nor where the beaches are separated from the land by what is now or has been a bay or lagoon. Pave- ments of boulders are not as commonly seen in front of modern shores as in front of some of those more elevated and ancient. Turning to the more typical form of the beach structure, as shown in the raised shores, there may be seen sand or gravel ridges, most frequently from 100 to sometimes 500 feet across, rising to 15 or 25 feet above a flat or very gently descending plain, whose surface is most commonly composed of fine clay. Sometimes this descent is so very gradual as to be inconspicuous; at other places the descent is quite sudden. The depression behind the ridge is generally less than that in front of it^ and here also the floor may be composed of clay. Where the beach is broad, it is apt to be broken up in a number of ridgelets (c, fig. 3). Indeed, some of the larger and more important beaches mark the recession of the waters by being separated into several ridges, often at considerable distances apart, each a few feet below the preceding, 32 Characteristics of Deserted Shores. where the lake floor is sloping very gently; but where the slope is more rapid, all unite into one large ridge. The beach has rarely a thickness of more than 15 or 20 feet, and rests upon the clay or drift deposits, which once constituted the floor of the former lake. As the plain recedes from the shore, the materials become finer and finer clay and freer from sand; but at varying distances, of sometimes a mile or more in front of the beaches, there may be found thin belts of sand resting upon the lake deposits. Again, the beaches may take the form of terraces of construction, resting against clay banks; or against these banks the ridges may abruptly (but not temporarily) end like the modern beaches {b, fig. 8). In measuring the comparative altitudes of a beach at different points the summit of a well marked ridge should be chosen, rather than that of the beach in the form of a terrace (a, fig. 4) against the shore or the junction of the coastal plain back of a cut terrace (c, fig. 7) and the bounding hills, as the exact water-level can here be only approximately determined. It is more accurate to make the calculations as to the former water-levels at the foot of the beaches, as the slope in front FiQ. 8.— Plan of Barrier Beach in front of a lagoon and overlooked by hills. 6=Lme of hills. s=Barrier Beach. The beach ends abruptly on the left. of them may be steep in one place, and in another very gentle, but the summit is easily recognized. Where the beach itself is absent, by tracing the coastal line, there will be found sooner, or later, a bar or spit in front of some river or extinct bay. In ascending from the modern lakes to the highlands, several old shores must be crossed. The country may be described as a series of terraces or steps, whose frontal margins are moulded into hills, and whose surfaces are plains, most commonly of clay, although sometimes of gravel or sand, at the back of which, there may be found the beach in some form. These gently rising terrace plains may each be several miles in width — and consequently the beaches several miles apait — or CHA.EA.CTERI8TIC8 OF DeSERTED ShORES. 33 they may be narrow with the beaches close together. In many regions, the old shores behind these plains rise and extend across the country as conspicuous ranges of hills. The plains themselves are occasionally eroded by streams, until the whole country is very broken. This is more likely to be the case with terraces of the greater altitudes, and here the more recent surface erosion has often rendered the ancient shore lines hard to follow. In crossing a series of beaches, the lowest is found to be composed of the finest gravel, or indeed perhaps of sand. In this case it is apt to be more or less heaped into dunes, by the action of winds. The ridges are often divided, bnt the branches unite again, or else send out spits ending abruptly. Occasionally the materials from which the beaches were formed was stony sand, in place of stony clay. Here, then, the extinct water-margins are difficult to determine, for there is no sharp lithological character, as where a beach crosses a clay plain — to mark the boundary between the sand beach, commonly heaped into hummocks or dunes, and the frontal plain composed of sand. Many of the upper beaches overlie drift deposits, but those of the lower elevations are more likely to rest upon stratified clay — the sedi- ments carried into the deeper waters whilst the lakes were at higher altitudes. The character of the materials underlying the beaches is commonly the same as that forming the surface of the plain in front of the ridges; but its structure is best shown in sections exposed by the subsequent erosion where streams cutting through the ridges cross the plain. When such streams have been large rivers, as has often been the case, there may be some trouble in tracing the continuity of the beach, especially across a broken country, as a portion of the valley may be older than the beach, which may swing around and skirt the embayment, or form a bar across it. Or again, the beach may be only represented by conical or other shaped sand or gravel hills, which were delta deposits at the mouth of former rivers. Such delta deposits may not rise to the level of the former body of water. With the varying conditions here set forth, which the shorc-lin>? undergo, the traveler, in coasting around the old lakes, can rarely pro ceed more than a few miles without meeting obstructions. When the beaches were a considerable distance apart, with perhai)s ohl}^ 50 or 100 feet of difference in their altitudes, there is a liability of get- ting off one series and upon another. Consequently it is often neces- sary to resort to accurate leveling, allowing for reasonable variations 5 •'34 Chakacteristics or Deserted Shores. in the height of the beach, and the differential elevation of the region, since the waters have receded from the former shores. In some regions the former expansions of the lakes were occupied by archipelagoes. Consequently, there is an absence of continuous beaches, and the explorer must depend upon following the plain, which formerly constituted the lake floor, finding here and there a fragment of the ancient beach, either upon the coast of the mainland or upon that of an island. Here again, it may be necessary to resort to accurate leveling to identify the beaches. Whilst steep coast-lines may be followed through wooded regions, it is most difficult to trace satisfactorily a beach across such a country. The greatest difficulties are found where the ancient beaches enter regions that are composed of hills of crystalline rocks, more or less wooded, and interspersed with numerous lakelets. In such places there are numerous gravel hills whose relationship to the old shores is not readilj^ discernable. In some places the surface of the beaches is composed of nearly clean gravel or sand; elsewhere, from some admixture of clay, it becomes more or less earthy soil, to a depth of two or four feet, somewhat obsuuring the beach structure. Again, there may be coarse stones resting upon its surface, as if these had been forced up after the beach had been formed, by a slight rise of the waters, or by the action of coast-ice, pushing them up. However, these must not be mistaken for the more ancient gravel beaches, covered with drift, such as frequently exist, and will be described elsewhere. The beaches, in the form of narrow belts of gravel or sand, crossing a flat country, were in many places used as trails by the Indian aborigines, and in some places these trails have been turned into roads, as they are always dry during the muddy seasons. These ridge- roads have attracted attention as ancient beaches for nearly a century. But the water long since withdrew from them owing to the elevation of the continent, which has been accompanied by their distortion from the water plain, on account of an increasing rise to the north and east. The great geological value of investigating the raised and ancient coa«t-lines lies, not only in gaining a knowledge of the former expan- sions of the lakes and their relationship to each other, but particularly in being able to make use of them, as old water-levels in order to measure the amount of deformation or warping of the earth's surface -caused by terrestrial movements, resulting in the development of the basins of the lakes themselves, and other features. While the old Boulder Pavements. 35 shore-lines record a great amount of unequal terrestrial movements, yet these movements have also left records in the older sea cliffs, Boulder Pavements and Fringes. In many localities of the northern part of our continent, the land surfnces are almost covered with loose boulders, varying from the size of cobble stones to masses commonly three or four feet long. Occa- sionally the blocks have a length of eight feet, but rarely longer. While some of the boulders are angular blocks of Paleozoic limestones and sandstones of local origin, a great proportion are Archrean rocks, which have been transported from the Canadian highlands, north of the great lakes, to distances of sometimes 300 or 400 miles. These crystalline rocks, although so hard and compact, have the angu- larities invariably removed. Blocks are frequently seen at altitudes of hundreds of feet above their original sources. Throughout the lake region, and the country north of the line of the southern limit of the drift, which is often fringed with them, the accumulation of boulders is not uniformly distributed. The country enclosed by that line is occupied by sheets and ridges of drift materials, through which the subjacent rocks occasionally protrude. Again, these plains and hills have their surfaces moulded by the action of the waves of vanished seas or shrunken lakes, often fashioning the region into a suc- cession of broad terrace flats and hilly coast lines. It is upon the surfaces of these moulded features that the boulders are found. Whilst there are vast areas where there is not a single stone to be seen, and others where only an occasional block occurs, as dropped down from some meteoric source, there are other localities literally so covered with large boulders as to prevent agricultural pur- suits. These boulder accumulations are superficial and do not pene- trate the subjacent earths. They occur along certain zones, outside of which they are not fonnd. The presence of these surface boulder accumulations has been most commonly explained alike by those who believe in the glacial origin of the drift and those who do not, as having been dropped by melting icebergs at the close of the drift epoch. A few glacialists regard these boulders as having been deposited from glaciers where they now rest. It has also been hinted that they have been left upon the hills, as the finer materials of the boulder drift have been washed away by atmospheric agencies; but it was only since the recent systematic studies of the high-level beaches, compared with modern I;ike shores, have been made that the natural explanation of boulder pavements and distribution of erratics become possible. 36 Boulder Pavements. There are three conditions under which boulder accumulations are found. The most important is where the boulders form pavements stretching as belts across a level country, usually in front of ridges which once constituted old shore-lines, or forming zones of stones rest- ing upon hillsides or capping the summits of ridges. Of less importance is the occurrence of blocks scattered sparsely and irregu- larly on the sides of hills. Lastly, occasionally erratics are found alike over the billy and over the flat country. That the boulders were brought from their original sources in the later Pleistocene daj s and dropped by either icebergs or glaciers where we now find them is an untenable hypothesis, for their birth places are now often covered with the older drift or are hundreds of feet below the elevations where they are now found. The relation of the boulders to the older drift are such that the erratics can commonly be recognized as of secondary origin, being derived from the earlier accumulations of boulder clay or sand. The manner in which the blocks have been brought to the surface has been by the removal of the finer earths from the drift, principally by the action of the waves or currents encroaching upon the hUls or ridges of such materials, charged with occasional boulders. Thus the coast-line has been moiilded into steep shores, in front of which there are the gently descending plains, once submerged — the floors of terraces since the recession of the waters (figs. 4 and 5). Ihus the boulders throughout the whole thickness of the drift, which were too large for transportation by the waves, were reduced to water-level and were accumulated upon the floor in the form of pave- ments or fringes along the former water margins. The removal of the earth beneath the boulders continued until they had settled to the maximum depth of wave action below the surface of the water, for a^ greater depths the fine earth would not have been removed from beneath the stones. The vertical range of the fringes is from 15 to 25 feet or more when the recession of the former waters was gradual, leaving a close succession of beaches. The width of the pavements varies from a few hundred feet to perhaps a half a mile, according as the slope is somewhat steep or very gradual. Where the finer materials were entirely washed out into deeper water, then the margins of the plains, at the foot of the old coast line, are simply fringed with boulders; but where the finer materials were assorted by the waves and currents, the sands and gravels have been formed into beaches, usually a few feet above the level of and behind the boulder belt. Duration of Niagara Falls. Plate II. ^ms:.. «^iAJ FIG. 9.— MODERN BOULDER PAVEMENT ON GEORGIAN BAY, Near the end of Blue Mountains of Collingwood, Ont. ,..-ftS«u^>- ^'''*^'~ -iiiS^\ FIG. 10. ANCIENT BOULDER PAVEMENT OF ALGONQUIN BEACH, Whose Crest Rises 189 Feet above Georgian Bay upon the N. E. Side of Blue Mountains of Collingwood, Ont. Coast-Ice. 37 But the story of the boulder pavements and fringes is not yet com- plete. Coast ice has also played an important part in the arrangement of the paving stones. The waves, acting upon the coast-ice wherein boulders have been entangled, cause the stones to be forced up into more regular zones, as to height, than would be affected by the residuary deposition alone, as just described. Blocks of large size can thus be moved, not merely by the heaving action of modern frosts, but by the action of coast-ice itself; for boulders upon the margins of the St. Lawrence river, weighing 10 tons, are known to have been shifted by the spring movements of a winter's ice. Again, the writer has seen upon some of the shores of Shoal lake, in Manitoba, situated in a flat drift-covered country, modern beaches composed of huge boulders, piled up by the waves of the lake acting upon the ice in which the stones were enclosed, as otherwise blocks four or six feet long could not be gathered from the shores of the lake and accumu- lated into beach ridges, nor could they have been residual pavements as above described, for no high shores of boulder clay occur into which the waves could have made encroachments. An excellent illustration of the modern formation of boulder pave- ments and fringes may be seen upon the shores of Georgian bay, between Thornbury and Collingwood, as shown in Plate II, fig. 9. There the lake waves are encroaching upon a shore composed of boulder clay. The larger stones standing out in the water are too heavy to be materially affected by the waves or ice action. Excellent illustrations of boulder zones are found a short distance from this locality, at an elevation of 187 feet above the lake, as shown in Plate 11, fig. 10. Other examples of fringes of boulders high above any modern waters may be seen a few miles beyond the eastern end of Lake Ontario. The same is true upon the northern side of the lake, as for example, back of Trenton and westward; these are parts of and in front of the finer gravels of an old beach, now more than 400 feet above the lake. Westward of Toronto, where the old shores are of Paleozoic rocks, in place of drift, the boulder pavements disappear from the front of the beach. Upon the steep hillsides, as long the Mahoning valley, near the crossing of the Ohio-Pennsylvania line, there are zones thicklj^ covered with boulders. There we find the records of old water-margins, as well as in the pavements associated with the well marked beaches and shore-cliffs facing the lake basins. The finer materials have been 403891 38 High Level Gravel Deposits. washed out of the associated drift to form bars, in the valley, which was once filled with water. On some of the higher hills between the southern part of Georgian bay and Lake Huron, to the west, the tops of ridges are covered with boulder pavements. These ridges were islands in a former expanded lake or sea, whose surfaces were encroached upon by the waves, until they were reduced to partially submerged reefs covered with great erratic blocks, as the finer mud was borne into deep water. That these were island shores may be seen from the boulder covered ridges, although miles apart, being reduced to a common altitude. On the hillsides, behind the fringes, there are only here and there irregularly deposited blocks t xposed by the action of rains. Besides, the meteoric effects upon any of the hills are small, compared with the encroachments of the waves, in exposing enough stones to make boulder pavements. The occasional erratic blocks often repo-ing upon fine lacustrine deposits are of little importance, and indicate only an accasional stone entangled in old coast-ice from an adjacent shore, when the region was covered with water, just as the boulders resting upon the sunken ships in the mouth of the Baltic have been deposited from the coast-ice moving out of the sea. The study of the relation of the pavements of boulders to beaches set at rest the speculation upon the origin of these fringes, and obviates the necessity for appealing to either icebergs or glaciers in later Pleis- tocene days to account for the erratics popularly called "hard headR," which are scattered over the country in the form of pavements or fringes ; for these are mostly seen only where they can now be referred to some old coast line, or a succession of shore lines, acted upon, in former days, by frost and coast-ice. High-Level Gravel Deposits in the Region of the Grea.t Lakes. Rather than rummage through the talus heaps of geological literature for the different kinds of gravel deposits which may represent beach structure, it is easier to go into the field of observation and investigate those forms which may be modified beaches, or be related to, or be mistaken for them. This method is the more satisfactory, as in geological literature different forms are confounded, and others are left unnoticed, or not considered in the light of the present investi- gation. The object of this chapter is to describe the various kinds of gravel deposits, which resemble or are related to beach structure, and not to consider their occasionally doubtful origin or distribution. Exclu- sive of the beds cf sand, which are intimately connected with the strati- BuEiED Beaches. 39 fied clays, or included in the drift accumulations themselves, and the ancient shores already described, the following groups of gravels and tjands should be noticed, some of which are covered with the stony clay of the upper till: A. The gravels and sands which are buried beneath the upper drift dep>osits. — These may be divided into (a) buried beaches ; and (J) more or less irregular beds and ridges of gravel and sand, often of earthy texture, having a more or less tumultuous structure, and resting beneath accumulations of the upper till. B. Surface accrmiidations of gravels and sands, forming ridges, mounds and plains. — These are in the form of (a) the so-called osars and kames; (b) other ridges and mounds resembling the last, but having a position corresponding to that of beaches, in front of more elevated plains or drift hills, or of the accumulations included in group A b ; and (c) gravel plains A a. — Hitherto, the buried beaches have not been distinguished from other beds of gravel and sand intercalated within the drift formations. As such accumulations, whose structure is the same as that of modern beaches, are only exposed in sections cut through the surface deposits by streams or artificial excavations, all of the knowledge that we can, at present, hope to acquire, is the recognition that there were beaches, now covered by drift, older than those upon the surface of the country. When beds of gravel and sand are met with in borings, it is not always possible to distinguish those which are buried beaches from others which are intercalated with drift deposits. The structure of the buried beaches does not show that tumultuous crumpling, so commonly seen in the next kind of accumulations (A b). In some places the gravels are found cemented into conglomerates. Thin laj ers of stony claj', constituting the upper till, which covers vast areas of the country throughout the lake region, often rest coraformably upon the undisturbed surfaces of the buried beaches, that may have a thickness of twenty feet or more. Excellent examples of buried beaches may be seen along the Au Sable river, southeast of Lake Huron, where the overljdng drift cla}' is only four or six feet thick. When the covering is thin, there is a liability of mistaking these older beds for those belonging to the beach epoch proper. A b. — The internal structure of this kind of gravel and sand deposits shows stratification, which may be regular in one place, but the beds soon become tumultuous, that is, the beds become ii regular, bent or twisted, and confused. The materials are apt to be somewhat earthy 40 Superficial Gkavels. Throughout these layers there may occur occasional boulders of large size, and pockets of gravel, whose outlines resemble those of boulders (as if the gravel had been cemented into masses by frost and then moulded into boulders, and afterwards deposited in the frozen state.) By the characters just given, these accumulations can be readily dis- tinguished from those of true beaches. They are commonly overlaid by a few feet (perhaps 10 or 20) of stony clay or other materials of the upper till. Occasionally the covering may reach several times this thickness. The extei-nal form of these deposits, with their clay mantle (which last is dependent upon the form of the underlying gravels), may be thit of undulating plains, or these undulations rising to the magnitude of ridges and hil's. In this case, the ridges rise in sue .ession one above the other, until they reach an altitude of 100 feet, or even more, above the plains which are commonly in front of them. They may occupy a breadth of several miles across the country. The ends of the ridges often overlap, and at other times send out spurs, and inclose kettle-like depressions, which are liable to be confounded with or not separated from those of the next group. These ridges occur asso- ciated with some of the so-called moraines of America. These slightly covered sand and gravel deposits are not so commorly devel- op* d below the altitude of 700 feet above the sea as at higher eleva- tions, for the lower country is more apt to consist of terraces, cut in the drift, and of silt deposits and beaches. But these accumu- lations cap the ridges of the great chain named the Oak hills, which extend for over 100 miles in length, parallel to the northern side of Lake Ontario, at an elevation of from 900 to 1,200 feet above the sea. Farther west, such are also the capping materials of the country, which is 1,700 feet above the sea. The same holds true for Michigan and other States. t>. — The gravels of this group are not only well water- worn but also well washed and free from earthy matter. Indeed, they are sometimes free from the finer sand. The pebbles are often coarser than in the lower beaches, in some cases foroiing accumulations of almost cobble Btones. There are occ jsional boulders in the mass, but these are more common upon the surface. The materials are mostly of local origin, with a small proportion of transported crystalline stones. None of the materials have been derived directly from the subjacent Paleozoic rocks, but secondarily from the assortment of the stony boulder clays. The gravels, with their accompanying beds of sand, when these are present, are stratified as in beaches, without anything of the tumultuous structure of the last group. Still, there may be a false bedding, as in OSAES AND KaMBS. ^^ beaches; and when the deposits assume the form of ridges, the layers may dip in opposite directions, as in barrier beaches. The materials of this groixp are never covered with drift deposits, but often rest upon the till, or against hills of the tumultuous accumulations already described. In external form, the gravel deposits differ greatly, and it is upon this character that they are divided into the three series. B a. Osars and Karnes.— The osars (Anglicized from the Swedish word asar, meaning gravel hills) being the term in America applied to very narrow gravel ridges (often only a few score yards in width at the base) or chains of mounds, winding in a more or less serpentine manner across a comparatively flat country, above which they rise at nearly as steep angles as the loose material will stand to a height of 40 or 60 feet. They are also defined as generally extending from a higher to a lower country and following the course of the greater val- leys —that i", at right angles to the coast lines. A beautiful example of an osar, as above described, is to be seen southeast of Lansing, Michigan. It trends into an inlet among the hills, oblique to the gen- eral direction of the ancient coast. Driving along the top of the ridge, which is scarcely wider than the road, it is seen to be composed of constantly and suddenly alternating stretches, each quite level, the one set being about 25 feet above the other. These so-called osars form a very limited proportion of the gravel ridges of this group. The term kame (the Scotch vernacular for gravel hill), according to its use in America, is described by Chamberlin as "asseaablages of conical hills and short irregular ridges of discordantly stratified gravel; between which are irregular depressions and symmetrical bowl shaped hollows that give to the whole a peculiar, tumultuous, billowy aspect. * * * These irregular accumulations are, however, more abundant in connection with deep, rapidly descending valleys, being especially abundant where they are joined by tributaries or where they make a bharp turn in open portions of their valleys, and especially where they debouch into an open plainer country. In such instances they are usually associated with gravel terraces and plains. Precisely similar accumulations are very common associates, if not constituents, of terminal moraines. * * * They are transverse to the slope of the surface, the course of the valleys and the direction of the drift movement."* From observation in nature, as also from the descrip- tion itself, it will be seen that the term kame is not specifically used, and that different kinds of gravel deposits are grouped under the same name. Indeed, from the above description, the term might be better applied to some of the deposits described above under group A b, which ♦ Third Annual Report of the U. S. Geological Survey, 1883, p.^300. 42 GbAVEL KiDGlS. are more or less covered with clay. However, there are conical and tapering ridges in many localities without a tumultuous structure, whose relations to each other are not easily discernable, that may be placed here under the name of kame. Some of the kames in the val- leys are doubtless river deposit?, and others are the remains of uncov- ered buried beaches of greater age, exposed by subsequent erosion. B h. — The internal structure of this series is similar to that of the other members of the group. The external form is that of intermit- tent ridges, sometimes rising to sixty feet above a frontal subaqueous coastal plain which is occupying the position as in front of a beach. The ridges may be replaced by cones, resembling delta deposits. The ridges are often scarcely less direct and scarcely more broken or more varying in height tl an beaches, especially when the subsequent erosion and unequal elevation, caused by terrestrial movements since the gravels were deposited, is taken into account. The ridges are often found to divide and enclose kettle-like depressions, sometimes dry and some- times containing ponds or lakelets, just like similar depressions along modern beaches, but on a larger scale. Branches and spurs add to the undulating appearance of the country. In front of these hills the plains may be covered with gravel. It is very difficult not to see in these ridges the remains of beaches belonging to former shore-lines. A single ridge of this character occurs behind a plain just north of Stouffville, Ontario, rising to a height of 75 feet above the plain, which is about 1,100 feet above the sea. This deposit rests against another and somewhat larger ridge of sand and gravel belonging to group Kh. Again, within a distance of 14 miles, stretching northwest- ward from a point near Flesherton (shown in fig. 11), there are three steps, each in the form of a slightly undulating plain, often paved with gravel, bounded by just such hills of gravel as are here described. These marginal ridges ai'e much indented with kettle depressions (^, h^ fig. 11), and are somewhat beueath the level of well-marked Fig. \\ — Section extending northward from Flesherton. 6-= Boulder pavement, g, g Ridges of Artemisia gravel, h. A- — Depressions beliind the gravel ridges. Artemisia Gbavels. 43 terraces, as if a somewhat oflf-shore deposit. The elevation of the coun- try above the eea descends from 1,600 to 1,200 feet. The ridges {g, g, fig. 11) border a mass of land that was rising out of, probably, the sea. The beach-like character of these accumulations is further brought out by the occurrence of zones of boulder pavements at levels below and immediately in front of the ridges {b, fig. 11). These boulder pavements, which do not enter the mass of the drift but only rest upon its surface, are too characteristic of the action of waves cut- ting into stony drift and of the accompanying action of coast-ice not to be regarded here as additional evidence of the coastal formation of the surface gravel ridges, described in this paragraph. "Artemisia gravel" is a name applied by the Canadian Geoleogical Survey to the gravels covering an area of 2,000 square miles of the highest land in Ontario, between the three lakes, Huron, Erie and Ontario, rising in places to 1,700 feet above the sea. But at that early date, geologists did not differentuate the various gravel accumula- tions. Indeed, the whole work upon the drift of Ontario was only pioneering, and now being somewhat antiquated and generalized, it needs to be revised and differentiated by modern investigations. Thus the term Artemisia includes sand, gravel, and even upper till deposits (the last, although occupying thousands of miles of the surface of the Province, was not formerly identified) of all kinds and ages mentioned in this chapter and in that on beaches. However, it was the accumulation of the gravels described in this group B5, in the township of Artemisia, that gave the name which was extended over such a wide range of materials and geological time as if all were one formation. At most the term should be restricted to the ridges occu- pying the position of very high-level beaches, just described. B <*. — Gravel plains are common in front of such high-level ridges as have been last described, representing the subaqueous floors when the waves beat upon the old shores. Some of them, however, may be the floors of terraces cut into the older gravel deposits. The plains are often very deeply eroded, owing to the high elevation of the coun- try and the long action of meteoric agencies upon the incoherent mate- rials. Thus, there sometimes remain of these plains onl}-- a succession of ridges, between ravines deeply excavated by the numerous streams and floods. Such plains occur in the typical region of the Artemisia gravel in Ontai-io, in Michigan, and in other States. CHAPTER IV Deformation of the Iroquois Beach and Birth of Lake Ontario . f Upon receding from the lake and ascending the high country which bounds the Ontario basin, an observer is attracted to the wonderfully plain shore-lines which record the former expansion of the waters. The terracep, beaches, scarps, and spits across the mouths of valleys clearly represent the deserted shores. But they are no longer horizontal lines as when laid down at the level of the former waters. As distinctive features, the beaches were so striking as to attract the attention of the aborigines, who used them as trails across an otherwise, sometimes, muddy country. The early white settlers, in turn, used them as high ways and hence we find the "lidge roads" about Ontario as well as about the upper lakes. But the recognition of the shore-like characters of the raised beaches, by the early writers,^ did not contribute much to the solution of the lake history. Nearly fifty years ago. Professor James Hall observed that the beaches of New York were not horizontal. But Mr. G. K. Gilbert was the first who connected and measured the deformation of the beaches upon the southern and eastern margins of Lake Ontario, and the writer upon the Canadian side of the lake to beyond Trenton, whence the same beach swings around towards the north and passes into a broken country. The writer has further carried the survey of the same beach to the northeastern portion of the Adirondacks. There are wide-spread remains of old shore-lines at altitudes so high above Lake Ontario, as to indicate that the same sheet of water * Reprinted from Amkr. Jour. Per., Vol. XL, pp. 443-451, Dec. 1890. t The forerunner of this paper —" The Iroquois Beach, a chapter in the Geological History of Lake Ontario" — was first read before the Philosophical Society of Washington, January, 1888, Proc. Phil. Sec. for 1888, and was subsequently amplified and published in full in the Trans' actions of the Royal Society of Canada for 1889. t For reference to early writers, see "Iroquois Beach," etc., Transactions Royal Society of Canada, 1889, page 121. Structure of thb Beach. 4:5 (Warren water) covered also the basins of the other and higher lakes. After the dismemberment of this greater sheet of water, the surface of that occupying the Ontario-St. Lawrence valley was gradually lowered, and fell several hundred feet, without pausing long enough to deeply cut out or straighten its changing shore lines. At last, this shrinkage of the waters came to a pause lasting until the shore-line became more pronounced than that of the modera lake. It is this shore-line that forms the basis of the present chapter, and constitutes that water-margin which the writer has named the "Iroqufiis Beach,"* in memory of the aborigines who trailed over its gravel ridges. The general structure of the ancient shore-lines is somewhat fully described in "Ancient Shores, Boulder Pavements, etc."t But let us here repeat some of the characteristics. Typically, the ancient beach consists of a ridge of gravel and sand rising sometimes to twenty-five or more above the frontal plain, which further descends lake ward (as in fig. 3, p 28). Back of the ridge, which rarely exceeds a width of 500 feet, and usually less, with a very narrow crest, there is often a lagoon-like depression. The beach may be broken into a number of ridges (b or c). The summit marks the heighth of the wave action. This barrier ridge may become a terrace, or it may pass into the form of a spit across some valley (A or b, fig 6). Again the ridge may be wanting, but the shore will be represented as a cut terrace (fig 4), in front of which a boulder pavement may frequently be seen iP). This pavement is also often found in front of gravel beaches. In places Avhere the former waters were gnawing away the drift shores, or where rocky promon- tories rose out of deep water, true beach structure is wanting, or only represented by benches. In the survey of the Iroquois Beach, the shore line has been followed by one or another of its characteristics, even across areas of broken physical features. The altitude of the highest ridge, where the beach is broken up into a series of ridges, is that which has been everywhere taken, for it is the one giving most accurate results. No elevations have been adopted except those of the summit of the crests (as in fig. 3), or of the spits (at h or b, fig. 6, p. 30). The measurements conse- quently represent the maximum height of wave-action, in place of the mean surface of the water, which was a few feet below. The writer's leveling has everywhere been done instrumentally. * The name was first printed In Science, Jan. 27, 1888, p. 49. tBy the writer, in Bulletin of the Geological Society of America, vol. i, 1879, p. 71. 46 Extent of the Iboquois Beach. The coast materials, out of which the Iroquois shores have been carved, are most boulder clay, or stratified clays or sands, deposited upon the floor of the lake when the waters were at higher levels. At a few places the shores rest against Paleozoic rocks, in which cases the materials of the gravel beach are more scanty, as the pebbles were mostly derived from the stony drift, or there may be an absence of the beach. Except in spits across old valleys, the thickness of the sand and gravel of the beach does not usually exceed 20 feet, but in front of valleys it may reach a thickness of 100 feet {h, fig. 6). The internal structure always shows stratification, with such sloping and false- bedding as are characteristic of beaches. There are frequent exposures which shows that the Iroquois Beach rests upon stratified storieless clay — the silt washed into the waters when the waves were encroaching upon older and higher shore-lines, and aseorting the boulder clay, which, at the higher elevations, formed the coast. Eastward of Watertown, the beach rests upon stratified sand in place of clay, as there was but little stony clay in the drift to furnish silt for the older lake floor. From near Trenton to the head of the lake, and thence around the southern and eastern borders to near Watertown, the Iroquois Beach is not hard to follow; but eastward of that point the features are more complex. The old coast of stony clay is there replaced by stony drift sand, and hence there is but little lithological distinction between the frontal plain and the older sandy drift shores. Moreover, suoh a coast is apt to be defaced by the sand being heaped into dunes. Again, in the region beyond Watertown, the Iroquois Beach is interrupted by promontories of Paleozoic limestones and shales, rising out of deep water, upon which at most only benches were cut. Farther, north- eastward, the beaches trend among bold headlands and islands of crystalline rocks. Wave action, which carves broad terraces out of drift materials, can cut only moderately well-marked benches out of limestones. But when the same intensity of wave force is applied to hard crystalline rocks, especially when interrupted by islands, the benches become less conspicuous than when excavated out of lime- stones, or they may become very obscure. Still, upon the flanks of the Adirondack Mountains, the Iroquois Beach can be followed and identified by the remains of barrier ridges, terraces, bowlder- pavements, benches, and above all by the occurrence of spits across old valleys. Altitudes of the Iroquois Bbaoh. 47 Combining the surveys of Mr. Gilbert and the writer, the position of the Iroquois Beach is shown on the accompanying map. O^via'bu^ .j/i,\ Fio. Vi. — Map of the Iroquois Gulf. The following table gives the elevation at salient points along the Iroquois Beach. The elevations given are those of the crest of Feet above the sea. Lake Ontario, surface of Hamilton Burlington Heights Waterdown Station Cooksville Station, about Carlton Station , Kingston Road, crossing railway 12 miles east of Toronto Whitby, 6 miles north of lake, near Colborne Station, 2 miles north of Trenton Station, 2i miles north of Lewiston, N. Y Rochester, N. Y Canastota, N. Y Cleveland, N. Y Constantia, N. Y Richland, N. Y Adams Centre, N. Y Prospect Farm, 4 miles east of Watertown Natural Bridge East Pitcairn, 1 mile northeast^of Fine 247 (U. S. Lake Survey) 362 (Sp encer). 355 " 368 (< 400 '« 417 " 459 .. 507 << 602 <( 632 (( 385 (Gilbert). 436 << 441 << 484 (( 489 it 563 " 657 " 730 (Sp encer). 829 n 942 (( 972 (< 48 Defokmation of the Iroquois Beach. the highest ridge, where the beach is broken into a number of ridgelets, having sometimes a vertical range of twenty- five feet or more. Thus we see that the Iroquois Beach has been deformed to the extent of 600 feet, between the western end of Lake Ontario and Fine, of which only 78 feet of rise occurs upon the southern side of the present lake, while the great proportion of the uplift is found west and north- west of the Adirondack Mountains. Upon the northern side of the lake, the eastern equivalent of uplift is more pronounced. At the western end of the lake, the mean maximum uplift is 1.60 feet per mile in a direction of N. 28 E. This rate increases towards the northeast. To give a mean rate of rise, at the eastern end of the lake, does not convey a correct idt-a, for the uplift increases in a progressive ratio. Thus in the region of Oneida Lake, the uplift is 3.5 feet per mile, while in the region of Watertown it amounts to 5 feet per mile; and farther northeastward the deformation reaches 6 feet per mile, in the direction of N. 60° E. This seems an extraordinary amount of measurable terrestrial movement, but the records are inscribed in the beach. It is not yet known where this upward movement ceases. Upon the Erie beaches, outside of the Ontario basin, Mr. Gilbert found a considerable amount of warping recorded at Crittenden, N. Y., over the horizon at the western end of the same lake. I have traced the Erie beaches around to the southeastern side of Lake Michigan. Combining our results, I find the measured uplift between the two regions amounts to 324 feet. But the beach, where last observed near Lake Michigan, is 45 feet above its surface. Indeed, it is there diffi- cult to trace, owing to the duny character of the sandy country. By the assistance of other beaches found in that region, the conclusion is readily arrived at that the shore-line under consideration must pass from 40 to 60 feet beneath the waters of the lake at Chicago. It is then evident that the terrestrial uplift, between Chicago and Crittenden, amounts to not less than 410 feet. Crittenden is nearly on the line of strike of the Iroquois beach (S. 62° E.), at its lowest point, at Hamilton The Erie beaches, eastward of the Niagara River, were deformed to the extent of 0.4 feet per mile before the Iroquois episode, the remainder of their uplift having been synchronous with that in the Ontario basin. But the pre-Iroquois diflferential uplift of the beaches farther west is reduced to almost zero, for the beaches south and west of Lake Erie have suffered very little deformation. Consequently a sufficient amount of deformation of the beaches has been measured to allow for inaccuracies when we take the elevation of the Iroquois Beach above the sea level (363 feet), as the amount of movement that must Focus OF Elevation. 49 be added to the Iroquois plain in order to represent the terrestrial uplift of the Ontario basin since the Iroquios f^hore was formed. Therefore, it is apparent that the great Iroquois Beach was constrmted approxhnately at sea level. The total amount of uplift since the episode will then be the height of the beach, at any place, measured above the sea level, which, at Fine, is 972 feet. Were the Erie beaches recognizable in the Adirondack wilderness near Fine, they should be found at altitudes of 1,600 feet and more above the sea. But this is a calculation outside of our subject, which is based upon measurements. The terrestrial movements recorded in the beaches have not been those of subsidence towards the west, but of uplift towards the east, in the same direction as those changes which have left unquestioned marine remains deposited at high altitudes in the St. Lawrence valley. One focus of the warping about the western end of Lake Ontario and about Georgian Bay appears to have been in the region of lat. 48° N., long. 76° W. Another focus of uplift is somewhere beyond the last point of rise measured in the Adirondacks. Thus, the axis between these foci appears to coincide, more or less, with the old Archrean axis of the continent, as suggested by Professor Dana. The uplift of the Iroquois Beach has been since the episode of the uppermost deposits of drift or till, for higher and older beaches than the Iroquois rest upon the newest stony clays of Ontario, Michigan and and other states. The Iroquois Beach rests upon the mud floors of the earlier sheets of water which covered the till deposits. The rate of northeastward regional uplift has been gradually diminishing, for we find other beaches, lower than the Iroquois, whose rate of riee is much reduced below that of the great beach. But the Iroquois plain was the great event in the history of the Ontario basin. In the rising of the land, after the Iroquois episode, there were pauses, but not of such duration as to permit of the formation of great shore-lines like that thus described. After the waters had fallen about 200 feet below the Iroquois plain, there was a conspicuous rest. This is recorded in a terrace near Watertown at 535 feet above the sea. At Oswego, we find a beach descending to near water level, at about 185 feet below the great Iroquois beach. Farther westward, it passes below the lake. The dip of the Iroquois Beach, between the region of Oswego and the western end of the lake, is about 78 feet; and accord- ingly we should find the remains of this younger shore- line (for a large proportion of the regional 'uplift has been effected since its 7 ^0 Birth of Lake Ontario. formation) submerged to 65 or 70 feet at the western end of the lake. Behind the modern bars and beaches, the water of Irondiquois Bay (a narrow river like channel) is IS feet deep; the Niagara River, 72 feet; and Burlington Bay, 78 feet. These conditions indicate that the lake covering these channels was at one time withdrawn, leaving only a few feet of water in the rivers which flowed through the otherwise dry •valleys. Here, then, in front of the bay, submerged or buried by more recent accumulations (upon re-submergence), is the position of this lower beach extending westward of Oswego, which was formed at a level now 70 feet below the f-urface of the western end of the lake. Indeed, the uniformly narrow Burlington Beach {b, tig, 6), with a length of five miles across the end of Lake Ontario, is thus easily explained as having originated as a small barrier, in front of the shallow river flowing down the Dundas valley and across the now sub- merged floor of Burlington Bay. With the more recent backing of the waters of the lake, this bar grew to the proportions of the modem beach, built out of materials derived from the older shores and not from river deposits. At the time when this young beach — now beneath the lake — was being formed, the waters had receded for only from three to live miles from what are now the western shores of Ontario, but they extended farther landward than at present upon its northern side, as shown by the raised beaches, and by the absence of submerged channels. The Niagara River was about four miles longer than now, cutting its way over a projecting point of shaly rocks. But this channel is at present filled, and is again further submerged beneath the lake. During the continued rise, the waters of the Ontario basin may have 'been even somewhat further shrunken at its western end, and the waves may have moulde^fl some of the submerged escarpments upon the southern side. The waters upon the southern side could have nowhere been more than about 200 feet below the present level, even if that amount of shrinkage, which represents most of the barrier hold- ing the basin above the sea, ever obtained. However, no important geographical event is recorded in any of the possible coast-lines sub- merged at levels below that just described. With the regional uplift, the barrier across the St. Lawrence valley •eventually cut off free communication with the sea, at a common level. This uplift was continued until the Iroquois Beach now rests at 972 feet above the sea at Fine, and the modern lake at 247 feet. 7%tcs the ^modern lake had its birth. This warping at the northwestern end of Origin of the Barrier Retaining the Lake. 51 the lake, during the later and since the Pleistocene period, has been enough not only to account for the rocky barrier holding the lake above the sea, but to account for all the barrier across the St. Lawrence valley closing the ancient basin of Ontario to a depth of nearly 500 feet below sea level.* In the Iroquois Beach no shells have been found. Only the remains of mammoth, elk and beaver have been met with.f Consequently, the question arises as to the freshness of the waters. Not far from the eastern end of Lake Ontario, the remains of a whale were found at 450 feet above the sea — at an elevation which would admit of the free access of oceanic waters into the Ontario basin. J Still no other marine or fresh-water fossils have been found in the beaches. It therefore appears to me that the absence of such organisms speaks no more in favor of fresh water conditions than of brackish or even salt when the Iroquois shores were being formed; and does not preclude the idea of free communications with the sea any more than when the whale came landward in waters 200 feet higher than the present lake surface. Indeed, I look upon the Ontario St. Lawrence valley, during the Iroquois episode, as resembling the Gulf of Obi, which is a sheet of water from 40 to 60 miles wide, and 600 to 700 miles long, into which BO much fresh water is dibcharging as to render even the Arctic Sea for 60 miles beyond the mouth of the gulf so fresh as to be almost potable,§ and sufficiently fresh to destroy marine life. The only dam that has been hypothecated as filling the St. Lawrence valley is that of a glacier. As the Iroquois Beach was at sea level, no dam ought to be required to hold up the water, but at most only to keep out the sea. However, I have followed the beach for 100 miles within the margin of the hypothecated barrier without find- ing the traces of an ending of the old shore markings upon the con- fines of the Adirondack wilderness. Even the coincidence of the shallow and small channel, discovered by Mr. Gilbert, connecting the Iroquois waters with the sea, by the Mohawk valley, or of the broader and lower valley of Lake Champlain, does not prove the necessity of a former barrier across the St. Lawrence valley ♦ See origin of the Basins of the Great Lakes, by J. W. Spencer, Q. J. G. 8., vol. xlvl. Part i. 1890. tCol. C. C. Grant of Hamilton has recently found other vertebrate remains, but not yet determined. tSir W. Dawson. Can. Nat., vol. x, p. 388. The remains are in the Redpath Museum at Montreal. $ Nordenskjold In " Voyage of the Vega," p. 140. 52 Hypothesis of Glacial Dams. any more than the narrow channels among the gigantic islands north of Hudson Bay would prove the former presence of a dam holding in the waters of that bay, were the whole country elevated. For a glacial dam to exist across the Adirondacks, even at the narrowest point, it would need to be 80 or 100 miles wide. If it had no greater depth than the water north of Fine used to have, the ice would need to be thick enough to fill a channel of 800 feet depth. As the differential uplift probably continues throughout the Adi- rondack region, we would need to be prepared to accept a dam of at least 1,300 feet in thicknes?, and a hundred miles across. *Apparent beaches in VermoLt at 2,100 feet above the sea (Hitchcock),** and the post-Pleistocene emergence of Mt. Desert, observed in the coastal markings to its summit of 1,500 feet (Shaler),f increase the probability of our regional uplift continuing throughout the Adirondacks. Any water proof dam in front of the Iroquois Beach would have had to endure throughout the long period of its formation. But all known glacial dams are small and evanescent. Yet the one suggested as closing up the Ontario's basin would have had to retain a greater sheet of open water than that of modern Laks Ontario, receiving not merely the waters of the then upper lakes, but also those of the melt- ing of the hypothecated glacial dam. It is questionable what thick- ness of ice would hold in the waters, for the modern glacial dams of Mt. St. Elias discharge beneath 500 feet of ice for a distance of eight milesj As soon as the waters fell below the Mohawk outlet, the dis charge of the glacial lake ought to have melted aLd lowered the ice on the one side and carved out terraces on the other, unless the river were 60 to 100 miles wide. And there are terraces upon the northern side of the Ottawa valley, as well as upon the flanks of the Adirondacks. There seem to me to be no phenomena in the later lake history of Ontario necessitating the existence of a dam across the St. Lawrence valley. In short, the Iroquois water was a gulf. The Adirondacks and New England formed greiat islands. The Iroquois episode com- menced almost synchronous with the birth of the Niagara Falls. And the history of Lake Ontario records interesting and great changes which now form a simple story. *NoTE.— Subsequent investigations confirm the absence of glacial dams. **Geology of Vermont. tQeology of Mt. Desert. Eighth Annual Report of U. 8. Qeol. Survey. ^Harold Topham in Proc. Roy. Qeog. Soc, 1889, p. 424. Bea.che8 North of the Adieondacks. 53 Appendix to the Iroquois Shore North of the Adirosda.ck3. la previous papers on the Iroquois shores of the Ontario basin, their position was definitely located only to a point near Belleville, on the northern side of Lake Ontario. But, from the general character of the country, I pointed out the necessity of extending the Iroquois water across a broad expanse of country to the highlands north of the Ottawa river, on the flanks of which shore deposits are known at various localities. I have a' so shown that the Iroquois water stood at or near sea-level; and in my working hypothesis considered the Iroquois water as an extension of the gulf of Saint Lawrence into the Ontario basin, although more or less obstructed by ice. Since the last paper was written, Mr. G. K. Gilbert and myself have revisited the region as far as a point 100 miles northeast of Watertown. Owing to Mr. Warren Upham's recent acceptance of the extension of the open Iroquois water as far as Quebec, it becomes desirable that the old shore line, so far as definitely surveyed, should be published. After a long stretch of unbroken continuity, the Iroquois beach is abruptly interrupted by rocky cliffs on the side of the escarpment about five miles east of Watertown. Beyond this point, owing to the broken continuity, the remmants of the ancient shore are more or less fragmentary. The old subaqueous plain extends up the broad Black river valley far above Carthage, with gravel deposits characterizing portions of its margin. The northeastward elevation of the Iroquois beach in this region rises at over six feet per mile. Beyond Car- thage, the country becomes more broken, being traversed by ridges of crystalline rocks, forming a late extension of the archipelago of the Thousand Islands at a higher level. The drift deposits become more sandy, with very little clay, and consequently^ are less favorable for the production of well defined beaches. The island character of this region is particularly unfavorable for the development of well defined shore markings. But wherever valleys enter the archipelago, their outlets are characterized by delta deposits of terraces, whose hypsometric position can be predicted in proceeding eastward. At Mr. Frank Wils)n's, four miles east of Watertown, the unques- tioned beach is broken into ridgelets between "730 and 704 feet, with a frontal gravel-bearing terrace at 682 feet. Below this horizon there is an extensive terrace plain east of Watertown at about 535 feet. At the mouth of Indian river, at Natural Bridge, these delta deposits form terraces, with more or less beach structure, at an elevation between 829 54 Beaches North cf the Adirondacks. and 802 feet, with a frontal gravel plain descending from 787 feet downward. In both cases, the waves, in carving out the lower terraces, have removed portions of the higher ridgelets. Between these limits there is no strongly marked terrace, but the lower is more confined to this regi )nal topography than the upper; and where gravelly, the pebbles are subordinate to the sand. For quantity and size of water- worn pebbles, the gravel deposits at Natural Bridge are physically the equivalents of those of the Iroquois beach to the southwestward. Above and below this level, at Natural Bridge, there are no fragments of ancient water lines liable to be mistaken for the Iroquois shore. The elevation of these deposits is that which would be expected from the measured warping recorded about Watertown. Beyond Natural Bridge there are extended gravel plains, in height conforming to the terraces at the old mouth of Indian river; but these are often more or less pitted. These plains appear to me as due to the presence of floebergs or other masses o± ice stranded upon the old shore. Even if they were shore deposits formed in glacial lakelets, their elevation is such as to show a common water level. They now face a lower descending country to the rorth\^estward, and are deformed by the gradual warping toward the northeast. At Pitcairn, the valley is 200 feet or more in depth, forming a deep channel in the late expansion of the Laurentian archipelago. High on the sides of the valle}^ zones of boulders, which are so often characteristic of old shore lines, are found at heights in keeping with the deformed Iroquois beach. A little north of East Pitcairn, there is a fine display of terraces, with beach s' ructure. These are partly in front of a now unimportant valley. There are several ridgelets, the highest being 942 feet; but the most important is 930 feet above tide. These ridgelets descend to a terrace or frontal plain 60 feet below. A short distance beyond, the terraces of Oswegatchee river are seen. Just north of Fine, they close around and connect a rocky island with the eastern side, and form a sort of barrier beach. This bar has an elevation of 972 feet. All of the above recorded terraces were leveled. The following are of barometric measurement. The rise in height in these beaches corresponds to the deformation of the Iroquois beach, increas- ing from five to six and seven feet for miles toward the northeast, which amount ought perhaps to be slightly modified, owing to imperfect Reprinted from the Boll. Geol. 8oc. Am , Vol. III., p. 48J-191, 1891. Tekkaoes North of the Adirondaoks. 55 identification in the crests of these terraces or the absence of some portions of the highest ridgelet?. The next great valley is that of the Grassy river. At Clifton Forge (Clarksboro), the old mouth of the valley is well defined by a beautiful gravel terrace at 1,055 feet (bar.), with an inferior terrace or ridge at 45 feet below. Lower than this no well marked gravel terrace occurs; but at 850 feet there is an extensive sand plain, forming a terrace con- fined to the valley. The terrace in the last valley is nearly due north of that at Fine, and appears to represent a warping of eight feet per mile, but probably the barometric measurement is responsible for the apparent increase in rate of elevation. Still, the northern uplift may probably exceed that to the northeast. The chain of observation was continued by Mr. Gilbert and myself to Racket river. The elevations were not satisfactorily obtained, as the changing weather greatly affected the barometer, especially above South Colton. At South Colton there is a sandy plain at about 940 feet (bar.), apparently corresponding to the plains below Clifton Forge and Fine. Racket river presents an interesting change of channel near Stark post-oflBce. Its old course was in a broad valley, now occupied by Coldwater creek as far as South Colton ; but after the Pleistocene revolution, it cut across hard rocks and deserted its old channel. Following up the Coldwater valley, we reached a broad sandy terrace underlain by gravel. This plain forms terraces extending northward along the sides of the valley. Its elevation is 1,215 (?bar.; the weather was very threatening). Other deposits were noted at 1,350 feet, which were probably older valley terraces. Again, on the brow of the plateau facing Potsdam, there was a plain at 1,160 feet with a boulder pavement in front of it. The value of these measure- ments is impaired that they are only important in identifying con- tinued elevations of the terrace plains near the late outlets of the valleys as far eastward as Racket river. In descending from the last plain there was no extensive valley terrace below the level of South Colton of magnitude corresponding to tho-ie at Watertown or at Clifton Forge. It might be noted that throughout this high region all of the pebbles are of local origin and none that could be identified as Canadian, i'he Paleozoic rocks were absent from the drift above South Colton and Parishville. Indeed, some of the apparent sand- stones arecleavable (juartzitic gneisses, and require close observation to prevent mistake. Along the whole northern flank of the Adirondacks, there is a great poverty of glaciated surfaces. Near Natural Bridge the direction of 56 Gla.ciation North of the Adieondaoks. the strijB was south 75° west and south 55*^ west. On the hills farther south the direction was south 20° to 25'' east, and near Harrisville south 10° west. Boulders were of large size. One, at a school house three miles southwest of South Colton, showed at least 6,000 cubic feet above surface of the ground. From the recent explorations, allowing for errors in observation and measurement, it appears that shore deposits occur at the mouths of all the valleys which entered the Laurentian archipelago of the Thousand Islands. Throughout a considerable range of altiiude, there is only one set of terraces or delta deposits, always occurring at the mouths of old valleys, with occasional connecting gravel plains or terraces of beach-like structure, composed of coarse pebbles, in magnitude com- parable to the physical development of the Iroquois beach farther westward; the lower terraces being mainly sandy and confined to the valleys; and the higher, if known at all, much above the possible alti- tude of the Iroquois plain. These terraces form sets of ridgelets rang- ing downward from their crests about 50 feet to the gravelly deposit of their frontal terraces. This holds true alike for the exposures of the Iroquois beach east of Watertown and for the recorded terraces at the mouth of the valley. The next great terrace plain below these gravel shores is about 200 feet and mostly sandy, alike near Watertown and alou P,(S s a a> §5 00 "cS C _C o ^ O OJ S ce a u ^ a s OWoo £0 c c6 C tS CO OD CO Ct 02 c -3 Co oca Map of the Niagara. River. 103 Ancient Topography and Basement. In the numerous writings upon the Niagara river one ancient topo- graphic feature has been overlooked and another exaggerated into importance which it does not possess. The ancient drainage ^of the 104 Ancient Tonawanda Valley. Erie basin was not by way of the Niagara, but by a channel 40 miles to the west.* Even at the end of the Lake Erie the borings show old channels deeper than the floor of the river across the Devonian escarp- nients.f The feature overlooked is the Tonawanda valley, a mile and a half in width, extending from the rapids above the falls to the Johnson ridge. Its basement is 80 or 90 feet below the northern bar- rier of Johnson's ridge. The rocky sub- surface of Goat Island was part of the ancient floor (see fig. 27). This depression is part of the ancient Tonawanda basin, which is now filled with drift (see fig. 24). The gorge through Johnson's ridge is modern with vertical walls, but half a mile to the west it falls away and the wells reveal the continua- tion of the Tonawanda depression extending northward. It is again made known by a well half a mile west of the whirlpool {w, fig. 19), in the line of the extension of the St. David's valley. This forms an embayment one and half miles wide and only three-quarters of a mile deep in the face of the Niagara escarpment. The modern river is simply crossing a portion of the old Tonawanda basin in the vicinity of the falls, and consequently it has here much les-s rock to excavate than through and north of Johnson's ridge. The other feature is the imaginary whirlpool — St. David's valley, supposed to have been the old course of the river. A.bove and below Fio. 18. — Map of the whirlpool ; bb, position of sectfon (in fig 17). the whirlpool alike, the gorge is of recent date as may be seen by the vertical walls shown in the several sections. The whirlpool ravi e has sloping V-shaped boundaries in its higher portion, which is an antique structure. The depression is so obstructed with drift, that gives rise to landslides that the old topography is much obscurtd. Yet a little stream has removed the fallen earth and exposed a natural section of * " Origin of the Basins of the Great Lakes," Q. J. G. 8. Lond., vol. xlvi, p. 523, 1890, and " Notes on the Origin and History of the Great Lakes," Proc. A. A. A. 8., vol. xxviii, 1888. + " The Life History of Niagara," by Julius Pohlman, Trans. Am. Inst. Min. Eng. TONAWANDA St. DaVId's YaLLET. 105 Clinton limestones, which cross the valley at an elevation of 115 feet above the surface of the whirlpool, or 160 feet above Lake Ontario, with Niagara shales showing for at least 20 feet higher. Thus the rocky barrier across the ravine is not less than 240 feet above the bot- tom of the cafion in the whirlpool. This barrier in the ravine is illus- trated in fig. 19, which should be compared with figures 22 and 23, in order to appreciate the insignificance of the whirlpool ravine,* Fig. 19.— Section across the whirlpool ravine, located at bb, fig. 2; W. well; R, stream. The form of the whirlpool cauldron requires explanation. At Mr. Shepherd's house, a short distance west of the whirlpool, there is a well 90 feet deep without reaching rock {w, fig. 19) and this shows the absence of Niagara limestones to a depth of more than 60 feet below the surface rocks of the western wall of the whirlpool. At that point the limestones rise 40 feet higher on the eastern side of the river than on the western, but the depression was leveled up with drift. Thus it appears that at this point the Niagat a river took possession of the eastern side of a drift-filled valley (Tonawanda-St. David's), and the whirlpool ravine was a little tributary to it. When the falls had receded to the whirlpool and penetrated the rocky barrier, the currents were able to remove the filling of the buried ravine, and this gave rise to the form of the cauldron, which deepened its basin to lower levels by the currents of the river acting upon the underlying soft shales, with the landslides obscuring the older features. It is evident that there was no preglacial Niagara river. The Niagara river crossed the broad shallow depression of the Tona- wanda drainage, at the falls and that adjacent to the whirlpool on a basement of drift, but elsewhere generally on hard limestones. Out of both of these materials, terraces were carved thus marking the old river level, before it sunk within the chasm. Discharge of the Niagara River. The Corps of Engineers, U. S. A., made the measurements of the outflow of the Great Lakes between June 27th and September 17th, * InRept. of meeting of Am. As. Ad. Sc. in Science, Sept., 1886, it is noted that Prof. E. W. Claypole found rocks in the ravine, without giving any details in explanation . Since this paper has been in type, Prof. James Hall informed me that Prof. J. W. Powell and himself had also seen the occurrence of the rocks, but no notice has been printed. The error has been even recently repeated by a writer in " Nature." 14 106 Discharge of the Niagaea. Kivek. 1868.* That of Lake Huron was 216,435 cubic feet per second; and of Lake Erie for the first part of the season, 304,307 cubic feet, and 258,586 feet for the second part. From these figures I have taken the maximum proportional discharge (as the volume is variable) of Lake Erie, which is found to gather y\ of the total drainage of the Niagara river, but the mean discharge is less than ■^\. This is an imjjortant factor in the following computations. Modem Recession of the Falls. The four surveys illustrated in figure 18 show the modern recession of the horstshoe cataract. During 48 years 275,400 square feet fell away. The mean width of the adjacent portions of the gorge (as opposite Goat Island) is 1,350 feet. Thus the mean recession would be 4.175 feet a year. The American falls have undermined 82,900 square feet of rock, which gives a mean rate of 0.64 foot a year. But the rate is not uniform. In 1819, the crest of the Canadian fall was Pio. 20. —The four surveys of the Canadian Falls showing the retreat of the cataract (In which some inaccuracies are apparent). (Kibbe.) very acute, it had become quite obtuse in 1842, acute in 1886, but it was broadening out again in 1890; thus there are cycles of slow and rapid retreat. The measured recession has probably obtained since the cataract cut its way through Johnson's ridge, for beneath the Tonawanda basin the * Report of Chief of Engineers for 1869, p. 582. Eecession of jSTiagaka. Falls. 107 limestones have a thickness of only 45-55 feet, as the upper 90 feet had been removed in pre-Pleistocene times. The capping limestone in Johnson's ridge was 140 feet thick. To the north the thickness was reduced. Along those portions of the chasm where the limestone i& heavier and the gorge narrower than in the pre-glacial depression, the stronger arches must have arrested the maximum rate of retreat, and on this account, I have reduced the measured mean rate of recession by an estimated amount of 10 per cent., or to 3.75 feet a year for the recession of the falls from the end of the caiion to Johnson's ridge, under conditions of the modern discharge and descent. The mean descent of the river was from the plain, now at 340 feet above Lake Ontario; but whilst passing the rapids of Johnson's ridge, 25 feet must be added to the declivity of the river. After the basin behind the river was reached, the water plain was reduced to about 320 feet, including 60 feet of descent above the falls in the form of rapids. The surface of the country has been deformed since the commencement of the cataract by a northward terrestrial up'ift to the extent of 12 or 15 feet, divided throughout the length of the gorge, where, as seen in the canon, the character of the different strata is remarkably uniform, except in the described depressions, across Johnson's ridge, and at the end of the chasm where the capping limestones were much thinner, but partly compensated for by the greater prominence of the hard Clinton and Medina layers. The following computations are based upon the mean rate of reces- sion, modified by the variations in the descent of the waters and their changing volumes, which have been discovered in the geological inves- tigations of the Great Lakes. Sketch of the Lake History and the Nativity of the Falls. This outline is taken from the chapters on the Lake History noted at the foot of the page.* At the commencement of the Lacustrine epoch, Warren water gulf covered most of the Lake region, and Forest beach was its last strand. Afterwards the waters sank 150 feet, thereby dismembering Warren water gulf into Algonquin Gulf (con- fining it to the basins of Superior, Michigan and Huron) with an outlet ♦"The Iroquois Beach, a chapter in the History of Lake Ontario," Trans. Roy. Sic Can. 1889, p. 132. " Deformation of the Iroquois Beach and Birth of Lake Ontario " Am. Jour. Sci., vol. xl, p. 443, 1890. "Deformation of Algonquin Beach and Birth of Lake Huron," Id. vol xl'i p. 12, 1891. "High Level Beaches in the region of the Great Lakes and their Deformation." Id , p. 201. "Deformation of the Lundy Beach and the Birth of Lake Erie," Id , vol xlvii, p 807, 1894. All by J. W. Spencer. " The HistDry of Niagara Biver," by G. K. Gilbert, Six. Rep. Com. State Res. Niag., 1891. 108 Effects of Changes of Water Levels. by way of the Ottawa valley, and Lundy gulf (occujDying the Erie basin and) extending into the Ontario valley. These two bodies of water appear to have^had a common level as if connected in some way across the Ontario basin, but their northeastern extensions are not known and involve unsettled questions that do not affect the history of Niagara. Again, the waters were lowered so that the Niagara River emptied the overflow of the Erie basin, without a fall into the Ontario valley. This condition did not last long, for the waters sank to a level (Iroquois beach) of 300 feetjbelow the Lundy (and also Algonquin) plain, and the falls commenced their descent with the waters of the Erie basin alone. The subsidence was accompanied by slight pauses, but waters remained for^a long'^time at the level of the Iroquois beach, which is now about] 135 feet above Lake Ontario at the end of the gorge. Again the waters subsided to the level about 80 feet beneath the present level of the head of Lake Ontario, and thereby lengthened the river to 12 miles beyond the end of the chasm. At this time the descent of the river after passing the rapids at Johnson's ridge was 420 feet. By the continued northeastern terrestrial elevation the waters of the Huron basin were turned from the Ottawa drainage into the Erie basin, whose northeastern rim was elevated so as to flood the lake. Later, the waters at the head of Lake Ontario were raised 80 feet to the present level. This differential movement was at zero at the head of Lake Erie; 2.5 feet per mile in the Niagara district; 4 ftet north- east of Lake Huroo, and 5 feet per mile at the outlet of Lake Ontario. At the nativity of the Niagara River there was no fall. A little later in the Iroquois episode the falls were very much like the modern American cataract, both in height and volume, but afterwards it increased in magnitude and went through the changes noted later. Laws of Erosion. When erosion is considered from'a theoretical point of view and the whole energy of the water is supposed to be expressed in the erosion, it varies as the mass of the water into the square of the velocity {wv^). Hence for a given river increase of the amount of its water or increase of the velocity along its course should be expressed by greater erosion. But erosion is not the only expression of the theoretical value of the energy of the river. Again, it is well known that the more rapid the descent of the stream the more the erosive effects are expended on the floor of the channel, in deepening and forming the U-shaped val- leys or gorges. On the other hand, the reduction in the slope causes Duration of Niagara Falls. Plate V THE WHIRLPOOL RAPIDS. Laws of Erosion. 109 the channel to become broader — a principle which has an important bearing in this study. While the observations are imperfect, owing to the variable conditions of erosion, still the attempt to ascertain the duration of the different episodes is the only natural sequence to the measurements of the modern recession of the falls, and it gives approxi- mate results, for without considering the changing episodes the rate of recession is of no geological interest. But this study may lead to further detailed investigations. Episodes of the Fdver and the Duration of each — Age of the Falls. First Episode. — From the history of the lakes and the river we learn that the early falls cascaded from the brow of the escarpment to the level of the Iroquois beach 200 feet below, (with the Erie drainage only y\ of the total discharge of the upper lakes). There is no indica- tion that the Erie rainfall was greater at that time than now. The length of the chasm excavated during the first episode is found in the Fig. 21.— Mip of the gorge at Foster's flits; F, location of the cross section fig. 20. Fig. 22 — Section of the gorge at Foster's flats (FT. flg. 17). Platform CF) of the old river floor projecting into the canon. Its section is shown in broken shading but with ravines descending from both sides of it. T, rock terrace surmounted by huge blocks of Niagara lime- stones; h, original river terrace; r, surface of river; L. 0., surface of Lake Ontario. Bottom of river about 80 feet below the surface of the lake. data furnished by the study of Foster's flats. Their location is shown at F, figure 17, and the structures are further illustrated in figures 21 and 22. The terrace (T) represents the former level of the river (about 190 feet above Lake Ontario). It is the only feature of the kind in the caiion. It is about 60-60 feet above the Iroquois le\'el to which the no Episodes of JS'iAGABA.'FtA.LLs. river descended. Thus the slope of the earlier and smaller streams was about half as great again as the modern river over the rapids at •this locality. The youthful river was broad and shallow, like and of about the same magnitude as the modern American channel and falls, acting evenly over the whole breadth and receding at about the same rate. The remnant of the platform shows how far the fall had receded before the physical change which threw the current to the eastern side of the channel. This change could be effected by increasing the height of the falls which would favor the deepening of the chasm at the •expense of the width, especially as the lower rocks are mostly shale. This change of breadth from a wide and shallow to a narrow and deep channel is shown along the lower part of the caiion and is illustrated by the contracted channel at the bottom of the caiion in a section just above the end of the gorge (fig. 23). p ■...,.., : ,T-«vt) H^ — „ 3uJ Nid2ir ,...,..,.... • 89 126 Index. PAGK. Terrestrial elevation 86 Tonawanda Valley, Ancient 104 Tonawanda — St. David's Valley 105 Topham, Harold 91 Trent Valley overflow 70, 71 Warping of beaches 22, 23 Warren, G. K 8, 21 Warren Gulf 57, 58, 60, 74-84 " dismemberment of 61,71,78,79, 92 " " high level shores. 74-82 " " shrinkage of 84 Whitaker. WilUam 25 White, I. C 89 CATALOGUE OF THE HUMBOLDT LIBRARY ofSCIENCE CONTAINING THE BEST SCIENTIFIC WORKS, AT POPULAR PRICES. - - • THE GREAT CLASSICS OF MODERN THOUGHT. - - - STRONG MEAT FOR THEM THAT ARE OF FULL AGE. Taper Covers, Price IS cents each. Double Numbers, 30 eenta each. These Books are complete and unabridged, tastefully gotten up, and are sold, on an. average, at one-tenth, the prices charged by other publishers No. 1. Light Science for Leisure Honrs. A series of familiar essays on astronomical and olher natural phenomena. By Ri- chard A. Proctor, F.R.A.S. No. 2. Forms of Water in Clouds and Rivers, Ice and Glaciers. (\'i iliustratioKs). By John Tyndall, F.R.S. No. 3. Physics and Politics. An application of the principles of Natural Science to Political Society. By Walter Bagehot, author of "The English Constitution." No. 4. Man's Place in Nature, (with numerous illustrations). By Thomas H. Huxley F.R.S. No. 5. No. 6. Education, Intellectual, Moral, and Physical. By Herbert Spencer. Town Geology. With Appendi.x on Coral and Coral Reefs. By Rev. Chas. Kingsley. No. 7. The Conservation of Energy, (with jiumerous illustrations). By Balfour Stewart, LL.D. No. 8. The Study of Languages, brought baclc to its true principles. By C. Marcel. No. 9. The Data of Ethics. By Herbert Spencer^ No. 10. The Theory of Sound in its Relation to Music, (numerous illustrations). By Prof. Pietro Blaserna. No. 11. ) The Naturalist on the River Ama- V zons. A record of 11 years of travel. No. 12. ) By Henry Walton Bates,F.L.S. (Dou- ble number. Not sold separately). No. 13. Mind and Body. The theories of their relation. By Alex. Bain, LL.D. No. 14. The Wonders of the Heavens, {thirty-two illustrations). By Camille f"lammarion. No. 15. Longevity. The means of prolonging life after middle age. By John Gardner M.D. No. 16. On The Origin of Species. By Thomas H. Huxley, F.R.S. No. 17. Progress: Its Law and Cause. With other disquisitions. By Herbert Spencer. No. 18 Lessons in Electricity, {sixty illustra- ■ ^zV».r). By John Tyndall, F.R.S. No. 19. Familiar Essays on Scientific Subjects. By Richard A. Proctor. No. 20. The Romance of Astronomy. By R. Kalley Miller, M.A. No. 21. The Physical Basis of Life, with other essays By Thomas H. Huxley, F.R.S. No. 22. No. 23. No. 24. No. 25. No. 26 No. 27. No. 28. No. 29. No. 30. No. 31. No. 32 No. 33. No. 34. No. 35. No. 36. No. 37. No. 38 No. 39. No. 40. No. 41 No. 42. No. 43 No. 44. No. 45. Seeing and Thinking. By William Kingdon Clifford, F.R.S. Scientific Sophisms. A review of cur- rent theories concerning Atoms, Apes and Men. By Samuel Wainwright, D.D, Popular Scientific Jjectnres,{iUust rated).. By I'rof. H. Helmholtz. The Origin of Nations. By Prof. Geo. Rawlmson, Oxford University. The Evolutionist at Large. By Grant Allen. The History of Landholding in Eiig» land. By Joseph Fisher, F.R.H.S. Fashion in Deformity, as illustrated in the customs of Barbarous and Civil- ized Races, {numerous illustrations). By William Henry Flower, F.R.S. Facts and Fictions of Zoology, (nu- merous illustrations) . By Andrew Wilson- Ph. D. The Study of Words. Part I. By- Richard Chenevix Trench. The Study of Words. Part II. Hereditary Traits aud other Essayff* By Richard A. Proctor. Tignettcs from Nature. By Grant Allen The Philosophy of Style. By Herbert Spencer. Oriental Religions. By John Caird* Pres. Univ. Glasgow, and Others. Lectures on Evolution. (Hlustrated), By Prof. T. H. Huxley. Six Lectures on Light. (Illustrated). By Prof. Tyndall. Geological Sketches. Part I. By Archibald Geikie, F.R.S. Geological Sketches. Part II. The Evidence of Organic Evolution By George J. Romanes. F.R.S. Current Discussion in Science. W. M. Williams, F. C. S. By History of the Science of Politic*^ By Frederick Pollock. Darwin and Humboldt. By Prof* Huxley, Prof. Agassiz, and otliers. The Dawn of History. Part I. By G. F. Keary, of the British Museum. The Dawn of History. Part II, THE HUMBOLDT LIBRARY OF SCIENCE. No. 40. No. 47. No. 48. No. 49. No. 50. No. 51. No. 52. No. 63 No. 54. No. 55. No. 66. No. 67. No. 68. No. 69. No. 60. No. «1. N*. 62. N*. 83. The Diseases of Memory. By Th. Ribot. Translated from the French byJ.Fita- gerald, M.A. The Childhood of EeUgion. ward Clodd, F.R.A.S. By Ed- Life in Xature. Hinton. (Ili'ustraied.) By James The Son; its Constitutionals Phenomena, its Condition. By Judge Nathan T. Carr, Columbus, Ind. Money and the Mechanism of Exchange. Part I. By. Prof. W. Stanley Jevons, F.R.S. Money and the Mechanism of Exchange. Part II. The Diseases of the Will. By Th. Ribot. Translated from the French By J. Fitzgerald. Animal Automatism, and other Essays, By Prof. T.H. Huxley. F.R.S. The Birth and Groirth of Myth. Edward Clodd, F.R.A.S. By The Scientific Basis of Morals, and other Essays. By William Kingdon Clifford, F.R.S. ninsions. Part I. By James Sully. ninsions. Part. II. The Origin of Species. (Double num. ber). Part I. By Charles Darwin. The Origin of Species. Double num- ber). Part II. The Childeod of the World. By Edward Clodd, Miscellaneons Essays. Proctor. By Richard A. The Beligiong of the Ancient World. By Prof. Geo. Rawlinson, Univ. of Oxford, (Double number). ProgressiTe Morality. By Thomas Fowler, LL.D., Preisdent of Corpus Christi Coll., Oxford. The distribution of Animals and Plants. By A. Russell Wallace and W. T. Thistleton Dyer. Conditions of mental Development, and other essays. By William Kingdon Clifford. Technical Education, and other essays. By Thomas H. Huxley, F.R.S. Tlie Black Death. An account of the. Great Pestilence of the 14th Century. By J.F.C. Hecker, M.D. Three Essays. By Herbert Spencer. Special number. Fetichism: A Contribution to Anthropo- logy and the History of Religion. By Fritz Schultze, Ph. D. Double number. Essays Speculatiye and Practical. By Herbert Spencer. Anthropology. By Daniel Wilson. Ph. D. With Appendix on Archaeology. By £. B. Tylor, F.R.S. No. 72. The Dancing Mania of the Middle Ages. By J.F.C. Hecker, M.D. No. 64. No. «6. No. 66. No. 67. No. 68, No. 69. No. 70. No. 71. No. 73. Erolution in History, Language and Science. Four Addresses delivered at the London Crystal Palace School of Art, Science and Literature No. 74. \ The Descent of Man, and Selection in No. 75. ( Relation to Sex. (Numerous Illustr'ns) No. 76. f By Charles Darwin. Nos T4, 75, 76 are No. 77. ; Single Nos. \ No. 77. is a double No. No. 78. Historical Slietch of the Distribution of Land in England. By William Lloyd Birkbeck, M.A. No. 79. Scientific Aspect of some Familiar Things. By W.M. Williams. No. 80. Charles Darwin. His Life and Work By Grant Allen. (Double number). No. 81. The Mystery of Matter, and the Philosophy of Ignorance. Two Essays by J. Allanson Picton. No. 82. Illusions of the Senses; and other Es' says. By Richard A. Proctor. No. 83. Profit-Sharing Between Capital a]i4 Labor. Six Essays. By Sedley Tay- lor, M.A. No. 84. Studies of Animated Nature. Four Es- says on Natural History. By W. S, Dallas, F. L. S. No. 85. The E;.8ential Nature of Religion. By J. Allanson Picton. No. 86. The T'nseen Unirerse, and the Philosophy of the Pure Sciences. By Prof. Wm. Kingdon Clifford, F. R. S. No. 87. The Morphine Habit. By Dr. B. Ball, of the Paris Faculty of Medicine. No. 88. Science and Crime and other Essays. By Andrew Wilson, F. R. S. E. No. 89. The Genesis of Science. By Herbeit Spencer. No. 80. Kotes on Earthquakes: with Fourteen. Miscellaneous lissays. By Richard A. Proctor. N«. 91. The Rise of Unirersities. ByS.S. Laurie, LL.D, (Double Number.) No. 92. The Formation of Yegetable Mould through the Action of Earth Worms. By Charles Darwin, LL.D., F. R. S. (Double Number.) No. 93. No. 94. Scientific Methods of Capital Punish- ment. By J. Mount Bleyer, M.D., (Special number.) The Factors of Organic Evolution. Herbert Spencer. By No. 95. The Diseases of Personality. By Th. Ribot. Translated from the French- by J. Fitzgerald, M. A. No. 96. A Half-Century of Science. By Thomas H Huxley, and Grant Allen. No, 97. The Pleasures of Life. By Skjohn LubbocK, Bart. No. 98 Cosmic Emotion: Also the Teachings of Science. By William Kingdon Clifiord. (Special number). No. 99. Nature Studies. By Prof. F. R. Eaton Lowe; Dr. Robert Brown, F. L. S.; Geo. G. Chisholm, F.R.G.S., and James Dallas, F.L.S. THE HUMBOLDT LIBRARY OF SCIENCE. Ko. 100. No. 101. No. 102. No. 103. No. 104. No. 105. No. 106, No. 107 No. 108. No. 109. No. 110. No. 111. No. 112. No. iia. By Henry Drummond, Science and Poetry, with other Essays. By Andrew Wilson, F. R. S. E. Esthetics: Dreams and Association of Ideas. By Jas. Sully and Geo. Croom Robertson . Ultimate Finance; A true theory of Co- operation. By William Nelson Black. •fhe Coming Slarery: The Sins of Legis- lators; The Great Political Supersti- tion. By Herbert Spencer. Tropical Africa. F. R. S. Freedom in Science and Teaching. By Ernst Haeckel, of the University of Jena. With a prefatory Note by Prof. Hu.xley. Force and Energy. A Theory of Dyna- mics. By Grant Allen. Ultimate Finance. A True Theory of Wealth. By William Nelson Black. English, Past and Present. Part I. By Richard Chenevix Trench. (Double number. English, Past and Present. Part II. By Richard Chenevix Trench. file Story of Creation. A Plain Acconnt of Evolution. By Edward Clodd. (Double number). Che Pleasures of Life. John Lubbock, Bart. Part II. By Sir Psychology of Attention. By Th. Ri- bot. Translated from the French by J. Fitzgerald, M. A. Hypnotism. Its History and Develop ment. By Fredrik Bjornstrbm, M. D., Head Physician of the Stockholm Hos- pital, ProfesBor of Psychiatry. Late Royal Swedish Medical Councillor. Authorized Translation from the Second Swedish Edition by Baron Nils Posse, M. G., Director of the Boston School of Gymnastics. (Double number.) No. 114. Christianity and Agnosticism. A Con- troyersy. Consisting of papers contri- buted to The Nineteenth Century by Henry Wace, D. D., Prof Thos. H. Huxley, The Bishop of Petersborough, W. H. Mallock, Mrs. Humphrey IVard, (Double number,) N». 115. Darwinism : An Exposition of the The- ory of Natural Selection, with some of its Applications. Part I. By Alfred Russel Wallace, LL. D., F. L. S., etc. Illustrated. (Double number.) No. 116. Darwinism : An Exposition of the The- ory of Natural Selection, with some of its Applications. Part II. Illustrated. (Double number.) No. 117. Modern Science and Mod. Thought. By b. Laing. Illustrated. Double number. No. 118. Modern Science and Mod. Thought. Part II. By S. Laing. No. 119. The Electric Light and The Storing of Electrical Energy. (Illustrated) Ger- ald Molloy, D.D., D.Sc. No. 120. The Modern Theory of Heat and The Sun as a Storehouse of energy. (Illus- trated ) Gerald Molloy, D.D., D. So. No. 121. No. 122. No. 123. No. 124. No. 125. No. 126. No. 127. No. 128. No. 129. No. 130. No. 131. No. 132. No. 133. No. 134. No. 135. No. 136. No. 137. No. 138. No. 139. No. 140. No. 141. No. 142. No. 143. No. 144. No. 145. Np. 146. No. 147. No. 148. No. 149. Utilitarianism. By John Stuart Mill. Upon the Origin of Alpine and Italias Lakes and upon Glacial Erosion. Maps and Illustrations. By Ramsey, Ball, Murchison, Studer, Favre, Whymper and Spencer. Parti. (Double number.v Upon the Origin of Alpine and Italian Lalies, Etc., Etc. Part II. The Quintessence of Socialism Prof. A. Schaffle. By By David &. Darwinism &, Politics. Ritchie, M. A. Administrative Nihilism. By Thomae Huxley, F. R. S. Physiognomy & Expression. By P. Mantegazza, Illustrated. Parti. (Dou- ble number.) Physiognomy & Expression. Part II. (Double number.) The Industrial Revolution. By Arnold Toynbee, Tutor of Baliol College, Ox- ford. With a short memoir by B. Jow^ ett. Parti. (Double number.) The Industrial Revolution. Part IT.. (Double number). The Origin of the Aryans. By Dr. Isaac Taylor. Illustrated. Part I. (Dou- ble number.) The Origin of the Aryans. Part II. (Double number.) The Evolution of Sex. By Prof. P. Geddes and J. Arthur Thomson. Illus. trated. Parti. (Double number.) The Evolution of Sex. Part II. (Do* ble number. The Law of Private Right. By Georga H.Smith. (Double number.) Capital. A Critical Analysis of CapitaK ist Production. By Karl Marx. Park I. (Double number.) Capital. Part II. (Double number.) CapitaL Part III. (Double number.)- CapitaL Part IV. (Double number.) Lightning. Thunderand Lightning Cob..- ductors. (Illustrated.) By Gerald M»W loy,D.D., D. Sc. T^Tiat is Music t With an appendix on How the Geometrical Lines have their Counterparts in Music. By Isaac L, Rice. Are the Effects of Use and Disuse I». herited i By William Piatt Ball. A Tindication of the Rights of Wom»». By Mary Wollstonecraft. With an In- troduction by Mrs. Henry Fawcet. Part 1. (Double number.) A Vindication of the Rights of Woraam. Part 1 1 . (Double num ber. ) Civilization ; Its Cause and Cure. By Edward Carpenter. Body and Mind. By William Kingdon Clifford. S«cial Diseases and Worse Remediea* By Thomas H. Huxley. F. R. S. The Soul of Man under Socialism. By Oscar Wilde. ElectricitT, the Science of the N?i^ teenth Century. By E. C. CaiDar* (Illus.) Parti. Double number. The same. Part 11. THE HUMBOLDT LIBRARY OF -SCIENCE. No. 150. No. 151. No. 152. No. 153 No. 154. No. 155. No. 156. No. 157. No. 158. No. 159. No. 160. No. 161. No. 162. No. 163. No. 164. Degeneration; A Chapter in Darwinism. Illustrated. By E. Ray Dankester, M. A., LL.D., F. R. S. (Mental Suggestion. By Dr. J. Ocho- f rovvic.z Parti. (Double number.) I The same. Part II. J (Double number.) ) The same. Part III. ( (Double number. ) ) The same. Part IV. X (Double number. Modern Science; The Science of the Future. By Edward Carpenter. Studies in Pessimism. By Schopen- hauer. J Flowers, Fruits and Leaves. Illustra- V ted. By Sir John Lubbock, F. R. S. \ (Double number.) Glimpses of Nature. Part I. Illustra- ted. By Dr. Andrew Wilson, F. R. S, E. (Double number.) No. 165. The same. Part II. No. 166. Problems of the Future. Part LI. By Samuel Lang. No. 167. Problems of the Future. Part II. (Double number.) No. 168. The same. Part III. (Double number.) No. 169. Uhe Moral Teachings of Science. By Arabella B. Buckley. No. 170. The Wisdom of Life. By Schopenhauer. (Double Number.) No. 171. The Mystery of Pain. By James Hin. ton. No. 172. ) What is Property 1 An inquiry into No. 173. ( the Principle of Right and of Govern- No. 174. r ment. By P. J. Proudhon. Four No. 175. ) Double numbers, $1.20. No. 176. The History and Scope of Zoology. By £. Ray Lankester. A NE^W SERIES. T^e ^ocial Science Ltoarij OF THE BEST AUTHORS. PUBLISHED MONTHLY AT POPULAR PRICES. Paper Cover, 25 cents each ; Cloth, extra, 75 cents each. NOW READY. No. 1. Six Centuries of Worlt and Wages. By James K. Thorold Rogers, M. P. Abridged, with charts and summary. By W. D. P. Bliss. Introduction by Prof. R. T. Ely. No. 2. The Socialism of John Stewart Mill. The only collection of Mill's Writings on Socialism. No. 3. The Socialism and Unsocialism of Thomas Carlyle. A collection of Car- lyle's social writings; together with Joseph Mazzini's famous essay protest- ing against Carlyle's views. Vol.1. . No. 4. The same. Vol.11. Na. 5. William Morris, Poet, Artist, Socialist. A selection from his writings together with a sketch of the man. Edited by Francis Watts Lee. No. 6. The Fabian Essays. American Editioa, with Introduction and Notes by H. G. Wilshire. No. 7. The Economics of Herbert Spencer. By W. C. Owen. No. 8. The Communism of John Rusliin. No. 9. Horace Greeley and other Pioneers of American Socialism. By Charles Soth- eran. Special Number, 3S cents, in Paper Cover ^ No. 10. IN PREPARATION. By DeLave* The Socialism of To-daj. leye. No. 11. The State U Relation to febor. By Jevons, iMlchie, etc. Individualism. A series of Essays bjr Donisthorpe, and others. No. 12. LIST OF BOUND BOOKS THE HUMBOLDT LIBRARY SERIES. The volumes of this series are printed on a superior quality of paper, and bound in extra cloth. They are from fifty to seventy-five per cent, cheaper than any other edition of the same books. STANDARD WORKS BY VARIOUS AUTHORS. A Vindication of the Rights of Woman. With Strictures on Political and Moral Subjects. By- Mary WoUstonecraft. New Edition, with an Introduction by Mrs. Henry Fawcett. Cloth 11.00 Electricity : the Science of the Nineteenth Cen- tury. A Sketch for General Readers. By E. M. Caillard, anther of "The Invisible Powers of Nature." With Illustrations. Clo'th, 75 cts Mental Suggestion. By J. Ochorowicz. Sometime Professor Extraordinarius of Psychology and Nature-Philosophy in the University of Lem- berg. With a Preface by Charles Richet. Translated from the French by J. Fitzgerald, M. A. Cloth $2.00 Flowers, Fruits, and Leaves. By Sir John Lub- 'bock, F.R.S., D.C.L., LL.D. With Ninety- five Illustrations. Cloth 75 cts Glimpses of Nature. By Andrew Wilson, F.R.S. E., F.L.S. With Thirty-five Illustrations. Cloth 75 cts Problems of the Future, and Essays. By Sam- uel Laing, author of "Modern Science and Modern Thought," &c. Cloth $1.25 The Naturalist on the River Amazons. A Rec- ord of Adventures, Habits of /Vnimals, Sketch- es of Brazilian and Indian Life, and 'Aspects of Nature under the Equator, during Eleven Years of Travel. By Henry Walter Bates, F. L.S., Assistant Secretary of the Royal Geo- graphical Society of England. New Edition, •Large Type, Illustrated. Cloth $1.00 The Rise and Early Constitution of UniTersitieg, with a Survey of Mediaeval Education. By S. S. Laurie, LL.D., Professor of the Institutes and H,istory of Education in the University of Edinburgh. Cloth 75 cts The Religions of the Ancient World : including- Egypt, Assyria and Babylonia, Persia, India, Phoenicia, Etruria, Greece, Rome. By George Rawlinson, M.A., Camden Professor of Ancient History, Oxford, and Canon of Canterbury. Author of "The Origin of Nations," "The Five Great Monarchies, &c. Cloth : 75 cts Fetichism. A Contribution to Anthropology and the History of Religion. By Fritz Schultze, Dr. Phil. Translated from the German by J. Fitzgerald, M. A. Cloth 75 cts Money and the Mechanism of Exchange. By W. Stanley Jevons, M.A., F.R.S. , Professor of Logic and Political Economy in the Owens Col- lege, Manchester, England. Cloth 75 cts On the Study of Words. By Richard Chenevix Trench, D.D., Archbishop of Dublin. Cloth 75 cts The Dawn of History. An Introduction to Pre- historic Study. Edited by C. F. Keary, M.A., of the British Museum. Cloth 75 cts Geological Sketches at Home and Abroad. By- Archibald Geikie, LL. D., F.R.S., Director- General of the Geological Surveys of Great I Britain and Ireland. Cloth 75 cts THE HUMBOLDT LIBRAE Y OF. SCIENCE. lUusions: A Psychological Study. By James Sully, author of "Sensation and Intuition," "Pessimism," &c. Cloth 75 cts The Pleasures of Life. Part I. and Part II. By Sir John Lubbock, Bart. Two Parts in One. Cloth 75 cts English, Past and Present. Part I. and Part II. By Richard Chenevix Trench, D.D., Arch- bishop of Dublin. Two parts in One. Cloth 75 cts The Story of Creation. A Plain Account of Evo- lution. By Edward Clodd, F.R.A.S. With over 80 illustrations 75 cts Hypnotism : Its History and Present Deyelop- ment. By Fredrik Bjornstrom, M.D., Head Physician of the Stockholm Hospital, Professor of Psychiatry, late Royal Swedish Medical Councillor. Cloth 75 cts Christianity and Agnosticism. A controversy consisting of papers by Henry Wace, D.D., Prebendary of St. Paul's Cathedral; Principal of King's College, London. Professor Thomas H. Huxley.— W. C. Magee, D.D., Bishop of Petersbororgh.— W. H. Mallock, Mrs. Humph- rey Ward. Cloth 75 cts Aarwinism:An Exposition of the Theory of Nat- ural Selection, with some of its applications. By Alfred Kussel Wallace, LL.D., F.L.S. With portrait of the author, colored map, and num- erous illustrations. Cloth |1 .25 The ablest living Darwinian writer.— C/»