.,ia;>f?^'l l Q f\p ]VI| An A D A umj/m>-H VOLUME OF KIVEE. Page. Sig-nificance of this section 173 Characteristics of Foster flats and associated features 174 Wilson terrace of Clinton limestone and its remains 181 Pot-holes in fallen blocks 182 Distinction between Niagara and Clinton blocks 182 Northern spur of Wilson terrace and cove behind 183 Union of Niagara and Clinton falls, their heights 187 Third or Medina cataract 188 Increased volume of the river 188 Eeview of former conclusions 194 CHAPTER XV. TERRACES ABOUT END OF GORGE AND LAKE LEVELS BELOW THE PRESENT. Eoy terrace and lake level at birth of falls (287 feet) 197 Eldridge flat (200 feet) 201 Bell terrace (174 feet) 201 Iroquois beach (137 feet) 203 Ijower terraces 204 Delta of Niagara river and further subsidence of lake 204 Lowest level of water in Ontario basin 208 Final backing of waters to their present level 208 CHAPTER XVI. GLACIATION AND DRIFT ADJACENT TO NIAGARA RIVER. Glaciation in Niagara district 211 Character of clayey and stony drift 212 Character of sand ridges 215 River deposits with shells 216 CHAPTER XVII. LAKE BASINS AND METEOROLOGICxiL CONDITIONS AFFECTING THE FALLS. Drainage area. The Erie ratio 217 Mean rainfall and evaporation of Lake basins 219 Modified conditions in the Erie basin 221 Humidity 221 Temperature 222 Velocity of the wind 222 Relationship of evaporation to rainfall 222 of Canada] CONTENTS Xxi CHAPTER XVIII. FLUCTUATIONS OF THE LAKES. Page. Lake Ei'ie. Note on jjosition of observation 223 Mean quinquennial fluctuations of Erie, 1850-190.") 225 Fluctuations of Erie and Huron compared 226 Lowering- of Lake Huron 227 Drainag-e by Chicago canal 228 Changes of level of Huron affecting that of Erie 228 Fluctuations of Erie and Ontario compared 229 Mean quinquennial fluctuation of Ontario, 1854-1905 260 Fluctuations of Lake Ontario and Saint Lawrence river com- pared 231 CHAPTER XIX. LOWERING OF THE LAKE OUTLETS. Stability' of the outlet of Lake Superior 233 Table of mean annual fluctuations of Ei-ie, Huron and Superior. . 234 Lowering of the Huron outlet 235 Table of mean annual fluctuations of Erie and Ontario 239 Lowering of the Erie outlet at the same rate as that of Ontario. 240 Lowering of the outlet of Ontario 242 Effects of lowering of Ontario on higher lakes 243 Effects of the lowei'ing of the lakes on the canals and harbours. 246 Corrected elevations of the Great lakes 246 CHAPTER XX. DISCHARGE OF NIAGARA AND OTHER RIVERS OF ST. LAWRENCE DRAINAGE. Notes on discharge 247 ]\Iean quinquennial discharge of St. Mar^', St. Clair, Niagara and St. Lawrence rivers 248 Variations in discharge of Niagara river 249 Variations in discharge of St. Clair river 250 Variations in discharge of St. Lawrence river 251 Variations in discharge of St. Mary river 252 Pro]5ortional drainage of Erie basin 252 CHAPTER XXI. POWER OF NIAGARA PART ONE. Variable discharge and present shrinkage 255 Range of horse-power of the falls 255 .Available power in recession 256 Net mechanical horse-power 256 Ratio of Canadian and American channels and falls 257 Franchises 259 Limitations of use 261 Power of Niagara river below the falls 261 Xxii FALLS OF NIAGAEA [Geol. Surv. CHAPTEE XXI.. PAET TWO. SHKI^'KAGE OF FALLS AND LOWEEIZVG OF LAKES BY POWER DIVEESIOjST. Page. Effects on Niagara falls by use of power 265 Lowering of lakes and canals, by use of power at the Falls.. .. 268 CHAPTER XXII. CIIAXGES OF ONTARIO LEVEL SHOWN IN IROQUOIS BEACH. Early observations concerning the Iroquois beach 277 Characteristics of Iroquois beach 280 Tilting of the land recorded by Iroquois beach 281 Low^r beach and increased descent of the Niagara 285 Sudden changes of level 285 CHAPTER XXIII. EFFECTS OF TILTED SHORES OF WARREN WATER ON THE NIAGARA DISTRICT. Notes on the study of Lake Warren 287 Elevation of the Forest beach 289 Tilting of land at eastern end of Lake Erie 291 Elevation of Forest plane above Iroquois 292 Warping after a long recession of Niagara falls 292 CHAPTER XXIV. SEPARATION OF ERIE AND HURON ON THE DISMEMBERMENT OF WARREN WATER SHRINKAGE OF LAKE ERIE. Dismemberment of Warren water 293 Separation of Huron and Erie drainage 293 Original discovery of the diversion of the Pluron drainage . . . . 294 Shrinkage of Lake Erie 297 CHAPTER XXV. NORTHEASTERN OUTLET OF ALGONQUIN LAKE (tHE THREE UPPER lakes). Northeastern Huron drainage and diversion from Niagara . . . . 299 Northeastern rise of Algonquin beach 299 Amount of tilting shown in beach 302 Barrier to Algonquin lake 302 of Canada] CONTENTS Xxiii CHAPTER XXVI. HEAD OF THE ST. CLAIK TKIBCTAEIES OF ALGONQUIN LAKE. Page. Terraces about Lake St. Clair 305 Depth of drift about the St. Clair outlet of Lake Huron 305 Head of St. Clair tributaries 306 Date of drowned St. Clair valley 307 CHAPTER XXVII. AUGMENTATION OF NIAGARA DISCHAKGE BY ACCESSION OF ALGON- QUIN WATEES THAT IS HURON DRAINAGE. IVipissing — Ottawa outlet 309 Anioimt of terrestrial tilting. Huron waters added to Lake Erie 311 Origin of Lake St. Clair 311 CHAPTER XXYIII. RECENT CHICAGO OVERFLOW. Recent Chicago overflow 313 CHAPTER XXIX. ST. LAWRENCE CHANNEL WITHOUT HURON DRAINAGE AND WARPING. Smaller channel of St. Lawrence river 317 Terrestrial deformation shown in the rapids 319 Ancient Niagara and present Ottawa rivers compared 320 CHAPTER XXX. TIME OF WARPING OR TILTING OF LAKE REGION. Post-Iroquois warping shown in that beach 321 Post-Iroquois warping in Niagara district 322 Warping of Algonquin beach 322 Tilting of Nipissing beach affecting Niagara falls 323 Effects of post-Iroquois uplift on Niagara river 323 Chicago overflow 324 Cause of earth movements — Fisher's theory 324 Xxiv FALLS OF NIAGARA '^^'^°^- ^"''^• CHAPTEE XXXI. NO PRESENT EARTH MOVEMENTS OR STABILITY OF THE LAKE REGION. Page. Oljservations about Lake Erie 327 Fluctuations at Port Colborne and Cleveland compared 328 Table of flvictuations between Port Colborne and Cleveland.. .. 329 Stability of earth's crust in Erie basin since fifty years ago.. .. 331 No earth movements in Ontario basin 331 Table of fluctuations at Toronto and St. Lawrence stations.. .. 332 Absence of earth movements about outlet of Lake Huron . . . . 335 Importance of terrestrial stability 336 Former supposition of future diversion of the Niagara to the Mississippi 336 CHAPTER XXXII. RECESSION OF NIAGARA FALLS. Preface 341 Present recession 342 Effective height of the falls in recession 342 Former changes in the height of the falls 346 Variation in the volume of the water 349 Differential discharge from the Erie basin 350 Laws of erosion 351 Character of origial river channel 351 Effects of Falls-Chippawa valley upon the recession 353 llecession of American falls 353 Effects of rock structure 354 Height and volume of falls at their birth 355 Increasing height of falls and establishment of second cataract. 356 Siibsidence of waters to Bell terrace 357 Iroquois shore level 358 Union of two upper cataracts above Foster flats 358 Medina or third falls and their great height 359 Increased volume of Niagara at Foster flats 361 Effective ending of Medina falls 361 Increased effective height of falls above Foster flats 363 Conclusions as to stages of recession 363 CHAPTER XXXIII. AGE OF NIAGARA FALLS. Episode of the modern stage 367 Duration of Erie stage 368 Total age of the Falls of Niagara 370 Former conjectures as to the ages of Niagara falls 371 Date of Accession of Huron waters to Niagara falls 376 of Canada] COXTEXTS XXV CHAPTEE XXXIV. FUTUKE RECESSIOX" OF THE FALLS. Page. Future recession of the falls 379 CHAPTER XXXV. ORIGIX" OF THE LAKE BASINS DEOWNED AXD BUEIED VALLEYS. Preface ?>91 Topography of the Lake basins 392 Features of the Ontario basin 393 Features of the Erie basin 396 Features of the Huron basin 397 Features of the Michigan basin 397 Buried valley between Georgian bay and Lake Ontario, etc 398 Former higher elevation of the continent -100 Southern tributaries of the ancient Laurentian valley (Ohio and others) 403 CHAPTER XXXVI. XIAGAEA PENINSULA. Preface 407 Features of Niagara peninsula 407 Geology of the peninsula 408 Elevation of the Erie plains 409 ' Short Hills ' and the sand ridges 409 Grand river — Dundas valley — ancient drainage 411 CHAPTER XXXVII. DISCOVEEY OF THE EEIE OUTLET. Deep channels about Lake Erie 413 Salina basin 413 Well-borings 414 Southern boundary of Cornif erous limestone 416 Northern boundary of Niagara limestone, Thorold depression . . 416 Features of the ' Short Hills ' district 417 Erigan canon 418 De Con and Swaze falls 423 Erigan channel and buried Erie outlet 423 Slope of the Erigan outlet of the Erie basin 424 Erigan tributaries 426 Crossing of the lake depressions 427 APPENDICES APPENDIX I. EARLY DESCEIPTIOKS OF NIAGARA FALT.S. Page. A, Hennepin's (1678) 431 B, Kalm's (1750) 432 C, Ellicotfs (1790) 440 APPENDIX II. SURVEY NOTES OF RECESSION OF FALES. Survey notes of recession of falls 4t3 APPENDIX III. WELLS BELOW ESCARPMENT. Wells below escarpment 447 APPENDIX IV. IMETEOROLOGICAL TABLES. Meteorological tables 448 Table I. Rainfall in Basin of Lake Superior 448 " II. " " " Huron-Michigan 449 III. " " " Erie-St. Clair 449 IV. " " " Ontario 450 V. Humidity 451 " VI. Temperature 451 " VII. Wind velocity 452 APPENDIX V. Table I. Fluctuation of Lake Erie (at Port Colborne) 453 II. " " " (at Cleveland) . . 454 "• III. " " Huron 455 V. " " " (at Charlotte, Oswego). 458 " IV. " " Ontario (at Toronto) 456 xxvii FALLS OF NIAGARA [Geol. Surv. APPENDIX VI. Table I. Discharge of Niagara river 459 II. " St. Clair river 460 " III. " St. Lawrence and St. Mai'y rivers . . . . 461 APPENDIX VII. NOTES ON ORIGIN OF GREAT LAKES. Notes on origin of Great Lakes 462 APPENDIX VIII. On the discovery of Niagara Falls and the name 466 In the foregoing tables the researches include a vast number of data, many obtained instrumentally, which are either new or revised de novo, whereby the chain of evidence is completed in solving many diflBcult prob- lems ; therefore, scientifically, the author assumes responsibility for the statements, opinions and phraseology of this report. The absence of cer- tain capitals, notably ' f ' in Falls when referring to Niagara ; the change of names as from St. David's, thus known in history, to St. David; the mixed form of numerals, when in series, etc., do not meet with the approval of the writer. Certain observations, set forth in some chapters elucidate sub- jects discussed in others, hence repetitions are unavoidable. Errata. — Some errors have escaped observation, and others indicated were not corrected, but the significance has not been obscured. The sense is changed where the word ' Huron ' occiirs in the fourth line from foot of page 262, which should be Erie ; so also ' to be ' on sixth line from foot of page 396 should read at nof. ILLUSTRATIONS Plate. Page. Map and Chart of Gorge of Niagara (Spencer) 1905. ' Canadian Falls ' of Niagara, 1899 Frontispiece. I. Map of Boundary, 1S19 15 II. ]\lap of Recession (Spencer) 1905, opposite 19 III. Map of River (ISIontresor) 1761 21 Ilia. View of Falls (Pierie) 1768 23 IV. Map about Falls (Stegman) 1799 25 V. Map of Falls (Hall) 1842 26 VI. View of Crest-line 29 VII. Goat Island shelf 31 VIIo. View of Falls (Van der Lyn) 1804 33 VIII. Map of Recession of American Falls (Kibbe) 37 IX. View of Both Falls 39 X. Position of Hennepin's cross-falls 43 XIa. End of Whirlpool with boat 67 XIb. Sounding, below Whirlpool (Boat) 67 XIIa. First Cascade, western end 77 XIIb. First Cascade, N.Y. channel 77 XIIIa. River Bed, drained 78 XIIIb. River Bed, deep crevice 78 XIV. Panorama of Upper Rapids and both Falls 81 XV. Profile view of American Falls, and terraces 113 XVIa. Hubbard Point, with terrace 117 XVIb. Opposite same, with corresponding terrace 117 XVIIa. Outlet of Gorge, cutting escarpment 121 XVIIb. Iroquois terrace, outlet of gorge 121 XVIII. Map of Pre-glacial Whirlpool-St. David Channel.. .. 131 XIX. Whirlpool Rapids 141 XX. Map of Pre-glacial Whirlpool Rapids 145 XXIa. Medina red sandstones and shales 157 XXIb. Original Banks of River 157 XXII. Map of Falls-Chippawa Buried Valley 164 XXIIIa. Southern end of Foster Flats looking down 171 XXIIIb. Southern end of Foster Flats looking up 171 XXIV. Map of Gorge of Foster Flats 176 XXVa. Wilson Point, Foster Flats 179 XXVb. Cove at head of Foster Flats 179 XXVIa. Rapids rounding Foster Flats 185 XXVIb. Gorge obstructed by Foster Flats 185 xxix FALLS OF NIAGARA [Geol. Surv. Plate. Page. XXVIIa. Pot-holes 189 XXVIIb. Thompson Point and Buttress 190 XXYIII. Eoy Terrace, at Birth of Falls 199 XXIX. Niagara River beyond end of gorg-e 205 XXX. Sand beds at Berrynian's farm 213 XXXIa. First Cascade, eastern end 265 XXXIb. First Cascade, above Sister Islands 263 XXXII. Canadian Falls, western and before curtailment . . 269 XXXIII. Map of Iroquois Beach 282 XXXIV. Map of Forest Beach 288 XXXV. Maps of Algonquin and Nipissing- Beaches 301 XXXVI. Outlet of Whirlpool 365 XXXVII. Canadian Falls, very high water 377 XXXVIIIa. Goat Island Shelf, being drained 381 XXXVIIIb. Carter Cove 381 XXXIXa. First Cascade, profile view from east 385 XXXIXb. First Cascade, being drained 385 XL. Map of Pre-glacial Rivers, Lake Region 387 XLI. Map of Erigan Channel (Buried Erie Outlet) . . . . 415 XLII. Map part of same enlarged 421 XLIII. Niagara Falls (Hennepin) 1078 430 Figure 1 Sketch map of the Falls 1764 (Montresor) 20 " 2 Profile section of Canadian Falls 48 " 3 Longitudinal section and slope of river 50 " 4 Ice jana showing stranded mass .. 58 " 5 Sounding" section; Table Rock to Goat Island shelf .. 58 " 6 " " Carter's Cove to Prospect Point . . 58 " 7 " " at Ui^iJer Arch bridge 59 " " " " at Cantilever bridge 60 9 " " at Whirlpool 64 " 10 " " below Whirlpool 70 " 11 " " below Foster's Flats 70 " 12 " " inside of gorge 71 " 13 " " across river from Queenston Pier beyond mouth of gorge 73 " 14 Cross section of gorge 600 feet inside 99 15 " " " at Foster's Flats 103 16 " " " above Foster's Flats 103 17 " " " at Whirlpool 104 18 " " " at Whirlpool Rapids 105 19 " " " at Hubbard's Point 107 20 " " " at Table Rock-Goat Island.. 108 " 21 " " Whirlpool Rapids at Cantilever bridge complete, with boring sec- tion 147 " 22 Longitudinal section at Whirlpool Rapids 154 of Canada] ILLUSTKATIONS XXxi Page. Figure 23 Long-itudinal section at Foster's Flats 175 " 24 jSIap of terraces at end of gorge 198 " 25 Map of terraces at Burlington Bay 280 " 26 Map of St. Clair tributaries (Drowned) 306 " 27 Map of Chicago outlet 314 " 28 Longitudinal section of recession of the Falls 343 " 29 Cross section of Erigan canon at DeCoii Falls 422 " 30 Lonsritudinal section showing rise of Erie outlet.. .. 425 CHAPTER I. PARTIAL SUMMARY. The International Boundary Line at Niagara falls was deter- mined by the signed map of the Commission of that time (1819). At one point this line is found within 235 feet of Goat island, thus throwing the whole of the great crescent of the falls within Canadian territory, while only the turn beyond the crescent, trending to Goat island, belongs to !N"ew York. Here the cataract is now often reduced to mere strings of water. The withdrawal of the water for power purposes threatens to leave exposed a strip of rock belonging to Canada, now covered by the rapids, on the eastern side of the falls. The quantity of water falling over this end adjacent to Goat island is too small a fac- tor in the whole volume to be considered. But the volume flow- ing down the American falls is approximately seven per cent of the whole. {See Chapters ii. and iii.) The propor- tion of the discharge at the rocky rim, which determines the flow of water down the Upper rapids, gives to Canada 75 to 80 per cent of the total discharge of the river. {See Chapter XXI.) The recession of the falls between 1842 and 1905 was 285 feet at the centre of the apex, or a mean retreat for the full Avidth of the gorge of 265 feet. jSTearly seven and three-quarter acres of the rocky floor of the Upper rapids fell during that time. The retreat takes place by rapid central recession fol- lowed by a cessation of the same with a rapid lateral enlarge- ment. Thus for twenty years there has been no apparently measurable central retreat. The mean recession has been found 1 2 FALLS OF NIAGARA [Geol. Surv. to be 4-2 feet per year, but during the last fifteen years tliis lias been greatly reduced. The discovery of tlie position of Hen- nepin falls in 1678 shows that this rate has prevailed since then. The length of the Canadian falls was 2,950 feet in 1900, but from this 415 feet have been taken by the curtail- ment on the Canadian side. The recession of the American falls is only at the rate of 0-60 foot a year. (See Chapter iii.) Its length including Luna island is about 1,000 feet. The height of l^iagara falls on the Canadian side is 158 feet, and 175 feet in the centre. The descent from the rim of Greens or First cascade to the edge is fifty-five feet, after the river has descended fourteen feet from Lake Erie. The heiglit from the rim across the river to the cauldron below is 212 feet. This is the gross head of water in relationship to the horse-power of those companies that take the water from above the Upper rapids, while for those below it is only about 160 feet. This feature is most important when considered in its connexion with the effects on the falls^ or the results that will arise from the diversion of water by the power companies, as well as the relative quantity of water used. For other features, and the slope of the river see Chapter iv. The total descent of the river from lake to lake is 326 -58 feet (mean of fluctuations from 1891 to 1905). The new sounding off the Goat island shelf at 192 feet reaches to ninety-two feet below tlie level of Lake Ontario. This depth is found near the head of the Whirlpool rapids, a mile and a-half Ixdow, but under the falls the dej^th is only seventy-two feet (to fallen rocks), with a shelf beyond at eighty-four feet or a little more. At the Cantilever bridge the depth is eighty-five feet, but the channel is refilled to a further depth of a hundred feet. At the Whirl- pool, which is fifty-one feet below the head of the rapids, the depth before reaching the middle of the river was found to be 126 feet, or about fourteen feet less than in the river of Canada.] PARTIAL SUMMARY above. Below the ^\'liirlp(»(il it is considerably less, as also is the case below Foster flats. But near the end of the canon a drowned cataract was bronglit to light where the river within the gorge rapidly deepens from fifty-three feet to nearly 150 feet below Lake Ontario level. The narrow channel as deep as 183 feet was also fonnd a third of a mile beyond the mouth of the gorge, althongh it has generally been silted to a much less depth. (See Chapter v.) The rock structure and the excavating power of Xiagara falls are treated in Chapter vii., and the characteristics of the gorge in its relationship to the receding falls are described in the two following chapters; the original Itanks and bed of the Xiagara river are described, in Chapter ix. The ^Vliirlpool-St. David buried channel extends from the Whirlpool to the edge of the escarpment, about two and a half miles beyond, and was erroneously thought to be the ancient course of the Xiagara river. The ancient caiion has now been explored by borings, the deepest of which reached to 269 feet, or to a level only a little above that of the Whirlpool. The operations were stopped at this point on account of the ex- pense, which was growing very heavy, l)ut in a general way we now know the effects of the buried valley on the river. To its great depth it gave rise to the Whirlpool, but not to the gorge above, as might have been the case. This was one of the most important points established by the present survey, as the effects it had produced in the recession of the falls had not been fully known. The head of the ancient stream, which at the Whirljiool began to ioYin a gorge, was only a short distance south of the railway bridge — at Lyell ridge. The now buried channel never drained the Erie basin, nor that of any creek above the Upper rapids. But about the Whirlpool, it originated the valley later filled with drift, which was quickly re-excavated when the falls had receded as far as its outlet. Some curious features were observed in the borings, FALLS OF NIAGARA [Geol. Surv. such as a breathing well at a depth of 226 feet, and the occur- rence of a log of white spruce at a depth of 186 feet, buried there long ago in the Glacial period. See Chapters x. and xii. The last-mentioned chapter treats of the Whirlpool ra])ids. In this locality the old river banks are well preserved, while the gorge itself is greatly contracted in breadth. Here, after the Glacial period, a small superficial valley diverted the ISTiagara river, after the falls had receded past the outlet of the Whirlpool, so that most of the water was concentrated and made a narrow caiion some 400 feet deep, which, upon the retreat of the falls above this point, has been refilled, thereby producing the Whirlpool rapids, which have been completed at a very recent date. The building of these rapids eventu- ally lowered the height of the falls above. The widening of the gorge above this point is a very noticeable feature, but this is too technical for a short summary. The curious basin at ^Niagara falls has affected the re- cession. Here the rock floor is about 110 feet lower than the rocky Lyell ridge, or about sixty feet lower than the river bed of pre-glacial times crossing that ridge. From the Lyell ridge southward, the rock valley, partly buried, deepens and widens to the falls. It long challenged explanation. By borings it has been found to be part .of a valley heading in Lyell ridge and growing in the southward direction. At the falls it is over sixty feet below the level of Lake Erie, and it is traced by bor- ings farther southwestward. (See Chapter xii.) Its explanation accordingly demanded the finding of a deep outlet from Erie basin. This was afterwards discovered by means of complete investigation of the visible and concealed features. It is described in Chapter xxxvii. The basin at the falls is now found to be part of a shallow valley extending a few miles to join the ancient outlet of the Erie basin, all the features of which are deeply covered by drift. Tliis exjilains the occur- of Canada] PARTIAL SUMIMARY rence of the Upper rapids, where now the Niagara river is descending over the left bank of the otherwise buried valley. It has a great deal to do with the recession of the falls. Here the river is turning its course nearly at right angles, now causing the falls in receding to climb, so to speak, the ancient lateral bank of a valley, so that the hard rock formations above are getting thicker, with the softer lower beds becoming more concealed, thus tending to reduce the rate of the recession. The investigation of the basin in the rocky floor brought to light the Falls-Chippaw^a valley. In the vicinity of the falls this is covered by a hundred feet or more of drift, which has been explored in many places by borings, so that the rocky floor beneath the mantle is everywhere found to be sinking towards the southwest. {See Chapter xiii.) Foster flats begin at about two-thirds of a mile below the Whirlpool, and extend for a like distance. It is one of the wildest and most beautiful places in eastern America. Here is a series of terraces of the greatest importance in the investiga- tions of the recession. These are floors of the Niagara river, showing that until the falls had reached this point, nearly three miles from the mouth of the gorge, there were two cataracts, one in advance of the other, each about 120 feet in height. Also it is found that the second one, which had been gaining on the upper, now became united with it. At this time the second channel was still high above the present bottom of the river. Lower down the gorge was the third cataract. It may be said that this cataract was at one time 300 feet high, and was low- ered by the subsequent backing of the waters of Lake Ontario into the gorge. The low^est cataract remained until long after the united upper ones had passed this section of the caiion, but subsequently joined it, and made one Niagara falls from the time that they reached the Whirlpool. This complex history of Niagara river is largely recorded at Foster flats, which is a wonderful place. Until lately only the Erie waters emptied by 6 FALLS OF NIAGAEA [Geol. Surv. the iSTiagara river. It was immediately after the falls had passed the head of Foster flats that the Huron drainage turned into the Erie basin, and henceforth the recession became very rapid. The features are described in Chapter xiv., and are referred to elsewhere. The terraces at the mouth of the canon, upon the flanks of the escarpment, furnish evidence of the lowering of the waters in the Ontario basin, and from these, along with the features at Smeaton ravine and at Foster flats, and the soundings in the river, the height of Niagara falls at different epochs in its early history is determined. (Chapter xv.) The glacial features of the region are described in Chapter XVT. The meteorological phenomena required investigation for several purposes, especially in relationship to the fluctuation of the lake levels, and to the discharge of the Erie basin com- pared with the whole volume of the J^iagara river, for in some cases the rainfall appeared at variance with the stages of water in the lakes. In short, they were necessary elements in studying the problem of the lowering of the lakes. (Chapters xvii. to XIX.) After 1890 there was a diminution of the rainfall, but since 1900 the mean has slightly increased, and raised the lakes. But this is too complex for a brief summary here. The fluctuations of the lake-levels have proved one of the most important studies. The levels are derived from the daily records kept at various stations since 1855 or 1860. The figures vary from year to year as well as from month to month, so that groups of years have to be taken for a fair average. Since 1890 all the lakes, except Superior, have been much lower than duriug the earlier years of observation. This has affected the calculation of the discharges. From these investigations, along with those of the meteorological conditions, it is found that the outlets of the lakes have been lowered (Chapters xviii. and XIX.), but that the lowering has lately been slightly ob- of Canada] PARTIAL SUMMAKY scured owing to the recent increased rainfall. The most accurate statement of the elevation above the sea is the mean of the last fifteen years (1891 to 1905), and not that of the whole period of record of about half a century, although the mean annual fluctuations and discharges are transcribed without introducing the desirable corrections except where special use is made of them. The lowering of Lakes Erie and Ontario has been about one foot and of Lake Huron about a foot and two-thirds. Lake Superior has risen owing to increased rainfall. The lowering of the outlets is permanent, though increased rainfall may con- ceal the fact even more than now. The lowering of the outlets aftects the calculations of the dis- charge of the rivers, causing those before 1891 to be too great, and producing anomalous conditions that were not understood until the recent discovery of this feature, an allowance for which brings all the results into harmony. We may therefore take either the mean discharge for the fifteen years (1891- 1905), or that of the whole i:)eriod, if the correction be made. In the former case it reduces the discharge given by the U. S. Engineers by about 15,000 cubic feet per second for ISTiagara, that is from about 219,000 to 204,000 cubic feet per second. Other figures given, which require correction, are 222,000 and 215,000 cubic feet. A relatively greater correction is required for Lake Huron. It was necessary to determine the relative dis- charge of the Erie basin as compared with that of the four LTpper lakes. The results derived from the rainfall of the basins, from their drainage areas, and the discharge of the fifteen years coincided very closely, but were entirely at vari- ance with the previously given mean discharge, until allow- ance was made for the lowering of the outlets. Finally I found that the Erie discharge was fifteen per cent of the whole volume of the j^iagara river. This factor was one very long sought for, on account of its bearing on the solution of the ao-e of ]^iao;ara falls. 8 FALLS OF NIAGARA f«e°l- ^urv. Another outcome of the investigations of the hike fluctua- tions is the establishment of the present stability of the earth's crust in the lake region during the last fifty years, where for- merly it was supposed to be rising. (Chapter xxxi). The discharge of the different rivers is given, and their relationship to the lake fluctuations, which are great. Thus on October 7, 1858, the corrected volume of Niagara was 292,000 cubic feet per second while, on February 28, 1902, it fell to 158,500 cubic feet, — the mean of that month being 175,000 cubic feet. (Chapter xx.) The mean discharge of 204,000 cubic feet gives a gross horse-power from above the Upper rapids of 4,900,000. It was reduced in February, 1902, to 4,200,000 horse-power, when low water prevailed, or for the lowest day to 3,800,000. But this does not represent the available force, as two of the power companies take the water from nearly fifty-five feet be- low the head. Again there are other great losses in the appli- cation so that the whole power cannot be used, and these will reduce the amount by 30 or 35 per cent. Accordingly the avail- able low water discharge is reduced to 2,600,000 horse-power; for only this amount can be considered in the power question. Seventy-five to eighty per cent of the power is on the Cana- dian side of the Boundary Line. But this is modified by the position from which it is taken. Thus the power taken above the First cascade affects alike both sides of the river, and also the flow of water over the falls, in so far as they are preserved or despoiled. The water taken below the rim affects the Cana- dian side almost exclusively, and since the encroachment here, the effect on the river occurs mostly where the water is the deepest, thus doing the least amount of harm. As much of the rim of Greens or First cascade is covered by only a thin sheet of water, the diversion of the present franchise power will gieatly modify the eastern side of the falls. In fact, if the full amount be utilized it must drain not merely 800 feet on of Canada] PAKTIAL SUMMARY 9 the eastern side of the Canadian falls, but some two hundred feet more on the western, so that the falls will be reduced from nearly 3,000 feet in width (in 1900) to 1,500 or 1,600 feet. It will further reduce the American falls from 1,000 feet (in- cluding Luna island) to a few disconnected streams. The full utilization is not in sight at the present moment, but that taken already has reduced the sheet of water on the Goat island shelf, so that in the near future it seems that the shrinkage of the water will leave only that portion of the greater falls which lies Avithin the Canadian domain. This will complicate the ques- tion of the use of the water. Another most important question that is arising from the di- version of the water above Greens or First cascade is the low- ering of the Upper lakes. From 20 to 25 per cent of the dis- charge will be taken from the basin of the river above the First cascade — this will lower the river from 3 '2 to 4 feet according to the stage of mean or low water, though at first the lowering of the river here will be partly lessened by the stronger current from above. Tlio increased flow means eventually a lowering of the lakes, as the run-off will be in excess of the rain supply. This condition will continue until equilibrium is again estab- lished with the lakes at a lower level. The word lakes is used, for on the reduction of Lake Erie the same results wnll affect Lakes Huron and Michigan. Even with the lowering of the mean level of the lakes by two feet, and it may reach three or four feet, the effects on navigation in the harbours and canals are sure to be serious. This feature has not hitherto been considered so far as I know, except to be discarded as unimportant, because the in- take of the water is below the lake outlet. This ignorance has arisen from not investigating the science of Niagara. The diversion of even 125,000 horse-power above the rim will affect the Upper lakes more than the use of an extra 5,000 cubic feet at Chicago. But the increased velocity of the discharge in 10 FALLS OF NIAGAKA f<^«°'- S"^^- leaving Lake Erie will not merely lower the water of the lake to equilibrium, but it will produce an extra scour of the clayey and rocky bottom where the current runs five miles an hour, and thus permanently further reduce the level of the lake, as has occurred to the extent of one foot since 1890. {See Chap- ters XIX. to XXI.) The raised and tilted shore lines, showing the recent move- ments in the earth's crust and former lake levels, are described in Chapters from xxii. onward. So far as Lake Erie was at first concerned in post-glacial times its area was only about one- sixth the present size, but upon the land rising faster at ISTiagara than farther westward the lake expanded. When the waters of Lake Ontario became lower the lake shore was about twelve miles from the mouth of the gorge, then the channel was ex- cavated to a depth of about 180 feet below the present level. Later, by tilting, the Ontario waters backed up and drowned the lowest part of the Niagara, and reduced the aggregate height of the falls from 500 feet in all its parts to the present descent of the river of 326 feet. The Huron drainage was towards the northeast, ultimately by Lake Nipissing, and the Ottawa. What is now Lake St. Clair became the headwaters of Lake LIuron, draining north- ward, as is shown in the drowned channels in the bottom of that lake. But the tilting of the basin turned the waters into the Erie drainage, and augmented the Niagara from 15 to 100 per cent. As these were late changes, only about 3,500 years ago, the absence of warping during the last fifty years becomes the more striking. During the time when only the Erie waters flowed down the Niagara the St. Lawrence was small, as is proved by the recently observed smaller inner channel found there. In the process of the tilting which sent the Huron dis- charge southward some waters overflowed by way of Chicago into the Mississippi drainage, but the lowering of the Huron barrier to the south lately caused the entire diversion of that discharge to the Niagara. of Canada] PARTIAL SUMMARY 11 Havino- investigated ami analysed the components wliicli have affected the recession of Niagara falls I have determined the changes in the volume of the discharge, in the height of the falls, the character of the original banks of the river and the variation of the features of the rocks and the rock surfaces, and the effective erosive power of the falls. Then, finding the character of the first and second cataracts, especially, I have been able to tell the story of the retreat of the Great Falls, and determine their approximate age. At first the falls were only thirty-five feet high with a vol- ume of 15 per cent of the present, falling directly into Lake Ontario. Thev receded for only a short distance before their height increased, but based upon proportional recession these conditions lasted 3,200 years. The falls were now separated into two cataracts, the second one growing in height^ they receded to Foster flats, where their union occurred at a point extremely well marked. Tt was on the upper of these falls that I was enabled to make the determination of the time required for the recession. Had this evidence not been available the difficulties might have proved insurmountable. The time re- quired for the recession of the double falls to Wilson point (in addition to the 3,200 years mentioned) is found to have been 31,600 years, and 700 years more to the head of Foster flats, the whole distance being nearly three miles. This was the length of the gorge excavated during the Erie Epoch. From now onward the recession was very rapid, modified at times, but in all re- quiring only about 3,500 years, so that the age of Niagara falls may be placed at about 39,000. Slight variations on one side or the other are probable, but under the conditions, all of which are now^ apparently known, the error in calculations will not exceed ten per cent. An account of the pre-glacial topography is added, though it might have preceded the order of this work on the whole re- cession of the falls. During the present survey the buried 12 FALLS OF NIAGARA [Geol. Surv. channels explored have thrown so much light on Xiagara that those bearing on the outlet of Lake Erie are of particular in- terest, and thej form a distinct chapter that is added to the work, as also another describing the features by which the origin of the Great Lake basins has been found. The Falls-Chippawa valley demanded a search for what proved the discovery of the Erigan valley and caiion, a few miles to the west. The canon was brought to light in the explorations of the ' Short hills/ where portions of it dissect the ^N^iagara escarpment and are exposed in the deep ravines. Southward there is no suggestion of such an ancient outlet for the Erie basin on the surface of the flat country only a few feet above the level of Lake Erie. But the records of well borings all the way to the lakes were obtained, and they have great economic value in the question of supply of water, and obtaining natural gas. These problems require much further investigation, which the limit of the survey did not permit. However, a great buried valley was discovered, with details that could not have been expected. Thus, not merely was a channel found deep enough to drain the present Erie basin, but also several smaller tributaries of it. This completes the most important evidence in the study of the ancient Erie basin, which formerlv received the drainage of even the Upper Ohio river. The history of these investigations required many years for elucidation and it embraces much of that bearing on the origin of the lake basins, which at last is pretty well knowm. See Chapters xxxv. to xxxvii. CHAPTER II. INTERNATIONAL BOUNDARY LINE AT NIAGARA FALLS Boundary line located by the Com- First Canadian Survey of Reces- mission of 1S19. sion of the falls. Crescent of the Great falls on Can- Deepest Inner Channel close to adian side of that line. Goat Island shelf. BOUNDARY LINE LOCATED BY THE COMMISSION OF 1819. The use of the terms ' Canadian falls ' and ' x'Vmerican falls ' dates back to the settlement of the country, and the names were adopted by the writers since the earlier part of the nineteenth centnry, as by Bakewell, Hall and others, bnt they have had no more significance than as convenient designa- tions of the two great cataracts. Indeed, the great natnral phenomena should be quite independent of political limitations. On almost every map on which this Boundary Line is indi- cated, even though in official publications, it has been erron- eously placed. The map issued by the United States State De- partment for the use of the Lighthouse Board, which is a photo- graphic copy of that of the Boundary Survey Commission, is an exception. The International Line on the recent map of the Interior Department of Canada is as correctly located as possible on one of such a small scale. Hitherto, the position of the Boundary has been of no importance as it was ab- solutely unapproachable and very few people knew or even cared where it Avas. Lately the question was raised in con- nexion with another survey of the falls (page 29), and accor- dingly it should be definitely understood. International ques- tions may arise, as the river below the falls is navigable ; and the withdrawal of the water from Goat island shelf, due to the 13 14 FALLS OF NIAGARA ^^^°^- ^"^'^■ power companies, will involve the necessity of the Boundary Line being well known. An article has just appeared under the title of 'Niagara Falls Already Kuined,'* wherein the writer, Mr. Alton D. Adams, advocates the saving of the American falls by carrying the point of deepening the channel to the International Bound- ary. But to cover the cost he suggests the further abstraction of water from the Canadian falls to the extent of over one-third of the whole volume of the river. The scientific aspect of the pro- position is germane to this paper — the fulfilling of it lies en- tirely within the jurisdiction of the State Department. The Boundary Line was established by the International Commission in 1819. Under the Treaty of Ghent the Boundary Line is drawn through the middle of Lake Ontario ' until it strikes the com- munication by water between that lake and Lake Erie, thence along the middle of said communication into Lake Erie through the middle of said lake,' etc. ; and the Treaty provided for a reference to commissioners to decide what islands should belong to each party. Their decision was to be final. The commission adopted rules, one of which was to follow the deeper channel and to compensate for islands assigned to either party. It was for compensation, as it were, that the Boundary Line was located near Groat island and Grand island, giving only water equivalent to Canada. Unless other concessions were made it was no compensation to draw the Boundary Line near Goat island and Grand island, for the middle of the river (under Treaty) would have divided Grand island nearly equally, while a corner of Goat island and even more river would have fallen to Canada. The Boundary map, here reproduced, is signed by Peter B. Porter and Anth. Barclay, commissioners, and William A. * The Technical World Magazine, pp. 115-124, 1905. of Canada] INTERNATIONAL BOUNDAEY 15 16 FALLS OF NIAGARA l^^®"^- ®"*'^'- Bird and David Thompson, surveyors. It is dated 1819. The present copy is on the same scale as the original, the official copy in the State Department Library of Washington, to which access for making the photograph, was obtained through the kindness of Mr. William MclSTair, chief librarian. The sketch of the crest of the falls in 1819 was only ap- proximate, but the established position of the Boundary Line in its relationship to the shore of Goat island has been trans- scribed to the map of the recession (Plate ii.) so that it will be understood in its bearing upon the present crest of the falls. 'CRESCENT OF THE GREAT FALLS ON THE CANADIAN SIDE OF THE LINE. At the end of Goat island shelf the Boundary Line is be- tween 235 and 260 feet from the island. Opposite the apex of the falls it is about 300 feet from the Goat island shore, while the apex is 400 feet within the Canadian side of the line. After swinging round nearly parallel to the southwestern shore of Goat island it bends towards the ISTew York side, above Goat island, and there passes onward between Navy island and Grand island. It hugs close the shore of Grand island, being only 200 feet away, with the Canadian bank from 1,200 to 2,300 feet beyond the Boundary Line. The placing of the Boundary Line gave the whole of the eastern channel, the American falls, and Goat island to the state of New York; but only from 235 to 300 feet of the river at the falls. This threw the crescent of the Canadian falls within the territorial boundary of Can- ada. FIRST CANADIAN SURVEY OF RECESSION OF NIAGARA FALLS. As the Boundary Line places the crest of the falls, where the recession is occurring most rapidly, entirely within Canadian jurisdiction, the preservation of Niagara falls becomes a special duty of the Canadian people. Four jirevious surveys have been of Canada] INTERNATIONAL BOUNDARY 17 made to ascertain the amount of recession of the falls, but the present survey made in October, 1904, and revised in August and E'ovember of 1905, is the first under Canadian authority. Even if the popular opinion, that the Boundary Line follows the deepest channel had here been correct, then below the falls that line would have been near Goat island as shown by the deep sounding- of 192 feet (see map). It was none too soon to make this survey, as the diversion of the water had already commenced, and will be greatly in- creased in the near future when there will be a marked differ- ence in the natural changes of the cataract. The effects of the artificial changes are only now beginning to be appreciated. The Goat island shelf is already losing the water from its sur- face. (See Chapter xxi.) Below the bank of Goat island, at the angle of the shelf, is a low flat bush-covered rock-surface, showing that the margin of the river was from thirty to fifty feet inside the present water edge. This would be mostly flooded were the river here two feet higher as it was formerly, for the inner was the true bank of the river. On the Canadian side, Table Rock has fallen away leaving only fragments of the terraces, as at T on map, plate ii., and at some other points, to establish the old shore line, now further obliterated by the embankment curtailing falls. On the ]3i'eliminary recession map the end of the Boundary Line is placed thirty feet too far west, but mea- suring to the natural bank it should still be thirty to fifty feet nearer Goat island, as shown on second edition of the Recession map, Plate ii. fitoliiijiraLS'nnjtj of Kaoalia l — i ,. \-^'// C5/\\Goat Island i\i % 5 'f" ^^~~"--~. '" ii^A "* ^! _j, ' il \ /-^ RErESSION LrNF.S -v «•""■'■"" ^' '(4 \v NIAGAI^ FALLS m\^ ;l 'N.A, To iUuBlrMp Rcpopl bv ' 'V ^^ ^\ ^-~-.,_._ ___ ,../'^|, 3- / ] \i^ ''€" j^ ^"^-x^-*.. ..„ , „.„, -__^il^^' w, ,.».„.. L «.,.,. "L'!., _ w'r./.r'j"''""" ^__ 1 CHAPTER III. NEW MEASUREMENTS OF THE RECESSION OF NIAGARA FALLS, Surveys of the falls. Rate of Recession and change of Montresor's Map of Niagara river Form. (1764): Stegman's (1799); Inter- Recession of American falls, national Boundary map, 1819. Slower retreat of the falls now and Hall's (1842); U. S. Lake (1875); in the future. Woodward's (1886) ; Kibbe's Position of the falls at the time of (1890); Spencer's (1904-5). Father Hennepin established. Results of the Surveys of the Can- adian falls. SURVEYS OF THE FALLS. Montresor's map. — On June 5, 1764, an order was given for the immediate survey of the JsTiagara river, by Capt. fFolin ]\[ontresor. King George's Chief Engineer in America. Tn his diary* he wrote, ' Directed an astronomical survey with the pLane table from J^iagara to the fort at Little Niagara (that is Fort Schlosser), for the several works carried on here since my arrival.' This map is reproduced in Plate iii. It should be noted that Fort Erie is not on the map, as it was not built until the end of the summer of that year. The map is from a photograph, of the original in the British Museumf, which was kindly furnished by Hon. Peter A. Porter, of Buffalo. This map, though inaccurate in detail, gives a fair representation of the river. It is especially interesting as showing the rounded apex of the falls as close upon the Canadian side, without the deep indentation, now due to the turn of the direction of the gorge. * Reprinted in N. Y. Historical Society publication for 1881. t The British Museum number on the map is CXXI., 73. 19 20 FALLS OF NIAGARA [Geol. Surv. Stegman's map. — A map of the vicinity of ISTiagara falls, on the scale of three chains to the inch, was made by John Stegman*, and bears date of January 17, 1799. It was made by compass bearings and is, therefore, not as accurate as it should be. This is further apparent where the intervening shore lines were sketched in, but the measurements between salient points often correspond closely with more recent ones. The curvature of the Canadian falls is too broad. But there are other points on the map of value in studying the recession. See Map, much reduced in size, Plate iv. Fig. 1. Sketch IVEap of the Falls in 1704 (Montresor'.s). Ellicott's Survey, said to have been made about 1789, has not been found. However, his description of the fallsf written in that year, is reprinted in Appendix I. Mr. Ellicott estimated the age of the falls at that early date as being about 55,440 years.:}: * ' Niagara Falls Park,' Court of Appeals for Ontario, Maps and Plans ' B ' to Appeal Book, 18li4. Only a few copies for use of court were printed. One may be seen in library at Osgoode Hall, another in the Public Library, Toronto; and the author possesses a copy. In this volume are also maps by Chewett, and others. t ' Massachusetts Magazine,' July, 1790, pp. 3S7-S. t ' Journal ' of William Maclay, Appleton's, 1890. of Canada] MEASUKKMENTS OE RECESSION" 21 22 FALLS OF NIAGARA [Geol. Surv. International Boundary Survey map of 1819. — On this map the crest line of the falls was sketched in, and has been transferred to map Plate ii. Difficnlty was found in locating the position of both ends of the crest line, but Stegman's map was of mnch aid in this matter. Otlier early maps. — Chewett's survey of 1831, that of Messrs. Burwell, Keating and llawkin of 1831, and others, add to our knowledo;e of the earlv determinations of the features of the falls^^ In Stegman's map the falls are a flattened crescent. In 1819 there was a very sharp apex, in contrast with the flat- tened form in the survey made by Prof. James Ilall.f Hall's, U.S. Lal-e, Woodward's, Kihhe's and Spencer's surveys. The first trigonometrical survey of E^iagara falls, made for establishing a basis of the measurement of recession, was that of Hall in 1842. The second measurement was made by the United States Lake Survey in 1875:|:. At this time another notch was beginning to form on one side of the centre, but the princi- pal recession was at the head and towards the western side. The third survey was that of Prof. P. S. Woodward in 1886. Dur- ing the last interval the recession was small where before it had been greatest. But there was an enormous enlargement of the apex which had appeared in 1875. It was now somewhat similar to the form of 1819. The survey of August S. Kibbeff in 1890 was the most detailed. The growth of the apex of 1886 had been suspended. The great recession was on its west- ern side as in 1875, but the apex had assumed a mnch more acute form. The next survey was begun in October, 1901, by myself, aided by Mr. James Goodwin, C.E., and Messrs. * See foot note, page 20. t ' Natural History' of New York. Vol. 4, 1S42 (on reduced scale). t Lake Survey chart of Niagara Falls, scale 1 to 10,000, 1875. tt Seventh Report Com. State Res., Niagara, N.Y., 1891. 23 of Canada] jmeasukements of recession 26 FALLS OF :S"IAGAKA [Geol. Surv. of Canada] :ilKASUKK]MEKTS OF KECESSIOiS' 27 Hoyle iuid IMePlicrsoii. Engineers of tlie Electrical Develop- ment ('()ni])any. In Anjinst and auain in Xoveniher, 1005, I re-snrveyed the crost-line with my own assistants ; bnt in December a remarkable rook-fall occurred which is added in a sketch line on the ma]). ]t is also shown on Plate vii.,""^ which gives the effect of the water diverted from Goat island sludf. In Appendix ii., reference data of the survey will be found. These are the five surveys which mark the four periods of measured recession. The fact that another survey has been made requires sonu^ mention. After my survey wsis first made in October, 100-I-, and T had l^een requested to contril)ute the results to the Commission of the X. Y. Reservation at Niagara falls (which I could not do), I found on my return to complete the investigations of the falls in June, 1905, that the United States Geological Survey had sent out a surveyor to duplicate my survey, which was then purely a scientific question. The gentleman in charge had the opportunity of seeing my survey of October, 190-i, and later I saw the result of his. Some dis- crepancies appeared, which led me to make the re-survey in ^November, 1905, which confirmed my previous observations as does also the photograph of the crest-line shown in Plate vi. However, this is an unimportant detail, as in a very few years all will be changed. From being a scientific question only, where my survey had the priority, it became an International one when the surveyor located the International Line very much nearer the Cana- dian side than where it had been established by the Boundary Commission in 1819. The apex of the falls is now situated about -100 feet west of the Boundary Line, thus placing the crescent within the Canadian territory. Apart from any bound- ary question, a survey of the falls can only be made from the Canadian side. My two surveys of August and Xovember, 1905, constitute them the last actually made, while that of October, * Fh'Otographed Dec. 9, 1905. 28 . TALLS OF NIAGARA f^eol. Surv. 1904, gives it priority after that of Kibbe. It -^vas shown in the first edition of the Recession map. to accompany Summary Report of the Geological Survey of Canada for 1905, no im- portant change having occurred in the meanwhile. RESULTS OF SURVEYS OF THE CANADIAN FALLS. Most writers have taken the total area of recession and divided it by the entire length of the crest line, namely: — 2,215 feet in 1812, 2,950 feet in 1880, 2,535 feet in 1905. This later shortening of the falls is due to diversion for power thus restricting the width of the river by filling in the bank so that the crest line has been reduced 415 feet. Calcula- tions based upon the perimeter show the enlargement of the cauldron, but to measure the retreat of the cataract the area which has fallen aw^ay should be divided by the width of the cataract which has produced the gorge. Just north of Table Rock House is a fragment of a terrace (T on map Plate ii.) which marks the former bank of the river. Between the original shore of the river here^ and the end of the Canadian falls at Goat island, is a natural cross- section which has escaped subsequent changes from the Aviden- ing of the gorge by frost action, and undermining of the cliff. Thus the true width of the river is shown to be nearly 1,200 feet across, not including the recentlv uncovered rock flat at the edge of Goat island. From a point 600 to YOO feet below Table Rock House, where a fragment of the old rocky bank can be found, the full width of the gorge reaching across the same Goat island shelf is also found to be 1,200 feet. Accordingly this figure will be adopted as the mean breadth of the chasm which is due to the recession of the cataract. KATE OF RECESSION AND CHANGE OF FORM. The recession of the Canadian falls from 1890 to 1905 I have found to be 39,832 square feet, or nearly one acre. This represents a recession of 32 -2 feet as above defined in fifteen 29 OS 31 r 7JB '^^Sa^^j^^^H ■Ni ^l^^^^^^^B ^3 'Wf^ ?*VL '^^^1 ^■:t^ r«iin.«H«j»«,-v-- - •»■ ■■' . '.^ ■■. G!)GS— 3 33 of Canada] ^MEASUREMENTS OF RECESSION 35 years, or aiinnallv 2 -2 feet. In the meantime there has been practically no medial recession ; bnt the remnants of the former western border of the apex have disappeared. The beginning of the new apex is suggested by a short new channel in the upper beds of rock down which the water shoots before break- ing over the falls. This is illustrated in Plate vi. A second channel is also beginning to appear, but not sufficiently to affect the outline of the crest in a pronounced manner. Since the above was observed the rocks have fallen away, so that by October, 1000, a widening was in progress reducing the width of even the little V-shaped trench. Adjacent to the apex, upon its eastern side, the manner of recession has recently taken upon itself the wedging-off of upper layers of limestone. This occurred after the instru- mental measurements of jSTovember, 1905, but prior to De- cember 9, 1905. Accordingly the cataract strikes a projecting shelf and rebounds to the abyss below. This is shown in Plate vii. The manner of rebounding was anticipated at one point in the shelf where the waters had been noticed to strike a ledge forty or fifty feet below. In October, 1906, the projecting shelf appeared to have a length of 200 feet or more. These changes seem to have affected the bursts of spray rising out of the cauldron. The elevation of the spray is scarcely greater, if as great, as the level of the wall top, while formerly the ex- plosions were more frequent, sending up columns to a great height. Plate vii. also shows the reduced volume of water on the Goat island shelf, occasioned by the diversion of part of the Avater, which at the time the picture Avas taken was equi- valent to the lowering of Lake Erie by three-quarters of a foot. Between 1875 and 1890, 97,735 feet fell away, making an average recession of 81 '44 feet, or 5 -42 feet per annum. The greatest fall occurred in January, 1889, and another in 1882. Attention should be called to the fact that during this time a 6968— 3i - 3G . FALLS OF NIAGARA t^^^'' S"'-V- great apex which was beginning in 1875 was completed in 1886 ; after which there was a rapid widening of the chasm. Indeed, since the date just mentioned, there has been no recession of the apex, the work being expended in broadening and straight- ening the crest line. Going back from 1875 to 1812, 180,000 square feet above the point of Table Rock House collapsed. During that period the most notable rock-falls occurred at Table Rock in 1816, and in 1850. The area here given represents a mean recession of 150 feet, or an annual retreat of 4 -54 feet. Owing to the deep incision of the crescent, shown by the survey of 1819, the recession of ISTiagara falls after this date must have been very rapid as a consequence of the double face of the great natural quarry. This is shown by the representation of the form of the crest of 1819 on map (Plate il), though it can only be considered as approximate. However, this form was soon afterward changed by the great fall of rock in 1823, which carried away a large mass of Table Rock then extending north- ward of that shown by Hall in 1842, though it left the platform in front of Table Rock House projecting fifty-eight feet until the rock fall of 1850. Before 1823 Table Rock extended some 600 or 800 feet to north of Table Rock House, or beyond the location of the Hennepin channel, though here it may have been as wide as it was known elsewhere. The survey of 1819 shows a form more favourable for rapid recession than during the succeeding periods, on account of the deep indentations of the crest line, like that which appeared in 1886, and is now beginning to repeat itself in the V-shaped apex mentioned. The representations of the surveys of 1764, 1799 and 1831, while not to be relied upon, show the crest lines to have been flattened. Between 1842 and 1904-05 there has been a disappearance of 317,642 square feet, or nearly seven and three-quarter acres, representing a total recession of 265 feet and a maximum of 285 feet, not taking into account the small superficial channel of CanailaJ MKASUREMENTS OF KECESSIOX 37 Plate VIII. Recession of American Falls (Kibbe). 38 FALLS OF NL^GAKA tGeol. Surv. now appearing; bnt this total medial recession was reached in fortj-fonr years without any addition in the subsequent nine- teen yearSj during which time the work of the falls has been that of rounding the irregularities of the crest line. This represent a mean annual recession of 4-2 feet. RECESSION OF THE AlMERICAlSr FALLS. The recession of Xiagara Falls is almost wholly determined by the Canadian falls, as they had already receded about 2,400 feet since they parted company with the American falls at the foot of Goat island, some 600 years ago. During all this time the American falls do not seem to have retreated more than 110 feet, if so much, in excess of the unknown widening of the gorge from frost action. In 1819 the International Boundary surveys represented the incision in the crest line almost as deep as we see it to-day. On account of the inferior recession of the American cataract as compared with the Canadian one, and on account of the small amount of recession shown between the surveys of Prof. Hall in 1812 and Mr. Kibbe in 1890, a re- survey was unnecessary. (Mr. Kibbe's survey is given in Plate VIIL The total breadth of this cataract in a straight line from Prospect point to Lima island is 855 feet, and from Luna island to Goat island is sixty feet. The amount which Mr. Kibbe found fallen between 1842 and 1890 was only 20, GOO square feet, which is an average recession of twenty-nine feet in forty- eight years, or 60 of a foot per annum. A curious phe- nomenon is seen here {See Kibbe's map), in that the wall of tlie gorge appears to liave receded faster than the shelf beneath the northern side of the American channel, which shelf pro- trudes as shown on map in Plate viii., and in profile on Plate XV. This at first suggests that frost action beiicatli the falls is very small, while at a ])oint exposed to s])iay it l)L'('()uies 39 of Canada] MEASUREMENTS OF RECESSION 41 exaggerated. But there seems another reason. In 1721, Charle- voix visited Niagara falls, and specially mentions several points as then jutting out. The projecting shelf may be the remains of the floor of the falls before the separation of the Canadian cataract. SLOWER RETRJ5AT OF THE FALLS NOW AND IN THE FUTURE. From the measurements obtained it is seen that there has been reduction in the recession of ISTiagara falls during the last fifteen years, giving rise to problems which require investi- gation. The rock structure is variable. The channels in the river had not hitherto been studied. The discharge of the river has been reduced, to some extent artificially, and to a small extent by meteorological conditions ; there has also been lower- ing of the lakes by scouring of the beds of their outlets. The course of the river is changing in relationship to the underly- ing rock formation. The measurement of the recession of the falls is the first step in the research. The diversion of the water for power purposes already shows a pronounced lowering on the Upper rapids. The great reduction in the rate of the recession during the last fifteen years now seems to be due more to the changing course of the receding channel and character of under- lying beds, than to other causes. In the immediate future there will be a large quantity of power brought into operation, and should the full capacity of the franchise be utilized the retard- ation of the recession from this cause must be very great. To generations 600 or 1,000 years hence this retardation may be of advantage, provided any of Niagara falls be left. On the other hand their grandeur will become a matter of history only, while at the present day they are seen by 600,000 to 1,000,000 peo- ple every year. From the scientific point of view it will not be possible in future to determine the rate of recession in its bearing on the 42 FALLS OF NIAGAKA t^^^^- ^"1^. past age of the gorge ; for a large volume of water will be so diverted from the falls as to greatly check the retreat. The present time affords the last opportunity of making the mea- surement of l^iagara falls in even their approximately natural condition. rOSITION OF THE FALLS AT TUE TIME OF FATHER HENNEPIN ESTABLISHED. Father Hennepin first saw ISTiagara falls in 1678. He men- tioned and illustrated a cross cascade {See Plate xliii.^ Ap- pendix 1), on the western side of the great cataract, without the separating rock appearing at the surface of the Avater. Could such a condition have obtained, and if so, where ? After I had recorded a remnant of the lower margin of the river re- maining just north of Table Rock House, Mr. James Wilson, who has reudered 'me invaluable aid throughout this work, called my attention to a depression since filled in north of Table Rock House, beyond which no lateral falls could have existed. He had the profiles made 'before the changes in the Park. While some of these features could still be found, the identi- fication of the position of the falls is due jointly to his oliserva- tions and my own researches. Near Table Rock House is a slight elevation of six or seven feet above the lowest marginal bed of tlie river shown in the terrace T, on Map ii. It is also well shown in the view, Plate X. (on opposite ])age). From the old profile it was found that north of this ridge at Table Rock House the low depression, already mentioned, begins to rise at -a point 450 feet north of that Imilding. About 250 feet beyond this a steeper old bank is found from -fifteen to twenty feet above the fioor of Henne- ])iii chauneh This point at 450 feet was the beginning of the inner edge of the old channel, the course of which was nearly parallel with the clay banks behind ; consequently the pi*esent brow of the gorge follows a line at a very acute angle to the ii CD §p; W be >; &0 a; 43 of Canada] MEASUEEMEXTS OF REOESSIOX 45 course of the deserted cliannel. That this eliannel did exist there is proof in the map of Steginan. made in 1799 (Plate IV.). He shows a pond 400 feet long and nowhere more than forty or fifty feet wide, connecting at its southern end with the river, but blocked at its northern extremity. This depression has a position identical with the drained one of later date. As it is known that 100 feet of the Table Rock clitf have fallen away since the time of Hall's survey, the channel just men- tioned should be projected somewhat farther as on the map. Here then are the remains of a channel behind a low elevation of rock which must have produced a cross fall when the great cataract was passing this point. Thus the position of the falls at the time of Hennepin in 1678 has been closely located. Al- lowing for the fallen shelves of Table Rock, and carrying the crest line of the great falls outward in a curve similar to that of Hall's time, it crosses to the northern face of Goat island shelf at the same distance in front of the present falls as would be found by extending them down the gorge 950 feet in accord- ance with the measured rate of recession. Had there been no measurements for determining the rate of recession, Hennepin's cross-fall carefully worked out would have furnished means for ascertaining the rate during the last 227 years (to 1905) carrying back the date 164 years before that of Hall, when no other white man than Hennepin had left a description of the ' Moccasin falls ' of the Indians, of the time of Champlain. Then there was no Goat island shelf, but the Canadian falls had the form of semi-circle without the re-entrant curve by which the sheet of w^ater has been since lengthened to 2,950 feet in 1900, before the curtailment of 415 feet by com- mercial men. Accordinglv the Canadian falls at this late date were much grander than in the time of Hennepin, when their diameter was only 1,200 feet, with a perimeter of 1,500 to 1,800 feet. The form was then a flattened curve with pro- 46 FALLS OF NLVGAEA t^^°>- ^^^'^ bably a slight indentation corresponding to the deep sounding at the angle of Goat island shelf. The insular rise north of Table Rock House with the de- pression beyond and behind the power-house may be seen in Plate X. Note 1. — Plate Ilia (page 23) is reproduced from a drawing made in 1768, by Lt. Wm. Pierie., Royal Artillery. The illustration was kindly furnished by Hon. Peter A. Porter, of Buffalo, N.Y. Next to that of Hennepin (1678) shown in Plate xliii., this is the oldest picture of the Falls of Niagara known to me. At first glance the picture appears very much distorted, with Goat island far too small and out of proportion. However, when taken along with Kalm's description in 1750 (See Ap. 1), the picture becomes a valuable record of the recession of the falls. It shows that the eastern limb of the crescent, covered with water, reached nearly across the face of the island; also that the same rocky shelf extended much further across the gorge than now, and that the re-entrant angle of the falls was then relatively small. Note 2. — Plate Vila, page 33, is reduced to one-fifth of its full size, from an engraving of a painting by John Van der Lyn, 1804. The engraving is dedicated to the Society of Pine Arts of New York. The picture shows Table Rock, since fallen, and also the greater breadth of Goat island shelf than now. (By kindness of W. K. Vanderbilt, jr., Esq.) CHAPTER IV. HEIGHT OF THE FALLS AND SLOPE OF NL\GARA RIVER. Descent of the Upper rapids. cade at outlet of Whirlpool, and Height of the falls. of Rapids at Foster flats. Descent of Whirlpool rapids, Cas- Table of slope of Niagara river. DESCENT OF THE TPPER KAPIDS. The descent of the river from Lake Erie to the rapids, where a ridge extends across the river producing the First cascade, is nearlv fourteen feet. From this determinino- bar- rier the Upper rajiids descend in different cascades fifty-five feet to the present edge of the water, as now diverted on the Cana- dian side ; but the descent of the rapids to the apex is con- siderably less, as it is sitnated several hnndred feet farther np the side of the lately bnried pre-glacial channel. This is an additional canse for the thinness of the sheet of water there. The depth of water is greatest along that part of the crest ex- tending from the Canadian edge to the present apex. This is dne to the transverse ancient Falls-Chippawa trough crossing the direction of the rapids. From Greens or First cascade the rapids extend abont 2,500 feet to the apex of the falls, bnt the distance is mnch greater toward the western edge, owing to the cnrvature of the river and the crescent shape of the cataract. THE JIEIGIIT OF THE FADES. Tlie height of the falls on the Canadian side at ordinary stages is 158 feet, bnt on the side adjacent to Goat island it is two feet more. These heights are to the level of the river at the power-honse a thonsand feet to the north, where the river 47 48 FALLS OF NIAGARA [Geol. Surv. is a mass of surging foam Unquestionably the water must pile up somewhat immediately at the foot of the falls so as to reduce slightly the height given. It would be difficult to deter- mine this exactly^ for at the end of the tunnel of the Electrical Development Company under the falls, some 600 feet from the edge of the river, one sees the impetuous irresistible torrent boiling and churning the ever-changing surface of the water. Fig. 2. Profile section of Canadian falls (liorizontal and vei'tical scales the same) R., surface of river ; L. O., surface of Lak3 Ontario : N. 1., Niagara lime- stones ; N. s., Niagara shale ; C. 1. , Clinton limestone ; R. M., Red Medina shale and sandstone ; g. M., Gray Medina sandstones ; M. s., Medina shale. At the apex the falls descend 175 feet. This superior height is due to its being farther up the rising bed of the channel and nearer its middle. There is then, in the middle, both an increased height of the cataract and a greater volume of water, wliicli facilitate the recession here luitil retarded by its pro- truding ledges. of Canada] SLOPE OF NIAGARA EIVER 49 Below the foaming cauldron, the surface of the river forms a stretch of comparatively smooth water, so that it is navigable for a mile and a half, to near Cantilever bridge, with a descent of a foot and a half or less. I have seen this section of the river, during the season of 1905, twelve feet above ordinary stages, when the Canadian eda'e of the river above the falls had risen three feet. Turning now to the 'New York channel, there is a barrier in the upj^er part of the river similar to that described above the Canadian falls. This is situated a little below the head of Goat island, and is also the First cascade of the eastern channel. The descent of this Upper rapid is forty-five feet. The American falls at its northern edge descends 167-5 feet, while upon the margin, adjacent to Luna island, it is 169 feet. Luna island produces a third but intermediate cataract of small size, from which, before takins: the final leap, the waters bound from one ledge to another. DESCENT OF WHIRLPOOL RAPIDS, OF CASCADE AT OUTLET OF WHIRLPOOL, AND OF RAPIDS AT FOSTER FLATS. Below the head of the Whirlpool rapids the river descends fifty-one and a half feet to the Whirlpool, and again seven feet to another smoother stretch above Foster flats. Twenty feet more of violent rapids reduce the river to the level as at the foot of Foster flats, and sixteen feet more of descent to the mouth of the gorge. These features are more fully shown in the accompanying table of the slope of the river, and also in the longitudinal section, figure 3. The level of the river in the gorge is constantly changing, and this differing at various points. It may vary from a few inches to two or three feet — these pulsations occur irregularly, varying from a fraction of a minute to a few minutes apart. 50 FALLS OF NIAGARA [Geol. Surv. 50 >, -F^' of Canada] slope of ?; i agar a river Table of slope of Xiagara River. 51 Lake Erie— At Pt. C.)lborne, 1860 1905 lS91-in05 I!t04 (Jan.-Nov.) At Cleveland, 1855-1905 1904 (Jan. -Nov.) Niagara river — At Bridgeburg Low . . High. 1871-1895 Mean. Tonawanda, N.Y., 1871-1895 Schlosser, N. Y Low.. High. 1871-1895 Chippawa, 1904 (Jan. -Nov.) At Mean Lake level Head of Rapids — Forebay Ontario Power Company, 1904 (Jan. - Nov.).. '. Rock floor at intake At elbow of loop behind Dufferin island ... . Top Canadian falls, central apex, opposite Monu- ment of Survey, T. P. 6 Edge of river opposite T. P. 6 Edge of falls at Goat island Edge of falls on Canadian side, ordinary stage (new shore line) Edge on Canadian side, verv high water Bench Marks, Table-Rock House Descent of Upper rapids, Canadian side from Ont. Co. forebay River surface below falls at end of Electrical Development timnel River surface 1,000 feet farther down (Ontario Power Company). . . River surface Very high. Descent of Canadian falls, centre 11 11 westein side 1. 11 eastern side Top of American falls, northern side 11 11 southern side, at Luna island Descent of Upper rapids. New York channel, from level of Ont. Co. forebaj' at western end of Green cascade Descent of American falls River s-\irface at Cantilever bridge, head of Whirlpool rai)ids River surface under Grand Trunk Railway bridge River surface eddy at foot of Whirlpool rapids proper River surface at Whirlpool Descent of Whirlpocjl rapids Above Sea. 'Above Lake I Ontario. Fall. Ft. Ft. 572-35 326-58 571 05 572-28 572-67 572-51 566-49 568-74 567-46 564-76 .561-69 566 19 562 561 18* 315 .562-50 558-75 313-7 552 306 532 521 + 275 520-5 274 505 - 5 259 504-3 258 507 3 511-13 346 + 346 ± 358 512-5 511 345 343 301 293 265 100- 100 ± 112 266-5 265 98-5 96 5 54-5 47 Ft. 55 175 158 160 50 167-5 51-5 This datum appears too high. 52 FALLS OF NIAGARA [Geol. Surv. Table of slope of Nugara River — Continued. Above Sea. Above Lake Ontario. Fall. Ft. Ft. Basin between Whirlpool and Foster flats |£; Foot of Foster flats, in eddy : Cable crossing (Ontario Power Company) . . ■" Mouth of gorge. ... .... River at Queenston Low .. High. 1875-189:. Mean. Lake Ontario — At Toronto, 1855-1905 1891-1905 ,. I Descent from Lake Erie to Lake Ontario (1891- 1905) : 28G 40 2ti5 19 2.-.6 10 249 3 24fi-20 20 250-71 4 7 247 -93 f- 20 245-86 245 -07 Ft. 320-58 t Mean of 1891 -1905 would be lower. The lake levels are reduced from the mean monthly tables in the Lake Survey reports, and unpublished records of Wel- land canal and Toronto harbour. The datum at Niagara falls, the profile of the Niagara Falls Park and River railway, the section at the Michigan Central Railway bridge, Mr, Jen- nings' measurement of the rapids, levels determined by the power companies, new levels concerning the falls taken by myself, wherein the first attempt at measuring the height of their centre appears, and Mr. White's ' Altitudes in Canada ' for other points of the slope of the river, have furnished the above information, which is the most precise obtainable. For convenience, unless otherwise specified, throughout this work Lake Erie will be considered as 326 feet above Lake Ontario ; and Lake Ontario as 246 feet above the sea, as^the elevation of 245 feet applies only since 1890. Attention should be called to the fact that the Canadian edge of the falls has been artificially carried back about 415 feet. This should have slightly increased the height, but as the water was diverted to a deeper part of the channel the des- cent was not materially changed. CHAPTER V. NEW SOUNDINGS IN THE GORGE OF NIAGARA RIVER. Former soundings in the river. Soundings in the Whirlpool. Position of the new soundings. Soundings below the Whirlpool out- Soundings under the Canadian falls let. (1906, first attempted). Soundings below Foster flats, and Soundings from the Canadian falls just inside of gorge. to near the Cantilever bridge. Soundings in the river beyond end Soundings and borings at Canti- of gorge. lever bridge. Soundings above the Upper rapids. Whirlpool rapids. Depth of water on the Upper rapids. rOKMER SOUNDINGS IN THE RIVER. Soundings of the depth of the river had been made where navigation demanded. Thus they were taken in the river from Lake Erie to as near the rapids above the falls as the boats dare go without risk. Soundings were also made in the river below the mouth of the gorge as far up as the landing stages of Queenston and Lewiston. Since before 1750 the earlj settlers of the country used small boats to cross the river below the line of the American falls^ as is done now by the Maid of the Mist. In this region soundings were also made by the United States Lake Survey, and published in 1875. Another line of sound- ings was made at the Michigan Central Railway bridge in 1S99. These have been published by Mr. P. W. Curry* from the data furnished by the engineers of the railway. No other soundings have ever been published. The present data at the bridge were kindly furnished me by Mr. II. Ibsen, bridge engi- neer, of the Michigan Central railway. The current at the Michigan Central bridge is very strong, and rendered the sounding of the depths a difficult undertaking. Trans. Can. Inst.. Toronto, vol. VII., p. 7, 1901 53 54 FALLS OF NIAGARA '^*^®°'- ^"'"^• It was, however, accomplished by using weights of GOO ]ionnds made in the form of a tad-pole, f supported on an axis, so as to offer the least resistance to the current. A large sized telegraph wire was used. Three out of four of the sinkers were eventu- ally lost. The soundings could not disclose the character of the gorge, as beneath the bottom of the river, the channel was re- filled with fallen blocks. The statement was made that no other soundings had been published, but it is said that a Mr. TvTissen, in a boat called The Fool-Killer, made soundings in the Whirlpool and elsewhere wliich were not published. Soundings, however, taken in swift currents bv ordinary methods are very unreliable. POSITION OF THE NEW SOUNDINGS. In my investigations, the question arose as to what was the character of the channel beneath the surface. Without a knowledge of the changing features I could not determine what work the river had done at the various points. No one knew the depth of the river in front of the American falls, or at any point above in the direction of the Canadian falls. All guesses as to the depth beneath the Canadian falls proved most erroneous, based as they were upon deep soundings (previous to my work), below the line of the American falls, two-thirds of a mile away, which were taken to indicate the depth of the river throughout its course, interrupted, however, by the Whirlpool rapids. The Whirlpool is situated along the course of an ancient buried channel. Its depth was a mystery. Just beyond the outlet of the Whirlpool the channel is modern. What had been accomplished by the ancient stream, and what by the modern river? Here was another ])oint where a knowledge of the depth of the river was necessary. Borings at the Michigan Central Railway bridge give proof as to the character of the old valley, t See also Rept. of Chief of Engineers, U.S.A., pt. VIII., 1900. of Canada] sou:n-dings in the gorge 55 tliong-li more or less filled as before mentioned, and the sound- ings at the bridge supplied information from which to draw con- clusions as to the depth of the Whirlpool rapids. Farther do'^ni the river below the Whirlpool .are Foster flats. Here there had been great changes in the river. The soundings below the mouth of the Whirlpool would show what the falls had done after they had passed Foster flats, and before reaching the Whirlpool gorge. The rapids in the nar- rows opposite Foster flats could only be regarded as a repetition of the Whirlpool rapids on a modified scale — occupying a channel partly refilled by the extraordinary mass of fallen rock which occurs at this point. It was necessary to know the character of the channel below the flats inside the gorge and near its end, as well as just beyond. The soundings of 1005 brought to light for the first time the nature of the floor of the gorge, leaving us still ignorant of the dej^th of the river under the Great Falls. To make any soundings here seemed at first impossible. Captain Carter of the Maid of the Mist and ]\lr. Wilson, with others, were ready to assist. Finally the soundings were successfully made in September, 1906, completely changing prevailing views of the recession. Other methods than those used elsewhere had to be adopted for sounding under the falls. A buoy about 3*5 feet long pointed below and loaded at the end with twenty-seven pounds of lead, (the whole weighing fifty pounds) was made. About six inches of it floated above the surface of the water. A rod on toi? carried a red flag. The lower end was covered with a lead shoe to record the scratches when l)ottoni should be struck. Within this buoy were two Tanner-Blish sounding tubes doubly protected from the jarring blows. The buoy Avas taken out in a tug from Fort Day through the kindness of Mr. Champagne, and placed overboard by my assistant, Mr. Walker. Different points chosen for sending off the buoy were determined by the 56 FALLS OF NIAGARA [Geol. Surv. distance from an anchored buoj. The course of the sounding buoy was observed by others. It struck the rapids of the First cascade at different trials in fifteen to twenty minutes. Another fifteen to twenty minutes carried it down the Upper rapids over the falls and some distance below, even as far as Carter cove, whe.re it was picked up by the steamer. SOUNDINGS UNDER THE FALLS FIRST ATTEMPTED. As shown on the large map and in a section (figure 5), the depth of the river is eighty-four feet at a point about a thou- sand feet from the falls. This point is near the middle of the gorge, with a deeper channel reaching to 192 feet nearer the eastern side. As the shelf was found for some distance doAvn the river it was thought to extend to the falls themselves, as also the deeper inner channel. The opinion as to the shelf was found to be correct, but not that as to the deeper channel. ISTear the centre of the apex, where the volume of water is largest, soundings under the falls reached depths of sixty-nine and seventy-two feet. From the markings on the lead shoe of the buoy it was found that the sixty-nine foot sounding struck hard rock of a boulder or fallen block, while the seventy-two foot sounding scraped on such a surface as would be shown in strik- ing shales. On the eastern edge of the apex the buoy struck rock with such force as to damage the end, showing that sound- ings could not be made there, as the water was checked by ledges breaking its descent. The width between the rock wall mentioned and the point of the seventy-two foot sounding -is too narrow to permit of a deep channel corresponding to that of 192 feet extending to the present site of the great channel. The effective depth below the falls, being somewhat less than that on the shelf (which has a general depth of 80 to 100 feet extending as far as Carter cove) may be due to the boulders or fallen blocks, or the floor may rise upwards as in the case of the cove behind the Wilson ridge at Foster flats. These variations of Canada] SOUNDIXGS IN THE GOKGE 57 of the depth have changed our idea of the mode of recession, and of the history of the channel itself. SOUXDIXGS FROM THE CANADIAN FALLS TO NEAR CANTILEVER BRIDGE. The soundings in this section of the gorge were made from the Maid of the Mist, and extended from a point in the foam as near the cataract as it was safe for the vessel to navigate, for a distance of a mile and a half, to a point near the Canti- lever bridge above the Whirlpool rapids. The soundings on the chart show extremely variable depths. This is due to a narrow and very deep channel penetrating the general floor, which floor near the falls is from 80 to 100 feet below the surface. Still there are points on it where reefs occur much nearer the surface of the water. An extensive one is found opposite the American falls, where at extremely low water great boulders may be seen near the surface. The current here is rapid, and on account of the danger in sailing over rocks a detailed exploration was not undertaken. On my large chart or map this reef is shown to extend about one- third of the way across the river, but it is supposed to reach farther than this distance from the Canadian side. Additional information is thrown upon this subject by my assistant, Mr. Claude E. Eldridge, in his sketch map of the ice-jam of March, 1906 (figure 4). In this figure, a stranded ice mass is seen about two-thirds away from the Canadian shore, nearly opposite Luna island. It is at present supposed that the floor rises so near the surface that at this point one of the great rock blocks protrudes to hold the ice. Indeed, near this point, I found a depth of only fifty-seven feet, and I have seen fallen blocks in one case standing fifty feet high, where the mass was lying on its edge. This observation is of importance as showing that the deep channel, which trenches the floor under the Goat island shelf, is very narrow. A cross section from the Goat 58 FALLS OF XIAGAKA [Geol. Surv. Fig. 4. Sketch ef Ice- jam, March, 1906, showing stranded mass at x (C. E. Eldridge.) island shelf is shown in figure 5. The deep channel is further illustrated in figure 6, which is a section between Carter cove and the shore below Prospect point. u. s./ \ 1 1 51 / / / y y L. ~ " ^ •c^_=_-_"- - -^ - ^-^> }A ."--- - -- -- ^ -- -_-.^ -- "_ ^ -^ / / y y M. s. Scale ° , ^°— ^- 200 1 300 \ \ Feet Fig. 5. Section from near Table Rock Hou>;e to the Goat Island shelf. Horizontal and vertical scale the same. R. S., river surface ; L. O., level of Lake Ontario ; g. M., band of Medina gray sandstone ; M. s., Medina shale. This legend applies to the following sections and s«' ^1 § iOO ^1/ ijarwrv^^ (k^Si^-" L. /' 'o: M 3, Scale ° 100 I V 200 I 1 1 ^ / , 3O0 Feet Fig. 7. Section « t the Upper Arch bridge. The discoverv of the reduced depth of the river under the falls, showing its effective excavating power, confirmed the idea that in this vicinity the height of the falls had been greater but is now reduced by the backing water in the gorge. SOUNDIlSrOS AND BORINGS AT CANTILEVER BRIDGE. At the east pier of the bridge, the Michigan Central railway bored with a diamond drill in 1899, to determine the character of the foundation. This is located in the upper end of the E'arrows of the gorge of the AVliirlpool rapids, and for the first time disclosed its character, showing a succession of clay, and boulders, mostly limestone, extending to a depth of 185 feet below^ the surface of the river, which is here about ninety- seven feet above Lake Ontario. {See Chapter xii., on Whirl- pool Rapids section.) Thus it is seen that deep channel in the rocks, extending from the falls through the wider portion of the caiion, still continues at the same depth of about eighty- 60 FALLS OF NIAGARA [Geol. Surv. seven feet below lake level into the l^arrows of the ^ATiirpool rapids. The first geological uses of this section were made by Mr. P. W. Currie, who published it in the ' Transactions of the Canadian Institute.' As may be seen in cross-section, figure S, this buried channel is refilled so that the present greatest depth of the river is eighty-six feet. R.M. y \ \ ' \ ' — 1 1 r.e/ > S. , / ^ ( / £^=i_M.^z^J^ ^-^. ■-': L. ^ / O M. s. Scalt 100 1 refilled « iith block J , ^r , ^y F«et, Fig. 8. Section of river at Cantilever bridge. THE WHIRLPOOL RAPIDS. From this point on the velocity of the current becomes greatly accelerated, as between here and the "Wliirlpool the de- scent is over fifty-one feet. While the water shoots with a smooth surface under the railway bridges, it soon becomes tempestuous in the rapids. Adjacent to them are some un- usually large masses, one of which is 120 feet long, and more than 60 feet wide, and 10 feet thick. Such have fallen re- cently, as is shown by the entire absence of an incipient caiion where Muddy creek tumbles over the soft rock on the "vestern side of the Narrows. It is with such blocks that the bottom of the channel is refilled, some reaching near the surface, pro- ducing boiling breakers. ISTo ordinary rock-mass could resist these wild currents, broken and interrupted by all manner of secondary currents, and succession of tossing billows rushing down at the rate of from twenty to thirty miles an hour. Even this velocity is indeterminable on account of the constant changes, and we know nothing about the under currents. The of Canada] SOUNDINGS IN THE GORGE 61 sounding lead might be lowered into the river from a cable, but there is every probability (hat it would be caught in some crevice among the rocks. No useful information would be derived from ascertaining the actual depth, as the borings at the Canti- lever bridge show that the channel in the rock formations had been excavated to its full depth, before the fallen blocks had obstructed the passage of the river. The maximum depth in the rapids is very much less than eighty-six feet. In spite of their wild character some daring spirits have navigated these rapids in barrels and specially constructed boats. In one case a boat was made wdth water-tight compart- ments and a keel of iron weighing 1,600 pounds. The navi- gator was strapped into it in a manner that enabled him to release himself. On one of his two voyages the boat upset in spite of the heavy keel, when the man was under water for over a minute, after which the boat righted itself and he escaped. On the other voyage his boat shot nearer the Canadian bank and did not upturn. A further illustration of the caprices of the current occurred lately. Five small, flat-bottomed boats, belonging to the Maid of the Mist, were cut loose at the landing of that steamer by some miscreant and turned adrift. All of these went through the rapids into the Whirlpool where they were recovered. One of them w^as upset, tw^o more were filled with water, while two others containing the oars went through in an almost dry condition. The first steamer, Maid of the Mist, shot these rapids under a full head of steam and escaped. SOUNDINGS IN THE WHIELPOOL. There has been much speculation and interest as to the depth of the Whirlpool. Mr. Nissen made some soundings in it from his boat, but these are not known. He could not di- rectly cross its course. For those who are not familiar with the Whirlpool let it be said that the waters come into it, crosses its outlet, circle round, and finally pass out almost entirely 62 FALLS OF NIAGARA tGeol. Surv. as undei'-eurrent'^. Thus any floating object carried into it may remain swirling round for days or even weeks. Where the upper current comes in contact with the lower, great tim- bers may be seen pitching down endwise, being dragged into the lower current, but most of these soon come to the surface again. There is always much material in the AAliirlpool con- stantly floating round its course, advancing or retreating from the shore with the intermittent surging. During the last season the quantity of logs was greatly in- creased by the removal of the temporary dams of the power companies, who dislodged their timbers by dynamite, thus send- ing them into the river. These caused the greatest difficulty in making the soundings. While a boat could sail round in the safer waters it could not cross the centre of the pool, where the currents meet, lest it be drawn endwise into the boiling vortices. The surface of the Whirlpool is forty-seven feet above Lake Ontario. In my paper on the ' Duration of Niagara Falls,' I had made sections of the river channel, of which the Whirlpool is an arm, assuming the depth as substantially the same as the deepest soundings above. From one of these soundings I con- cluded that the river and Whirlpool reached a depth of froin 90 to 100 feet below the level of Lake Ontario, which would indicate channel 136 to 146 feet deep. I have heard it esti- mated at 300 feet, but the great current is an under-tow, so that its depth was not determinable, except by measurement with specially devised appliances. My method of sounding here at the Whirlpool, as at other points, was by swinging a cable across a stretch of water of 1,150 feet. This distance was considerably increased as it was neces- sary to carry the cable to a height of sixty or eighty feet above the water in order to allow for curvature, and to prevent its fall- ing into the Avater lest it be caught by the drifting wood. This cable of seven strands is the same as that used on the naviga- tional sounding instruments of Commander Tanner. It is sup- of Canada] SOUJ^DINGS IN THE GORGE 63 posed to stand a strain of 600 pounds, althougli at one time it was subjected to a strain of 900 pounds without breaking. But at different times it was broken. It was suspended at the two ends on movable drums, with a ]iullev wheel clamped to it, which was shifted to any desired point over the river. Through this wheel a second cable was operated, having at one end the sounding lead, and at the drum a recording meter. By this means the position, and by the use of hydrostatic tubes, the depth to which the sounding lead reached could be measured. Two ' leads ' were used, one of which was twelve and a half and the other thirty pounds in Aveight. They were made of lead with a form offering least resistance possible, yet these were beaten about and hurled against the rocks by the current as if struck with a sledge hammer. The depths were accurately determined by the Tanner-Blish tube of small bore, with rings closely ground on the inside. These tubes are twenty-four inches long with a rubber cap on their upper end. With the closed end above, one was placed in a brass tube upon support- ing springs so as to relieve the jar, and this again was inserted in the sounding leads. Protected from the currents these little tubes quietly recorded the depth of the water ; for the pres- sure compressed the air into a smaller space, and the rising water wet the inside of the tube. So far as moistened, the ground rings, which appear white when dry, become trans- parent and sharply mark the height to which the water has ascended. By measuring this on the proper scale the depth in fathoms is immediately read off. In a very few cases the rough- ness of the current dashing the leads against the rocks rendered the readings not quite certain. In such cases the readings were rejected. When the friction brake of the drum was released, the weight was allowed to sink as rapidly as possible in order to reach bottom along the shortest line, as otherwise the current carried it down. In some cases this line when measured did 64 FALLS OF JSriAGAKA [Geol. Surv. not materially exceed the depth recorded in the gauges. Some- times the lead was carried far out of position, but in every case the sudden striking of the l)ottom and releasing of tension on the cable was instantly felt at the drum, and the slack of the wire began to uncoil. Accordingly there was no doubt when the of Canada] SOUNDINGS IN TILE WHIRLPOOL 65 bottom was touched. In this way I made a line of soundings between Thompson point, at the outlet of the Whirlpool, and the little gorge of Colt ravine, in directions at right angles to the axis of the Whirlpool. This line is on the edge of the Whirl- pool proper, where it joins the true river. (See the large map. The section is illustrated in figure 9 (p. 64). IvTine soundings were made along this line, besides scattered ones. Close under Thompson point a depth of seventy-five feet was obtained. This increased to 102 feet at a point about 500 feet from shore. On the western side of the Wliirlpool there is a shelf submerged to a depth of only eighteen feet at a point about 150 feet from the shore. Just beyond there is a sudden deepening to seventy-eight and eighty-seven feet. In sounding under Thompson point the inward currents were moderately strong, but at the point where the depth was 102 feet there was a neutral zone, so that the sounding cable paid out only three feet more than the hydrostatic depth. Between this sounding and one of eighty-seven feet upon the other side a very re- markable result was obtained. Long after the drum Avas ex- pected to have ceased uncoiling the weight and cable still con- tinued t(3 run out. At the surface the current did not greatly deflect the cable, but it was caught by an under current which carried it some 200 feet (beyond the point where it entered the water), into the channel of the river proper — -that is to say, in the course of the river as if there had been no Wliirlpool. The whole feature, however, is, generally speaking, regarded as the Whirlpool. It is an important distinction in the study of the mechanics of the river. On account of the vortices in the currents, and the dilficulty of swinging the cable, it liad seemed impracticable to determine the depth at this point. The current favoured the sounding here, which reached a depth of 126 feet, indicating that the channel was deeper than the Whirlpool proper, and that the lead had reached a point seventy-nine feet below the level of Lake 5 66 FALLS OF XIAGARA t^eol. Surv. Ontario. This is a depth of only eight to fifteen feet less than that of the river above Whirlpool rapids, or that of the refilled Whirlpool rapid channel revealed at Cantilever bridge. Yet the lead had not been drawn into the deepest portion of the channel. Putting all these things together it now appears for the first time that tlie forces, which excavated the Whirlpool rapids section, acted to the same depth as in the gorge above, and in the Whirlpool beloAv the rapids. Learning, by several failures, how to sound the Whirlpool, I now think it possible to make soundings in the outer channel, but for the discovery of an additional ten feet or more in depth the result would not be commensurate with a very high cost. In the last mentioned sounding the lead was peculiarly checked by battering against numerous small points, as if it struck a ragged edge of the Medina shale, and not scraped bright by occasional heavy blows on hard limestone as had occurred elsewhere ; so that it is prob- able that the pre-giacial channel does not exceed a depth of more than fifty feet below lake level, if so much. Before leaving the Whirlpool it may be said that the sus- pension cable was carried round the j^ool in a boat, illustrated in Plate si. a. Some additional soundings were also taken, wdiich are shown on the chart, reaching from twenty-four to ninety feet in depth. SOUNDINGS BELOW THE OUTLET OF THE WHIRLPOOL. Below the Whirlpool outlet is a descent of about seven feet, where the waters again become smoother before passing on to the rapids at Foster flats. From a point on the Gorge Rail- way track I attempted to send a cable across the river in a boat, as shown in Plate xi. b, opposite this page. Attached at one end to the boat the cable was paid out from the, shore ; the boat reached the opposite side and was caught by an assistant, but before it could be landed the current swept it down so that Platk XI. A. View of end of the Whirlpool, with boat carrying the cable acrotss for sounding purposes. Plate XI. b. View of Boat crossing the rapids, just below outlet of Whirlpool, and carrying a cable across river for sounding purposes. 5i 67 of Canada] SOUNDINGS IN THE GOEGE 69 the cable had to be cut in order to save the men, who were carried past the vortex of a small whirlpool to where there was imminent danger. A second attempt was made by having a long rope lying in the boat t(» which the cable was attached so that it pai M. 8. Feet Fig. 10. Section of the river, quarter of mile below Whii-lpool outlet and above Foster flats. I nmst here pav tribute to the bravery and skill of James Scott and Alexander Leger in the navigation of this hitherto unexplored part of the river, for without them it would probably have still remained unknown. Fred. Scott was the third man. SOUXDIXGS BELOW^ FOSTER FLATS,. AT\^D JUST IXSIDE THE GORGE. It was necessary to ascertain the character of the river beloA\- Fosrer iials. The Ontario Power Company had suc- ceeded in getting a cable across the river swung from the top of the gorge. This was kindly placed at my disposal by Mr. Banker Payne, the general manager. Here the river has lost much of its depth, being onlv sixty-three feet. This is at a level ten feet above the lake ; thus makino- the soundings \ ^X, A*. S. y^ y ;a. VI =^^^ ■i^^d-zz-'z^-^l L. o. Scale 100 I I r 200 300 1 M. 3. Feet. 1 21 jy Fig. 11. Section of river below I'"()ster Hats and a mile and a-fjuarter within gorge. of Canada] SOUNDINGS IN THE GORGK 71 reach to tlic dejDtli of about fifty-tliree feet below the surface of Lake Ontario. We find the depth a little less here than in the reach between Foster fiats and the outlet of the Whirlpool. But at this point, which is a mile and a quarter above the mouth of the gorge, there is still a strong current. At a point about a half mile within the gorge and opposite the second bridge of the Gorge railway I made several sound- ings reaching to 138, 111: and 150 feet respectively. This deeper channel is on the ISTew York side of the river. A quarter of a mile above this place one or two soundings were taken to a depth of 120 feet, while a few hundred yards above an- other one gave sixty-nine feet, though complete sections across the river were not undertaken. At these points the river can- not be more than three or four feet above Lake Ontario, and accordingly the channel here is very much deeper than in the upper part of the river. L. \.>^- ^y o. \ ^^ 1 1 ^1 •51 M. 8. 1 1 Scale ? r- 100 - 1 zoo .1 _^_S00 Feet. Fig. 12. Ssction across the river, a quarter of mile within the end of gorge. SOUNDINGS IN THE RIVEK BEYOND THE END OF THE GORGE. At the mouth of the gorge, three soundings, less than 200 feet apart, were made from the Suspension bridge. The deepest of these was ninety-nine feet. At the time they were taken I did not suspect a deep channel both above and below the bridge. It is possible that intermediate soundings may show a con- tinuation of the unfilled channel of 150 feet in depth. A third of a mile below the end of the gorge, at a point nearly opposite the Queenston dock the floor in the middle of the river 72 FALLS OF NL\GARA f^^°'- S"'"''" reached ninety feet, but passing toward the ]S[ew York side, a narrow gorge was found which had a great depth. Here a considerable number of leads were cast, reaching to 120, 135, 150 171 and even 183 feet, showing a remarkable chasm. The discovery of this channel was due to the suggestion of my boatman, James Humphries, who could not find bottom when fishing. The Lake Survey soundings below here reached only to a depth of ninety-six feet. This is at a le\'el two to three feet above Lake Ontario. The current here runs at the rate of four or five miles an hour, so that it is difficult to hold a small boat in one position. This drowned gorge is narrow, perhaps not over 200 or 300 feet in width. It has also jDrecipitous walls, not merely shown by the rapid change of depth, but on pulling up the sounding lead on the western side;, the upper part was bent over the inner tube in a manner to indicate a blow which it could only have received in being drawn up against an overhanging ledge. Again on sounding nearer the eastern side the lead was allowed to drag in order to ascertain the maximum depth. In bringing it up the weight caught so that it had to be again released by tiie men rowing backward. Here, at a third of a mile beyond the portal of the great l^iagara gorge, is a submerged canon trenching the outer channel. Combining the evidence here with that inside the gorge I have discovered wliat is now a drowned falls or rapid, formed when the lake level was 180 feet or more lower than now. This w^as an entirely unsuspected feature in the physics of the river {See figure 3, page 50), and was formed while the Niagara was small, else the inner channel would have been very much broader. A little farther down, opposite Lewiston, the river shows a depth nowliere exceeding ninety-six feet, and below this point it is silted up so that the old channel is entirely obscured, with the modern river varying from twenty to sixty feet. In Lake Ontario, beyond the mouth of the river, the evidence of of Canada] SOUNDINGS 73 this silting- is observed in the fan-shaped delta deposit covered by about twenty feet of water though at its outer edge by about sixty feet, l)eyond which is a sudden descent of the hike floor to eventually 400 feet, which, however, is a pre-glacial trough, and affords no indication that the Xiagara channel was once deep. Indeed it is onh' the deep channel described which shows 74 FALT.S OF NIAGARA f^^^^" ^"'■''■ that the waters of Lake Ontario were ever lower than ninety- six feet, after the birth of Niagara river, and it is even possible that it may now be partly refilled. SOUNDIIS^GS ABOVE UPPER RAPIDS. The soundings in the river above the falls were made by the United States Lake Survey, and along a line between Chippawa and Grasse island the mean depths are reduced to sixteen feet. At a point below this section one sounding reached twenty-two feet. Southward the river increases to twenty feet in depth, and near the northern end of Grand island it is thirty- three feet. The depth of the upper river is extremely variable. At the Liternational bridge, about two miles below Lake Erie, it reaches fifty-three feet, which is much below the rock barrier at the head of the rapids above Niagara falls. At a mile and a quarter above the bridge it is only seventeen to twenty-four feet. DEPTH OF THE WATER OX THE UPPER RAPIDS. This is determined by the ledge of rocks producing the Greens or First cascade. Erom the Goat island shelf outward, for a distance of 400 feet or more to the channel separating the outermost Sister island, the depth of water at present ordinary stages may be taken at an average of only one foot, as shown in Plates xii. a and b, page 77. (See also Plates XXXI. A and b). Above the outermost Sister island and beyond, where there is a fall of six or seven feet, Plate xxxi. B, the depth is l)etween two and three feet and remains so for an unmeasured distance of perhaps 400 feet, beyond which the crest shows a rocky ridi^e with the water again re- duced so that at times the ledge appears almost bare. This condition may tlien extend more than half way across the river, beyond which to the Canadian shore the river is much deeper, reaching perhaps in places to twelve or even fifteen feet, as, in of Canada] SOUXDI.XGS 75 fi channel near the lower cascade, at the Electrical Development fore-hay, where crevices five feet deep were fonnd in the floor. {8ee Plates xiii. a and b). Aceordinjily hx-al, narrow, deep channels may occur. For a distance of nearly 1,700 feet from Goat island, the mean depth now can scarcely exceed two feet, though before the artificial division, the water was probablv a foot more. This shallowness on the First cascade is not so noticeable when the river is above mean stage. The average depth of the re- maining distance of 1,600 feet across to the Canadian bank is assumed to be nine feet or less, while Mr. James Wilson places it at less than seven 'feet, although there are deeper channels, as mentioned. Except in some channels of the First cascade east of Goat island, the depth of the river does not exceed a mean value of more than three or four feet, and less than a foot and a half on the American falls. (See chapter xsi.) Accord- ingly eighty per cent of the water here is found to pass on the Canadian side of the Boundary Line. Below the First cascade, the depth varies greatly, as seen on the shoals amone,- the breakers. Plate XII. a. \'it-w iif (Greens ur First Cascade across Ontario Power Co. Forebay, (from hill above western .side of Upper Rapids). Pi..\TE XII. p.. View of First Cascade, New York Channel (from Goat Island). 77 Plate XIII. a. View of Bed of River at Electrical Development Company's Forebay when water was drained off. The joints in the bed rock opened by solution of the limestone, leaving separated masses still in place. Plate XIII. View of deep crevice opened in bedrock by the river currents at Electrical Development Company's Forebay. • 78 CHAPTEE VI. ROCK STRUCTURE AFFECTING THE RECESSION OF THE FALLS. Rock structure of floor of Upper Mode of recession of falls in the rapids, joints, etc. rock formations. Effect of Falls -Chippawa valley Depth of the excavating powers of upon Upper rapids. the present falls. ROCK STRUCTURE OF FLOOR OF UPPER RAPIDS, JOINTS, ETC. . From an examination of the receding crest lines one sees that the rate of recession of the Canadian falls varies according to its form — whether a regular crescent, or one indented by a V-shaped apex, these being alternating features. The structure of the rock beneath the rapids in a manner facilitates the work of the falls. The rock floor under these ^ rapids, above the cataract, is everywhere composed of hard dolomitic limestone in beds of variable thickness — some being only a few inches thick while others are massive, especially nearer the brink of the falls. This rock floor is everywhere jointed, and these joints being opened up produce crevices as shown in Plates XIII. A and B, which are photographs of the river bed tem- porarily drained by the Electrical Development Company, These opened joints give rise to the greatest erosive effect near the edge of the falls, as they expose blocks of rock to tre- mendous force of the current, already accelerated by a descent of fifty feet, so that the frontal masses are wedged off and further expose the lower beds to the action of the falls. Such results are well illustrated by the JsTovember fall referred to before (See Plate vii., page 31). At other points on Goat island shelf, the water is seen to impinge upon the separated blocks, and rebound, producing pleasing effects of the rapids. As recognized over sixty years ago by Sir Charles Lyell, 80 FALLS OF NIAGARA ^^^°^- ^u^-v- these rapids above the falls have not been produced by the Niagara river, but are a sloping surface dating back to pre- glacial times. I have discovered that the rocky floor of the rapids was formerly the site of a broad shallow pre-glacial valley, now reopened and forming the floor of Queen Victoria Park on the Canadian side, but still buried at Goat island on the Xew York side. For the cause of this Falls-basin, filled and heaped over with drift deposits, search for many years was made; but not until my recent investigations was it re- warded by the discovery that this basin, bounded on one side by the rocks beneath the rapids, connects with the buried valley extending southwestward around the rapids towards Chippawa village. The rock ledge which determines the height of the river is almost horizontal on account of crossing the strike of the beds at a right angle. It is this structure which gives such uniform depth to the rim of the upper basin as shown in Plates xii. and XXXI. a and b. The rim at the First cascade is a conspicuous feature from the head of Goat island to a point above Dufferin islands, yet upon approaching the western side it is reduced in height sev- eral feet as the rocky floor passes under the drift deposits near the islands. Indeed, had the drift ridge which formed the banks of the original Xiagara river here been a little lower, the course of the channel would have been diverted half a mile or so to the west where the bed rock is lower than at Greens cas- cade. This change would have caused a reopening of the buried valley, with the lowering of Lake Erie and the formation of a cascade at its outlet. Then the rapids above the falls would not have been formed. EFFECT OF FALLS-CHIPPAWA VALLEY UPOX THE UPPER RAPIDS. This Falls-Chippawa valley just referred to has loft its im- pression upon the rapids. Looking down upon them from the -Bfu CO o -a o of Canada] UPPER KAPIDS 83 bills above, as shown in Plate xiv. (on page 82), one sees the smoother waters in front of the Canadian-Xiagara Company's plant, and below the rapids at the Electrical Development Company's works. These waters occupy a transverse channel trending from the widest part of the Park obliquely to near the head of the present apex of the falls. The waters rushing down the river are diverted at this cross channel toward the Goat island shelf, thus piling up the greatest volume of water to accelerate the formation of the apex in advance of the reced- ing falls. Still, as seen by the slope of the river, the increased height at the apex shows that the falls, which hitherto had been receding in the bed of the Falls-Chippawa valley, are now crossing the side of it. This fact doubtless accounts for much of the recent reduced rate of recession, and partly for the fact that the apex has scarcely advanced in twenty years. The greatest force of the falls has been expended along a line close to the present edge of the Goat island shelf, as is shown by the deep channel of 192 feet, which I found off Goat island, while the middle of the river below the falls is only eighty-four feet deep. The fall of rock which occurred in jSTovember, on the eastern side of the apex, caused a widening of the channel, a feature which is still continuing. After the survey of 1842 there was a great broadening of the gorge under the western part of the falls. MODE OE RECESSION OF THE FALLS IN THE ROCK FORMATIONS. In the near future much encroachment should be expected upon the upper part of the Goat island shelf from the un- dermining action of the main current of the river. On the other hand, the end of the shelf, being protected by great talus blocks, with but little water falling over it, may remain a per- manent feature. The removal of the heavy blocks of rock immediately under 84 FALLS OF NIAGARA ^Gecl. Surv. the bed of tlie river above the falls should not be overlooked, as there is a popular idea that the recession is entirely due to undermining by the removal of the soft shales beneath the lime- stone strata. In a certain case one bed alone has a thickness of eighteen feet, and is of the most durable character. Again, near the base of the limestone is a very massive bed projecting, on which the water is falling as already mentioned. The I^iagara limestones at the crest of the falls have a thickness of eighty to eighty-eight feet. They form a coping over the Niagara shales with a uniform thickness of sixty feet. These last rest upon a very hard Ijand of Clinton limestone from eight to ten feet thick, below which are the thinner layers of Clinton limestone — the whole here amounting to about tAventy feet. These rocks partly pass below the surface of the river at the falls. (See figure 2, j)age 48.) The Electrical Development Company has excavated a tun- nel under the falls which terminates at a point 580 feet in a direct course from the present shore line. Its bed rests in the Clinton limestone. Here may be witnessed the imposing effects of the under side of the falling sheet of water, now plunging into the boiling cauldron. On the western side is a talus- bearing ledge, but east of the portal the ledge has been broken away. At this ])ortal the top of the limestones is twelve feet above water level, or 112 feet above Lake Ontario. Beneath the Clinton limestones are five or six feet of Clinton shales. (See figure 2, page 48.) For the character of the rock beneath the surface of the water it is necessary to examine the exposures farther down the gorge. This can easily be done, as the general character- istics are remarkably regular after allowing for the dip of the strata. The massive Clinton limestone is of a uniformly persistent character. Beneath the Clinton beds occurs a series of ]\Iediner beds of limestones and shales. I believe that the first differential measurements for the pur- pose of determining the effect of the strata upon the recession of the falls were made by myself and appeared in ' Dura- tion of Xiagara Falls/ published in 1894," but the measure- ments given were partly barometric, and not sufficiently accurate for the greater detail of the present work, which is based upon instrumental determinations.f The measurements of the diiferent rock formations are taken from various points so as to bring out the characteristics of the beds, as to their thickness, and their dip or slope, which have an effect upon the recession of the falls. These data are plainly expressed in the following table, which is also the basis for showing the rock formations in the various sections. They also serve for determining the direction and amount of dip of the different strata throughout the course of the Xiagara pora-e. • Amer. Jour, of Sci., Vol. XLVIII., page 457, 1894. t These determinations were made by levelling at various points by this survey and by triangulation of certain inaccessible strata in the gorge by Mr. George A. Rucker, and by measurements of different power com- panies. 89 90 FALLS OF NIAGAEA [Geol. Surv. Tables of Elevations of Stkata in Niagara Gorge. Above Lake Ontario. Surface of ground (adjacent) Surface S'iagara limestone near or at edge of cano Top of Niagara shale. . . . Top Clinton thick band of limestone Top Clinton shale Top Medina series Top Medina gray band sand stone , Base of gray band and top of Medina red shale. . . . East ID b£ O 'o 'o Mouth Side. 333 345 322 322 309 (301) 248 240 225 219 (218) (212) 135( + 6) 133 (120) 111 m ^1 w S^S ., — ( . 01 s -a be o cc cS'C .K-^ >i Q. ° & . l^A -S'SX! rf'^ ^X'Ji ^ 1^ 340 330 345 348 326 312(+10) 305 + 25 (269) 256 209 (189) (184) 195 (176) (170) 330 (256): 195 (176) (169) 70 56 291 236 175 (155) (150) 63 (41) 300 (236) 175 (155) (150) (63) (41) Above Lake Ontario. aj o w o p^ o ■l O 374 272 262 186 126 105 99 (15) (-5) Canadiiin Niagara Co's. shaft near N.F.P.R. Co. house. Canadian Niagara Co., No. 2, north end of pit. Electrical Development Co., soiith end pit. Niagara Power Co., N.Y., 4,700 feet from portal of tunnel . Niagara Power Co. near intake. N iagara Power Co. ,8,000 feet trom portal of tunnel. Surface of ground (adjacent) Surface of rock (adjacent) . 374 264 254 175 116 96 90 (5) (-15) 322 (312) "i78 118 316 (305) 164 102 324 312 Surface Niagara limestone near edge of canon Top Niagara shale Top Clinton thick band lime- stone . . Top Clinton shale 257 166 104 271 160 (100) Top Aledina series Toi) Medina gray band sand- stone Base of gray band and top of Top of Clinton thick band at Portal of Niagara Falls Power Company 145 feet above Lake Ontario. Figures in parentheses are interpolated; others observed. From tbe mouth of gorge to Whirlpool and to Grand Trunk Railway bridge dip is 1 in 200. South 60° west. Between Ontario Power Company's shaft (near their power house). Intake of New York Niagara Company, and Electrical Development Company's pit, the dip is 1 in 160. South 25° east. Between Ontario Power Company's power house, the north end Cana- dian Niagara Company's pit and the south end Electrical Development Com- pany's pit, the dip is 1 in 110. South 25° west. of Canada] ROCK STRUCTURE 91 DIP OF STRATA. From the information given in the tables, and also shown in figured sections, it can be determined that the average dip of the strata from tbe mouth of the gorge to the railway bridges, or above, has a direction of south 60° west, amounting to one in two hundred or only twenty-six feet per mile. Before col- lecting all this information regarding the dip of the rocks it had been observed that the beds were disturbed at Hubbard point, a short distance above the railway bridges. (See Plate XVI. B, on page 117.) Indeed this point was a critical one in the history of the falls, as will be seen further on. From here southward there is a general average dip of one in one hundred and sixty, south 25° east, or about thirty-three feet per mile. From exposures in Victoria Park, and from the borings of the power companies, a local dip of one in one hundred and ten, south 25° west, is found. It is a thrust or twist, the axis of which crosses the gorge in the vicinity of Hubbard point, where the highest ridge of I^iagara limestone is found. From these very dry details, a reason appears for the ex- istence of the ledge at the First cascade extending from Goat island to the Canadian shore. It has substantially a hori- zontal surface, except where broken near the western end by the beginning of the ancient valley passing under tlie adjacent drift hills. This gives rise to a broad and relatively shallow river of uniform depth. This ridge forms the First cascade and is in reality the rim of the Erie basin, over which the waters are passing out. EFFECTS OF DIPPING STRATA ON THE RIVER. From Niagara falls in a stretch of slightly mure than two miles to the railway bridges the strata rise sixty feet, with variations in the dip of the beds owing to the change of direc- tion of the gorge. From the railway bridges to the Whirlpool 92 FALLS OF NL4.GAKA t««°l- ^^I'v. the beds are horizontal so far as one sees, but in reality they dip slightly crosswise of the river, tending to throw the deeper waters on the Canadian side. After passing this mile of dis- tance the river turns at right angles at the mouth of the ^^Hiirl- pool, and obliquely crosses the sloping beds. Below this, the gorge extends for a mile and a half to Devils Hole, in which, distance the beds rise thirty-five feet. Here the caiion turns at a shar]:) angle and extends in a direct line to its mouth two miles below. Along this portion of the gorge the strata rise forty feet more, but at its mouth they occur about eight feet higher on the eastern side than on the western side. The character of the rocks has, in a general wav, been described when considering the formations beneath the falls. There is a remarkable regularity in the thickness and character of the different formations, except that the upper limestone grows thinner on passing northward towards the edge of the ' mountain.' On the other hand, where the uppci- limestone should have been thickest, its surface has been deeply denuded, forming ancient valleys, buried by drift where this has not been removed. These valleys have had more to do with modifying the rate of recession of the falls than the variable thickness of the lime- stone. Beneath the capping bed of Xiagara limestone there appears to be no reason for the rate of recession of the falls being modified by the changing character of the underlying beds themselves. On proceeding downward, the lower strata, in the canon rise, besides which the slope of the river descends. Con- sequently, the eifect upou the falls of the various harder or softer beds is determinable at any point, and really depends upon the height of the falls at any particular jdace, while the volume of water remains constant. At the present time all of the Niagara and Clinton forma- tion have been penetrated. Even beneath the falls themselves the hard bed of Clinton limestone, having a thickness of eight of Canada] ^^^^^^ STKUCTURE 93 or nine feet in one layer, is only twelve feet above the surface of the river ; but here all the underlying beds have been re- moved to the hard l);ind of ^Medina gray sandstone which forms part of the floor of the river. This bed rises and the river descends so that it constitutes the formation at the mouth of the A^liirlpool both above and below its surface. From this point on it can be seen to rise until it has a height of 133 feet above Lake Ontario at the western side of the mouth of the gorge. Everywhere below the Whirlpool the river flows on the shaly beds of the Medina series. But this was not always the case in the history of the river, for at times both the hard Clinton limestone and the Medina gray sandstone gave rise to sec- ondary falls which later were united with the main cataract. There is a tendency in these formations to produce three cata- racts, as at DeCou falls a few miles to the west, and at Swazee beyond in the ' Short hills.' A more notable case occur? in the Genesee river, at Rochester, where there are three cataracts. The narrows of the channel passing by Foster flats are occupied l)y impetuous rapids beloAV the level of the sand- stone, and these are due to obstructions by great rocks of limestones and sandstones which have fallen into the river. Even the Whirlpool rapids might be expected to owe their origin in part to the Medina band producing a cataract, but here the still deeper channel refilled by the fallen blocks of limestone has been found, showing such was not the case. The Clinton formation beneath the sixty feet of Niagara shale is composed of about twenty feet of limestone overlying five or six feet of its oavu shale. As the upper member of this is generally composed of a very compact limestone in a single bed of eight or nine feet in thickness^ it gives rise to a very strong lopographic feature, so that everywhere in the gorge it stands out and forms a shelf receiving the crumbling debris from the overhanging beds. That this formed the floor of a sec- 94 FALLS OF NIAGARA t^-^«°'- S"'^" ondarj falls there is ample proof in the Wilson terrace and Smeaton ravine, features of so ninch importance that thej will be considered in separate chapters. The large terrace of Foster flats is a remnant of the old river bottom underlaid by the Medina sandstone, from which farther down the gorge the third cataract descended. IRKEGULAKITIES OF CAPPING STRATA OF NIAGARA LIMESTONE. The capping band of Xiagara limestone at the outlet of the canon is reduced to a few feet in thickness^ although the formation attains 140 feet or more at the head of the rapids above the falls. As stated before, this thickness is greatly re- duced in the trough adjacent to the falls in Queen Victoria Park on one side, and at Goat island on the other. This Falls- Chippawa trough was cut to a depth of sixty or eighty feet into the rock surface of the country, while in the vicinity of Hub- bard point the ISTiagara limestones attain their greatest eleva- tion, forming Lyell ridge transverse to the course of the river, to a height of from sixty to one hundred and ten feet above the trough mentioned. Although the ridge is high, the canon has dissected it. Add to this irregular surface feature the buried Whirlpool-St. David valley, there might be found causes for great variation in the recession of the falls due to topo- graphic features. Below the ^¥hirlpool and also near the end of the caiion other transverse rocky ridges, of no considerable height, have to be crossed. From what has been observed there is little in the structure of the rocks which would give rise to great variation in the recession of the falls during any period of constant height or moderately uniform discharge in the volume of the river. The inferior height of the falls at the mouth of the gorge should compensate for the effects of thinner hard capping rocks. The variations in the recession of Niagara, when considering the question as a whole, are dependent upon the changing volume of Canada] ROCK STKUCTUKE 95 and descent of the river, and upon the crossing of higher rock, or buried valleys, due to the ancient topography of the region. The rock structure, as old geological formations, might be reviewed as a separate subject apart from its effect upon Niagara river. Knowing the general character of the forma- tions in relation to the gorge, the next step is to see what work has been done by the river at different points regardless of the causes. CHAPTER VIII. CHARACTER OF THE GORGE EXCAVATED BY NIAGARA RR^R. Preface. Smeaton ravine. Deep channel beyond end of gorge Foster reach and Devils Hole. and drowned rapids or falls. Whirlpool rapids reach. Lower reach of the gorge sections. Falls reach. PREFACE. Since men began to think seriously as to the origin of the gorge which is headed hj the cataract of jSTiagara falls, few people supposed the canon was due to other causes than exca- vation by the falls. Even as far back as 1789, at a time when the antiquity of the earth was almost entirely discredited, Andrew Ellicott hazarded an assumption as to the time that the falls had taken to recede from the mouth of the canon to their site at that time.* Over half a century later a disting- uished geologist attributed the formation of the gorge to fault- ing and Assuring of the earth's crust; but this was at a time when the theory of faults was resorted -to as a cloak for ignorance. The idea that Niagara falls had a beginning pre-supposed that the waters of Erie were once scarcely above the level of Lake Ontario, which were afterwards lowered. As far back as 1837 Mr. Thomas Eoyf measured old beach lines about Lake Ontario which indicated that the water of that lake had been as high as that of Lake Erie. Later Prof. R. Bell of the Geological Survey described, at the head of Lake Ontario, the occurrence of high level beaches.:}: This was further amplified by the writer in 1882.** *f ocality cited paste 20. t In ' Geology of Canada,' 1863. j In ' Geology of Canada,' 1863. ** ' Geology of Region about western end of Lake Ontario.' J. W. Spencer, Can. Nat., Vol. X., 1S82. 7 97 98 FALLS OF NIAGARA t^^^'' S"^^" The S2>ecific application of the former higher lake level to the reduction of the descent of the falls, and the subsequent calculations of the diminished excavating power, did not appear until in mj paper, ' Duration of ISTiagara Falls,' which was published in 1894, '^ In the paper referred to I gave but little prominence to the height of the falls at their birth, but mainly considered the level of the lake when it was 135 feet higher than now. The omissions in that paper are now filled in. In antici- pation it may be said that I made a survey of the river channel and marked its boundaries as they were just before the birth of the falls. DEEP CHANNEL BEYOND THE GORGE DROWNED RAPIDS. The first section across l^iagara river is taken at a j^oint about 1,800 feet below the outlet of the gorge. Here the chan- nel was not made by the cascading of the waters from the table land of the Niagara plateau, or from the Medina sandstones now protruding from the side of the escarpment, as the section is much beyond the former limit of the falls. At this point the banks of the river arc from sixty to seventy-five feet in height, with higher slopes of the land immediately beyond, rising to 100 feet or more. The rock formation is Medina red shale. The breadth of the river is about 1,200 feet, and the western half of the floor deepens to ninety feet below the sur- face. (See figure 13, page 73.) This floor is trenched by a narrow gorge reaching to 183 feet below the river surface or over 300 feet below the adjacent plain of red shales. This inner channel is from 200 to 300 feet in width. The small river flowed in this when the waters of Ontario stood at 180 feet or perhaps more, below their present level. The subsiding of the Ontario waters brought into existence a lower fall cascading from the Medina sHnd- stone, and the now drowned rapids, below the new falls, exca- * Am. Jour. Sci. cited before. of Canada] CIIARACTKR OF GORGE 99 vated the eliainiol to this depth. The deepest known sounding's below this point did not exceed ninety-six feet, so that the discovery of tliis river cliannel is very important as proving that at a time long after the birth of the ISTiagara falls the descent of the river became 180 feet greater than now, or perliaps even more, as the subsequent refilling may have obscured the deepest part. THE LOWER REACH OF THE GORGE SECTIONS. The iSTiagara gorge commences at Queenston and Lewaston where the Suspension bridge crosses its outlet. At about 600 feet within it the section shown in fio'ure 14 is situated. Fig. 14. Section of Niagara gorge, about 600 feet within its end. R. S., river surface ; L. O., level of Lake Ontario ; N. 1., Niagara limestone ; N.s., Niagara shale; (1 1., Clinton limestone ; R. M., red Medina shale and sand- stone ; g. M., gray Medina sandstone ; M. s., Medina shale. Longitudinal and vertical scales the same. It is 1,350 feet wide. Upon the western side is Roy terrace, under Brock's Monument, at a height of 285 feet, and having a breadth of about 300 feet, with the plateau rising to 340 or 345 feet beyond. The terrace was the river floor at the birth OjL- Kiagara falls, and its height and features will be described in Chapter xv. The river at the bridge is over 600 feet wide, and 750 feet just above. It reaches to a depth of 150 feet at a point a quarter of a mile within the canon, while the surface of the river cannot be more than three or four feet above Lake Ontario. At Brock's J\Ionument the ground is 329 feet above the lake. The talus slopes are 300 to 400 feet or more in width, extending from the base of the Niagara limestone (which is 7i " 100 FALLS OF NIAGARA tGeol. Surv. here only twelve feet thick, increasing to twenty or thirty feet immediately beyond the edge of the chasm), and sloping to the water's edge. At this point there are little ravines in the upper beds on each side, but these do not pass beyond the stage of gullies. At about 7,000 feet from its mouth the cable of the Ontario Power Company crosses the river. Here the canon is 1,145 feet wide, while the river has a breadth of only 480 feet. The ground at the edge of the gorge is 318 feet above Lake Ontario, while the depth of the river is sixty-three feet, or fifty-three feet below lake level. This first reach of the ISTiagara caiion extends above iN^iagara University in a nearly direct line for a distance of 9,000 feet from the mouth of the gorge (I — I on map), having a southward course. Here it bends sharply w^est- ward. SMEATON RAVINE. The occurrence of this curious feature for a long time seemed inexplicable, but it eventually proved to be a record in filling an intermediate gap in the earlier history of the Falls of Niagara. It is situated 4,000 feet above the mouth of the gorge and is a true caiion 500 feet long and 150 feet wide, bounded by the perpendicular walls of JSTiagara limestones, over the shales, which in the deeper part of the ravine are covered by talus, sloping downward to the band of Clinton limestone, at the depth of about 90 feet below the surface of the country. This band of limestone, however, has itself been incised for a distance of fifty feet within the brow of the gorge, but being at lower level it is actually trenched for a length of 200 feet. From the Clinton band is now a little fall of thirty-five feet with rapids below to fifty feet, or to a level of 175 feet above the lake. At one time Smeaton ravine must have been supplied by a stream of nearly fifty feet in width. To-day there is only an of Canada] SMEATON RAVINE 101 intermittent supply of water such as occurs after heavy rain- falls. Even this has not been able to make a shallow channel on the rock surface exposed above. It is shown on the map how the western banic of the original river sweeps around a lake-like expansion and just incloses Smeaton ravine. Also that above this is an insular bank of river deposits, or a bar, in mid-lake, which was sketched on the map after levelling its height, so that it may extend a little farther than shown. When l^iagara falls had receded to Smeaton ravine the stream was flowing behind the island shown on the map. This produced a cross-fall. As ^Niagara falls receded the expanded river channel became drained, after they had passed some distance above the island, owing to the supply of ^\'ater for the cross-fall being cut off. This did not probably occur until Niagara falls had reached the angle of the caiion^ some 4,000 feet above Smeaton ravine. At any rate the falls must have cut back a long distance before the cross-fall ceased to flow. The modern American falls have cut back not more than 200 feet in 600 years, since the two cataracts parted company, with seven per cent of the whole river discharge descending 167 feet, or lately somewhat more. The Smeaton fall had only a very small proportion of fifteen per cent of the present volume. There is another side to its history. The Clinton lime- stone in the ravine is only trenched at its lower end. This shows that the level of the river was at about that of the lime- stones, else Smeaton fall would have had a greater descent, and also that the second cataract of the great river had just reached this point when the Smeaton fall ceased to exist, leaving only the deep lateral caiion. In this ravine is preserved the evidence of the height of the upper cataract during a mid-portion of the Erie stage of Niagara falls, while the terraces at the mouth of the gorge and at Foster flats record its height in the earliest and latest parts of it. 102 FALLS OF NIAGARA t^^^l- ^"i"^'" THE FOSTER REACH DEVILS HOLE, From the point beyond the bend of the river mentioned, Reach ininiber 2 (II — II on map), extends to the outlet of the Whirlpool a farther distance of 8^800 feet, or 9,400 feet to the centre of the Whirlpool gorge. The ^^^lirlpool expands so that in an oblique direction the line mentioned would extend another thousand feet in length to the opposite edge of the gorge. This addition, however, cannot be made to the medial length of the canon. This second Reach is perhaps the most important of any stretch along the course of the I^iagara river, and at the same time is the most complex. The original ISTiagara river, outside of the gorge, has been surveyed by myself throughout this region. It widened out into a small lake along the upper part of Reach number 1 (see large map), as has been described when treating of the Smeaton ravine. So also the gorge in the second Reach broadens to even 1,750 feet and incloses Foster flats, but this is independent of the outer surface channel. Some of the changes in the physics of the river were described in a paper published in 1894.* Devils Hole, which is a cave some five by seven feet at its outlet, in base of Niagara limestone, opens into a strongly carved lateral canon, but smaller than Smeaton ravine. Foster flats is 3,600 feet long, the lower end beginning at 2,300 feet above the end of Reach number 1. At its upper end is a remarkable little bay called Cripson or Fisherman Eddy. The river is reduced to a breadth of 280 feet at the narrowest part of Foster rapids, although it has a width of 900 feet immediately above. At Foster flats there is a remnant of a terrace proper which once formed the floor of the river. Here is also Wilson terrace, where there was a cataract from the Clinton limestone. Over- hanging it is Wintergreen flat, a remnant of the old floor * ' Duration of Niagara FaUs,' Am. Jour. Sci., Vol. XLVIIL, p. 464, 1894. of Canada] XIAGARA GOKGE lO:; of the ]Sriag';ira rivei' frmn wliicli the main falls formerly cas- caded. Figure 15 is a section at this point. S'^l° '>- »° 30,° r. Fig. 15. Section across gorge at Foster flats. (Legend as before). W. T., Wintergreen terrace ; W. P., Wilson point, a projection of the terrace of Cliton limestone, from which a ridge surmounted by huge blocks e.xtends to the river. In the gorg-e below Foster flats the soundings show a maxi- mum depth of sixty-three feet, or fifty-three feet beneath the level of Lake Ontario. Accordingly the channel here is six feet higher than above the Foster flats. The deepest point be- low may not have been found, but although so nearly that of the section below the outlet of the Whirlpool, the history of the channels at the two points has been entirely different. ISTor is there here a deeper inner channel. It is quite possible that the real bed of the channel in front of Foster flats has a less depth than the river above, but it is re-occupied with great blocks of fallen rocks, such as those on the bank adjacent to Fig. 16. Section of the gorge a quarter of mile below mouth of Whirlpool. these narrows, at a point a short distance from their head, where the piles of detached blocks extend to the river edge, indicating sufficient material for any amount of obstruction, and now sufficient for holdino; the water twentv to twentv-one 104 FALLS OF JSriAGARA [Geol. Surv. feet higher than below the flats (see figures 15 and 28). Owing to the velocity of the current among these boulders the depth of the rapids is immaterial, as it has no bearing on the excavation of this part of the caiion. Above Foster flats the elliptical form of the gorge contracts to one of uniform breadth of about 1,250 feet. The section of the caiion, between here and the Whirl- pool, is given in figure 16. WHIRLPOOL-RAPIDS REACH. The old river banks are shown at the mouth of the Whirl- pool. These, however, come under the study of those of the original river channel. At the mouth of the Wliirlpool, the gorge is 900 feet in width, and here the river is only 460 feet wide. The Whirlpool gorge is widened out to the maxi- mum breadth of 1,750 feet, while that of the cauldron at the water's edge is 1,150 feet across. The basin is an extension of a thousand feet to what would be the natural river course, now forming an eddy or tributary to the river proper, appearing as if it had been the old channel of the river itself. The head of the Whirlpool is bounded by banks of clay, sand and gravel to a height of nearly 300 feet above its surface. (See Plate xi. a on pag'e 67.) A cross section is shown in nVs 3CI "m"^ It } ^^'^IZ t M.£ ~~ — N ^^^ 0. / M, s- ?- _l!)o_ V, f«i^ . . -^^rr^- -^^ "^ Fiof. 17. Section across* gorge at Whirlpool, between Thompson point and Colt I'avine. Broken line shows river bottom outside this line. figure 17, where the maximum depth of the Whirlpool is 102 feet, but in the river just outside 126 feet were found without the greatest depth of the river being measured (p. 64). This shows a depth of seventy-nine feet below lake level, or twenty feet more than that of the river between the Whirlpool outlet and °' Canada] NIAGARA GORGE 105 bead of Foster flats ; and only eig-ht to fifteen feet less than the greatest depth of the npper channel, which might also be expected here if the lead could be sunk 200 or 300 feet beyond the last point sounded. Immediately above the Whirlpool (at Sinclair point), the gorge is reduced to a breadth of 1,000 feet, but just beyond there is a short expansion to 1,200 feet. Then the gorge rapidly contracts to form the N^arrows of the Whirlpool rapids. Here the chasm is reduced to Y50 feet or less, while the channel itself is contracted to 350 feet in width. ENt ^ V rr-T-' ^^ ^ (yj ) , ~ 1 t '■— E= ^ /^~-~ 1—- 1 — «— . ■ — • t I . 1 V y_ -. _^ „. TCt^^K'. ^" Vr/JC r r= ! -H 1 1 v^ ^ - -,r^ L Vq' M 5 Sea,,, " !°° ir ",° F U 11 cz fillt Fig. 18. Sf-ction across Whirlpool rajjids at narrowest point, a third of a mile below Grand Trunk bridge. (Legend as before), t t, Original river limits ; a, lower pre-glacial rock terrace. All these features are in Reach iN'o. 3, which extends from the end of the Whirlpool in a sweeping curve to a point 500 feet above Cantilever bridge, or for a total length of 6,200 feet from the middle of the river proper at the Whirlpool (iii.-iv.). In passing upward from Reach 2 to Reach 3 the course of the gorge bends no less than 110° toward the southeast. Here is a section of the river which has attracted much attention. More than sixty years ago Sir Charles Lyell thought that he* recognized in the A^T^iirlpool extension of the gorge the course of an ancient I^iagara river. That it was the course of an ancient stream is certain, but it was not that of a Niagara river draining the Erie basin in pre-glacial times. f At the Cantilever bridge the river shows a maximum depth * ' Travels in North America ' (1841-42), Sir Charles Lyell. Also Proc. Geol. Soc, Lon., Vol. IIL, 1841, lb.. Vol. IV., 1843. t ' Discovery of Preglacial Outlet of Lake Erie,' Proc Amer. Phil. Soc, Phila., Vol. XIX., 1881. 106 FAT.I.S OF NIAGARA t°eol. Surv. of eiglity-six feet and a buried cliannel extends to 185 feet below its surface. (See figure 21, Chapiter sii.^,) Above the surface of the water the wall of the gorge rises 208 or 210 feet. Along this reach, while the gorge is reduced to such a narrow pro- portion, the old banks are strongly marked and show that the river had a breadth of 1,500 feet before the waters were concentrated within the narrow chasm. This feature proves that the constricted gorge was not entirely due to the shrink- age of the river at this point. • These matters will be discused when considering the channel of the Whirlpool rapids. THE FALLS REACH. Above the railway bridges is another bend in the course of the gorge where it turns 40° westward, and extends thence in a direct course to the crest of the Canadian falls. It sud- denly widens out from the end of Reach 3 (here 840 feet broad) to an average breadth of 1,300 feet. This width is, however, varied slightly, l)eing reduced to 1,200 feet at Hub- bard point, and increased somewhat opposite the American falls, as would be expected. Above the American falls it is again reduced so that the gorge has a breadth of 1,200 feet between Goat island and the western walls below the line of the Canadian falls. This is Reach Xo. 4 (above iv. on map), and has a length of 12,000 feet, but it bends towards the apex so that the length may be slightly increased. Thus the length of the caiion is about 36,600 feet from the Queenston Suspen- sion bridge or about 36,200 feet from the brow of the escarp- ment. After passing the end of Reach 3 the sudden widening of the gorge indicates a marked change in the history of the falls. Besides the uniform breadth the greatest depth to near their site is constant. This suggests that there was no material variation in the effective height of the falls in this portion of * See Chapter XII. on Whirlpool rapids, showing section of boring. of Canada] NIAGAKA GOEGE 107 the cafion. At Swift Drift point, 3,700 feet, and IInl)l)ar(I point, 4,500 feet above Cantilever bridge, the gorge is sliglitlj narrowed to less than 1,100 feet with a basin 200 feet wider between these points. Here the rocks are distorted, bnt frag- ments of the old river banks show that the river maintained the fnll breadth of 1,300 feet, although the gorge itself is somewhat narrower. -E=--__ Hubbard's Pt (^ t— -— •- ^^" — xT— \«- *— • I — -' ^ / 1 ,,,».^. .^L ^. *>^.,..^__^ J^'— " t?--. g . M '--'- = - -c---- - --- -- - --_-_- -^^^"^^.r^'^ (= =-^ _ ^ _ _ ^ ^- i ^=._^-- --^ -- sc»i. 5 !S2 i?a — aspf.ei FisT- 19. Section across Gorge at Hubbard point. (Legend as before.) r., Remnant of terrace once a pre-glacial floor of the trough through Lyell ridge. It wonld seem that there has been an average uniform reces- sion of the falls throughout Reach No. 4 until approaching their present site. The surface of the river is about 100 feet above Lake Ontario, and the depth from 186 to 102 feet. The river itself has a breadth varying from 750 to 1,000 feet; in front of Hubbard point it is nearly 800 feet, though at Swift Drift point it is scarcely more than 500 feet wide. These are im- portant features in considering the origin of the Wliirlpool rapids gorge, and also the Falls-Chippawa trough. At this point, also, it is found that the river crosses the highest ridge of limestone occurring anywhere along its course. The width of the canon in front of the American falls reaches its maximum of 1,600 feet, which is due to frost action, and the broadening effects of the American channel. (See figure 6, page 58.) Above the Upper Arch bridge, near Carter cove, (see Plate xxxviii. b) is the end of the drowned terrace which extends thence to the falls (see figures 3, 4 and 6), though incised by a very deep channel. 108 FALLS OF NIAGAL'A [Geol. Surv. The last cross-section of the gorge (figure 20), is one from Goat island, over the Goat island shelf, and extends to near Table Rock. This shows the terraced river bottom with the deep channel, near the eastern side of the gorge. {8ee also longitudinal section, figure 3, page 50.) Goat Uland ^g M -^"f. Fig. 20. Section across gorge from Table Rock House to Goat Isl.and shelf, showing submerged terrace and deep channel ; also, rock floor of Falls- Chippawa basin at Table Rock House. This deep channel opens out into the wilder one beloAV the American falls and Carter cove, after passing the upper sub- merged terrace. The height of the rock wall here is about 160 feet above the river, while it is nearly 220 feet at Hubbard point, and the rock surface on Lyell ridge a short distance back from the chasm rises to 270 feet. CHAPTER IX. ORIGINAL BANKS AND BED OF NIAGARA RIVER. Outlet of Lake Erie and the Upper From Smeaton ravine to end of reach of Niagara river. caiion. Cafioin reach above ajid below the From mouth of gorge to Lake Whirlpool. Ontario. OUTLET OF LAKE ERIE AND UPPER REACH OF RIVER. The character of the river above the falls has changed very- little since the separation of the waters of Lake Ontario from those of Erie, when Niagara became an established river. Although this is the case the river is not such as should be ex- pected from the features of the imderlying rock surfaces. It is found that its depth reaches fifty-three feet, at a point two miles below the outlet of Lake Erie. This depth is reduced to from seventeen to twenty-four feet through a ridge of Cornifer- ous limestone, which rises sixty feet on both sides of the river, half a mile away from its present banks, while some miles west- ward the rocky rim on the northern side of Lake Erie is very much lower. The Corniferous ridges had been dissected long before the birth of the river, as is further shown beneath the International bridge where one of the piers rests upon a glaci- ated surface at the depth of forty-five feet. The river at this point is 1,850 feet across, but at a recent stage, when five to twelve feet higher, it was nearly double that breadth, as shown by the now raised flats at the town of Fort Erie. This terrace belongs to the history of the river when the stream was first coursing across the rolling country with poorly defined banks, flooding the estuaries and leaving a suspicion of terraces at many points at about 340 feet above Lake Ontario. 109 110 FALLS OF NL\GAKA ^^^^o^- ^urv. From Fort Erie (12,000 feet from the lake), the distance to the rim above the Upper rapids along the western channel is seventeen miles, while along the eastern channel it is somewhat farther. Its breadth varies from 2,000 to 3,000 feet, and while its depth is rendered irregular bj bars, the soundings show it to be thirty-three feet or more at points throughout the south- ern half of its course, although it is reduced to some twenty feet among the bars which cross the river at the mouth of the Chippawa creek. The country is a low plain, much of it rising no more than from ten to twenty feet above the lake. Indeed, if the surface of the river did not descend to nearly fourteen feet below lake level, at the head of the Upper rapids, it would turn the creeks at Chippawa and Tonawanda, with their tributaries and other low places, into extensive estuaries. This low country is slightly modified at a point from two to three miles south of the Upper rapids where the land rises nearly fifty feet above it, or forty feet above the lake. Simi- larly elevated ground appears on Grand island. There must have been a slight depression in this ridge, like that in the Corniferous belt near Lake Erie, else the waters would have been diverted to a more western course. Below Grand and N^avy islands the two arms of the river unite, and although it somewhat widens it is again reduced just below Chippawa. A greater breadth formerly obtained when the waters passed over the projecting flat point now forming the site of N'iagara Falls, 'New York. The rocky barrier obstructing the upper Niagara basin is shown by the line of Greens or First cascade of the Upper rapids, (see Plate xu. b, on page 77). The now uncovered rock floor composed of Magara limestone, as seen at the upper end of Goat island, is 313 feet above Lake Ontario, while at the lower end of the island the underlying rock surface is only 2G5 feet. The floor of the river at the Ontario Company's in- take, in front of Duiferin islands, is at 306 feet. Where the of Canada] ORIGINAL BAXKS 111 river sends a branch round Dufferin islands the rock passes nnder the drift, which at the southernmost bridge is 299 feet, while at the elbows behind the island, the slope descends to 286 feet with the buried valley still further declining west- ward nnder the drift, as shown by borings beyond the river nnder the floor of the Park adjacent. Throughout this region there are heavy hills of drift which rise above the former river banks behind Victoria Park to a height of from 375 feet to 400 feet above Lake Ontario, while about a mile distant in Luiidys Lane a point of 465 feet is attained. To the east of the river behind the point of land occupied by the city of iSTiagara Falls the altitude reaches 360 feet or more while on the lower plains there are some pronounced hilly ele- vations. These form a third ridge across the course of the original river, in which there was an ancient depression that permitted the first stream of Xiagara water to flow from Lake Erie to the lower basin. It must have been less than thirty-six feet above Lake Erie else the waters of the upper lake would have coursed through a channel in the vicinity of Welland canal. The difference of altitude at the two localities was perhaps even less than five feet, but this slight amount caused the Xiagara river to have its present location, in place of the outlet of Erie being situated near the Welland canal. CANON' REACH ABOVE AND BELOW THE WHIRLPOOL. The feature of the Upper reach of the river, with a breadth of more than a mile, trending almost westward, abruptly terminating at Goat island and passing over the side of a narrow transverse valley at right angles to it shows a remark- al)le change in the physical conditions. Ilere at the end of the upper broad valley the American falls plunge over the eastern side of the gorge, on one side of the precipitous walls in front of Goat island, on the other the Canadian falls, in front of the widest angle of Victoria Park. 112 FALLS OF NIAGARA f^^o^- ^urv. This view is represented in the panorama on Plate xiv. (on page 81). In it is also seen the eastern bank of the Falls- Chippawa valley, which has been uncovered by the rapids cross- ing it. The smoother water shows the deepest part of the pre- giacial channel. The apex of the falls is now cutting back and across the old bank, and this tends to reduce the recession of J^iagara falls as mentioned on page 41. So long ago as 1841 Sir Charles Lyell attributed the rapids to ancient topography buried beneath the drift. The modern Niagara river, after establishing its course along the Upper reach, swung round over the buried valley, whence its general course was determined by the low surface depressions, irrespective of the materials which formed its bed. So long as its lower extension was blocked by a rocky barrier the buried valley produced no effect upon the features of the river. The floor of its margin had a level now represented by an elevation of 316 above the lake, as may be seen adjacent to the intake of the Niagara Power Company at Niagara Falls, but the surface of Lake Erie is now fourteen feet higher than Greens or First Cascade. This old terrace floor passes round an island in the southern part of the town, where there is a sharply cut terrace between it and the river, opposite Goat island. Through the town this old shore line has been obliterated by artificial grading, but it reappears beyond, and forms a distinctive feature of a terrace, where the Gorge Railway car shed is located, as shown on map, diagonally across from Hubbard point. (See also Plate xvi. B.) The banks on this eastern side were not generally high. On Goat island a deposit of river gravel with fluviatile shells occurs near its southeastern quarter having now the same level of 316 feet. These deposits were described by Prof. James Hall and Sir Charles Lyell more than sixty years ago. Prof. Hall had also measured the terraces at this locality and compared them with others at Whirlpool point. The terraces at Plate XV. Profile view of American Falls and Goat Island shelf, with terrace on Canadian side. (Winter scene). 113 of Canada] ORIGIXAL BANKS 115 lower levels did not belong to the original river banks, but marked the lowering of the waters at more recent date. Adjacent to the falls on the Canadian shore were, at that time, high banks of the old river rising to 380 feet above Lake Ontario, back of which the country rose in Lundys Lane, a mile distant, to a point 465 feet above same datum. The terraces on the ISTew York side have been mentioned first, for the reason that here is an extensive floor of the original river bed, wdiile on the Canadian side, in the subsequent lowering of the waters and the removal of the material from the buried valley, the river has undermined the old banks, and all higher terraces have been washed down to the present river level in the vicinity of Niagara falls, to a depth of sixty feet below Hubbard point. Thus it is that the high blufl^s bounding Victoria Park have a heio-ht of from 100 to 120 feet. These bluffs are well shown in Plate 5v. (on preceding page), which is a winter scene. rrom the northern part of the Park reservation the differ- ent terraces begin to be recognizable, and continue until they reach the vicinity of Hubbard point, where they are abruptly cut ofl^ by the encroachment of the western wall of the gorge. Here at Hubbard point is a remnant of the old floor projecting outside of the bank of the gorge for a breadth of 150 feet, and a length somewhat greater. This is shown on Plate xvi. a. Behind this fragment of the old floor, whose inner margin is at an elevation of 318 feet, there is a steep bank of a few feet. Back of this the hills of limestone ri^e in a short dis- tance to a height of 370 feet at the circle in Wesley Park, where they are covered by only two or three feet of soil. On the opposite side the land has an equal height back of the brewery, Plate xvi. b, and at the site of the new post office it is covered by only a few feet of earth. Here the river crosses the highest limestone ridge in the whole Niagara district; so high that the river could never have flowed by this course had the ridge not been much lowered by the pre- 8i 116 FALLS OF NL-iGARA 1^^^"' ^'''''''■ glacial depression, where at Hubbard point the old glaciated surface was found. (See Plate xvi. a, also map.) It was this barrier at Hubbard point which determined the height and the level of the river until after the falls had receded past it, when the loose earth was rapidly removed from the Falls-Chippawa basin to a depth of sixty feet. The lower terraces are features of the subsequent history of the river. The gorge at Hubbard point is reduced to a breadth of 1,150 feet and to 1,050 feet at Swift Drift point to the north. There is no evidence that the lower pre-glacial valley to the south cut to a lower level at the narrows of Hubbard point. Moreover, from Hubbard point southward the boundaries of the old river rapidly diverge, showing that the pre-glacial valley had a south- ward and not northward direction. {See Plate sxii.. Chapter XIII.) Below Hubbard point, for a half mile or more^ the terraces on both sides are cut off, as the modern caiion is wider ^han the original valley. (See large map.) Just south of Cantilever bridge the gorge rapidly contracts and at the same time the old river banks grow wider apart. This feature is showm on map and in figure 18 (page 105), which is a cross-section about 1,700 feet north of the Grand Trunin Railway bridge, where also the lower terraces appear. The great terrace at this locality is strongly marked as shown on the large map and on Plate xx. The bank has been deeply carved out by the currents^ thus leaving a steep bljiff. The inner edge of the floor has a height of 316 feet, with bluffs behind over twenty feet higher. Its distance from the caiion edge is 530 to 560 feet, but this also includes the lower terrace from Y5 to 125 feet wide, the rock floor of which is only 280 feet above Lake Ontario. The upper terrace, upon reaching the vicinity of the cove between the Whirlpool rapids and the Whirlpool, has been cut off by the re- ceding walls of the gorge. On the eastern side of the river this terrace is equally well Plate XVI. a. View of Hubbard Point (terrace in front of house) at col between pre-glacial valleys. Plate XVI. • BJr'ffBP 1^:^^ iO^ ite ^i^Q ^^g^m^^ ii!^SE wbL 1 P^i liH W^ ■"^P i*^'^ i j View opposite Hubbard Point (with correspondinof terrace in front of point. 117 of Canada] OKIGINAL BAXKS 119 defined by the very sharp banks below the old Mount Eagle hotel {see Plate xxi. b), which have a height of more than twenty feet above the old floor, with an altitude of 316 feet at its inner edge. Here also are two lower terraces nearer the river, belonging to a later epoch, but the main ter- race mentioned extends from the gorge for a breadth of 510 feet This, added to the width of the gorge, about 750 feet, and that of the terrace floor on the western side, shows that the original river reached, at this point, 1,780 feet in width; but as this section is somewhat oblique the general breadth is about 1,500 feet. While the western bank has been abruptly cut off at the Whirlpool, the eastern one is still intact, but swings round so that the remains of the old floor have a breadth outside of the present gorge of only 150 feet at A^liirlpool point. It continues for a distance of a few hundred yards beyond the outlet of the Whirlpool, where it also is truncated at the edge of the gorge. At the point mentioned above the Whirlpool, where the ter- race is abruptly terminated, the old river had a breadth re- duced to 1,250 feet. Its bed was composed of drift which filled the buried Whirlpool gorge, on the other side of which the terrace reappears, showing a river breadth of 1,500 feet. At a quarter of a mile beyond Thompson point the banks show the breadth of the old channel to have been 1,400 feet and just beyond the terraces are suddenly cut off. At Thomp- son point, as well as at Whirlpool point opposite, the lower terraces coalesce, and are not so distinctly separable as above this locality. Indeed, from below the Whirlpool outlet to the mouth of the gorge no lower terraces remain like those above. The absence is striking. Below the Whirlpool, where the terraces on the Canadian side end, there is a rock ridge which was crossed by the Niagara river when it commenced flowing. It surface eleva- tion is about 330 feet. However, a few hundred yards 120 FALLS OF NIAGARA ^^''°^- ^"rv. beyond, a renmant of the old floor is again seen at Win- tergreen flat, where the inner edge has an elevation of 316 feet, bounded by a bank fourteen feet higher. This flat, which was the bottom of the old river, has, at one place, a breadth of 500 feet, with its floor declining from 312 to 306 feet. This shows that the depth of the river was considerable. Opposite here, on the ISTew York side, the encroachments of the gorge have re- moved the river bank. A short distance below Wintergreen flat the floor of the old river re-appears in a lake-like expansion, with an islet covered with river deposits of rounded and flat stones in loamy soil. Here its edge is not always sharply defined, as the washings of the hillside and the cultivation of the soil have obscured the original water line. These deposits up to an elevation of 324 feet approximately represent the old water surface. Their elevation at the edge of the caiion is 318 feet above Lake Ontario. In some places the river deposit has a considerable depth. It is ten feet deep where the Ontario Power Company's cables cross the river. I^ortli of Foster flats the river expanded into a little lake with a breadth of about 3,800 feet. The more sluggish character of the currents, and the smaller depth of water here, abundantly accounts for the less sharply defined western bank, while on the eastern side, which is now the out- side of the bend of the river, from Devils Hole to beyond the cable crossing of the Ontario Power Company, the bank is more sharply defined back of tlio terrace floor which is from 100 to 250 feet wide. FROM SMEATON RAVINE TO THE END OF CANON. Just beyond Smeaton ravine the old river banks contract to the brink of the gorge on both sides, where they are again trun- cated. Here the river cuts through a rock ridge with an altitude of about 330 feet, covered with a clay floor rising Plate XVII. a. View of outlet of gorge cutting tlie Niagara escarpment (about 330 feet high with Brock's Monument, to the right). Plate XVII. b. View of Iroquois terrace, at end of gorge, east side, here cut out of Medina sandstone. 121 of Canada] ORIGINAL BANKS 123 beyond to 340 or 350 feet. This was the barrier which made the expansion or lagoon north of Foster flats. It was crossed by the river for a distance of 3,000 feet. Beyond this last ridge the distance to the end of the gorge' is 200 yards with the delta deposit at the end. {8ee Plate xvii. A.) Here the old river floor was about 322 feet higher than the present surface of Lake Ontario. FROM THE MOUTH OF GORGE TO LAKE ONTARIO. At the birth of I^iagara falls there was no river below the escarpment, for the water cascaded directly into Lake Ontario. As the waters receded the river cut across the Medina shale for two miles. Beyond that point it flowed over drift clays and other material, exposing in places the Medina shale. At its lowest level the waters in the Ontario basin receded many miles from the present shore, when the deep channel at Queenston was excavated. {8ee Plate xxix., Chapter xv.) From this account of the river banks it will be seen that the early [Niagara river can be traced from Lake Erie to what was then the margin of the lower lake, now Lake Ontario, at sub- stantially a level which indicated but little slope of its surface ; much less than that of the present upper reach. At first it may have been little more than a strait with ill-defined margins. CHAPTER X. WHIRLPOOL-ST. DAVID BURIED VALLEY. Speculation as to its origin. Borings in the channel ; origin of Surface features from Bowman modern Whirlpool. creek to edge of Escarpment. Character of drift in deep channel. Characteristics about Whirlpool Fossil wood buried in drift. gorge„ its breadth. Breathing well. SPECULATION^ AS TO ITS ORIGIN. In treating of the features of the Niagara canon the ex- tension of it now occupied by the Whirlpool has been partly described (page 104). The lateral walls are precipitous, the same as the other portions of the gorge, but at its head only drift banks are seen. {See Plate xi. a, on page 67.) The heavy band of Medina gray sandstone, {see Plate xxvii. b), having a thickness of about twenty feet, reaches to ten or fifteen feet above the surface of the pool. At the head of the Whirlpool the gray band is absent for a breadth of 1,000 feet, while at a point between the Whirlpool and the Whirlpool eddy, it is re- moved for less than 600 feet. As late as 1841 Prof. James Hall* had regarded the cove of the Whirlpool as an eddy in the course of the river, but in that year Sir Charles Lyell first recognized it as a part of a buried valley, and connected it with the embayment in the escarpment of Saint David, f supposing it to have been the course of a pre-glacial Niagara river. In 1881 I first showed that the Niagara river was not the outlet of the pre-glacial Erie basin. I then supposed that the channel was interglacial. * • Natural History of New York,' Part IV., Vol. IV., 1842. t ' Travels in North America,' 1841. 125 126 FALLS OF NIAGARA [Geol. Surv. This idea was repeated as late as 1901 by Dr. G. K. Gilbert.* Until the present investigation of the refilled Whirlpool rapids, and the discovery of the Falls-Chippawa valley, its true character could not have been known. The idea of any portion of the channel being interglacial I withdrew in 1887, and even at that time regarded the AVhirpool rapids section as not representing the ancient course of any considerable stream. At that time I said, ' we are led to the conclusions that the course of the N"iagara river above the AVliirlpool and below is mostly of modern origin throughout, and not to any extent a drift-filled gorge re-excavated since the ice age.' ' The Saint David valley represents only the water course or water courses of local drainage before the ice age. '** For some years little attention was given to this channel. In 1886, f Prof, E. W. -Claypole pointed out the occurrence of rock high up in Bowman ravine at the head of the Whirlpool cauldron. Again, in 1894, I measured the height to which these beds reached, using the United States Lake Survey topographic chart for the position of the stream. This map was very incorrect, showing the gully as extending from the middle of the Whirl- pool, so that one had to infer that the buried channel was only about half the depth now found. Doubt, however, existed, and at a meeting of the American Association for the Advancement of Science, Prof. H. S. Williams suggested boring to settle the question, and offered the first subscription. This was not acted upon until 1905, when the work was carried out by the Geologi- cal Survey of Canada. The accompanying corrected map is from that of Mr. E. Gardner, Provincial land surveyor, ex- tended by myself. In 1811 Lyell distinguished between the precipitous cliffs of the modern gorge and the more gradual slopes of the buried channel. These last are capped with limestone having polished * Atlas sheet of Niagara. U. S. Geol. Survey. ** Amer. Naturalist, Vol. XXI., p. 270, 1887. t Report upon meeting of Amer. Ass. Ad. Sci., Sept., 18S6. of Canada] WIIIKLPOOL-ST. DAVID VALLEY 127 rounded edges, but their slope never exceeds 45°^ and rarely more than dO°, while that of the lower shale is much less. SURFACE FEATURES FROM BOWMAN CREEK TO EDGE OF ESCARPME2«rT. The loose materials at the end of the Whirlpool have been carried awaj by Bowman creek, which has made a deep ravine for three-quarters of a mile, in a direct line, in the very slightly "undulating plain of from 340 to 350 feet elevation above the lake. If it were not for the deep gully of Bowman creek and its branches, there would be no trace of the buried valley upon the surface of the country. Beyond the head of this creek no further evidence of the valley appears above the ground which rises slightly higher. However, at about two miles from the headwater of the Whirlpool, the country becomes broken by several deep ravines among hills of drift covering the face of the Niagara escarpment, where it bends behind the village of St. David, but there is no St. David valley proper. The country to the northeast of the buried valley is slightly rolling, as far as the brow of the escarpment, and covered with a clay surface. It does not rise more than 350 feet above the lake. But at its mouth and westward are deep deposits of sand and gravel rising in Berryman hill (Plate xxx.) to the unusual altitude of 442 feet. Prof. James Hall, in 1842, noted the occurrence of these materials to a depth of 150 feet in a well. Beneath the surface deposits of this region, outside of the buried valley, the limestone floor usually rises to about 330 feet above Lake Ontario. To the northeastward of the buried valley^ the country is only slightly rolling to the brow of the escarpment, and it is covered with a clay surface. Nearer the mouth of the buried valley there are deep deposits of sand and gravel. Prom the mouth of the Niagara gorge the rock face of the escarpment extends westward for about two miles and a 128 FALLS OF NIAGARA [Geol. Surv. quarter (see large map), wliere it suddenly turns southward for three-quarters of a mile. Here it is covered by gravel hills mentioned, or continues in and forms the eastern wall of the buried gorge. As an escarpment the rocks are not shown farther than where the Michigan Central railway crosses the Town Line road. West of this point for a mile and a half the face of the l^iagara escarpment is almost entirely obscured by rolling drift hills, although an upper terrace, with a rocky foundation, begins to appear. However, at nearly two miles from the point mentioned, a spur of the escarpment emerges from the sand. hills. The upper portion is really an insular mass in front of a rocky terrace exposed when the waves of lake were making the Roy terrace 287 feet higher than now, at the date of the birth of N^iagara falls. This j)romontory led Sir Charles Lyell to think there was an embayment or expansion of the buried Whirlpool-St. David valley, and to compare it with the outlet of the gorge (see Plate XVII. A, on page 121) ; suggesting that the former had a great breadth, while the latter is reduced to a quarter of a mile. But this turn in the escarpment is not the enlarged valley of the buried channel, although many have followed him in this view. jSTor was this strange, for over a consider- able portion of this section it was only by making local inquiries and careful search that rock exposures were found. Thus an obscure quarry on Mr. Thomas Berryman's farm was found hidden in a field below the railway, and away from any public road. It has an altitude of 320 feet, and consequently is on the brow of the buried escarpment. The occurrence of rock everywhere to the west was subse- quently established, thus reducing the possible width of any buried valley. By digging along- the course of the stream at one or two places to the east rock was found. At a point a few hundred feet west of the railway crossing of the St. David road is the head of the creek. It emerges as an enormous of Canada] WHIRLPOOL GOKGE 129 spring from beneath the railway, where it tiows out of a now covered rocky cavern, as I am informed. Accordingly this feature, establishes the occurrence of rock at a lieight of 320 feet above Lake Ontario and, further, restricts the possible size of any buried valley. Descending the hillside a few hundred yards to a point behind the old brewery building, and now almost entirely coverecl by earth, is an old quarry pit showing the surface of thin slabs of limestone at 254 feet above the lake. This point is at the outlet of the buried "Whirlpool channel, not at the top of ,the escarpment, but farther down its slope. The outlet of the valley, between the approximately determined eastern edge and the ledge at the brewery, at this level, does not exceed 1,000 feet, if so much, while at the 320 foot level it is approximately 1,700 feet. Thus, at last, the confines of the old valley are delimited at its mouth. The amount of interest which has been taken in this valley has justified this detailed study, as also that of the buried gorge, for it was necessary to determine what part the ancient trench had in the recession of the falls. Otherwise it would be only one of the many features of pre-glacial topography, which is generally characterized by rounded outlines. Indeed, until making this investigation, I should have considered the ancient valley much broader without knowing its depth, as did Lyell. OHAKACTERISTICS ABOUT WHIRLPOOL GORGE ITS P.READTII. From Whirlpool or DeVeaux point, on the ISTew York side, directly across the river to Sinclair point, separating the W'hirl- pool rapids eddy from the AVliirlpool, the ^Xiagara gorge has a breadth of only 1,000 feet. At the broadest part of the Whirl- pool the gorge is 1,750 feet wide. In a little ravine on the northeastern side, Avhere the wall of the gorge passes under the drift at the head of the pool, one may closely determine its position on the eastern side. The distance from here to the. 130 FALLS OF I^IAGAKA ^^^"^^ ^^rv. exposed western pre-glacial wall with its rounded edges is 1,400 feet ; consequently the difference in these measurements represents widening of the gorge due to the action of l^iagara river. (See map, Plate xviii. opposite.) The western wall of the Whirlpool has not receded to any considerable extent, for immediately beyond the end of the Whirlpool fragments of the old rounded limestone edges ap- pear, where their upper surface has an altitude of 290 feet; but this surface, being forty feet below the ridge penetrated by the river beyond the outlet of the Whirlpool, indicates that there was a much broader upper valley than the buried gorge itself. The absence of limestone above 290 to 300 feet for a breadth of possibly nine hundred yards shows here a rapid broaden- ing of the pre-glacial superficial valley trending northward from Lvell ridge. At the quarry of Bo^vman ravine the glaciated edges ap- pear. The same feature is seen a few hundred yards still farther on, across the Electrical Railway embankment ; and again on lot 42, where Bowman creek produces the Harvie waterfall in passing over the side of the old gorge into the deep ravine. Here the rock surface is 301 feet above Lake Ontario, and beneath forty-six feet of drift. From all these exposures the line of the western wall is discovered, and here it trends much nearer the north than was supposed. Midway between this last exposure and the brewery, where well I^o. 4 (see map p. 131) was just sunk, the rock is reached at seventy-five feet below the surface, or 292 feet above Lake Ontario. This shows it to be witliin the valley adjacent to the Wliirlpool channel, but not within the gorge itself. At well jSTo. 3, which my driller sunk during the recent operations, the rock was found to be thirteen feet higher, or at 304 feet above the lake. Accordingly, these wells are at the edge of the trough, and by connecting them with the cor- responding height at the brewery, the western wall of the of Canada] WHIKT.POOT.-ST. DAVID GORGE 131 huried gorge is established, showing a westward turn near well No. 4. Of the location of the eastern wall we are not quite so well informed. The rock wall of the gorge is found near C on map (Plate XX.). The northern end of the outlet is located near the point where the Michigan Central Kailway branch Plate XVIII. Map showing buried Whirlpool-St. David channel. Upper figures give depth of borings, represented by circles ; the lower ones give height of well-bottom above Lake Ontario. 132 FALLS OF NIAGARA ^^^°^- ^urv. to Queenston crosses tlie Town Line road between jSTiagara and Stanford. Just beyond it the line of railway skirts the escarpment on its descent to the lower plain and exposes the rocky wall. These terminal points established, they can be connected with a line cnrving- as on the western side. At the intervening schoolhonse, where well ISTo. 5 is situated (see map), a depth of 120 feet did not reach rock, with the surface at an elevation of 370 feet. This point is in and near the eastern margin. On lot 42 I sunk another well to ninety feet without reaching rock, and this was at a distance of 1^140 feet from the western wall. From all these observations I have been able to limit the breadth of the buried valley, which is 1,400 feet immediately beyond the end of the Whirlpool, to a breadth of 1,600 feet or perhaps 1,800 feet just before breaking through the face of the escarpment. BOEINGS IN THE WHIRLPOOL-ST. DAVID CHANNEL ORIGIN OF MODERN WHIRLPOOL. Although not quite complete, a survey has at last been made of the Whirlpool-St. David channel, after a period of much writing upon the subject, dating back for sixty-five years to the time when Sir Charles Lyell first recognized its char- acter and suggested that the Whirlpool was due to the exist- ence of this buried valley. The origin of the Whirlpool is now established as never before. The modern Whirlpool is the result of the reopening of a fragment of a buried valley after the falls had broken through its side at the present outlet. It is due to the Geological Survey of Canada that this ques- tion has now been so completely established, and I have to thank my assistant, ]Mr. Claude E. Eldridge, for the final supervision of the extremely trying operations, which enabled us to reach tlio depth mentioned later in this chapter. On lot 42 — where the dee]iest well was sunk, the position chosen was 030 feet from the rock exposure of the western wall, of Canadii] OKIOIX OF WHIRLPOOL 133 with tlio iiitcM-iiicdiato deep ]5(»winaii ravine between. This position was selected so as to be somewhat nearer the weste,rn wall of the buried valley than its supposed, middle. The total depth attained was :209 feet^, or seventy-one feet above Lake Ontario. Accordingly the well reached to a point only twenty- four feet above the Whirlpool level. Difficulties here arose so that the borings had to be discontinued. That rock is absent in the buried channel t<> a (Icptli much below the level of the Whirlpool is unquestionable, particularly does it so appear on account of the absence of the, thick band of Medina sandstone above the level of the water at the end of the Whirlpool. The present survey establishes the ex- istence of the pre-glacial canon, with its upper edges rounded in a more pronounced form than was expected. As only soft beds of shale occur below the sandstone it is imma- terial to the investigation of the recession oi Niagara falls whether the ancient gorge reached scarcely to the level of Lake Ontario or below it. CHARACTER OF DRIFT IIT DEEP CHAN^STEL. The following section represents the material penetrated at the deep well : — Feet. Red clay with a few angular pebbles. ... 40 Hounded gravel 2 Brownish sandy loam (dry) 38 Small angular gravel in red clay matrix. -f Loam with fine angular gravel 10 Angular gravel with some binding clay (very tough boring) 26 Bluish clayey sand with some angttlar fragments 60 Fine washed sand, gray colour, with re- mains of white spruce wood -2 Clayey sand 11 Bluish clavev sand 20 134 FALLS OF NIAGARA ■^*^®°'- ®"^■'^■• Fed. Angular gravel with earthy binding, peb- bles size of pease or beans (extremely difficult boring) 19 Loamy sand with small angular pebbles. 27 Extremely fine quicksand (angular) with clay binding, when mixed with water easily flows, but soon sets, so that it cannot be taken up by pumps. . . . 2+ 268-| + In this last bed the casing was bent, and at the same time the quicksand flowed in and fllled the tube for several feet. Throughout the whole depth the dryness of the materials, not holding even the water poured into the boring, rendered opera- tions very difliGult. Water occurred in the lowest quicksand only. The materials found in the well show a strong contrast with the stratified sand and waterworn gravel covering the edge of the Niagara escarpment and filling the old mouth of the gorge where, in sand pits, they have been exposed for fifty feet or more. In Bowman ravine, cut through the deposits in the Whirlpool channel to a de])tli of nearly 300 feet, very few boulders were seen. So, also, at the end of the Whirlpool, the accumulation of boulders, left after the 300 feet of drift have been washed away, is relatively very small. These show that the few larger stones in the drift are almost exclusively granites, quartzites, and other transported crystalline rocks with rarely a boulder of Niagara limestone from the adjacent walls. FOSSIL WOOD BURIED IN DRIFT. The fossil wood occurring at a depth of 186 feet, as men- tioned in table, was kindly determined by Prof. D. P. Pen- hallow of McGill University, who found it to belong to Picea alba, or white spruce, in a fair state of preservation, although buried there for probably more than one hundred thousand of Canada] WHIRT.POOL-ST. DAVID VALLEY 135 years. Of it he says : ^ This species is known in the Pleis- tocene, where it has been recognized in only a few instances ; the hlack spruce, Picea nigra, being, on the other hand, very common and well defined types of such deposits.' BKEATHING WELL. Boring Xo. 1 was found to be a breathing well. Upon reacliing a depth of 226 feet, when the end of the pipe was in the coarse gravel bed, an inward suction was observed. At this point it was deemed necessary to dynamite the, end of the casing in order to insert a smaller tube as the outer casing had become bent in the gravel. Immediately after the explosion, and for another day or two, a strong inward draught continued. Subsequently, at a time when the water in the Whirlpool was particularly high, there was an outward current amounting to a blast. So strong were these currents it hardly seemed possible that such could occur except in proximity to fissures in ad- jacent rock. However, as the continued borings showed the absence of rock, this breathing appeared to be due to the character of the porous gravel, not associated with caverns, and depending upon atmospheric conditions, and the height of the river in the Whirlpool. As the deeper borings were made when the temperature was below freezing point, water poured into the inner casing froze into solid ice at the depth of 226 feet, on account of the suction of the cold air between the outer and inner casing. CHAPTER XI. ST. DAVID CHANNEL BELOW THE ESCARPMENT. Pre-glacial surface shown by Survey of the banks of Niagara depths of wells. river below gorge section. PEE-GLACIAL SUKFACE SHOWX BY DEPTHS OF WELLS. As shown upon the large map the trend of the buried Whirlpool-St. Da^•id valley, as it approaches the escarpment, has a direction much more to the northward than was formerly- supposed. It then curves westward as if.it had originally left the plateau region a mile or more west of the present site, with a narrow ridge of rock, between it and the lake valley, subse- quently removed by wave or atmospheric action, which thus produced the indentation in the jSTiagara escarpment. The topography here is not that of an ordinary embayment at the mouth of a valley, and its structure is so peculiar as to challenge attention. There is nothing in the topography suffffestina; a channel in this northwestern direction, and a northeastern one would require a sharj) turn in the course of the river. As a consequence an examination of the wells to the north of St. David was made. The records will l>e given in Appendix iii. Below the brewery alreadv mentioned (page 129), the sur- face slopes to the Bell terrace plain, with its edge situated beyond the hamlet of St. David, so that this has a breadth of a mile or more. Xearer the Niagara river the Bell terrace becomes quite narrow and is underlaid by Medina red shales and sandstones, which are only slightly covered with surface accumulation. In the vicinity of St. David these rocks are wanting', but thev are replaced bv deposits of sand and gravel 137 138 FALLS OF NIAGARA [Geol. Surv. forming a 'terrace of a mean height of 168 feet above the lake, while a loAver point at St. David village is 164 feet. On this platform are several wells in sand and gravel reaching to a depth of sixt}^ feet, where water is obtained. Adjacent to the road, a qnarter of a mile south of the village, at Mr. Woodward's, there is a well of 130 fe,et, but below sixty feet very little water was obtained. It is not certain whether the well terminates in drift deposits or Medina shale, but even if the drift reaches this depth the bottom of the well is still forty feet above Lake Ontario (or only seven feet below the surface of the Whirlpool). Another well, on lot 96, on this same terrace, is ninety feet deep. Several other wells to the north of St. David show the presence of rock to a height which would preclude a buried channel in that direction ; but such might have skirted the K^iagara escarpment to the northwest. Mr. J. F. Scovell* called attention to the existence of a buried channel between St. David and the ISTiagara river, on account of finding some deep wells, and also because of the absence of Medina shale at certain points on the western bank of the ISTiagara river. The result of an examination of the wells shows that such is the case. On lot 34, a well is said to have a depth of eighty-one feet without reaching rock, while at another one, just sunk, rock occurs at forty feet. On lot 64 there is an absence of rock to a depth of ninety feet, while a quarter of a mile to the north (lot G^), rock is found at sixty feet. On lot 80 the well did not reach rocks at sixty-eight feet. ISTearer the river (lots 19 and 20) no rock was found at a depth of eighty feet. These features would indicate here a channel about twenty feet below the lake level. SURVEY OF BANKS OF NIAGARA RIVER. These consist of Medina red shales for a distance of about two miles north of Queenston village, but on passing Field point * Proc. Amer. Asso. Ad. Sci., Vol. XXXIX., pp. 245-246, 1891. of Canada] BANKS OF XIAGARA RIVER 139 such beds disappear benei.th the water. The banks everywhere form steep bkiffs from forty to sixty feet, and even higher on receding from the river. (See Plate xxix.) In the cove of Hogshollow only bhie ehiy and other drift material appears. About a mile and a half below Field point, at a place just north of Slinglands point, the shales rise to the surface. Lake Ontario is only about three and a half miles farther away. Thus along the west bank of the river a buried valley is found to occur. The eastern banks of the river below a point two miles from the mouth of the gorge consist mostly of drift clay. Whatever shales underlaid the old topographic features they were largely reduced to a level below that of the lake, along the lower course of the river, especially on the eastern side. From the data thus obtained it appears that there was an old channel coursing from St. David to the Niagara, but so far there is nothing to show that it had any gi*eat depth, nor, on the other hand, that there is not a deep one. Even this may have belonged to a local drainage, and may not have come from the WTiirlpool-St. David valley. An old channel in the softer shale should have been relatively much wider than the chasm cut through the jSTiagara limestone. On tlie other hand, a buried valley may possibly be found extending northwestw^ard (but not northward) to the lake, but only one or two wells in this direction have been examined. There may be a complexity of the drainage features of pre- glacial times in this region, which are not yet understood, but they are unimportant if they do not sliow^ a depth throwing liglit upon the bottom slope of- the AVhirlpool-St. David gorge. 141 CHAPTER XII. NARROWS OF WHIRLPOOL RAPIDS. Importance of this section. Amount of excavation and refilling Survey of features about Whirlpool of the Narrows. rapids. Character of the ancient valley at Borings at Cantilever bridge with Narrows. Table of Section. Whirlpool rapids Narrows, rock ex- Accumulations in Channels of the cavation by modern river. Whirlpool and Narrows com- Widening of gorge above rapids. pared. IMPORTANCE OF THIS SECTION. This section of IS^iagara river, which is a continuation of the Whirlpool-St. David gorge, is so very important as to re- quire special attention. The reason of this lies in the problem whether at this locality a drift-filled valley existed, so that the Niagara river simply removed the unconsolidated drift, or whether the Whirlpool headed in an amphitheatre of hard rock, requiring the falls to excavate the gorge out of solid rock, as it is doing to-day. If the falls encountered drift, it would rapidly be removed, and the present comparatively slow rate of recession would not be applicable here. The older investigators of the falls did not consider this question, but thought there was a uniform rate of retreat from the mouth of the gorge to the present site. Dr. Julius Pohlman,^ one of the earliest writers in the renaissance of Niagara studies, thought that the burie,d channel reached far up toward the present site of the falls, and accordingly believed that the time required for the cataract to re-open the drift-filled valley, and recede past the Narrows, was one of short duration. This being the case, there appeared no limit to the length of the pre-glacial channel, even to near the present site of the falls *Proc. Amer. Asso. Sci., Vol. xxxii, p. 202, 1S83. lb. Vol. xxxv, 1887. 143 14:i FALLS OF iXIAGAKA LGeol. Surv. themselves; nor to tlie sliortness of the period since the cataract broke through the barrier at the Whirlpool, which is midway in the length of the gorge. From this one point alone will be seen the extraordinary need of discovering the true character of the "Whirlpool rapids. But there was another problem of equal magnitude, namely, to account for the enlargement of the canon immediately above the ^Xarrows. Hence this was one of the most critical sections in the investigation of Xiagara. Several have written upon the subject, myself among the first, and I fully appre^^iated the difficulties of this part of the river. It was not until 1899 that the invaluable borings at the Canti- lever bridge were made, and not until now were the necessary soundings taken to throw light upon the question. Throughout the field Avork I did not consult my older writings, but made the re-investigations in the light of new facts, which are pre- sented with a closing discussion on the problem. SURVEY OF THE FEATURES ABOUT WHIRLPOOL RAPIDS. The survey shows that the pre-glacial trench at the northern end of the Whirlpool is 1,400 feet wide. At Sinclair point, between the Wliirlpool and ^^Tiirlpool eddy, 2,000 feet to the south, the ancient trench, together with its modern enlarge- ment, is now 1,000 feet wide. This eddy is in a cove 1,500 feet long and 1,300 feet wide, and its upper end merges into the Xarrows, eventually reduced to about 750 feet wide. De- scending to the water level the Medina gray band is incised by 1,000 feet at the end of the Whirlpool. Under Sinclair point, just referred to, the l:)readth of the river is only GOO feet, and this, to some extent, includes the enlargement of the gorge by the modern river. At the eddy it increases to 850 feet, while in the Xarrows it is reduced to .350 feet. South of Whirlpdol jxiint ( Phite xx.), the eastern walls of the gorge have fallen, so as to widen the ]n'e-glacial valley. This feature, together with the encircling form left at Sinclair point of Canada] WHIRLPOOL EAPIDS 145 Plate XX. Mo C an tile vei: T3ricl ge Co PRE GLACIAL CHAINl^EL "^^ WTLLRLPOOL - RAPID S Bjr J.W. Spencer Swift Drift PL LYE 1000 1500 ■ \ Hi^bbarcl's Pt. 370 L L ^■'O RID Map of Whirlpool rapids. The Eddy and Narrows section inside the great river banks (g r t) ; the lower pre-glacial terrace. P., truncated at both ends, but restored by broken lines (e e b) ; ai:>pr(.ximate edge of pre-glacial valley represented by dotted lines heading near b b, though the shallower channel continued toward Lyell ridge, opposite Swift Drift point ; b— b, position of section figure 18. 10 146 FALLS OF XIAGAKA [Geol. Surv. shows the sucklen contraction of the buried caiion producing here the head of an amphitheatre. Comparing this structure with modern topography the head of the deep ancient gorge shouhl be looked for at Sinclair point below the eddy; but what was the character of the tribu- tary ra\dne ? To answer this question one need not go far afield. At a point from five-eighths to seven-eighths of a mile south of the Cantilever bridge (see map, Plate xx.) there is the Lyell ridge of Xiagara limestone, rising on both sides of the river to 370 feet above our datum. In ancient times this was trenched, as at Hubbard point, down to 317 feet above Lake Ontario, with a resulting rounded topography. At a short distance nortli of IIubl)ard point the present gorge at Swift Drift point is contracted to a breadth of scarcely more than 1,000 feet. In the cove between these points there is no evidence of the drainage in either direction. This impression slowly grew upon me day by day, living with the falls for a long time, and learning almost every feature. Indeed, on the eastern side, at a point where the Gorge railway descends, I could not help fancying I saw the head of the Whirlpool rapids ravine in a little cove adjacent to which the rocks showed polished and rounde(l surfaces (opposite to b 1) on Plate xx.). The greater l)readth of the canon is the result of subse- quent excavations by the falls. The ISTarrows of the rapids are inside another valley depression. Thus, at the railway bridges, the floor on the western side has an altitude of 308 feet, Avhile on the eastern side it is slightly less, with the rocks beyond rising twenty feet or more. A 'third of a mile to the northward there is another inner valley with a glaciated bed at 280 feet (111 the Canadian side, while on the Xew York side the pre- ghacial trough is not represented, as the wall is at 291: feet, with tlic ]>resent gorge 750 feet wich- l)etween them. Here the buried valley has been uncovered by the modern river, whose outer banks were more than twentv feet hio-h (above their of Canada] WJIJKT.POOI. RAIMDS U1 margin at olG feet), and 1,500 feet apart where the strong currents carved out sharply marked terraces before the river slirunk into a narrow pre-glacial channel now f(jund to he shallow. (A^Ve Plate xxi. b.) BOEINGS AT CANTILEVEK BRIDGE, The boring at the eastern pier of the Cantilever bridge, by a diamond drill, and the recovery of the core, brought*to light the character of the refilled Whirlpool rapids section. It was thus proved that the buried rock was absent to a depth of 185 feet below the surface of the river, or eighty-seven feet below that of Lake Ontario (page 60). Out of the total thick- ness of the 185 feet, 12-1 feet were found to consist of blocks and boulders of limestone, and sixteen feet of sandstone, all with clay between. A table of these borings is given on a fol- lowing page. Some of the blocks had a thickness of ten or twelve feet, as in the vicinity of seventy-five feet below the river surface. The talus slope extends beyond the overhanging cliffs of ISTiagara limestone for a distance of 220 feet on one side, and 222 feet on the other, with the breadth of the river 410 feet, and a maximum depth of eighty-six feet (figure 21). Fip. 21. Section across the cafion at Cantilever bridge. Tiiis is 100 feet longer than at narrowest part of gorge. N. 1., Niagara limestone; N. s., Niagara shale ; C. 1., Clinton limestone ; R. M. Red Medina sliale and sand- .stones : g. M., Medina gray sandstone; M. s., Medina .shale ; L. O., Level of Lake Ontai-io ; R., level of river; W., boring on eastern side; N C B B C N , approximate form of refilled channel; Pre-glacial channel between limits of W— W and R— R ; t— t., terraces showing boundary of Niagara river before it sunk into Whirlpool Rapids channel. If this section be now examined, as shown on a true scale, with the uniformity of the declination on both sides, one is lOi 148 FALLS OF NIAGARA ^Geol. Surv. struck by the great steepness of the covered slope from the edge of the cliff to the bottom of the well ; also with the fact that the slope of the western side should be the same as on the eastern. This leaves the impression that the bottom of the gorge must have been kept open by a more vigorous current than now obtains, or by one excavating the bed of the channel and re- moving all the debris as rapidly as it fell into the current. How niMch deeper the middle of the buried channel is than the bottom of the well cannot be known, but its depth reaches to almost the same level below the surface of Lake Ontario as the river channel above this point. Accordingly, a much greater depth can not be expected. TABLE OF BORINGS AT EASTERN PEIR OF CANTILEVER BRIDGE. Feet. Surface of ground* to 253-6 Lime rock (fallen block) " 251'25 Boulder " 247 Black slate " 243-,9 Cavity " 243 Boulders " 238-6 Lime rock '• 237-25 CLay and boulders " 234 Clay " 231 Bastard lime rock " 229-5 Clay " 228 Bastard lime rock " 225 Clay " 223 Bastard lime rock ■ . " 222 Clay and boulders •' 218 Bastard lime rock " 317 Clay " 213 Bastard lime rock " 211-5 Clay and boulders " 208 Lime rock " 206-5 Clay and boulders '• 208 Lime rock " 202 Clay and boulders " 198 Boulders " 197 Clay " 178'5 (Bottom of casing) " 176 Lime rock " 175 Clay and boulders " 171 ♦Surface of ground about eight feet above the river, which is nearly ninety-nine feet above Lake Ontario. Datum assumed at 300 feet, which is the top of the coping of the tower here. The figures are taken directly from the Section of the Railway Engineers. It should be noted that some of these masses of limestone are large; but they are not in their natural place, as both these and the blocks of sandstone mentioned occur at depths in the channel far below the levels at which their beds occur. In short, they are refilling a channel which has been cut out of red Medina shales. of CanadaJ WJIIKI.I'OOL IJAPIDS 149 Feet. Lime rock to 161 Clay " 159 Lime rock " 147 6 Clay and boulders " 139 Sand rock " 135 Clay and boulders " 131 Sand rock ' 128 Clay '• 127 Sand rock '• 121 Clay and boulders ' 119 Sand rock " 116 Clay " 111 Sand rock " 110 Clay and boulders " 93 Sand rock " 92 Clay and boulders " P2 Sand rock '' 81 Clay and boulders " 80 Clay and boulders and little red shale " 62 This is bottom of channel. Red shale bed rock " 57 ACCUMULATIONS IN" THE CHANNELS OF WHIRLPOOL AND RAPIDS COMPARED. Glancing now at the character of the material in the refilled gorge here, and comparing it with that filling the Whirlpool- St. David channel, the contrast is striking. In the latter mentioned trench the accumulations are sands, angular gravels, a little loam, rarely a boulder, and such boulders as do occur are found to be composed of crystalline rocks. A little red sandy clay covers the surface. This drift material is now known to an actual depth of 269 feet. Only one or two blocks of ISTiagara limestones were observed on the sides of these deposits ; so, also, the boulders which have accu- mulated at the end of the Whirlpool are, almost entirely, trans- ported crystalline rocks. The contrast of this glacial material with that in the boring at the Cantilever bridge, which is com- posed of fallen Niagara rocks, is conclusive evidence of the newness of the Whirlpool rapids gorge to the full depth of 185 feet below the surface of the river. AMOUNT OF EXCAVATION AND REFILLING AT THE NARROWS. Here is a caiion of unequalled narrowness reaching to a depth of 395 feet below the adjacent floor of the outer valley. 150 FALLS OF XIAGAKA [Geol. Surv. In the section it i^^ fonnd that ninety-two feet are composed of hard limestone, twenty-five feet of softer red sandstones, not conntino' the thinner hiyers which are perishable, heing non- resistant, (as shown in Plate xxi. a, on page 157), and twenty feet of very hard, gray sandstone, wnth the base of the lowest of these beds forty-one feet above Lake Ontario. The lowest 128 feet of the i-efilled gorge is excavated out of soft shale, which conld not snpply any of the limestones or sandstones found throughout the lower depth in the 1)oring. When the full force of the current was diverted into the smaller pre-glacial channel most of the shale beds were swept away, and also some of the thinner layers of limestone and the friable sandstone, although some shales remained — to the ex- tent of forty-five feet out of 255 feet — the total of such beds found here. Some shales produce clays in the interstices be- tween the boulders when protected from the stronger currents. Tt is now possible to compare the cross-section of the modern chasm with that of the gorge at its maximum develop- ment on one hand, and on the other, with the small pre-glacial valley which gave rise to the narrow canon at the Whirlpool rapids. As shown in figure 21, the area of the cross-section of the whole channel, including the buried portion brought to light by boring, is 248,000 squ.are feet; that of the modern gorge is 156,000 square feet; thus leaving the l)uried portion repres- ented by 92,000 square feet. Of this last amount, 60,000 square feet consists of hard limestone or standstone blocks or boulders. The area of the Niagara and Clinton limestones, and the gray band of Medina sandstone, which have fallen from their natural position into the gorge, together with that portion re- moved in ])rc-glacial days, amounts to 87,000 square feet. The difference between these figures represents 18,000 square feet removed from the natural bedding in the gorge in pre-glacial times. That is to sav that the ancient vallev had a cross-section of Canada] WHIRLPOOL RAPIDS 151 of this area; wliieh if distributod with a breadth of GOO feet its depth would be thirty feet, and this form would be some- what favoured by the occurrence of the lower shelf at 280 feet beside the Narrows. This form of section is represented at (a) figure 18 (page 105). On the other hand, if the width were 400 feat, about that of the river to-day, the depth should have been forty-five feet as shown at II R (figure 21). In these calculations some errors rise from treating the mass of fallen boulders, having interstices filled with clay, as if com- posed of solid rock in its natural bedding. Allowing for this structure would leave a decrease of the fallen mass, and an increas3 in the size of the pre-glacial valley. On the other hand, with the removal of the clayey matter by the strong current of the river, much material from the thinner beds of limestone and sandstone would be carried away, and allowance here would decrease the arei of the pre-glacial gorge. However, whether the ancient channel were a little larger or a little smaller than here represented, the analysis shows that at this point a deep gorge did not exist. Accordinglv, this Ijuried, partly refilled canon, nearly 400 feet deep, was excavated by the modern falls when the river surface was somewhat lower than now. The ancient valley diverted into it the waters of the river, so that the concentrated force would remove the debris and leave walls as steep as those discovered hy the deep boring. Upon the recession of the falls beyond this section, favoured by the backing of the water into the gorge below, the bottom of the channel could not be kept open ; accordingly, below the bottom of the river, which is eighty-six feet deep at the (Canti- lever bridge, the channel is reoccupied liy fallen rocks as shown by the borings at the bridge. Just below the bridge the A^^iirl- pool rapids descend impetuously fifty-two feet to the Whirl- pool. Unquestionably these rapids are formed by blocks fallen into the channel from the overhanging brink of the chasm, though the greater part of the refilling occurred earlier, without 152 FALLS OF NIAGARA [Geol. Surv. producing the present rapids wliicli have been completed only after the falls had receded to near their present position, which accounts for the freshness of the walls, and the absence of any indentation where Muddy creek runs down the western wall at the ]Srarrows. CHAEACTEE OF THE ANCIENT VALLEY AT NAEEOWS. So far attention has been primarily given to the modem character of the channel at the Whirlpool rapids. Thanks are due to Mr. P. W. Currie'" for calling attention to these boulders and fallen blocks found in the borings. This channel is exca- vated in an outer valley which has an elevation of 308 feet above Lake Ontario on its western side at the railway bridges. From a quarter to a half mile northward there is a fragment of a lower rock cut terrace, having a width of 125 feet, and an elevation of 280 feet. That it is ancient is shown by its gla- ciated surface. It is represented at (a), figure 18 (page 105). Here is the narrowest part of the gorge — only 750 feet wide, or a hundred feet narrower than at Cantilever bridge, where there is no remnant of the lower terrace. (See map, Plate XX.) The lower terrace does not remain on the eastern side where the wall of the gorge rises fourteen feet higher. Such removal has been occasioned by the currents encroaching on and undermining the eastern more than the western side. Still the breadth of this inner valley, with its depth, seems to have been sufficient to produce a cross-section of 18,000 square feet with the pre-giacial channel unable to carry the whole volume of the river. This conclusion is derived from the fact that the inner channel is situated within a broader outer one bounded by steep banks shown in the view and on Plate xxi. b, and at 1 1 figure IS, These letters represent the margin of the great river. The outer channel continues to Sinclair point, where the western * Trans. Can. Inst., Vol. VII., pp. 7-14, 1891. of Canada] WHIRLPOOL KAPIDS 153 bank is abruptly terminated at tbe brink of the chasm at the "WliirlpooL Above Sinclair point, and extending past the eddy section (on maj) Plate xx.), the falls excavated a- canon mnch wider. This broad amphitheatre at the eddy was not the pro- duct of the falls acting within a narrow chasm, but it shows that much of the water flowed over the whole end of the outer channel before the main stream was contracted in width. A channel with a section as estimated above, could take the greater part of the volume of the river, while a portion of it was being diverted by way of the Chicago overflow into the Mississippi drainage. This condition continued until the falls had receded in the ancient valley to a point, where it was too small to direct the forces of the cataract, after the return of the full volume of the river, when the water overflowed the sides of the chasm and recommenced the broadening of it. Thus it appears that the pre-glacial trench at the "Wliirlpool rapids, beginning in the amphitheatre at the head of the Whirl- pool, somewhere between Sinclair point and a few hundred feet to the south of it, extended upward and soutliAvard through the section of the Narrows. In order to permit part of the discharge to still flow over the floor of the main outer valley, and at the same time divert most of the force into the main channel, its depth must have been less than the thickness of the Niagara limestone, so much less that at first there was a great fall at its lower end occupying the whole outer valley. This occupancy of the floor of the pre-glacial channel was like the lateral discharge of the waters over the wall of the falls be- yond the present apex. In this way was formed the cove at the eddy, broader than between the Sinclair and Whirlpool points. But the greater force of the river in the pre-glacial val- ley must have deepened the channel sufficiently to withdraw the water from the outer and broader one. Such a feature is shown by the contraction of the eddy section at the end of the Narrows, when the river no longer spread its waters over the broad 154 FALLS OF NIAOAKA [Geol. Surv. i/i_- ■■■■/ \r' fe =''?' I- K "^ of Canada] WHIRLPOOT. KAPIDS 155 channel. Thus is satisfactorily explained the features of the eddv section. A characteristic of such an old gorge would he the descent from the head of the slope by steps over the harder jSTiagara, Clinton, or Medina sandstones^ which are represented in the longitudinal section iigure 22. It is not to he supposed that the ancient valley terminated abruptly at the upper end of the Xarrows, for the pre-glacial land surface beyond this point continued to rise for a distance of three-quarters of a mile to Lyell ridge, the rocky summit of which has an altitude of 370 feet. This ridge formed the divide between the Whirljiool valley and the Falls-Chippawa basin trending in the opposite direction (see Plate xxii.). Even in pre-glacial times the full height of the ridge Avas inter- rupted, as the ancient erosion at the divide had reduced it by lifty feet to a height of 320 feet at Hubbard point. In studying the features upon the ground there was always a strong impression left that the stream proper headed in a little cove opposite Swift Drift point, where the Gorge railway com- mences its descent, and that the basin or cove between Swift Drift point and TIul)liar(l ]:)oint was one like many upon the summit of divides which drain in both directions. WIIIRLPOOI, EAPIDS NAREOW^S ROCK EXCAVATION BY MODERN EIVER. From careful measurements from soundings, and particu- larly from the borings at Cantilever bridge, the results show that the pre-glacial AVhirlpool gorge did not extend much above Sinclair point, though a small channel led into it. Therefore, the falls, except at the AVhirlpool, have had to excavate all of the gorge out of solid rock, as they are doing to-day. It has been mentioned that various observers thought that a buried channel may have reached to nearly the site of the present falls. This was not, however, my earlier impression. I did not then closely define mv views beyond stating that ' at 156 FALLS OF NIAGARA [Geol. Surv. this point the Niagara river took possession of the eastern side of a drift filled valley.'" Bnt later my opinion was made quite clear, namely, that the channel of the modern river ' maintained a fnll breadth and depth, but the constriction applies to the gorge alone/ that is, the inner one which was filled with drift. ' The depth of the depression was the greater in the centre, and the river took possession of a deeper portion, and upon the removal of the drift, sunk within the gorge.' ' This shallow- buried valley began in Johnson ridge ' (that is the Lyell ridge at Hubbard point) ' just above the railway bridges, and extended to the "Whirlpool, whose cauldron is only a deeper extension of the same channel. 'f The views expressed in 1894 were only an avant courier to the present w^ork, and it is no little satisfac- tion that they have been so fully sustained. Subsequent to my contribution just mentioned Mr. F. B. Taylor, who has been one of the principal investigators in the later history of the lakes, contributed a paper upon this subject in 1898. In this he thought that there was an amphi- theatre at the head of the Whirlpool, located at Sinclair point, but he did not consider the importance of the little valley which supplied the waters for the small ancient cataract. With this modification, Mr. Taylor's conclusions agree with mine formerly expressed. The head of the amphitheatre evidently descended by a series of steps which extended into the eddy section. The origin of the jSTarrows of the Whirlpool rapids section is now set forth, namely, that the larger part of the river was concentrated into the narrow pre- glacial valley, while another portion of the lake drainage over- flowed by way of Chicago. So long as a considerable portion of the river was concentrated into the channel it would not be widened except by the undermining of its sides. Eventually this ancient valley was not large enough to take *Amer. Jour. Sc, Vol. 3, xlviii, pp. 455-472, 1894. tib. Vol. 4, vi.. 439-450. 1898. Plate XXI. a. View of Section of Medina Red shales and sandstones, on Gorge Roads, just below G.T.R. bridge. Plate XXI. is. View of Original River bank, outside of the Narrows of Gorge, in front of Mount Eagle. ]57 of Canada] WHIRLPOOL KAPIDS 159 the volume of the river, increased by the portion of water tem- porarily diverted at Chicago. In order to account for the Whirlpool Narrows, Mr. Taylor formulated a temporary with- drawal of the Huron waters while the falls were passinc; the rapids. He thought the temporary diversion was occasioned by certain clianges of the outlet of Algonquin hike, before the waters were finally backed southward by the rising of the bar- rier at Lake Kipissing. I had not found anything to sustain this view, because the old shore line of Algonquin lake cut off the overflow of Lake LIuron to the south until the barrier was l)roken through by the rise of land at the ISTipissing outlet. E[is hypothesis required reduction of volume as complete as before the original addition of the Huron drainage. Mr. Taylor's temporary diversion proved only partial. It was due to the overflow at Chicago into the Mississippi drain- age, and not to glacial dams as he supposed {see chapter xxvii.). The amount of diversion was suificient to cause the withdrawal of the water from the shallower portion of the channel, at the AVhirlpool rapids so that the falls produced a caiion 750 feet wide, in place of 1,200 feet as elsewhere. WIDENING OF THE GOEGE ABOVE THE RAPIDS. As the outer channel at the Narrows is 1,500 feet wide, with the gorge reduced to 750 feet, and as there was a small temporary reduction of Huron discharge, which reduced the volume of water here, the question arises as to what caused the rapid enlargement of the gorge to a breadth of 1,300 to 1,100 feet above the Whirlpool rapid ISTarrows. This width should be reduced somewhat on account of frost action and under cutting of the cliffs by the river, so that the original banks above here have been undermined and completely fallen away; however, a little farther up, at Hubbard point, they again appear, show- ing the river to have been 1,200 feet wide. As the pre-glacial channel became too narrow to direct the course of the river, 160 FALLS OF jSriAGARA [Geol. Surv. now enlarged again, it would flow around the head of this channel and over the side of the narrower chasm, as it did at an earlier time round Wintergreen flat, and now over Goat is- land shelf to the side of the apex of the present falls, and break over the sides until it could widen a gorge sufficiently large to receive the volume of the falls. To emphasize this feature the river was not shooting through the confined channel as that be- neath the Cantilever bridge to-day but through one of moderate depth, and was breaking over a large periphery with the conse- quent undermining effects. From that time to the present, facts now known show that there has been no great change in the character of the gorge, except that there has been a large amount of undermining of the walls and falling in of the original river banks. CHAPTER XIII. FALLS-CHIPPAWA BURIED VALLEY. Well borings, depth of drift, and Falls-Chippawa valley south of absence of channel west of gorge. Victoria Park. Valley indicated by deep wells ad- Late reversal of drainage at the jacent to falls. falls. Southward enlargement of Falls- Lower terraces of Falls basin. Chippawa valley. Results of finding the Falls-Chip- Origin of Upper rapids and Great pawa valley. basin of the falls. WELL BOEINGS, DEPTH OF DRIFT AND ABSENCE OF CHANNEL WEST OF GORGE. Reference has already been made on former pages, to the ap- parent rock basin at the falls as noticed by Hall, Lvell and Julins Pohlman. Later, Mr. Scovell ascertained that there were deep wells to the west of the Clifton House, and thought that a buried valley extended from the present site of the falls to St. David valley. Learning that there were deep wells I provisionally followed Mr. Scovell, and I regarded the Whirl- pool St. David valley as a tributary to a greater but shallower channel to the west (189-i). The valley above the falls is wide ; the outlet of the "\^Tiirl- pool-St. David valley was supposed to be broad, though it is now found to be narrow. Then, again, the river at the Upper rapids was nearly in the direct course of Pohlman's Tona- wanda drainage. Consequently these broader features favoured the location of the supposed valley and provided a working hypothesis, especially as the caiion-like form of the Wliirlpool- St. David trench had not been surveyed. On commencing to re-investigate the physics of Niagara falls the first work to be done was to determine how this falls basin could have been originated by streams flowing west- 11 161 162 FALI.S OF NIAGARA t^^"'' «"'■^• ward of the present course of the river toward St. David. In order to ascertain the origin of this basin, which is excavated to sixty feet in depth out of hard limestone, the depths of a very large number of wells were obtained ; some of which were ninety feet before reaching the rock surfaces. The buried rods floor was then determined by taking the levels of the wells. The result showed that the pre-giaeial surface was too high for a trench extending from the falls northward upon the western side of tlt^ river. The rock floor of the country west of the river, from the town of Xiagara Falls and along Lundys Lane, was found tc have a general elevation of 340 feet above Lake Ontario, while the pre-glacial surface of the basin at the falls was reduced to 258 feet. VALLEY INDICATED IN DEEP WHEELS ADJACENT TO FALLS. To the southward deep wells were found. West of Victoria Park, on the property of the Carmelite monks (well 'No. 6), borings reached a depth through the drift of ninety-five feet, showing that the floor at a point 700 feet back of the upper edge of the drift hills was -reduced to 277 feet above Lake Ontario. This proved to be at a point on the side of a buried channel. A number of other wells had been sunk, which are shown on the maps. As a result of the sinking of these wells a deep trough was found between the rocky floor south of Drum- mondville and the rising floor adjacent to the Upper rapids, and at Chippawa. The breadth of the upper part of the trough was most plainly brought to light, as it was crossed by the pipe- line of the Ontario Power Company in an oblique direction, exposing depressions in the rock for a distance of 2,400 feet, or along a direct section across the old valley of about 1,700 feet, with the rock bed absent above the level of 259 feet. On either side the rocky rim rose to 320 feet or more. Accordingly the re-opened basin at Niagara falls extended southward under the drift hills. of Canada] FAT^LS-CIIirPAWA VALLEY 1 63 It was then ascertained from otlier deep wells, some three miles distant, that the rocky floor iinder the flat snrface of the comitry was reduced to a similar depth as that of the basin at the falls. This suggested a trench. It now seemed that a valley had been discovered which furnished a large strc^uu for exca- vating the Whirli^ool-St. David valley. This was before fully analysing the evidence bearing upon the Whirlpool rapids. Indeed weeks elapsed, with additional data added from time to time, before any attempt was made to carry new investi- gations in the direction where they subsequently led. Had I then published, a serious error would have been made. The importance of the features at Hubbard point would have been overlooked, and it would have been stated that the ancient Chip- pawa creek had been turned into the Whirlpool-St. David chan- nel in conformity with the present topograph}'. SOUTHWARD EISTLARGEMEIN'T OF FAELS-CIIIPPAWA VALLEY, At Hubbard point, also opposite Swift Drift point be- yond, the floor of a ])re-glacial valley was 317 feet above level of Lake Ontario. This depression was made in the limestone formations of the Lyell ridge, which here crosses the course of the gorge, and rises back from the river to 370 feet. If one stands at the eastern end of the Upper Arch bridge and looks at the broad valley, bounded on one side by the Upper rapids, and on the other by the terrace-hill of Victoria Park, and then turns in the opposite direction, the terraces on the Canadian side are seen to swdng round to near Hubbard point, and appear to meet the corresponding ones on the ISTew York side, which, close to the end of the bridge, are underlaid by rock at or near the surface. The view presents an amphitheatre nearly closed at the northern end and widening out to the south, suggesting a former topography draining southwestward. Applying measurements : Up the Uerry road leading from Clifton Hotel the rock was everywhere shown to be near the surface, for it 11* 164 FALLS OF iSriAGARA [Geol. Surv. Plate XXTT. FALLS- CHI PPAWA VALLEY By J.'W''. Speiicci* Swift I>riftPt., Hubbaxd'aPtJ Map of ]*^alls-Chippawa™pre-glacial valley. of Canada! FALLS-CIIIPPAWA VAT.LEY 165 was reeceiitlv exposed in the sewer-cnts. On the map (Phite XXII.), the bounclaiy of this ancient basin was taken at the level of 320 feet above Lake Ontario, as the rocky floor above rises somewhat slowly, while at Hubbard point the old floor has an altitude of 317 to 320 feet. Opposite the Ferry road, across to the corresponding point in Prospect Park, the valley has increased from 1,200 feet at Hubbard point to a breadth of quite 2,500 feet, though it is only 3,500 feet to the south of Hubbard point. Another tra- verse of 3,000 feet brings us opposite the southern corner of Goat island. From the rocky shore near the upper end of the island, across the gorge to the terrace behind Victoria Park, is a distance of 3,500 feet; but as the Canadian rocky rim here is buried, the additional amounts that must be added to this breadth are not definitely known. At another section 3,000 feet southward at the Carmelite monastery, the distance, from the rocky edge of the Upper rapids at Goat island to the well CNo. 6), is about 4,700 feet. As this well is within the valley the additional distance to the buried western side is not known. Thus the picture of the widening of the basin to the southward, although head- ing only at Hubbard point, is supported by actual measure- ments, and the enlargement broadens from 1,200 feet to over 4,800 feet in a distance of less than two miles, while the depth increases about sixty feet in this distance. At even 270 feet above datum its breadth in the Park is 1,600 feet. As before pointed out, nowhere north of the immediate vicinity of this channel, as at the Upper Arch bridge or anywhere north- ward, are there any remains of a pre-glacial floor lower than 280 feet. This broadening and deepening to the southwest- ward, with a slope descending from the highest ridge at Hub- bard point, forced the conclusion that the Falls-Chippawa valley was a pre-glacial feature with the drainage in the oppo- site direction to that of the present day. 166 FALLS OF NIAGARA ^^^o^' ^^rv. Parentheticallv let it b? stated that at this hioher Lyell ridge at Hubbard point, while its surface was due to atmos- pheric action on the N^iagara limestones, the erosive features may have been determined bv a physical disturbance of the strata {see Plate xvi.^ P'lge ^^^), which on its northern side has a general southwestward dip and on the other side a south- eastward. OEIGIN OF UPPER EAPIDS AND FALLS BASIN. Tims was discovered the Falls-Cliippawa trough or basin which revealed the origin of the Upper rapids of the ISTiagara river, in that their slope belonged to the pre-glacial surface of the southoastern side of the Falls-Chippawa depression. FALLS-CHIPPAWA VALLEY SOUTH OF VICTORIA PARK. The establishment of this Falls-Chippawa valley was not completed by the data already given. There was a space of three miles southwest of Victoria Park that required investi- gation. At Chippawa village, in well Xo. 8, the rock floor has an elevation of 298 feet. The floor here is actually lower than at the intake near the upper end of the Dufferin island channel, where it is at 300 feet. A little below this point, at the elbow of Dufferin channel, the old surface is 286 feet, and in Logan's well (Xo. 9), nearby, rock is two feet lower. At another Avell (I^o. 7), a short distance to the westward, the rock surface is reduced to 205 feet, or only seven feet higher than the lowest ]iart of the floor of the channel in the south- western cove of the Park, which was formerly covered by a gravel bed, since removed in laying the water pipes of one of the power companies. Thus the old southeastern side of the Falls-Chippawa basin is established. The Carmelite well (Xo. 6), near Loretto convent, shows the rock floor at 279 feet; this is on the lower part of the rising northwestern boundary. The definite limits of this boundary have not been ascertained, of Canada] FALLS-CITIPPAWA VALLEY 167 ]>ut witliin a short (Hstancc the rock may be ex])octe(l at 330 or 340 feet. About three-quarters of a mile Avest of tlie Carmelite well, at another (Xo. 13), the depth is said to be 100 feet. If tliis be the case the rock floor is reduced to 266 feet; and even if somewhat higher it indicates an embayment of large size into which the Falls-Chippawa basin expands on receding from xs^iagara river. At a mile to the southwest of this latter point, the Malone well (jSTo. 20), reaches rock floor at 312 feet above Lake Ontario, while the Grev well (Xo. 21), three- quarters of a mile farther southward, shows the rock at a level of 267 feet; so that the boundary line is somewhere between these two last mentioned points. (See Plate xxii. and also large map where wells are indicated.) At half a mile south of the Grey well is the Ferry well (Xo. 22). It is situated just north of the creek. The rock sur- face is at 259 feet. Thus it is found that the embayment sweeps southwestw^ard into a greater basin between the ridge of Niagara limestone on one side, and the Corniferous lime- stone on the other. Eeturning now to the cove south of the convent, the south- eastern side may be located by the Glasgow well fXo. 10") where the rock surface is at 282 feet, or sixteen feet below that at Chippawa village. Again in the middle of this buried embayment, at iJMontrose Roundhouse, the rock appears at 256 feet (woll ISTo. 23), and in Kister's well nearby (Xo. 21), at 251 feet. Xear the middle of the valley are two Clark wells, one of which (Iso. 11), shows a depth of 112 feet to the rock. Some of the quicksand from the bottom had been kept by the owner, Mr. Patrick Clark. Here the buried floor is at 232 feet. In a southwestern direction, nearly half a mile beyond the creek, is the well of ]\Ir. Emanuel Read (Xo. 25), which reaches rock at a level of 236 feet above Lake Ontario. 168 FALLS OF NIAGARA [Geol. Surv. LATE EEVEKSAL OF DRAINAGE AT THE FALLS. Tlie last mentioned wells show the deeper part of a buried channel, which is nearly 100 feet below the level of Lake Erie, whose elevation is 326 feet above Lake Ontario. This feature opens up another question, namely, what became of the drain- age of the Falls-Chippawa basin before the mantle of drift was spread over the ISTiagara peninsula? Following up this investigation a new outlet for the Erie basin has been dis- covered, and will form a sequel to the chapters on Niagara falls. The reversal of the course of drainage of the Niagara dis- trict has been established by well borings; still, had one not blindly gazed upon the topography about the falls it would have been of itself apparent even without measurements already set forth in a former part of this chapter. The whole basin from Hubbard point clearly suggested a topography opening to the south, which now anyone can see at a glance, as shown in map, Plate xxii. Had not the drift hills adjacent to the Upper rapids been too high the water would have coursed a mile westward, when it would have swept around the triangle between Victoria Park and Chippawa village. In this case the river would have estab- lished its course in the deeper channel without forming the Upper rapids. The tendency of the river to sweep round and remove these drift hills is shown in the broad flat of Victoria Park, and in the cove which contains Dufferin islands. LOWER TERRACES OF FALLS-BASIN. The lower terraces of the Falls-basin, below the one des- cribed at about 317 feet (page 115), are of minor importance as they only mark the lowering of the stages of the river during the recession of the falls. In Victoria Park the undermin- iuff of the hiii'her terraces has led to the formation of one high bluff, yet on approaching its northern end a terrace is distinctly separated, at an elevation of 308 to 310 feet. It of Canada] FALLS-CHIPPAWA VALLEY 169 continues about a third of a mile north of the Upper Arch bridge where it is cut off at the edge of the chasm. It is com- posed of commingled river gravels and earthy deposits. Its equivalent occurs at Prospect Park and on Goat island, and its level is that of the rock floor at the Narrows, where crossed by the railway bridges. This last feature suggests that this ter- race was formed in quieter water held back by the higher floor in the vicinity of Hubbard point. Another terrace is shown at 292 feet, having a lower rock floor in front at 280 feet. The lower rock floor in the vicinity of Table Rock House at 265 feet, and a still lower fragment at 258 feet, are remains of the bed of the Falls-Chip- pawa basin, uncovered by the modern river. EESULTS OF FINDING THE FALLS-CHIPPAWA VALLEY. The account here given of the Palls-Chippawa channel is, I believe, the first suggestion of its existence. It certainly is the first proof of its discovery, with the topographic features opening to the southwest. These, as described in this chapter, establish the origin of the Falls-basin and that of the Upper rapids, and reveal the existence of the deeper trough out of which the falls are now beginning to emerge, with consequent retardation of their recession. They also show the former reversal of drainage in this district. Lastly, the investigations here described led to the finding of the great ancient outlet of the Erie basin, which, as earlier discoveries showed, received the pre-glaeial drainage of the Upper Ohio river. Plate XXIII. a. View of southern end of Foster Flats, obstructing gorge, with rapids cutting round, (from Whirlpool Point). Plate XXIII. View of upper end of Foster Flats, and Rapids, with outlet of Whiilpool in the distance, (from near De\ils Hole). 171 CHAPTER XIV. FOSTER FIATS— RECORDING CHANGES OF HEIGHT OF FALLS AND VOLUME OF RIVER. Significance of this section. Northern spur of Wilson terrace Characteristics of Foster flats and and cove behind. associated features. Union of Niagara and Clinton falls, Wilson terrace of Clinton lime- their heights. stone and its remains. Third or Medina cataract, making Pot-holes in fallen blocks a channel at Foster flats. Distinction between Niagara and Increased volume of the river. Clinton blocks. Review of former conclusions. SIGNIFICANCE OF TITIS SECTION. Midway between the Whirlpool and the bend of the river beyond the Devils Hole is an extensive terrace projecting far into the gorge. This is surmounted by other terraces and huge fallen masses of rock, the whole producing a sloping promontory extending into the canon at a level lower than the floor of the country trenched by Niagara river. This is shown in Plate XXIII. A. Standing upon the highest of these terraces, called Winter- green flat, which is a fragment of the late floor of the river, and looking northward, there appears to be a densely wooded flat obstructing the river beneath. This is shown in Plate xxvi. b. The name ^ Foster flats ' was derived from that of a set- tler of early days, and locally the main terrace is known by this appellation, which also appeared on charts of the river many years ago. Lately, however, the Park Commission called it the 'Niagara, glen. In the true sense of the word it is not a glen as it occurs on one side of the river only, in the form of a ter- race surmounted by others or flats wherein the whole scientific importance lies. The term ' bottom ' could also be appropri- ately used. Foster flats was not mentioned in scientific litera- 173 174: FALLS OF NIAGARA [Geol. Surv. tnre until I drew attention to their bearing upon the history of the Falls, 1894.* Foster flats furnish by far the most im- portant records in the history of Xiagara throughout the gorge. CrL\RACTERISTICS OF FOSTER FLATS AND ASSOCIATED FEATURES. Foster flats (see large map) begin at 11,000 feet from the mouth of the gorge', in a little eddy at the point of the wedge where tho river is about twenty feet above the lake. Then they broaden , out and extend ' up the canon for a distance of 3,600 feet. At first the terrace in front of the talus has a breadth increasing from 60 to 100 feet in a distance of less than 300 feet from the end of the flat, with an elevation 'of about twenty feet above the river. Here is a sudden rise to thirty-five or forty feet, whereupon the terrace rapidly broadens out to 300 feet. On its lower surface there are almost no fallen masses until reaching 600 or 700 feet from the lower end. (See map Plate XXIV., also figure 23.) 1 -Mof.. r.N. 1 -_i_^ _:^^int,erAr?^^ £J%i?_"^r'_ ■- — • — ' ^ ' ' < ■MO ^ 0. Sc^l. ? J$5 «!! iOO KjOp^^^ Fig. 2.3. Longitudinal section of Foster flcats. R.— river surface above flats (below Flats river is 20 feet lower). L. O.— level of Lake Ontario ; X. X.— approximate bed of channel partly refilled with blocks producing Foster rapids ; a b c d e f g, profile 200 to 300 feet from edge of river ; showing steps nearer the river at f and g. Spur of fallen masses at d with valley c at flat-level. W. P., Wilson point (or end of spur) of Clinton limestone with trail of blocks in front extending to river rapids. W^. t., Wilson terrace of Clinton limestone (C. 1.), 100-200 feet wide, ending at lower end in a spur of fallen masses, g. M.— gray Medina band forming i)art of floor of flats ; N. 1., Niagara limestone wall behind, with spur (Winter- ■ green flat) |n-ojecting 400 to 500 feet between x and y. At a thousand feet from the northern end, the plain has a breadth of nearly 800 feet, reaching from the river to the foot of the talus slope. Here is the end of the medial ridge, which * 'Duration of Niagara Falls,' by J. W. Spencer, cited before. of Canada] TERRACES AT FOSTER FLATS 17{ extends from luulor Wiiiterg'reen cliffs, and incloses a cove oi several hundred feet in length, the floor of which rises from seventy-five feet above Lake Ontario at end of spur, to 115 feet at its head, where it is characterized by fallen blocks. Wilsdu ]-idge, traversing the flats^ is 150 to 200 feet from the river, with the terrace in front, nearly fifty feet above the water. From the end of the ridge to the Pavilion is about 800 feet. ]\Iidway is an abrupt rise of ten feet to the higher terraced floor, which in front is seventy-five feet above the lake, rising- backward somewhat rapidly to the foot of the Wilson ridge. At a short distance above the Pavilion the terrace is sur- mounted by higher ground partly composed of natural strata, but these are mostly buried by fallen blocks ^\hich represent a tongue of debris extending from the higher ledges cross- wise of the river. Many of these blocks rest upon large round- ed boulders and stones which belonged to the bottom of the river. Beyond this spur is an embayment along the course of the intermittent drainage streams, with the floor having a height corresponding to that of the terrace. Farther on is another spur. Here the disturbed strata extend directly to the cliff above the water, and are covered by great fallen blocks. The end of the ridge is seen in Plate xxv. a, which also shows the rapids in front of the flats. Passing the second spur, or Wilson point, the slope rapidly descends to the surface of the main step or terrace, which extends to the upper edge of the flats 600 feet beyond Wilson spur, with a height of seventy-five feet above Lake Ontario. The two cross spurs shown on map are composed of blocks derived from Wintergreen cliffs, and from Wilson terrace. Indeed, it is from the blocks of these spurs that the river is obstructed so as to give rise to rapids situated here. The descent of the river from above Foster flats to just below the Pavilion, is about fifteen feet, while between here 176 FALLS OF ISriAGAEA [Geol. Surv. of Canada] TERRACES AT FOSTER FLATS 17Y and the lower end of the Flats it is only five feet. (See map, Plate XXIV.) The terrace at the upper end of the old river bottom shows that the forces of the river were still at work here at the same level as that which obtained while the falls were at the lower part of the flats, but a great change occurred immediately above them. As the terrace bends round from the nar- rows to the foot of the cliffs, close to the talus banks, it is indented by a trench leading to Fisherman eddy (also called Cripson eddy), cut through the underlying thick Medina beds; so that although there is no permanent stream the rains and frost action have excavated a large gTilly out of the hard rocks. While the main floor is generally underlaid by the Medina band of gray sandstone having a thickness of fourteen feet, with spurs covered by fallen blocks surmounting it in front of Wintergreen cliffs, yet the lower end of it has a level below the surface of the Medina sandstone. This shows that its origin and preservation are not entirely due to the dur- able rocks contained in it. So, also, with regard to the shales and red and mottled sandstones of the overlying spurs. Indeed these spurs are the remains of the Wilson terrace (W. P. figure 23) which once extended across the river, when the inter- mediate falls from the Clinton beds had not yet united with the upper Niagara falls, which union is found to have taken place here. The cul de sac, behind Wilson terrace, continues broad, heading in the marginal w^all of the canon from which Wintergreen flat projects, but the apex is beyond the limit of the old river bank. On descending the cove and receding from the cliffs the blocks become more scattered and almost disappear, in strong contrast to those at the head — (one of which measured fifty by forty by twenty feet) — showing how the river grinds up or dissolves the great masses which fall into it. Here on the terrace is a feature of the river floor showing what would be 12 178 FALI.S OF XIAGAEA [Geol. Surv. expected if the present channel conh;l be drained. This floor was the old bottom of the river when the waters were descend- ing upon it from the second cataract. The northern face of Wintergreen flat shows a broad V-shaped incision, over the edge of which the river once cas- caded. In front of the old river bank at Wintergreen flat its inner edge has an altitude of 316 to 318 feet, with a floor sloping downward to 305 feet at the northeastern angle. Thus the river must have reached a depth of over ten feet. The pre- sent form does not show the shape of the crescent when ISTiagara falls were located here. The northeastern angle of Winter- gi-een is undercut as if by currents rounding the point. The precipitous clifl^ is fortv-five feet to the talus, which now slopes sixty-five feet to the AYilson terrace, at an elevation of 195 feet above Lake Ontario. The point, 310 feet on map, is still more under-cut, as this was the last spur at the apex of the falls of smaller volume^ before the union of the upper ISTiagara and Clinton cataracts. This feature locates the middle, of the original river and gorge, when the upper falls were cas- cading on the Clinton or Wilson floor, as is more plainly seen in the' field than on the map. WILSON TERRACE OF CLINTO]Nr LIMESTONE AND ITS REMAINS. This terrace is due to the massive bed of Clinton limestone of eight or ten feet in thickness resting on another ten feet of thinner limestones. The upper bed is extremely durable, while the lower beds are somewhat yielding. Again, these strata rest upon the more perishable red shales and sandstones of the Medina series. Here occurs a conspicuous flat which I have called the Wilson terrace, in honour of Mr. James Wilson, who turned these wilds among the huge masses of fallen rocks into a wonderful park, unlike any other in eastern America. In front of tlie talus below Wintergreen cliffs the Wilson terrace has a breadth of 100 to 200 feet. Below the eastern angle Plate XXV. a, View of head of Foster Flats, and of Wilson Point, where the upper and middle cataracts united. Plate XXV. b. View of Cove and widened channel, where Foster Flats are abruptly 124 cut oft at their head. 179 of Canada] TKERACKS AT FOSTER FLATS 181 of the upper cliffs (310 feet above the lake), the spur of Wilson terrace (map and Plate xxxv. a), extends outwards across the course of the river, and terminates 400 feet from the edge of the cliffs with detached masses beyond. On the southern side of this spur the terrace is bounded by cliffs twenty feet or more in height, in front of which is a talus strewn with blocks of fallen Clinton limestone. These are now lying outward where they were broken off as the currents undermined them by the reiuoval of the softer strata beneath. It seems strange to see the effect of the river undercutting in front of it instead of on its margin, but this spur is a remarkable record of the union here of the secondary falls and the upper ISTiagara falls. In front of the main face of Wintergreen cliffs, there is a great assemblage of fallen blocks below the level of Wilson terrace. One of these blocks, though now split in t^vo, is eighty feet long, twenty feet wide above the ground, and eight or nine feet thick. It lies crosswise of the course of the river. It is the fallen edge of the Wilson terrace undermined from the lower side of the river, when the falls were still cascading over it, and it lies only a short distance below its natural position. In this locality are found an immense number of these great blocks, and they show that the Wilson terrace must have ex- tended across the river, being the floor of the channel into which the upper falls cascaded, but located nearer the Canadian side than the present channel of Foster rapids, as shown in Plate XXVI. A and other figures. POT-IIOLES IINT FALEElSr BLOCKS. Here are found pot-holes in fallen blocks, formed benealli the falls, after the blocks had broken loose and were resting on the slopes below, as shown by the holes which now have a vertical position. Three of these pot-holes in one block have a diameter of eighteen inches with a depth of three feet or more. Anotlier 182 FALLS OF NIAGAEA [^«°l- ^urv. one had a diameter of from three to four feet, and a depth of eight feet with the bottom broken out. (Plate xxvii. a.) A littk^ specimen of a pot-hole illustrates their mode of formation. A lead sinker for a fishing- line with a conical shape and a length of three inches and diameter of one inch had been lost in the Upper rapids. When recovered it was found to have made a pot hole only a quarter of an inch larger than its own diameter. The sinker itself was very little reduced in size. It seems that the sand became imbedded in the lead and this acted as a grinding tool to bore out the pot-hole. So it is that the round polished pot-holes are made by the sand being ground between the stones and the rock itself, and these pot-holes are due more to such an abrasion than to direct action of the pebbles, which would soon be ground away. DISTINCTION BETWEEN NIAGARA AND CLINTON BLOCKS. The Clinton blocks can easily be recognized. They are never more than eight or ten feet thick, while those derived from the Niagara cliffs may have a much greater thickness. The Clin- ton blocks have more or less the form of great slabs and can thus be distinguished from the thicker and more cubical liiagara ones. Minute examination, however, would show- other characteristics for distinguishing the two kinds. Upon the Wilson terrace there are a great many fallen blocks from the iS[iagara cliffs. So, also, beyond the remaining terrace among the Clinton masses, even down to the river edge, the ISTiagara blocks occur in great numbers. These show that Wintergreen flat must also have extended much farther toward the middle of the river after the recession of the Upper falls than now appears. NORTHERN SPUR OF WILSON TERRACE AND COVE BEHIND. From under the northeastern point of Wintergreen cliffs the Wilson terrace tapers to a point 200 or 300 feet distant of Canada] TERRACES AT FOSTER FLATS ISi The base of this wedge is about 100 feet across. Towards the point sonic transverse fissures dissect tlie mass. Beyond we tind these surface blocks first separated, though in their natural position, and farther on they are tilted. This feature extends for another 300 or 400 feet. This is the narrow medial ridffe on Foster flats. While some blocks of Niagara limestone occur upon the surface of the spur near its base, yet the fallen blocks of the ridge are those of the thick Clinton band which had been undermined on the removal of the underlying strata. That this ridge is part of the old Clinton floor, of which remnants extend as far southward as Wilson point, there is no question ; but the occurrence of this cove behind the spur of Wilson terrace already mentioned, presents difliculties. The Wilson floor was the bottom of the river receiving the waters of the upper cataract ; Foster flats in the same manner received the falls from the Clinton limestone, but in the cove the evidence of an intermediate water fall from the Clinton bed does not appear. The apex of the cove in the northern side of Wintergreen flat is beyond or outside of the original banks of Xiagara river, where the floor was 316 feet above the lake. Outward of this at a distance of 400 feet the floor is lowered by ten feet (see map Plate xxiv.) ; consequently this cove, heading where the water was shallowest, could not have been made by the current of river. Its origin is uot apparent. This feature is at variance with the physics of the normal falls, and cannot be overlooked. An island surmounted on the ridge, as suggested by Mr. Gilbert, would have had to be in front of the main course of the river, as it was passing the deepest channel, where Wintergreen flat is now located, with the apex of the lateral falls not in the middle, but on the outer edge of the gorge, where the water must have been shallowest. This is a very strong reason for discarding the idea of an island of such extreme narrowness as to be less than its height above the Clinton rocks, besides which the necessary debris of IsTiagara 184 FALLS OF NL-VGARA ^^''°^- ^urv. limestone does not appear. The occurrence could scarcely be expected as it would not be consistent with the Clinton floor of the intermediate falls. The present channel is an encroachment of 400 feet or more upon the eastern side of the gorge. Accordingly, the river must have passed over the transverse spur of Wilson point with the Wilson terrace as the northern spur, for a remnant of the floor is still preserved beneath Wintergreen flat. This con- clusion is confirmed by the features sho\Am. in Plates xxiii.^ sxiv. A and xxvi. a, and xxvi. b, indicating that the outer point of Wintergreen flat was near the middle of the river, which, from the views, is seen to be diverted entirely out of its course to the eastern edge of the gorge. At the spur of Wintergreen flat (310 on map) is a projecting point show- ing the place where, at one time, was the apex of the falls which cascaded to the Clinton floor below, with a lateral shelf, like that of Goat island, over which the waters were still flowing hundreds of feet in advance of the apex. The intermediate falls from the Clinton limestone became rapids in places, over the harder upper layers of the Medina series, as shown by the rising ground at the cove, and also by features well shown at the lower end of Smeaton ravine. These rapids soon passed to the softer layers until the channel reached the Medina gray sandstone band. IIXION OF NIAGARA AND CLINTOX FALLS THEIR HEIGHTS. The Wintergreen flat was the floor of tlio upper falls. The Wilson flat, which was the floor of tlie intermediate cataract, indicates the separation of the two falls, the latter descending from the Clinton limestone band into the river flowing over Foster flats and covering the floor of Medina sandstone. Behind the northern spur, and also above the terrace south of Wilson point, the absence of the Clinton limestone bed of the middle falls shows changed conditions. Plate XXVI. a. \ If u ot (Jliaiinel witli Foster Kapids, pa.ssing around Foster Flats, and under-cutting the eastern side of gorge. Plate XXVI. b. View looking down on >>()rtli Knd ot I'ostcr Flats and River, from outermost point of Wintergreen flat. 185 of Canada] HEIGHT OF CATARACTS 187 AVitli the Wilson terrace abruptly ending, the Clinton and Xiagara cataracts linally united at this point, but they did not excavate more deeply than before their union. At the time of the union the course of the current was obliquely from the Canadian to the jSTew York side, and it is found to have been diverted by the walls of Wintergreen flat and Wilson terrace. A higher buttress nearly opposite 310 on the map, occurs on the N^ew York side. This, however, has not the form of a terrace, but is the sloping edge of the plateau above (Eldridge). This feature of the falls breaking through a barrier and leaving buttresses on both sides is repeated at the present outlet of the Whirlpool (Plate xxvii. b, on page 190), where they have considerably reduced the present breadth of the caiion. The union of the two cascades easily explains the cliffs facing up the river on the southwestern side of Wilson point. These were undermined by the currents sweeping against them and causing the blocks of Clinton limestone to fall out- %vard in a direction opposite to that of the course of the river. The upper cataract had a height of 120 feet measured from the edge of the old channel at Wintergreen flat to Wilson terrace and from this floor to Foster flats, the Clinton or second cataract fell another 120 feet (taking the mean height of the flats which rise somewhat innnediately below). During the epoch of the Clinton falls the volume of the water was only 15 per cent of the 'total present discharge (Chapter xx.). As the retarded river below Foster flats has now a depth of only sixty- three feet the former depth of the water may be taken as not exceeding ten feet or less with free drainage ; lint the barrier at Wilson point may have temporarily increased the depth of the river above, after the addition of the Huron drainage. The distance from the mouth of the gorge to Wilson point, w^hich is the last remnant of the Clinton floor before the union of the two cataracts, is 13,4:00 feet. The united falls now 188 FALLS OF NLVCxAEA ^^^°^- S"^^' receded another 600 feet to the head of Foster flats, when the great change in the physics of the river occurred. The lowering of the Ontario waters on this section had as yet no other effect than that of affording an easy slope for the drainage. THIRD OE MEDIKA CATAKACT CHANNELLING FOSTER FLATS. At Foster flats another record of great change in the features of ISTiagara river is found. With the emergence of the floors of the river, by the lowering of the Ontario level, thus drawing off the water from the Clinton channel, a third fall came into existence, due to the occurrence of the heavy band of Medina gray sandstone. The volume of the river was not yet increased, so that a small channel was sufficient to take the stream with rapid descent. The narrows at Foster rapids have been formed by obstruc- tions of fallen blocks derived mostly from the eastern wall, as the river has been creeping that way and undermining that side (see Plates xxvi. a and b). Some blocks have rolled off the end of Wilson point. This blocking of the channel has oc- curred since the backing of the Ontario waters into the ISTiagara gorge at a later date. INCREASED VOLUME OF THE RIVER. AfterHhe upper united falls had passed Wilson point they descended 240 feet directly to the floor of Medina sandstone, yet they could not penetrate it. So they continued until they reached the head of Foster flats. It is immaterial Iioav far the third fall lagged behind the united upper ones. The flats are 35 feet higher than the river to the south of them, which is 99 feet deep. The modern falls with their present height have an effective erosive force of 80 to 100 feet below the river surface. ISTow it is found that the force of the river was sufiiciently increased to penetrate the original floor to a depth of 135 feet. This actual deepening corresponds to the increased Plate XXVII. a. Large pot-hole, with bottom broken through, formed in block of limestone after it had collapsed beneath the Falls when thej' were at Foster Flats. 189 Plate XXVII. b. View of Thomi)son Point and buttress below, (looking across outlet of Whirlpool). 1!)0 of Canada] TIIIED FALLS 191 mechanical power of the falls, due to increased volume, when the falls were passing this section. The penetration of the flats bj the third cataract seems to have been when the falls were at the Whirlpool, for here the surface is lowered some 30 feet below the level of the flats, but the Whirlpool reaches to an in- creased depth of about this amount, below the floor of the chan- nel outside of the pool, — ^wath the consequent lowering of its surface, so that the falls w^ere here able to excavate to the greater depth, than is now seen further down. The sudden widening of the channel occurs immediately- above the head of Foster flats, as may be seen in the map and in Plate xxv. b, as well as also the deepening observed in the soundings between here and the Whirlpool, where a complete section of the broad U-shaped channel was obtained. The river in front of Foster flats is about 300 feet wide, while just above it attains 900 feet, but beyond this local enlargement the mean width of the river is only 650 feet. The united upper and middle falls together had receded only 600 feet when the great enlargement took place. As there was no increase in the height of the falls, this enlargement could only have been due to a greater volume of water, such as the addition of the Huron dis- charge to the Erie drainage and the Niagara river, which actually occurred. {See Chapter xxvii.j Opposite the upper end of Foster flats, upon the eastern side of the gorge, is found an indentation in the cliff. It extends for about a thousand feet to a rocky promontory, and the general outline suggests a former connexion between it and Wintergreen flat. The conclusion arrived at, concerning these features, is that this is the locality where the increased volume of the river be- came effective, producing falls oblique to the general course of the river, as the waters were cascading into the smaller and more ancient channel which was insufficiently large to take the increased discharge, so that this was carried down and passed over the side of the chasm into the channel below. Such a 192 FALLS OF NIAGARA t^eol. Surv. condition necessarily in time excavated a broader caiion which assumed a normal form bj the recession of the falls. The distance between the upper edges of the caiion is ab- normal, reaching to 1,700 feet. This gave the impression that an extraordinary volume of water had been required to excavate such a great breadth. Below the Whirlpool the old banks of the river were 1,500 feet apart, while the gorge has a breadth of 1,300 feet. Above Wintergreen flat the former channel widens to 1,650 feet, while at Wintergreen point the gorge itself is reduced to 1,200, and farther down its breadth is 1,400 feet. The elliptical form (as shown on the map) is not a usual feature. That the course of the ancient river was over Wilson point is unquestionable. The deeper small channel on the side, with the Medina falls, directed the course of the now enlarged river, with the increased quantity of water pouring into the western side, obliquely, as just mentioned. This naturally crowded the gorge upon the eastern bank, as shown in Plates xxvi. a and b, and accounts for the widened cahon here, beginning to expand out of its natural course into an elliptical form just above the head of Foster flats, and the residual upper terrace of Winter- green. These features, both in the gorge and at its outer edges, establish the position of the falls when they received the in- creased volume of water from Lake Huron. The slope of the banks, in the curve of the river below Foster flats, on the eastern side of the gorge, is much less than in front of them, where there has been more recent undermining of the lower formations. On the Canadian side frost action is shown both above and below Wintergreen flat where the walls have fal'en away from 50 to 100 feet, cutting off abruptly the original river bank which extends from back of the Wliirlpool to a point 1,400 feet from the head of Wintergreen. Below Wintergreen the terrace has also fallen away so that the wall of the gorge is now back of the position of the old shore line. of Canada] INCREASED VOLUME OF RIVER 193 At Wintergreeii, the original banks are well preserved. A few hundred feet farther on the old shore line reappears and recedes from the present brink, showing what was a lagoon or expansion of the river. (See large map.) Upon the revision of the physics of the Whirlpool gorge, no evidence was found to show that there had been an increased volume in the river in this section, but the old banks were strongly cut terraces indicating great width the river, not merely to the Whirlpool, where the western shore has been undermined and fallen away, but also to below Thompson point, and between it and the head of Wintergreen flat. To repeat, the strongest proof of the augmentation of the river at this point is found in the penetration of the former floor of the channel above Foster flats, of which they are a remnant. These have a height of seventy-five feet above Lake Ontario, so that the soundings above them, in the channel 650 feet wide, reaching to fifty-nine fe?t below the lake, show that the increased discharge was able to penetrate 135 feet into the solid rock floor which received the earlier volume, and carried it to the third cataract situated farther down the gorge. The surface of th^ river has been lowered thirty-five feet or more since the receding third fall subsequently penetrated Foster flats. Indeed this depth of penetration corresponds to the height of the falls at this locality, in the changing physics of the river. REVIEW OF FORMER CONCLUSIONS. This revision of the study of Foster flats is made de novo, eleven years after the first writing on the subject. That tbe falls had receded, under a reduced descent of the river and volume of water, to the point was then shown. But between this locality and .the head of Whirlpool rapids difliculties then arose, to which only suggestive hypotheses could be applied. Owing to the lapse of time I was not guided by any precon- ceived theories in the present revision of the work. ISTor did I 13 194 FALLS OF NIAGARA t^^^'^- ^urv. refresh mv memory bv reading my earlier contributions, so that the present interpretation has not been the result of previous impressions. It is satisfactory to me, to find how well the early pioneering hypotheses have withstood the test of time and re- investigation. In 1891 I concluded that the great increase in the volume of the water took place when the falls were at the head of Foster flats, stating that : — ' After passing Foster flats the chasm shows the effect of a greatly increased force, for the gorge is again widened and the terrace below washed away. . . . The magnitude of the erosion indicates an increased discharge which was produced by the turning of the waters of the Huron basin and adding them to the jSTiagara drainage. The effects of the greatly increased volume of water were to widen the chasm and cut away part of Foster platform, but leaving enough to tell their history (page -t66f ). This is one of the discoveries) in the physics of the river, furnishing a time element in com- puting the age of the falls. AVhile three cascades were distinctly recognized, the upper and middle falls were not then known to have united here, for the detailed structure at the flats had not been studied so closely as during the present investigations. The depth of the river at Lewiston was thought to be only ninety-six feet, so that thte great descent of the Medina falls and rapids was not known. In 1888* I discovered that the Huron waters only lately turned into the Erie drainage. This paper was soon followed by others. One of these was an essay on Niagara falls, pub- lished within a fev,' months, by Mr. G. K. Gilbert, in which he has based his studies of the changing physics of the falls mostly upon my work. Two of the most important discoveries, which he adopted, are the change in the Huron drainage, and the posi- tion of the falls when their volume was increased. Writing of t Duration of Niagara Falls, cited before. * Proc. Amer. Ass. Ad. Sci., Vol. XXXVIL, pp. 197-199, 1SS8. of Canada! CHANGES AT FOSTER FLATS 195 these in 1895, lie says: When the falls had receded to just beyond the end of Wintergreen flat (that is the high level ter- race above Foster flats), ' probably at abont this time the whole amount of water in the river was increased in a manner to be considered later.' (p. 222. f) This had been explained by me in 'Duration of ^Niagara Falls' (p. 446). Again he says: 'By the tipping of the basin, the lake (Huron) was made gradually to expand towards the west and south, till the last waters reach- ed the pass at the head of the St. Clair river. Soon afterwards the water ceased flowing through l^orth Bay outlet' (p. 229). This was his form of adoption of the change of the Huron drainage found by me in 1888. A little later Mr. F. B. Taylor:]: worked out an explanation for the Whirlpool gorge. He thought that the waters of Lake Huron turned into Erie as I had set forth, but that they were again withdrawn, owing to changes in ice dam, w^hen the falls were passing the Whirlpool rapids and that later when ISTiagara falls were just south of Cantilever bridge, the full volume of the Huron discharge was again added. Had it not been for the new evidence of the character of the Whirlpool rapids, the pro- blem might remain unsolved. t National Geographic Monograph Series, No. 7, New York. 1895. t Bun. Geol. See. Amer., Vol. IX., pp. 59-84, 1898. 13J CHAPTER XV. TERRACES ABOUT END OF GORGE, AND LAKE EEVEL BELOW THE PRESENT. Roy terrace and lake level at birth Delta of Niagara river and further of falls (287 feet). subsidence of lake. Eldridge flat (200 feet). Lowest level of water in Ontario Bell terrace (174 feet). basin. Iroquois beach (137 feet). Final backing of waters to their Lower terraces. present level. EOY TERRACE AND LAKE AT BIRTH OF FALLS. The main liaiiks of the river at the birth of the falls crossed a ridffe extending to half a mile south of Brock's Monument and Hall point, where the old shore line has been removed bj the falling- of the brink (see large map and figure 24). Beyond this ridge an indistinct water line occurs at a corre- sponding height, showing the temporary level of water in the Ontario basin at about 322 feet above its present level. Xear the point where Brock stream crosses it, about 500 feet inside the escarpment on the western side of the gorge, is a well marked bench which I shall call Boy terrace in honour of Mr. Thomas Roy, who made a study of the Canadian terraces ad- jacent to Lake Ontario as long ago as 1837.* The inner edge of this terrace has an elevation of 287 feet, and 284: feet farther out, while there is a subordinate one near brink of gorge at 275 feet for the inner and 268 feet for the outer edge. This is the flat in front of the shelter of the Electric railway at Brock's Monument. Its maximum breadth may be taken at 300 feet, with the rocky banlvs behind rapidly rising to Brock's Monument forty-two feet higher. The total * Cited in Geology of Canada, 1863. 197 198 FALLS OF .NIAGAKA [Geol. Surv. length of this terrace is 900 feet, but the last 450 feet are be- yond the month of the gorge and are composed of delta deposit of river stones of varions sizes, while the sonthern part is nnder- laid bv rock. The northern side of the delta terrace forms a steep slope to a lower terrace. This marks the floor of the Fig. 24. Map of terraces at end of Niagaia gorge. , (Reduced from large map). river at the time when the falls had their birth. (Plate xxvitt. on opposite page.) On the rocky face of Brock cliff the Roy terrace soon becomes indistinguishable. Thongh this old water level has not been continuously explored it appears to occur at many points along the brow of the Niagara escarpment, where the "bJo^S O ST) sac -co fti 3 ^ Oi C c ^ 12 1 t:^ 3 V 2> ci ^- St3 " S S 3 = S i '-- o - --^ 2i 199 of Canada] HEIGHT OF FAI.LS AT THEIR BIRTU 201 front part of it seems to liave been cnt down by the old lake waves for twenty or thirty feet below the rising ground back of it. Still the recognition is easiest where the waves carved shore lines out of the soft drift material. About half a mile west of the road, from St. David south- ward, an old shore line becomes conspicuous, bounded by low rocky bluffs. At a mile and a half farther on, but still east of the Welland canal, a spur of the Xiagara escarpment stands out prominently above the plain below. The col connecting this outlying spur with the escarpment behind forms the floor of the terrace just to the eastward. Its height is 287 feet. In the region of De Cou falls (^ thirteen miles from Niagara river) at 283 feet, and on the opposite side of the great valley of Twelve-mile creek, the inner side of the terrace, with a rocky bluff behind it, has an elevation suggesting that it is at the same level, as at Roy terrace. The feature is repeated above Jordan, about seventeen miles from Queenston, with apparently the same elevation. But no survey has been made that can give the proper extension of this terrace. Suffice it to say that Roy terrace at the mouth of Xiagara river plainly indicates that the height of Xiagara falls, at their birth, was al)out thirty five feet. EEDRIDOE FEAT. On the eastern side of the mouth ©f the gorge there is a cut terrace about 100 feet wide and 200 feet high, shown on the map and named Eldridge flat. Its preservation is no doubt due to the harder beds of the Medina red sandstones distributed among shales, but its form is that of an old shore line, which represents a pause in the lowering of the Ontario waters. BEEL TERRACE. Bell terrace is the next. On this is located the road from Queenston to St. David and westward. At the crossing of this road over the Electric railway its height is 174 feet. With a breadth of a few hundred feet it extends to St. David, where 202 FALLS OF KLVGAKA ^^^ol- ^urv the average height is 168 feet, although it is somewhat lower at the corner of the hotel. At St. David it widens ont to half a mile or more, back of which are the drift hills in front of the iN^iagara escarpment. Behind Merritton it is more difficult to separate from the topography, but it reappears as a distinct feature at other points, as east of Jordan. I have noticed the same terrace at points ten miles east of Hamilton. It is some- times narrow, rising rapidly to the hills behind, or even cut away at points on the side of the escarpment. Again, it is broad. From ]Sriagara river westward its edge forms a steep bluff behind the old Iroquois shore line, which is thirty-five or forty feet below. The Bell terrace is seen on the one hand rising gently to the l)luffs behind, and on the other descending abruptly to the Iroquois shore line in front, this makes the terrace a strongly marked feature in front of the Niagara escarpment. Adjacent to ^Niagara river, at a few feet below its surface, the Bell terrace is underlaid by red sandstones and shales of the Medina series, and these, in part, no doubt, give rise to the persistency of the prominent features. But it is not everywhere on such a foundation. At St. David, and for some distance westward, the wells, to a depth of sixty feet in sands and gravel, show that the Medina sandstones are wanting, and that the terrace form is due to wave action on superficial deposits. Here is evidence of its being 'the floor in front of an old shore line. The inner edge of this terrace rises more rapidly than that of the Iroquois plain in front, which may be due to the washings from the hills. This old terrace, which I have named in honour of Dr. llobert Bell, who had made most of the studies in this region before 1880,-^ marks a pause of considerable duration in the sinking of the waters in the Ontario basin, and its occurrence, as connected with the study of the recession of the falls, is important. * Geology of Canada, 1863. Report upon Surface Geology. of Canada] IKOQUOIS BEACH 203 IROQUOIS BEACH. Again the waters sunk to the Iroquois shore line, which, at the mouth of the Niagara river, has an elevation of 135 to 137 feet. This is the most conspicuous and perfectly preserved strand ahout Lake Ontario. On the western side of the Niagara canon it is simply a cut terrace in front of the steep hank in front of the Bell terrace; it is carved out of a clayey soil. From this line the old lake hottom almost imperceptihly slopes outward. While the gravel heach characterized it else- where it is entirely wanting here, with the result that the determination of the exact water line is not practicable within the range of two or three feet, as the washes from the hills behind have somewhat modified its inner edge. With this qualification by the use of the level I found it to have a height of 130 feet at Queenston. {See figure 24, page 198, also map Plate XXXIII.) On the New York side, the old shore is more strongly marked than on the western, as the joint action of the river and lake currents have washed away the talus slopes of the ' moun- tainside,' and left a floor of Medina sandstone at 135 feet. In front of this terrace is a steep slope to a much lower one, beyond which a sand and gravel ridge rises at the village of Lewiston, with features such as characterize the Iroquois beach. It is a few hundred feet wide and slopes about twenty feet to the plain in front. At the end of the spur above the river its height is 126 feet, but where it joins the main Iroquois shore it has risen to 137 feet, where the crest of the spit is taken at a slight elevation above the water line. Looking at these features from the western side of the river one would suppose they were two shore lines a few feet apart ; one of sand and gravel, the other of rock. But the former declines gradually to the end, and the slope of the spit, deter- mined instrumentally, indicates that it is one of those spits commonly formed beneath the water level where rivers enter 204 FALLS OF NIAGARA [^«°^- ^^rv. lakes or seas. The well developed Iroquois beach was one of long duration, sufficiently so to allow the building up of gravel ridges often 300 to 500 feet w^ide, and fifteen or twenty feet thick. This beach marked the longest pause in the lowering of the waters of the Ontario basin, while the falls w^ere receding to Foster flats, which, together with other beaches and Smeaton ravine, record the story of the early Niagara falls, THE LOWER TERRACES. Behind the Iroquois spit a lower terrace is seen at seventy-five feet. Farther down the river, two miles below, at Field point, the old river bed and bank are well shown where the floor is composed of gravel overlying shale or qlay deposits. Here the effects of the river -were strongly marked, in channelling out the surface, or depositing bars. The height of the raised river bottom is about forty feet at the upper end, and somewhat less at the lower, with the surface chan- nels increasing in depth. Opposite, on the eastern side, is a cove forming a deep indentation in the generally high bank with an elevation corresponding in height, while the edge, below Field point, is marked by a lower water line of twenty feet or less. At the mouth of the river is a somewhat extensive river flat in front of the town of I^iagara-on-the-Lake. Its height is only three to five feet above the water, behind which is a steep bank. The whole story is not told, as the lake continued to recede much below the present level. But this terrace was made after the lake level rose again from the extreme subsid- ence, but before the recent lowering. DELTA OF NIAGARA RIVER AND FURTHER SUBSIDENCE OF LAKE. It is an undetermined question whether these lower ter- races record only the original subsidence of the lake waters, or whether aftci- the (iriginal shrinking of the lake to a smaller area, with the subsequent rise, the lake surface may not have Pi 205 of Canada] GREAT SUBSIDENCE OF THE LAKE 207 been slightly higher than now, so that the lower terraces mark a second subsidence, due to the cutting down of the St. Law- rence outlet. Lower water lines of the Ontario basin are now drowned. The Niagara river below Lcwiston has been re- filled with river deposits so that its depth is now reduced in places to only thirty or forty feet. jSlear the lake the river is seventy-two feet deep. Beyond the mouth there is a fan- shaped delta extending nearly four and a half miles, at first covered with only twenty to thirty feet of water, but with forty-eight to sixty feet near the edge. Beyond this margin, there is an abrupt descent to 100 or 150 feet, with the lake further increasing in depth beyond. The submergence indi- cated here is a repetition of the features about Burlington bay, at the end of Lake Ontario, which has a depth of seventy- eight feet behind a beach five miles from the end of the lake mentioned long ago by myself." With regard to the delta of the ISTiagara river^ to which Mr. Gilbert called attention as evidence that the lake level was from 100 to 200 feet lower than now, it may be said that as the delta deposits at their very edge rise to within about sixty feet of the surface, and as there are no indentations in the isobaths beyond, they do not of themselves furnish any evidence of greater subsidence of the lake than this amount, although the accumulations may have been heaped up in deeper waters immediately beyond, now covering the topography of the pre-glacial basin of the lake. From other proof a lower water line might have been in- ferred, such as that of the ninety-six foot sounding at Lewiston. Until the present time this depth of the river below the end of the gorge had furnished all the evidence of the former greatest emergence of the Niagara channel. Now, however, the deep sounding to 183 feet affords the first demonstration of the late great subsidence of the lake waters affecting Niagara river. * ' Geology of the Region about the Western End of Lake Ontario, by J. W. Spencer, cited before. 208 FALLS OF NIAGAEA ^Geol. Surv. The shore line represented bv the delta, I have taken as the equivalent of one now raised and tilted at tiie eastern end of the lake, which Dr. Gilbert stated passed under the lake w^aters at Oswego, The lake level, so far as it affected Niagara river, must have sunk t(^ that of the deepest sounding, thus leng-thening the channel by eleven miles and a half, which below the sur- face drift is entirely excavated out of soft Medina shale. The sinking of the waters below the level of Iroquois beach brought into existence the falls from the Medina sandstone, descending 320 feet to the deep river channel jnentioned — a much greater height than was suggested by previous evidence. This is shown by soundings, to have been diminished only thirty-three feet, at a third of a mile within the gorge if indeed the channel here is not partly refilled. LOWEST LEVELS OF WATER IX ONTARIO BASIX. If the deformation of Iroquois beach be considered as a whole it is found that between the head of Lake Ontario and Prospect farm, east of Watertown, the differential rise amount- ed to 367 feet. This w^arping of the land must have affected the rocky rim of the Ontario basin, seventy miles below the outlet of the lake; so that the waters, if the continent stood sufficiently high, could have been lowered another 500 feet. But of this amount only a little more than the depth of the great sounding is necessary to account for the deep channel of the Niagara river unless it be further refilled. FINAL BACKING OF THE WATERS TO THEIR PRESENT LEVEL. After the time when the waters in the Ontario basin shrunk far away from the Niagara shore the tilting of the land at the outlet of Ontario caused the lake to rise again, thereby over- flowing low lands at its head, and submerging part of the Niagara gorge. The final movement was at a late date. Thus of Canada] EISE OF LAKE SURFACE 209 was impeded the free flow of the river with the consequent effect upon its slope. On the return of the waters to what height did they rise ? Some years ago I brought forward evidence suggesting that it rose to about seventy-five feet, as represented in terraces at Queenston, Lewiston, and by certain gravels at the Devils Hole.* It is possible that these last may have fallen from the deposits of upper river gravel ; consequently the evidence here is doubtful. The rounded boulders beneath fallen blocks on Foster flats cannot be accepted as proof. At the outlet of Lake Ontario well marked terraces occur on both sides up to sixty feet or even higher, in keeping with those along the sides of the lower Xiagara river. Whether these shall be found dip- ping as if passing under the lake in going westward, or resting almost horizontally (as is suggested at many points), is un- certain without a complete survey. But the probability seems to be that the terraces of seventy-five feet were formed, while the water was originally sinking in the lake basin. This backing of the water would reduce the descent of the river, with some effects upon the lower rapids. To whatever height the lake rose after its late subsidence, the subsequent fall of the waters has been accomplished by the Saint Lawrence river deepening its channel. • ' Another Episode in the History of Niagara Falls,' Am. Jour. Sci., Vol. VI., pp. 439-450, 1898. 14 CHAPTER XVI. GLACIATION AND DRIFT ADJACENT TO NIAGARA RIVEK. Glaciation in Niagara district. Character of sand ridges. Cliaracter of clayey and stony drift. River deposits with shells. GLACIATION IN NIAGARA DISTRICT. The glaciation, or polished and scratched surfaces of the rocks, establish the character of the country before the glacial period was closed, and sharply defines the work performed by the modern river, which at different places has re-exposed the oJd drainage surface. Thus, in the excavation for the aqueduct pipe of one of the power companies in the Victoria Park, the southern edge of the buried Falls-Chippawa valley was found scratched and polished. Again, on rising from the lower ter- race at the Whirlpool rapids, a few feet to the higher floor the rocky bank of the ancient valley is found polished. This proves that the broad channel at the l^arrows was pre-glacial. So also at the Niagara quarry, on Monroe's farm, and beyond at Harvie fall, the Whirlpool-St. David valley shows its upper edge to have been rounded and polished. At the last mentioned place the direction is S. 10° W. just outside of the trench, the trend of which here is :N'. 20° W. During the summer of 1905 extensive excavations for sewers were being made on Ferry road and on the street be- yond, in ISTiagara Falls Centre, which showed that the upper rocks had been planed off to a remarkably level surface, highly polished and slightly scratched. Here the elevation of the sur- face rock is 354 feet; the direction of the striations is S. 45° W., approximately in the same direction as this section of the 14i 112 212 FALLS OF NIAGARA t*^eol. Surv. river, but as the old valley bends more to the south the course of the scratches becomes oblique to the axis. At the Queenston quarrv, on the mountain point southeast of St. David, the main striation is S. 60° W. with fainter lines S. 60° E. and S. Here were observed some deep grooves. It may be stated that the course of the glaciation rarely coincides with the trend of the ISTiagara escarpment or other surface features, but is oblique to them, often at high angles, thus show- ing that these were not produced bv the action of the ice, which only polished their surfaces. The glaciation is of the highest value in determining which depressions are old and which new. At various points adjacent to the river, now concealed, the po- lished surfaces have been observed ; as for instance at Hubbard point and opposite to it. CHARACTER OF CLAYEY AND STONY DRIFT. The drift along the river above Dufferin islands is composed of clay mingled with small stones and a few boulders. In iSTiagara Falls Centre, at the head of Ferry road, upon digging sewer-trenches, drift twenty-four feet deep was found lying over the polished rock. Here it was a reddish sandy clay with very few pebbles and occasional angular stones of granite and other material. In the bluffs overlooking Victoria Park, rising to 100 feet or more in height, some of the layers are sand, but these occupy the buried valley. The ridge of Lundy Lane trends westward from Drum- inondville (or l^iagara Falls South), for a distance of some- what more than two miles. Xear the monument to the battle of Lundy Lane the high drift ridge is composed of sand and gravel. In a well nearly a mile west, at a point where the surface has an altitude of 425 feet, the depth to rock is ninety feet, mostly through sands and gravel. At well ]S[o. 2, over the buried Whirl])Ool-St. David valley, the surface clay has a thickness of eight feet, quicksand fifteen feet, bluish clay Plate XXX. Stratified sand-beds at Benyiuan's farm. 213 of Canada] CHAKACTEE OF DRIFT 215 fifteen feet, red clay fifteen feet, with sands or angular gravels below, as was described in the deep well 'No. 1 reaching 269 feet. These materials usually have some clay mixture. At well ISTo. 4 the bottom layer of drift resting upon rock consisted of a sort of quicksand of extremely fine texture and angular form, which, with water, easily flowed, but set so quickly that it was difiicult to pump it out of the casing. At the Whirlpool the clays have a thickness of forty feet. CHARACTEE OF THE SAND RIDGES. Speaking generally, the plateau of the Niagara district has a remarkably level or slightly undulating surface covered with red clay soil such as characterizes the drift at ISTiagara Falls Centre, shown in the sewer, or at Well ISTo. 2. The ridge at Lundy Lane is an exceptional feature. A smaller hill of like character occurs south of the post office of South End, but there is still another higher hill — the Berryman — rising high above the immediate brow of the escarpment just west of the outlet of the Whirlpool-St. David trench, where the face of the escarpment is covered by drift hills. From the summit of this hill a declining ridge extends in both directions ; that covering the buried valley disappears within a half mile to the east, being indented with deep valleys. Extensive gravel pits here show the structure of the materials to a depth of fifty feet ; also a well to 100 feet. They are com- posed of sand and gravel with w^ell marked stratification, but in false bedding dipping in both directions : — that is, outward to- wards the low country of Lake Ontario, and inward towards the plateau. This structure is shown in Plate xsx. The pebbles are mostly from one to three inches long and form only a subordinate part of the whole. Many are com- posed of granite and quartzite, but the greater portion are lime- stone or reddish sandstone fragments which have been trans- ported from the Hudson river or the Trenton formations on 216 FAI.LS OF KIAGAEA [Geol. Surv. the northern side of Lake Ontario, with rarely a fragment of Niagara limestone. These deposits overlap the clays which cover the angular gravels of the buried channels. That these accumulations transported across the lake were deposited in water is unquestionable. They even occur to a height of 75 to 100 feet above the summit of the escarpment, so that they could not have been redeposited from otlier drift hills in this region. The transporting ice which supplied the material seems to have been attracted by the pre-glacial valley here, and to have deposited its load in melting, where the currents could stratify it. A repetition of this phenomenon is seen at Font hill, where similar deposits occur building up a higher mound which blocks the head of the ' Short Hills ' valley. (The buried Erigan chan- nel.) The Lundy ridge also is adjacent to the buried Falls-" Chippawa valley, and the sands have considerable depth. EIVER DEPOSITS WITH SHELLS. As pointed out long ago by Lyell and Hall, there are river deposits containing fresh water shells. Hall describes these on Goat island, and Lyell those which occurred in Victoria Park back of the late Cedar island, before the gravel pit was recently obliterated. The same shell deposits occur on the high point just at the outlet of the Whirlpool and elsewhere. They were accumulated in the quieter waters when the river covered these now drained terraces. CHAPTER XVII. LAKE BASINS AND METEOROLOGICAL CONDITIONS AFFECTING THE FALLS. Drainage area. The Erie ratio. Humidity. Mean rainfall and evaporation from Temperature. Lake basins. Velocity of the wind. Modified conditions in the Erie Relationship of evaporation to basin. rainfall. DRAINAGE AREA THE ERIE RATIO. In the investigation of the plijsics of Niagara falls the problems of the fluctuations of the lakes, of the discharge, and of the changing areas of the lake basins supplying the waters of the Niagara are fundamental questions and require presenta- tion here, although the data have been derived from the already collected meteorological notes, some of which consist of tables that will be presented in Appendix iv. Drainage Areas."^ Area of watershed or drainage basin of Lake Superior: Sq. Miles. In Ontario 30,780 In Minnesota 6,800 In Wisconsin 3,160 In Michigan 7,860 Water surface 31,800 80,400 The U. S. Lake Survey gives the total area as 76,100 in- cluding a water surface of 32,100 square miles. • Altitudes in Canada, James White, pp. 182-186, 1901. Also Rept. Engi- neers, U.S.A., Appendix FFF., pp. 2861-2, 1903. 217 218 FALLS OF NIAGARA t^eol. Surv. Area of watershed or drainage basin of Lake Huron: Sq. Miles. In Ontario 35,400 In Michigan 16,700 Water surface 23,200 75,300 This area inchides. Georgian bay 5,G00 I^orth channel 1,600 St. Mary river 150 Sagina\v*baj 1,050 Area of islands 1,700 10,100 Area of watershed or drainage basin of Lake Michigan: On adjacent land 40,200 Water surface 22,300 62,500 Area of watershed or drainage basin of Lake St. Clair and river : In Ontario 4,160 In Michigan 2,160 Water surface 445 6,765 Area of watershed or drainage basin of Lake Erie : In Ontario 5,480 In Ohio 11,950 In Michigan 2,990 In 'New York 2,210 In Indiana 1,270 In Pennsylvania 580 Water surf ace 10,000 34,480 of Canada] AREA OF LAKE BASINS 219 The U. S. Engineers give the total area of the Erie and St, Clair drainage basin at 40,800 square miles, including 10,600 square miles of water surface. Area of M'atcr^hed or drainage basin of Lake Ontario: In Ontario 11,255 In I^ew York 14,275 Water surface 7,450 32,980 From the foregoing tables it will be seen that the total area of the ISTiagara drainage basin is 259,445 square miles (White's tables) or 254,700 square miles (U.S. Lake Survey). The area of the Erie-St. Clair basin is 41,245 square miles (^Vhite), or 40,800 square miles (U.S. Lake Survey). In studying tlie physics of the falls it will be seen that this separation of the Erie drainage from that of the total amount of the Upper lakes is necessary, for at one time the !N^iagara river received only the water from the Erie basin, and part of that from the St. Clair. Thus the drainage area of Erie compared with the entire area drained by ISTiagara is sixteen per cent, or about one-sixth of the whole region. MEAN xVNNUAL EAINFALL AND EVAPORATION IN THE LAKE BASINS. The mean annual rainfall of the different lakes has been computed for the period between the years 1882-1905 ; and both the mean annual and mean monthly rainfall will be found in Api>endix iv." The average rainfall has been found to be in the basin of: — *Rept. Lake Survey, 1903, pp. 2878-9; for data since 1898 Annual Rept. of Weather Bureau. 220 FALLS OF NIAGARA [Geol. Surv. 1882-98. t 1882-05. Lake Superior. . . . 126.27 inches. i 27. 26 inches. Huron & Michigan 32.12 St. Clair-Ene |34.0S Ontario 36.87 32.00 34.46 36.87 1882-90. 26.17 inches. 33.06 ., a 34.12 M 37.31 ., 1891-00. 1901-05. 27 . 06 inches. 29 . 64 inches. 31.04 „ 6 32.40 „ c 33.75 , [36.23 ,, 36.00 M *38.20 M Rainfjjll and the evaporation based on the discharge (1882- 1898) in terms of cubic feet of discharge per second: — For Superior basin . Michigan-Huron m . St. Clair-Erie Ontario m . Rainfall. t 147,164 32.5,857 102,308 89,557 Evaporation. 69,954 (sf). 203,831 .t . 77,820 .- . 57,507 " . Evaporation. 13.75 (inches.) 20.29 26.10 23.67 (s-f means cubic feet per second.) The evaporation in inches is corrected for the fall of lake level at the close of the period of observation. The first period of rainfall given above is adopted from the Lake Survey. The other groups of years are selected on account of surveys of the recession of the falls having been made in 1890 and 1905, and the renewed rainfall since 1900. The earlier re- cords are too incomplete at certain points to satisfactorily extend the table. In fact, one should expect that the average rainfall in the Superior basin errs on the side of being too high, on account of the diminished precipitation north of Lake Superior, where there are only a few stations, some of which have not been organized many years. For instance, at Xipigon, a large river discharges into Lake Superior, here it is 20 -24 inches^ while to the northeast, at Martin :8alls, it is reduced to 14- 16 inches. The same feature occurs north of Lake Huron. t Engineers' Rept., 1903, p. 2860. o, b, c. Corrected by adding Huron rainfall in the means of the two Lakes. % Engineers' Rept., 1903, pp. 2860-1. of Canada] RATIO OF ERIE BASIN 221 Accordingly, the future revision of the rainfall may somewhat reduce the average discharges of the basins of Superior and Huron as compared with those of Michigan and Erie. This question also has a bearing upon the recession of Xiagara falls. From the above table it will be seen that the average rainfall in the different basins varies, but this alone does not account for the reduced height of the lakes since 1900. MODIFIED CONDITIONS IN ERIE BASIN. Although the drainage area of the higher lakes is larger than that of Erie, the proportional rainfall is less. The pro- portion of the Erie precipitation to the whole amount in the l,ake region, exclusive of Lake Ontario, is 17-7 per cent, or slightly greater than that of the Erie area compared with the whole, which is 16 per cent. The determination of these pro- portions will be found to have a direct bearing on the physical changes of Xiagara falls. The evaporation depends upon the humidity, temperature, wind, the proportion of lake surface compared with the whole drainage basin, and whether the lakes themselves have varied in size. It has been found that Lake Erie at its birth was of very small size. It had an area of 1,000 to 1,500 square miles compared with 10,000 square miles at the present day. Conse- quently the difference would have to be treated as a land area suffering less evaporation in former times. Humidity. Upon investigating the question of humidity the average amount from 18S2 to 1898 was found to be, for — Lake Erie 73-6 per cent. Huron and Michigan 76 "4 " Superior 76-4 " Ontario 74-9 " In Appendix iv., the monthly humidities are also given for Lake Erie. 222 FALLS OF ^'lAGAEA [^G^ol. Surv. Temperature. Evaporation depends upon the temj)eratiire, so it is given, for— Lake Erie (1882-'98) was 48-0 F Superior (18S2-'98) 35-9 F Huron and Michigan 42-08 F Ontario 441 F See Appendix iv. Velocity of Wind. For the same period the average velocity of the wind, for — Lake Erie 10-4 Superior 9 -05 Huron-Michigan 10-3 Ontario 10-7 See Appendix iv. Belationsliip of Evaporation to Rainfall. The question arises : — Can the amount of evaporation be determined from meteorological conditions alone ? On this subject, Prof. Alfred J. Henry and Prof. ISTorman B. Conger have made some calculations for various points on all lakes. But different stations vary greatly, with the evaporation on the southwestern side of the lakes being much greater than on the northern or eastern sides. The results of these calculations based upon the shore stations are too high for mid lake, because the determining temperature of evaporation (temperature of the surface waters) is, at some seasons of the year, from 10° to 20° higher than obtains in mid lake." Thus no valuable information throwing light upon this problem from the calculations of the shore stations is obtain- able. The differential character of the meteorological pheno- mena is of value in formino: impressions of changing conditions at the same stations in different localities. * Weather Bureau Bulletin No. 213, p. 22, 1899. CHAPTER XVIII. FLUCTUATIONS OF THE LAKES. Lake Erie. Xote on positions of Changes of level of Huron affect- observation, ing that of Erie. Mean quinquennial fluctuations of Fluctuations of Erie and Ontario Erie, 1850-1905. compared. Fluctuations of Erie and Huron Mean quinquennial fluctuaiion of compared. Ontario, 1854-1905. Lovi^ering of Lake Huron. Fluctuations of Lake Ontario and Drainage by Chicago canal. Saint Lawrence river compared. LAKE ERIE KOTE ON POSITIOIs'S OF OBSERVATION. The fluctuations of the lakes are primarily dependent upon the meteorological conditions such as those already mentioned. But in times past the more wooded portions of the country of the lake region must have modified the fluctuations, with possibly other causes not considered. The data covers fifty years or more of continuous observations^ with fragmentary in- formation recorded many years still farther back. Firsts they throw light upon the variable discharge; secondly, upon the present stability of the JSTiagara region, so far as dift'erential movements of the earth's crust are concerned ; and thirdly, the deepening of the outlets of the lakes, kc. The two most important points about Lake Erie for observ- ing the fluctuations are at the intake of the Welland canal, Port Colborne," and at Cleveland. f Observations at Bufi^alo * From the records obtained through the kindness of Mr. J. L. Weller, Resident Engineer, Welland canal, and of Mr. James White, Geographer, Interior Department, Ottawa, and Mr. Butler, Deputy Minister of Canals, Ottawa. t The fluctuations at Cleveland are from Report of U.S. Engineers' Lake Survey for 1903, 1904 and 1905. 223 224 FAI.LS OF NIAGARA f^^^^- S^'"^" and at the western end of the hike show greater variations than at Clevehmd, where the waters are less likely to be piled up or lowered bj the winds. Besides this, Cleveland and Port Col- borne are along the line showing the greatest amount of former differential change of level indicated by the raised beaches. Around the head of Lake Erie as far as Cleveland there has been very little unequal elevation in the earth movements, while from Cleveland northeastward an unequal elevation of the deserted beaches is strongly marked by a rise of 120 feet in a distance of IG-i miles to Fonthill, Ontario; or of 100 feet in a distance of 14-4 miles to Sheridan, IST.Y. This last distance is in a more easterly direction. Indeed, nearly all of the rise occurs east- ward of Madison, Ohio, which is forty miles east of Cleveland. From the daily fluctuations at Port Colborne, in its relation to the sill of Lock ISTo. 27 of the Welland canal, the monthly and annual averages of the lake have been calculated since 1849. So, also, the same information has been obtained with regard to the fluctuations at Cleveland since 1854. Authentic continuous records at Buffalo date back to only 1887. From these tables the average daily fluctuations during periods of five years, fifteen years, and other groups of years have been com- piled in order to compare them with the surveys of the reces-" sion of Xiagara falls. For the mean height of Lake Erie at Buffalo, before con- tinuous observations were luade, an approximate result can be obtained by adding 0-12 to the levels at Cleveland. These mean levels are compiled from U. S. Engineers' tables without deduction of -33 of a foot required by the latest precise levelling to sea level datum. The necessary correction has been made for Port Colborne. of Canada] PLUCTUATIOA'S OF LAKE ERIE 225 MEAN QUINQUENNIAL FLUCTUATIONS OF ERIE. 1850-55. 1855-59 1856-60. 1861-65. 1866-70 , 1871-75. 1861-75 . 1800-75. 1876-80. 1881-85 . 1886-90. 1876-90. 1891-95. 1896-00. 1901 05. 1877-05 1855-05 1891-05 Fluct.\tioxs of L.\kk Erie .\t Pt. Colborne. Cleveland. Pt. Colborne below Cleveland. 572 73* 573 26 573-55 - -29 573-40 573-63 - -23 573 25 573-18 + -07 572-60 572-66 - -06 572-28 572 42 - 14 572-71 572 SO 572-75 572-85 - 04 - 05 572 -75 573 02 - -27 572-92 573-18 - -26 572-80 572 93 - 13 572-82 571-66 571-57 571-76 572-20 572-49 571-60 573-04 571-92 571-86 572 04 572-46 .572-67 571-94 - 22 - 26 - -29 - -28 - -26 - 18 - .28 The Buffalo record for the short period shows differential fluctuations similar to those at Port Colborne, when compared with Cleveland, As the monthly and annual fluctuations are * Elevations refer to mean tide at New York. From the last determination of lake levels by precise levelling, a difference of 108 foot was found between Canadian altitudes and those of engineers of U..S.A. Of this amount, 0-75 of a foot were required to be added to Canadian tables, which has here been done; and 0"33 of a foot taken from U.S. levels of Lakes Erie and Ontario, but owing to the small amount this latter has not been deducted. If corrected, the differences between the levels, at the two points, would be very small, but would show a con.stant change from i)lus ( + ) to minus ( - ) signs. (See James White's Dictionary of Altitudes, p. viii, 1903.) 15 226 FALI>S OF XIAGAKA [Geol. Surv. given in Appendix v., the investigator will derive fuller in- formation by comparing special years, eitlier for tlie discharge of Xiagara river at any particular time, the lowering of the lakes, or the differential fluctuations in regard to their significance. FLUCTUATIONS OF ERIE AND HUEON COMPARED. Upon the fluctuations of Lake Huron depend the discharge of St. Clair river or outlet of the lake. From the differential fluctuations of the two lakes the question of the lowering of the Huron outlet is partly determined ; but the problem of the stability of the region is not assisted by a comparison of these lake levels, as any rise in the upper lake would be indetermin- able on account of the lowering of the outlet. In Appendix v., the mean monthly and annual fluctuations of Lake LIuron are given. Only the annual fluctuations of Lake Superior are given in table in next chapter, but these are not directly connected with the recession of the falls. On account of the lowering of the lake, outlets the mean level (1891-1905) should be adopted as the standard height, though the general average has been followed, where not other- wise expressed. The correction of ■ — ^0-33 has not been made in the following table as it occurs in both columns. MEAN QUINQUENNIAL FLUCTUATIONS OF HURON AND ERIE. Fluctiiation.s. 185.5-60.. 1861-65.. 1866 70.. 1871-75.. 187G-80 . 1881-85.. 1886- yo . 1891-95.. 1896-00. . 1901-05.. 1855-1889 1890-1905 1855-1905 Sand Beach. (L. Huron). Cleveland. (L. Erie.) 582 582 581 581 582 582 581 580 580 580 582 580 581 •58 573 • •25 573 •41 572- •67 572 • •04 573 • •42 573 • •9.3 .572 • •4.3 571 • •17 57r •54 572- •08 573 • •42 572 • •55 572 • L. Erie below L. Huion. -9 -9 -8 -9 -9 -9 -9 -8 -8 -8 of Canada] FLUCTUATIOXS OF LAKE ERIE 227 In eoiinoxiou with tlii^^ tomporarv fall of Lake Ilnron it should be stated that the canal across St. Clair flats was com- menced in ISGG and opened for navigation in 1871.* During these operations a thirteen foot straight canal was excavated across shallows of six feet (see figure 2G, Chapter xxvi.). This may have occasioned the temporary fall of Lake Huron, which in years corresponded to those of the princi])al dredgings. After 1886, and completed by 1902, the canal was further deepened to eighteen feet, and since then to twenty or twenty-one feet. But on turning to the tables, we find the most remarkable lower- ing of the other lakes as having occurred just after 1889 and continuing to the present time. LOWEEIXa OF LAKE IIUROX. If the mean annual differences be taken in place of the quinquennial, as in this chapter {see Appendix v.), the re- markable lowering of Lake Huron is seen to have commenced in 1889, reaching almost its maximum in 1890; in 1891 the lake rose somewhat with subsequent minor fluctuations, but the general effect has been to lower it •66 of a foot more than Lake Erie. On the other hand, from 189-1 to 1905 inclusive, Lake Superior had risen over -ll of a foot when compared with the previous elevations between 1860 and 1893. This lowering of Huron was shown by Mr. Russell, to whose writings I am indebted. He says: — 'The mean of 1873- 1892 indicates an increase of 11,355 cubic feet per second in the datum discharge of St. Clair river. The change in dis- charge in one foot of rise in the surface of the river is 14,21 < cubic feet per second. A lowering of the body of the river by almost -80 of a foot is therefore indicated since 1893 as com- pared with the twenty years preceding' (page 4118, 1904). On turning to the rainfall tables in Appendix iv., it will l)e * Appendix 00., Rept. of Engineers, U.S.A., 1905, pp. 598. 228 FALLS OF NIAGARA ^^^o^' S"»'^- seen that the mean annual rainfall of Huron-Michigan basin has diminished bj more than 1-5 inches during the years 1891- 1905, when compared with the period from 1882-1890; while the rainfall of both Superior and Erie has increased. This will not account for the fall in the Huron basin, but as the period under consideration embraces a number of years its effect on Lake Erie would have produced proportional results during most of the time. Even considering the fluctuations it is imjiossible to account for the differential subsiding of Huron other than by a lowering of the outlet, probably somewhat affected by the excavation of St. Clair canal, and to somq degree by the Chicago drainage canal, but mostly by scouring at the bottom of the channel. A closer analysis might shpw a further slight deepening of the St. Clair outlet. My calcula- tions amounting to 0-66 of a foot include the years 190-i and 1905 of higher w^ater, not entering into Mr. Russell's calcula- tion (0*80 of a foot lowered). Doubtless this difference and even more occurs where the data do not permit of more precise calculations. This does not cover the full amount of lowering of the lake, but to this must be added that of Lake Erie. (See next chapter.) DRAINAGE BY CHICAGO CANAL. In this connexion it should be stated that the full allowance of the canal (10,000 cubic feet per second) should theoretically lower Lake Huron 0*521 of a foot; Lake St. Clair 0*455 of a foot; and Lake Erie '379 of a foot, requiring 4*86 years to effect nine-tenths of this change. The canal was opened Janu- ary, 1900. During 1901 the average flow was 4,270 cubic feet. During 1905 the canal withdrew about 5,000 cubic feet per second. CHANGES IN THE HURON LEVEL AFFECTING THAT OF ERIE. The rise of the lakes is not simultaneous, and if this amounts to one foot in Lake Huron it would eventually raise of Canada] FLUCTITATTONS OF LAKE ERIE 229 Lake Erie bj -727 of a foot. It would require seventy-two days to accomplish five-tenths of this rise in Lake Erie, and for nine-tenths it would take 239 days.f A rise of one foot in Huron corresponds to 0-602 of a foot in St. Clair. A rise of one foot in IJrie produces 0-346 of a foot of back water effects in Lake Huron. From these facts it will be seen that a com- parison of the lake levels for single years might produce large errors, and only observations extending over several years can be relied upon. In former times the enormous changes of level in the On- tario basin have played a fundamental part in the recession of Niagara falls. Some of the later changes occur from the cutting down or scouring of the outlets. These have a measurable amount. FLUCTUATIONS OF ERIE AND ONTARIO COMPARED. The changes of Niagara river closely correspond to those of the levels of Lake Erie. The surface of the river above the First cascade of the Upper rapids is about fourteen feet lower than the lake. The differential changes of level between Lake Erie and Lake Ontario are shown in the following table. MEAN QUINQUENNIAL FLUCTUATIONS OF ERIE AND ONTARIO. Fluctuations. Lake Erie Lake Ontario Ontario, below at Pt. Colborne. at Toronto. Erie. 1854-60 573 23 246 78 - 326-45 1861-65 573-25 246-55 - 326-70 1866-70 572-60 246-10 - 326-50 1871 75 572-28 245-40 - 320 88 1876-80 572-75 245-93 - 326-82 1881-85 572-92 246-22 - .'526-70 1886-90 572 -8C 571-66 246-41 245 -06 - 326-39 1891-95 - 326-60 1896-00 571-57 244-68 - 326-89 1900-05 571-76 245-49 - 326-27 1854-90 572-83 246-24 - 326-59 1891-05 571-66 245 -07 - 326-59 1855-05 572-48 245-80 - 326-62 t Thomas Russell, in Appendix EEE., Kept. Lake Survey, U.S. Engineers, 1904, p. 4131. 230 FALLS OF KIACtARA [Geol. Surv. AVhile the mean elevation of Lake Ontario during fifty years is given, that of 1891 to 1905 is now the more accurate, as it allows for the lowering of the outlet after 1890. MEAN QUIKQUENI^IAL FLUCTUATIONS OF ONTARIO. Daily records of lake level have been kept at Toronto, Os- wego, and Charlotte for periods of over fifty years. At other points the information is more fragmentary, and it is to be regretted that the data over a long period cannot be obtained at the outlet of the lake itself. Important conclusions as to fluc- tuations can be arrived at by an analysis of the records of the three points named, and at the head of the first of the Saint Lawrence canals, sixty-six miles below the outlet of the lake, where the river surface is lowered a little more than one foot only. This last point will be discussed in a succeeding paragraph. QUINQUENNIALf FLUCTUATIONS AT TORONTO, OSWEGO AND CHARLOTTE*. Years. Toronto. Oswego above Toronto. Oswego.* Charlotte.* Charlotte above Toi-onto. 1856-00 246-89 240-55 2-10-10 245-40 245 93 240 22 240-41 245-00 244 08 245-49 245-86 240-24 "245-07 -f -28 -+- -75 -f -51 -f- -43 + -46 -t- -44 4- -42 + 5.S -f- -45 247 -17 247-30 240 01 245-83 240 -.39 240-60 246-83 245 -.'19 245 13 247 23 247 24 246-88 245-91 240-56 246 09 240-74 245-01 24-) -14 245-80 240 38 + -34 1861-05 1866-70 -1- -09 + -78 1871-75 1870-80 + -57 -1- 03 1881-85 + -47 1880-90 1891-95 1890-00 1901 05 + -33 + -55 -t- -46 + -39 1855 05 + -52 1854-90 1850-00 1891 05 + ■•47' 245-52 + -45 1 f Mean annual fluctuations will appear in Appendix V. Table 5. * The correction by precise levelling to Greenbush, N.Y., has been applied to Toronto levels, but not to those of Oswego and Charlotte, which would reduce them by0-33of a foot, thus diminishing the unavoidable difference in determining the heights at different stations. of Canada] FLUCTUATIONS OF LAKE ONTARIO 231 The variability of levels between Toronto and Osweg-o is remarkably small. Since 18G0 the maxinnini (juinquennial amount has been only 0-11 of a foot, though some mean annual fluctuations are greater. The uicau difference of the reading in these years is '47. With the difference for the (luinqucnnial ])Oi'iod eudiut;,- in 1870 at +'51, and that of 1900 at +'45, if there be any change at all it would suggest that Oswego had moved very slightly in one direction ; while on the other hand if the former period be compared with that ending in 1895 (-|-53), the opposite movement might be inferred. Thus only one conclu- sion can be arrived at, namely: that the small variation found in the levels at the two points is "svithin the range of mean arithmetical error and is of no value. It is remarkable how small the variations are found to be when spread over any con- siderable period. Comparing the levels at Charlotte with the other two points the fluctuations are found to be slightly greater ; but this station is situated at the mouth of the Genesee river, which might give rise to temporary higher water. On observing the fluctuations of the periods ending in 1860 and in 1865, the variations from the average ( +-47) are found to be exaggerated on both sides. Whether this arises from wind, less perfect observations, or other conditions, has not been determined. Finally, the mean fluctuation of Lake Ontario at the different points mentioned may be taken as a remarkably constant figure. FLUCTUATIONS OF LAKE ONTARIO AND SAINT LAWRENCE RIVEK COMPARED. Having found the constancy in the fluctuation at points on Lake Ontario, the question arises, first, what is the ratio in the fluctuation between Lake Ontario and Calops rapids, sixty- six miles beyond the outlet of the lake, at a point above that where the river begins to descend. Records of the fluctuations have here been kept at old Lock No. 27 of St. Lawrence canals. 232 FALLS OF NIAGARA [Geol. Surv. A comparison of tliese points slioiild throw liglit on the question of changes of level of land. Application of these data will be found in Chapters xix., XX., XXI. and xxx. From this table it may be seen that the Galops rapids can be taken as another point for comparing the fluctuations of Lake Ontario. The period of observation is here shorter, cover- ing only thirty years, and the mean fall is 1 "10 feet. MEAN QUINQUENNIAL FLUCTUATIONS AT TORONTO, LOCK 27 GALOPS RAPIDS, LOCK 15 CORNWALL, AND LOCK 14 VALLEYFIELD. Years. Lake at Toronto 246-55 246-10 245-40 245-93 246-22 246-41 245 06 244-68 245-49 246-02 246-18 245 08 244-87 246 -10 Lock 27. Lock 27 below Ontario. Lock 15. Lock 15 below Ontario. Lock 14. 155 08 154-44 153-97 154 26 154-27 154-48 153-91 153-38 153-56 154-50 154-33 153-62 153-47 154-42 Lock 14 below Ontario. 1861-05 -91-47 1866-'70 -91 -91 -91 -91 -91 -91 -91 -91 66 1871-75 ls76-'80 1881- "85 1886-'90 1891-'95 1896-'00 1901- '05 244-35' 244-86 245 10 243-89 243-53 244 21 -i-58 -1-3G -1-31 -1-17 - 1 - 15 -1-28 157-58 157 -06 157-25 157-69 156-61 155-92 -87-82 -88 87 -88-97 -88-72 -88-45 -88-76 23 67 95 93 15 30 93 1861-75 1876-'90 1891-05 244-77 243-88 243-88 -1-41 -1 20 -1-16 157-33 -88-85 -91-85 -91-46 1891-'00 1861-'90 156-26 -88-6 -91-40 -91-6S ' CHAPTER XIX. LOWERING OF THE LAKE OUTLETS. stability of the outlet of Lake Lowering of the outlet of Ontario. Superior. Effects of lowering of Ontario on Table of mean annual fluctuations higher lakes. of Erie, Huron and Superior. Effects of the lowering of the lakes Lowering of the Huron outlet. on the canals and harbours. Table of mean annua] fluctuations Corrected elevations of the Great of Erie and Ontario. lakes. Lowering of the Erie outlet at the same rate as that of Ontario. STABILITY OF THE OUTLET OF LAKE SUPERIOR. Tables of the quinquennial fluctuations of the lakes have been given, but in order to further investigate the question of the lowering of the outlets the mean annual fluctuations of the lakes are found in the following table, — for Superior at Sault Ste. Marie, for Lake Huron at Sand Beach near the outlet of the lake, and for Lake Erie at Cleveland. The fluctuations of Lake Superior since 1893 show an average increased height amounting to 0-41 of a foot as com- pared with the years 1860-'93. This is an absolute rise of the water surface, in excess of any lowering of the outlet over the crystalline rocks, the data for determining which are not at hand, if indeed it be measurable. The rise is in conformity with the mean rainfall, which increased by 1 "Y inches, between 1891- 1905, compared with the mean of the period of 1882-'90. Of itself this evidence is not conclusive as to the present constancy of the outlet of Lake Superior, but it sufficiently separates the question from that concerning the lowering of the Huron outlet, so that the scouring of the outlet of Lake Superior may be considered so slow as to be immeasurable. 233 234 FALLS OF NIAGARA [Geol. Surv. TABLE OF MEAN ANNUAL FLUCTUATIONS OF ERIE, HURON AND SUPERIOR. Years. Lake Erie. Cleveland. j Sand Beach 1 above ' Cleveland. Lake Huron. Sand Beach. Lake Superior. Sault Ste. Marie. 1855 1856 1357 1858 1859 1860 1855-60. 1861 1862 1863 1864 1865 1861-65 1866 1867 1868 1869 1870 1866-70 1871 1872 1873 1874 1875 1871-75 1876 1877 1878 1879 1880 1876-80 1881 1882 1883 1884 1885 1881-85. 573 572 573 574 574 573 573-54 573-58 573-69 573-40 572 79 572-44 573 18 -68 581- -07 581- -23 582- -87 583- •02 583- -37 582- 904 -78 -95 -55 09 -28 86 582 58 -28 -07 -94 05 00 582 582 582 581 581 907 582-25 .572-58 572-60 572-23 572-65 573-28 572-66 572 571 572 572 .572 .572-41 8-54 8-96 8-84 8-54 8-82 8 75 581 581 581 581 582 -12 -56 07 -19 10 .581-41 -69 9- ■73 9- 43 9- -94 8- ■28 9- -46 -37 08 -97 -42 582 581 581 581 581 -15 -10 51 -91 -70 573 02 572 573 573 573 573 -61 -4S 26 33 -24 573 18 9-26 581- 9-05 582- 9-58 582- 8-92 582- 8-80 .581- 8-72 581- •74 -45 -20 -32 •49 602-53 602 602 601 601 602 •58 -30 -94 ■62 03 602 -09 602 602 602 602 602 02 24 ■05 ■38 •07 602 15 601 601 602 602 602 601-96 002-52 602 09 001-61 600 96 601-41 9-24 582-42 601 68 of Canada] T^OWEKIKG OF LAKE OUTLETS 235 TABLE OF MEAN ANNUAL FLUCTUATIONS OF ERIE, HURON AND SUPERIOR— Concluded. Years. Lake Erie. Cleveland. Sand Beach above Cleveland. Lake Huron. Sand Beach. Lake Siijierior. Sault Ste. Marie. 1886 1887 1888 1889 1890 ... 188f>-90... 1891 1892 1893 1894 1895 ... 1891-95... 1896 .... 1897 .... 1898 . .. 1899 1900 1896-1900. 1901 1002 1903 1904 1905 1901-05... 573 35 9-73 583 -08 601-51 573 29 9 13 582-42 601-44 572 00 918 581-78 601 73 572-37 8-91 581-28 601 77 573 05 8-06 581-11 601 62 572-93 9 -00 581-93 601-61 572 15 8-39 580-54 601-24 572 13 8-24 580-37 601-19 572 08 8-52 580 -60 601 -.51 572-09 8-67 580-76 002-14 571-17 8-73 579-90 580-43 602 29 571-92 8-51 601-67 571-39 8-20 579-59 602 15 571-96 8-24 580-20 602 20 572-13 8-22 580-35 601-77 571-90 8-46 580-36 602-26 571-94 8-43 580-37 580-17 602-34 571-86 571-38 8-31 602 14 9-25 .580-63 602-37 571-84 8-37 580-21 602 31 572-37 7 99 ,^80-36 602 -49t 572-45 8-39 580-84 602 -71t 572 16 8-60 580-95 602 •34t 572-04 8-50 580-59 602 44 t Marquette. LOWERING OF THE OUTLET OF LAKE HURON. The Huron outlet has been partly considered (page 227), with the apparent sudden fall of the lake surface^ commencing in 1889^ which has remained lower since that time. The average lower water of Huron subsequent to that date com- pared with the mean level before amounts to 1-06 feet. If, however, the subsidence of Huron waters be compared with the "fluctuations at Cleveland it is found that the lake there has also 236 FALLS OF XIAGAEA [Geol. Surv. fallen one foot. As this has been continuous over a period of fifteen years the mean difference, which amounts to 0-66 of a foot, is not a temporary oscillation of level such as may be observed when even individual years are compared. From the evidence at hand, I, therefore, agree with Mr. Russell that this must be a permanent low^ering of the Huron outlet. The records of fluctuations before 1855 are too incoiuplete to be able to form definite conclusions from theiu. Between January, 1846, and August, 1851, only a few scattered monthlv records of the levels of Lake Erie are obtainable. Between January, 181:6, and August, 1849, there is a full monthly record for Huron — the mean height during this time was 580-44 feet. The Erie record for the first eight months of 1846 gives a mean elevation of 571 "49 feet, while that of Lake Huron is 581 '00 feet. Accordingly the difference of level between Erie and Huron was 9 "51 feet, which is greater than the average of any quinquennial period since 1854. This fragmental record goes to show that, as far back as 1846, the level of Lake Huron had not fallen as compared with Lake Erie. Again, there is a record from May to October (inclusive) for the year 1840, but this shows a relatively low^er level for Lake Huron, while dur- ing the high water of June and July, 1838, there was an in- creased differential height for the lake. So also a still further increased height for Huron is found in June, 1819, when the difference was 10-40 feet. Individual months or even seasons taken separately are of little value, for, as has already been show^n, any considerable sudden rise in Lake Huron will re- quire from months to even years to be equalized in Lake Erie. From the incomplete records now obtainable back to 1819*, there is nothing to show that for any considerable period the low differential elevation of Huron had occurred prior to the period between 1890-1905. * Compiled In Report of U. S. Engineers on the Lake Surveys. of Canada] LOWERING OF LAKE OUTLETS 237 Although the greater rise of Superior began in 1894, yet on account of the general subsidence of the other lak.es occur- ring just after 1890, this latter date may be taken as the break in the records without involving any considerable error. Be- tween 1861 and 1890 the mean height of Lake Superior -was 601 '67 feet above tide, while since that time its height has risen to 602 -08 feet. At the same time Lake Huron fell from 581-95 feet to 580-38 feet (above sea level), or an absolute fall of 1-57 feet; and Lake Erie subsided from 572-90 to 571-94: feet, or 0-96 of a foot. During 1891-1905 the mean annual rainfall of Superior basin increased 1-75 inches, and that of the Erie basin rose 0-46 of an inch. On the other hand the rainfall of the Huron-Michigan basin diminished by 1-57 inches. So far as the discharges of the rivers are concerned the recorded diminution of the Michigan-Huron rainfall is more than half compensated for by the greater outflow from Lake Superior and the slight rise in the Erie basin. Consequently the reduced overflow of Lake Huron^ due to the change of rainfall, should come within the limits of 1,400 to 1,800 cubic feet per second ; which would also affect the discharge of Lake Erie. Evidences of the differential lowering of Lake Huron have been given (page 227). This is now further confirmed by showing the small effects produced by the changes of rainfall, as also by comparison of the outlet of Lake Superior. Changes due to evaporation are not determinable from the data, in which no important modifications are recognizable. The fluctuations show that the principal lowering of the outlet of Lake Huron occurred between 1889 and 1902, — measurable to the extent of 0-6 to 0-66, or even 0-80 of a foot according to the years used in the calculations. But further slight changes might also be found in more detailed studies of other years. In addition to the differential lowering of Lake Huron whatever subsidence 23S FALLS OF .\L\GARA ^^^°^- ^^'''''■ has occurred in Lake Erie must be added to tliat of Lake Huron. About one foot is found to have obtained there in the same years. A small portion of this lowering only is attributable to meteorological changes. (See lowering of the Ontario level.) As the positions of the gauges for recording the lake levels are stationary, a fall in the gauge readings might indicate a rise of tlie land and not a lowering of the outlet. In the case of the Huron outlet the materials composed of silt or sand would be immediately scoured out by any increased height of the water, or the same materials would be deposited at the bottom of the channel by a slight lowering of tbe river level. On the oilier hand, the outlet of Lake Erie can be affected in a twofold luanner. The great rim which determines its level is the limestone ledge at the First cataract of the L'pper rapids. But at the outlet of Lake Erie there is a slightly higher barrier which has continued to exist, retarding the waters in the upper river. This feature is well shown between Fort Erie and Black Eock. While at the western end of the International bridge the barrier is composed of rock, yet in the centre and at the eastern end such is replaced by clay deposits. Accordingly a lowering of the lake could take place here in a similar manner to that due to a scouring of the outlet of Lake Huron. At a mile and a half southward, however, a rocky barrier seems to cross the river at a depth of seventeen to twentj'-four feet, so that this is at the same plane as the rim above the Upper rapids. In chapter xxx., it will bo shown that at the present time, compared with fifty years ago, there has been absolutely no change in the water level due to earth movements, so that the fall of Lake Erie is mainlv due to a lowerinii' of the outlet. of Canada] T,OWEKING OF LAKE OUTLETS 239 TABLE OF MEAN ANNUAL FLUCTUATIONS OF ERIE AND ONTARIO. Years. Tm-onto. Pt. Colborne. Toronto below Pt. Colborne. 1856 1857 1858 1859 1860 246-56 247 07 247-40 24714 246-31 572-57 573-06 .573-91 573-95 573-52 320 01 325-9;) 320-51 320-81 327-21 216-89 573 40 326-51 ■ 1861 1862 1803 1804 1865 247 -05 240-92 240-50 246-29 240 03 573-56 573-71 573 -50 .573 -07 572-41 326-51 326-79 327 00 320-78 326-38 246-55 573-25 :<26-70 I860 245 62 240-44 245 17 240 06 247 -20 572-72 572-44 ,572-23 .572-49 573 -13 327 10 1807 326 00 1808 327 00 1869 1870 320-43 325-93 246 10 572 00 320-50 1871 , 1872 1873 1874 1875 245 84 244-41 245-53 246-28 244-96 245-40 572-60 .571-64 572-25 572-71 572-21 326 70 327 23 326-72 326-43 327-25 572-28 326-88 1876 246-76 245-58 246 10 245 70 245-51 573-59 572-52 572-94 572-25 572-45 326-83 1877 1879 320-94 326-84 326-55 1880 320-94 245-93 572-75 326-82 1881 245-14 246-13 246 31 240 ■ 96 240-59 240-22 247 31 240 77 245-56 245 67 246 73 572-21 .573-10 573 11 .573-12 573 09 572-92 /3 573 -4fl 573-16 572-38 572-10 572-80 327 07 1882 1883 326-97 320-80 1884 326-16 1885 326-50 326-70 1886 326 15 1887 1888 1889 1890 326-39 320-82 320-49 320 -07 246-41 572 •7>3 326-38 240 FALLS OF NIAGARA [Geol. Surv. TABLE OF MEAN ANNUAL FLUCrUATIONS OF ERIE AND ONTARIO — Concluded. Years. At Toronto. At Pt. Colborne. Toronto below ' Pt. Colborne. 1891 245 77 571 88 326 11 1892 244 93 571 88 326 85 1893 245 49 571 85 326 36 1894 245 243 31 81 571 570 80 93 326 327 m 1895 12 245 0{j 571 66 326 327 327 326 60 1896 244 244 245 06 44 03 571 571 571 08 66 88 0^ 1897 9,'>, 1898 85 1899 244 97 571 63 326 66 1900 1^44 91 571 63 326 72 244 68 1 571 57 326-89 1901 244 73 571 05 326 32 1902 245 01 571 70 326 69 1903 245 54 572 02 326 48 1904 246 29 572-21 325 92 1905 . 245 91 571-84 325 93 245 49 571 76 32i^ 27 LOWERING OF THE OUTLET OF ERIE AT THE SAME RATE AS THAT OF ONTARIO. To ascertain whether the deepening of Niagara channel is measurable or not it is necessary to compare the levels of Erie and Ontario. The records at Port Colborne on Lake Erie (datum of which is slightly different from that of Cleveland) are here compared with those of Lake Ontario at Toronto, as shown in the accompanying table and in the quinquennial table, Chapter xviii., page 229. The great drop of Erie occurred in 1891 and a sudden lowering of Lake Ontario took place at nearly the same time. The mean level of Erie at Port Colborne between 1854 and 1890, see p. 229, was 572-83 feet above sea level, and between 1891 and 1905 it was 571-66 feet; thus showing a sinking of Lake Erie amounting to 1-17 feet. The mean height of Lake Ontario between 1854 and 1890 was 246-24, and between 1891 of Canada] LOWEEING OF LAKE OUTLETS 241 and 1905 it was 245-07 ; also showing a difference of 1-17 feet. Thus it will be seen that there has been no change (in the mean yjf many years) of lake fluctuations. The conclusion is that the lowering- of the Erie outlet is at the same rate as at the St. Lawrence outlet of Lake Ontario. At Niagara falls^ and at Galops rapids of the St. Lawrence, the two rivers flow over lime- stone which mostly determined the rims of the lake basins above these points; although n limited amount of drift affects the St. Lawrence as well as the outlet of Erie as above mentioned. While silts have seggregated around the piers of the Interna- tional bridge, yet in channels between some of them, the river has deepened its bed. Thus in the channel of the fourth span, some thirty years ago, the depth was forty-tw^o feet, while it has since scoured to a depth of fifty-three feet. At Galops rapids the depth to the rocky rim does not exceed thirteen feet, which in June, 1902, reached 2-09 feet or 1-25 feet below Ogdensburg, a few miles above (page 2793, Rep. Eng., 1902). There the bed of the river is composed of clay. AVhile the rocky rims are the principal barriers to the lakes, limiting the power of the rivers in the scouring of the clay beds, yet part of the lowering of the lakes appears to be due to changes in the clayey sections, and it is not surprising to find that the amount of lowering of the two lakes has been the same. The wear of the rocky floor of the river, adjacent to the head of Goat island seems at first insignificant; but after passing the Eirst cascade between two and six feet of rock have been removed by the modern river, which has acted upon the rapids only since the falls receded from 1,000 feet south of Hubbard point, or 7,000 feet from the present crest of the falls ; that is after the falls had begun to remove the drift from the buried Falls-Chippawa valley, with the appearance of the Upper rapids themselves. This occurred less than fifteen hundred years ago, and shows that the lowering of the river, through the 16 242 FALLS OF NIAGAKA [Geol. Surv. rocky barrier, lias been very slow, and accomplislied at irregular intervals. LOWEKING OF THE OUTLET OF LAKE ONTARIO. From the table of the quinquennial fluctuations of Lake On- tario and the St. Lawrence river already given (Chapter XVIII.) and the annual fluctuations (Chapter xxxi.), the follow- ing table has been computed. Years. s § a; 246-18 244-87 245-49 Lock 27. 244 77 243-71 244-21 ^ 6 1-1 1-41 1 16 1-28 Lock 15 157-33 156-26 ^ 6 .2H ^^ 88-85 88-61 Lock 14. 154 33 153-47 153-56 TiOck 14 bolow Lake at To- ronto. 1876-1890 1891-1900 1901-1905 91-85 91-40 91-93 Although this table is not complete for all the years of observation at the four stations, yet it is sufficient to show that it would be only slightly changed by the addition of the averages of fifteen years preceding that of 1876. From the quinquennial table (page 232), it is found that the level of Lake Ontario at Toronto between 1861 and 1890 is only 0*08 of a foot higher than the average of the later fifteen years, so also Lock 14 shows that the mean of thirty years from 1861 to 1890 exceeds that of fifteen years ending w4th 1890 by only 0-08. Consequently the earlier records at other points are unnecessary. On analyzing the table here given it is found that the mean water surface fell between the periods ending 1891 and 1900: at Toronto, 1-31; at Lock 27, 1-00; at Lock 15, 1-07, and at Lock 14, 0-86 feet. After 1900 the water rose again at Toronto, 0*62; at Lock 27, "50 ; and at Lock 14, -09 of a foot. With this rise of the broad surface of Lake Ontario, amounting to more than half a foot, the rise of the river level should have been commensurate in place of an insignificant quantity. The conclusion is that the ot Canada] LOWEKING OF LAKE OUTLETS 243 bed of the St. ]^aAvreiice rapids channel adjacent to Lock 14 was reduced 0-62 less 00, or 0-53 of a foot at least. The level of the lake fell 0-45 of a foot during the decade 1891-1900 more than the river at Lock 14. These results added give the per- manent lowering of the lake outlet at and above Lock 14 at -DS of a foot, although on account of the increased rainfall (1901- '05) the actual present subsidence in the lake surface has been reduced to only -69 of a foot (1 -31 — -62) lower than between 1876 and 1890. Turning now to the discharge of the St. Lawrence river (page 248 and Appendix vi., Table 3) the mean volume be- tween 1876 and 1890 was 260,700 cubic feet per second, but from 1891 to 1905 it was reduced to 234,000 cubic feet— a dif- ference of 26,400 cubic feet per second. As a fall of one foot of the lake level corresponds to a diminution of discharge amounting to 25,761 cubic feet per second, the actual measure- ment shows an amount corresponding to the lowering of lake level bj slightly more than one foot. As this was not due to decreased rainfall, for in late years the precipitation has in- creased, it was evidently caused by a lowering of the channel, thereby permanently reducing the surface of the lake to a lower leveL The lake level during 1891-1905 was 1-10 feet lower than in the years 1876-'90, while the lowering determined by the discharge is shown to be approximately 1 -05 feet, the mean of fifteen years. The increased rainfall during 1901-'05 has apparently counteracted this lowering to 0-69 of a foot, thus reducing it from 098 of a foot — a condition which canntot be expected to continue. EFFECT OF LOWERING OF ONTARIO ON HIGHER LAKES. The mean annual fluctuations given in table accompanying Chapter xxxi. show that this sudden subsidence of the water occurred between 1890 and 1902, with the subsequent fluctua- tions which are fairly expressed by taking the mean of several 244 FALLS OF NIAGARA ^^^°^- ^"^^ years. Changeable winds prevailing over several years may to some extent give rise to variations ; so also groupings of differ- ent years, but the results are all in the same direction — show- ing a lowering of the lake outlets, though now partly counter- balanced by the present excessive rainfall. The mean height of the water during thirty years until 1890 shows no extraordinary changes, while immediately following there was a sudden lowering at all the lakes except Superior. The slope between Toronto and Lock 27 being only slightly more than one foot, and most of this situated within twelve miles of the Lock, it may be considered that the whole dis- tance belongs to the lake level. Between Locks 27 and 15 the rapids descend nearly eighty-nine feet to the expansion of Lake St. Francis. Between Locks 15 and 14, which last is beyond the lower end of this lake, there is only a slight fall. Beauharnois canal, with Lock 14 at its head, extends (alongside Coteau, Cedar, Split Bock and Cascade rapids) to Lake St. Louis, which is more than eighty feet below Lake St. Francis. Some streams from the northern side of the Adirondack moun- tains enter the St. Lawrence above Lock 15, yet in the mean averages here, no effects appear, as it is seen that the lowering of the water is the same as that at Lock 27. Lake St. Louis receives the large and variable Ottawa river, producing very great fluctuations of ten feet or more. Such changes increase or reduce the head of the waters on these rapids, which would theoretically retard the outflow above Lock 14 at times of high water in the Ottawa river, if the amount were not too small for consideration on account of the great descent of the rapids. Turning now to the other lakes as well, the cause of this gen- eral low stage of water, also prevailing in the Erie and Huron basins, cannot be attributable to any great extent to the inferior rainfall throughout the whole period; for after some years of reduced precipitation (1891-1900) the mean rainfall for the following five years increased in the whole lake region, so that of Canada] LOWEIMXG OF LAKE OUTLETS 245 the precipitation was greater than that Avhich prevailed not merely during the years of low water (1890-1900), but also during higher water preceding 1891. Yet it has not raised the lakes to their former stages; and with the next cycle of low rainfall the reduced levels will become more apparent. Accord- ingly with the evidence before me there seems no other suf- ficient explanation for this lowering of the lakes than that due to the scouring of the outlets, making itself manifest by the sudden sinking of the water, particularly within a period of two years (1890-92). In the meanwhile the oulet of Lake Superior has not been measurably reduced, and the increased rainfall has raised its level. The rise of about 0-62 of a foot in Lake Ontario during 1901-'05 does not appear in Lake Erie, where the change is only 15 of a foot, while the height of Lake Huron is raised by •21: of a foot, although the increase in rainfall should have affected the level of Lake Erie rather than that of Huron. This inferior rise of Lake Erie in contrast to that of all the other lakes is doubtless attributable to tlie^ artificial diversion of the waters of the river, already begun to a small extent. (See Chapter xxi., Part ii.) A uniform rate of lowering could not be expected. The impact occurring on the various strata would weaken their resistance and cause them to give way at irregular times. When the next lowering of the outlets will occur cannot be foreseen, but the forces are always at work wearing down the rocky barriers of the lake. As has been shown the amount of lowering in the lake is about one foot. On referring to the discharge tables it will be found that the mean Erie overflow, as given, between 1860 and 1890, Avas 21,800 cubic feet per second more than during the period 1891-1905. This is equal to about 0-95 of a foot. This includes the slight rise caused by increased rainfall for the last five years. Eor the absolute sinking of Lake Huron one foot must be 246 • FALLS OF NIAGARA ^^^°^- S"^^" added to 0-66, or 1-66 feet as the total amount of lowering of Lakes Huron-Michigan which has occurred. The shoaling is very noticeable in the shallow channels among the islands. EFFECTS OF THE LOWERING OF THE LAKES UPON CANALS AND HARBOURS. About 1893 much concern w\as occasioned by the shoaling of the harbours^ the cause of which was not explained at that time. Low water was characteristic of the decade ending 1900. Since that time the water levels have been higher with the in- creased rainfall which is partly counteracting the effects of the loAvered outlets. With the return of normal dimin- ished rainfall the low^ering of the lakes will be more apparent, in the relationship of their levels to the harbours and canals. The sudden shoaling of the canals by one foot, when they were made for even fourteen feet draught, is a serious problem which must enter into the engineering features of the future, and also the shoaling of the harbours, the effects of which are variable with the oscillations of rainfall. Added to this shoaling of the harbours and canals will be the future lowering of Lake Erie, due to the diversion of the waters for power purposes. (See Chapter xxi., Part ii.) CORRECTED ELEVATIONS OF THE GREAT LAKES. As the outlets of the lakes have been lowered since 1890, the mean elevations that should be adopted are those between 1891 and 1905 inclusive, thus making: — Feet above mean sea level at New York. Lake Ontario 245-08 Lake Erie 571-66 Lake Huron 580-09 Lake Superior 601-94 (This elevation of Superior is the mean during 1861-1905.) CHAPTER XX. DISCHARGE OF NIAGARA AND OTHER RIVERS OF ST. LAWRENCE DRAINAGE. Notes on discharge. Variations in discharge of St. Mean quinquennial discharge of St. Lawrence river Mary, St. Clair, Niagara and St. Variations in discharge of bt Lawrence rivers. Mary river. . t..-^ >,„.,„ Variations in discharge of Niagara Proportional drainage of Erie basin. river. Variations in discharge of St. Clair river. NOTES ON DISCHAEGE. In the earlier investigations of Niagara river* the only available measurements of the discharge of the St. Clair and Niagara rivers were extremely fragmentary, and those obtain- able were made in 1868. f Very much more satisfactory in- formation is now at hand, as extensive determinations have been made for different stages of the lakes, and published in reports for the years from 1900 to 1905,:}: so that the discharge of any day since the fluctuations of the lakes were first recorded some fifty years ago can be calculated. The discharges for the Niao-ara river are given in the report for 1903. The mean monthly and annual discharges have been transcribed to Ap- pendix V. of the present report. AVliile the fluctuations of the lakes involve many problems bearing on the lake history, discharges of Lakes Huron and Erie only produce a direct modification of the recession of Nia- gara falls. The present conditions have not always prevailed. Thus, in 1888, it was first discovered that Lake Huron did not ** Duration of Niagara falls,' J. W. Spencer, Am. Jour. Sci.. Vol. XLVIII., p. 461, 1894. t Report of Chief of Engineers in 1869, p. 582. ± lb. 1900-'5. 247 248 FALLS OF NIAGARA [Geol. Surv. empty into the ^Niagara drainage until a recent date, con- sequently it was necessary to ascertain the difference of volume of discharge of the St. Clair and ISTiagara rivers. As the surveys of ISTiagara falls were made in 1875, 1890, and 1905, the periods of discharge in the following table have been chosen so as to correspond with the years of the surveys. From the table it will be seen that there was a sudden shrink- age of the water after 1890. The discharge of Niagara in 1890 was given at 228,974 cubic feet per second, while in 1891 the amount was only 208,908 cubic feet. The lowering of the water of the St. Clair river began to appear in 1889, but only in 1891 did it attain the present low level. MEAN QLTINQUENNIAL DISCHARGE OF ST. MARY, ST. CLAIR, NIAGARA, AND ST. LAWRENCE RIVERS. Years. St. Mary. St. Clair. Niagara. St. Lawrence. 1860-1865 1866-1870 1871-1S75 84,238 83,985 80,531 215,290 194,773 203,623 233,182 220,483 214,819 277,608 260,629 240,694 1860-1875 205,2.32 223,476 260,767 1876-1880 1881-1885...., 1886-1890 75,550 74,844 68,299 208,341 215,865 206.249 228,644 232.124 226,554 254,866 261,802 265,193 1876-1890 210,152 229,107 260,621 1891-1895 1896-1900 65,922 80,391 173,832 168,891 173,665 (a) 203,832 202,480 206,167 234,691 222,989 1901-1905 244,428 1891-190.J 204,167 234,036 1891-1902 1860-1892 1860-1890 1860-19ft5 171,746 197,603 207,357 * 202,185 219,499 225,967 218,859* 251,9.36' 260,694 (a) for 1901-02 only. * The discharge of Niagara river given in the 1904 report (p. 4058), is 215,200 cubic feet per second as the mean from 1860 to 1903, and for the St. Clair river 205,400 cubic feet. As will at once be seen the difference of these discharges is entirely too small and could not have been used in determining the relative physical features of Niagara river. of Canada] DISCHARGE OF NIAGARA RIVER 249 The tables are adapted from the Engineers' Report of 1903, based upon the mean measurements at an open section of the Niagara river and at the International bridge, two miles within the river. VARIATIONS IN DISCHARGE OF NIAGARA RIVER. The change in discharge for one foot of fluctuation of Lake Erie is (for datum discharge) equal to 23,205 cubic feet per second. This includes 1,200 cubic feet per second discharge for the Erie canal, and 1,100 feet per second for the Welland canal. (Thomas Russell, p. 1116, 1891.) The equation for the variation of the river stands thus : — Discharge in cubic feet per second (s-f) equals 158,500 +22,462 (0—570). C is the elevation of Lake Erie at Cleveland, (s. f. is cubic feet per second.) In the discharge tables the volumes of the various years are based upon the fluctuations of the lakes. The meter determina- tions for various stages were made during some of the years between 1898 and 1902. Thus it will be seen that the measure- ments were only taken after the lowering of the lake outlet, which feature was not recognized in computing the discharge prior to 1891. This omission accounts for the larger volume of the river between 1860 and 1890 compared with that of sub- sequent years, giving an excessive discharge between 21,000 and 23,000 cubic feet per second. If the Niagara channel has been uniformly deepened to the equivalent of one foot since 1890, one foot should have been taken from the lake levels in calculating all discharges prior to the lowering of the outlet ; which has not been done, but in the tables, the discharges are given as if the outlets were as high as when the lake-levels w^ere recorded prior to 1891, although the water-meter determinations were made since the lowering of the outlets. By comparing the mean discharges from 1860 250 FALLS OF NIAGARA f^eol. Surv. to 1890 with that from 1891 to 1905, the volume of the Nia- gara river for the two periods is reduced from 225,967 to 204,104: cubic feet per second, a difference of 21,803 cubic feet, or a lowering of about 0-95 of a foot. Transposing the case, with the bed of the channel as formerly nearly a foot higher, the mean discharge before and after 1890 would be ap- proximately the same. Thus is explained the extraordinary dis- crepancy that demanded investigation, as the meteorological variations were entirely unable to account for it. In short, the discharge of ISTiagara river in the future can only be taken as that of the m^an since 1890, which further re-determinations may slightly correct. Accordingly the mean volume is taken at about 204,000 cubic feet in place of 219,000 for the whole period from 1860 to 1905. This materially reduces the esti- mate of the power of Niagara falls. The Lake Survey has lately reduced their estimate to 215,200 cubic feet (their me- thod being unpublished), but nowhere have they corrected the discharges for the lowering of the outlets as mentioned. With my correction applied, the mean discharge would be furthei" reduced to 200,000 cubic feet per second. VARIATION IN DISCHARGE OF ST. CLAIR RIVER. A change of discharge in Saint Clair river for one foot of fall of Lake Huron is 19,238 cubic feet* per second (page 4120). Wliile the formula for the discharge is simple yet it was instru- mentally determined for only a limited range of height, as in the case of Niagara river, after the lowering of the St. Clair outlet. Consequently, similar corrections in the discharge tables must be applied for the years preceding 1891. Again, ice action on which observations have been made during only three winters, is not fully understood ; thus results are less satisfactory than at Niagara river. * Report of U. S. Engineers, page 4120, 1904. of Canada] DISCHARGE OF ST. CLAIR RIVER 251 The table (page 248) gives the mean discharge of Lake Huron from 1860 to 1890 at 207,357 cubic feet per second, and that from 1891 to 1902 at 171,746 cubic feet. As has been shown (Chapters xviii. and xix.), the Huron outlet has been lowered by 0-66 of a foot more than the surface of Lake Erie, to which one foot must be added. Applying this correction to the mean discharge given before 1891, the result shows 176,957 cubic feet per second, or about 5_,000 cubic feet per second more than the mean discharge of the twelve years 1891-1902. This excess would be reduced by including the discharge of the last three years; and further allowing for the imperfec- tions of observation, the discrepancy of calculation before 1891 compared with determinations since that date is greatly re- duced. Thus it becomes apparent that the differences between the discharges of the two periods is due to the omission in consider- ing the former higher bed of the St. Clair river. This analysis confirms the conclusion as to the recent lowering of the Huron outlet, and the necessity of adopting the discharge values since 1890 as the standard. VARIATIONS IN THE DISCHARGE OF THE ST. LAWRENCE RIVER. A formula has been found for extending the discharge calculations, and it may be added here, along with those of the other rivers. The meter measurements were taken in 1900-'02, at the narrow^est part of the river below Cardinal (that is below Lock 27). The discharge in cubic feet per second = 94683 -\- 25761 (level of Lake Ontario at Oswego — 240) ; or, in other words, the variation of one foot changes the volume by nearly 25,761 feet. Fluctuations of level of one foot at the Lock 27 repre- sent nearly 26,787 cubic feet, in the discharge measurements. 252 PALT.S OF NIAGARA LGeol. Surv. VARIATIONS IN THE DISCHARGE OP ST. MARY RIVER. The formula for tlie variations of the discharge of Lake Superior may be applied in the study of the relative fluctua- tions of that lake and Lake Huron. The pier bridge across the St. Mary has increased the velocity, but the formula since 1898 is: — Discharge in cubic feet per second = 48,235 -|- 17,656 (level of Lake Superior — 600). PROPORTIONAL DRAINAGE OF THE ERIE BASIN. A special object in studying the discharge of the Huron basin is to determine the differences between the outflow of the Huron and Erie basins. Tables given in the Reports of En- gineers for 1903 and 1904 vary the mean discharge of the rivers by a small percentage, which when reduced to the dif- ferential discharge of the Erie basin makes the apparent dis- crepancy here very much greater than is permissible. Concern- ing Erie alone, Mr. Russell shows that of the IN'iagara discharge, when it reaches 222,400 cubic feet per second, 31,100 cubic feet per second should come from the Erie basin, if the rain- fall of the region were proportional to the respective areas — that is the ISTiagara discharge would exceed the St. Clair by this amount. But the calculation from the dischar;!;e formula as given in the 1903 report shows only 21,300 cubic feet per second as the average between the years 1860 and 1892 inclusive. This is ten per cent of the total overflow of the Erie basin. In the report of 1904 this percentage Avas reduced so much indeed that IMr. Russell says : ' It would require a very great and totally inadmissible evaporation from the lake to have the differences as small as observations indicated (page 4125).' 'The most that can be done then with the water heights, discharge, and rainfall is to see what a reasonable value of the land run-off and the lake evaporation will give for the difference of the Detroit and l^iagara rivers.' of Canada] ERIE DRAINAGE 253 By taking the run-off of the year at six-tenths, and the evap- oration of Lake Erie at three feet in a year, Mr. Russell finds the Erie discharge "would correspond to 37,255 c. f. per second. Combining the result with that from the Niagara river he shows it would raise the lake surface by 1 -61 feet more than without it. He concludes by saying that the evaporation indicates that the winter run-off is probably more than six-tenths of the rain- fall, and the summer run-off somewhat less. This would make tlie proportional overflow from the Erie basin to be 16-7 per cent of the total discharge of Niagara river. For the Huron-Michigan basin the run-off from the land is taken at six-tenths of the rainfall, and the evaporation from the lake surface alone may be taken at thirty-three inches for the year ; while in Lake Superior region, with the same run-off from the land, the evaporation from the water is less than eighteen inches. Considering the discrepancies between the meteorological and the gauge determinations, Mr. Russell says : ' The adopted datum discharge for the St. Clair and Detroit rivers is possibly too large, and the discharge of Niagara too small by a very con- siderable quantity ; 10,000 cubic feet per second or more ' (pag8 4130). In the Report of the Engineers attention is also called to the fact that the gauge readings were taken during years nf low water. In the above, great discrepancies appeared between the mean discharge of the Erie basin and the total volume of the Niagara river, when comparing the tables with other evidence bearing upon the subject; for from the table the mean run-off from the Erie basin appears to be only 10 per cent of the total discharge of all the Upi^er lakes, while the drainage areas and the meteor- ological phenomena indicated much larger proportions. From the corrected discharges it is found that the run-off of the Erie basin is 15 per cent of the total discharge of Niagara river. 254 FALLS OF NIAGARA t^^o'- S^^^- This amount of 15 per cent as derived from the discharge values is close to that obtained from the meteorological calcula- tion, which is 17-7 per cent (page 221) ; and from the pro- portional part of the drainage area of 16 per cent. The discharge value is probably the most correct on account of the variable rainfall in the different lake basins. There will be other meteorological variations, such as a greater velocity of wind in the Erie than in the Superior basin, a higher temperature in Erie than in the other basins, as well as a greater evaporation. Finally, the adoption of 15 per cent as the proportional part of the Niagara discharge derived from the Erie basin may be used in the variable reces- sion of Niagara falls during the period of changing volume. This problem involved one of the most complex questions in the present work, as the approximately correct determinations of the age of Niagara falls hing-ed upon this investigation. CHAPTER XXI. POWER OF NIAGARA— PART ONE. Variable discharge and present shrinkage. Range of horse-power of the falls. Available power in recession. Net mechanical horse-power. Ratio of Canadian and American channels and falls. Franchises. Limitations of use. Power of Niagara river below the falls. VARIABLE DISCHARGE AND PRESENT SHRINKAGE. The descent of the Nia,2;ara river from the First cascade of the Upper rapids to Lake Ontario is 312 feet. The surface of the river below the falls may be taken at 100 feet. Thus the power of Niagara represents the product of a fall of 212 feet and the variable discharge volume of any particular day. Corrected Cubic feet per second. Average discliarge 1860-1800 1891-1905 1860-1905 18.58 (October 7) 1862 (for the year) 1862 (month of June) 1895 (for the year) 1902 (month of Februarj') . . 1902 (Feb'y. 28th) Reducing these figures to horse power the following results are obtained : — Cubic feet cubic feet per second. per second. *226,000 204,000 204,000 204,000 *219,000 204,000 314,000 292,000 248,000 221,000 260,000 238,000 187,000 187,000 175,000 17.5,000 158,500t 158,500t RANGE OF HORSE-POWER OF THE FALLS, Average gross horse power 1860-1890 1891-1005 1860-1905 185S (October 7) 1862 (for the year) 1862 (for month of June). . 1895 (for the year) 1902 (month of Febiuaryl 5,444,000 H.P. 4,015,000 5,276,00'J 7,563,000 5,854,000 6,264.000 4,50.5,000 4,216,000 1002 (Feb'y, 28th) 13,818,000 Corrected 4,914.000 H.P. 4,915,000 „ 4,748,000 M 7,033,000 „ 0, .326, 000 „ 5,736,000 M 4,. 505, 000 H 4,216,000 „ 3,818,000 n From these figures it may be seen that the work of Niagara varies enormously, * These are averages taken from the discharge tables, but on account of the lowering of the lake outlets, not allowed for here, the correct mean discharge is that of the period of 1891-1905. t Water at Port Colborne only 569-58 above the sea. 255 256 FALLS OF NIAGARA ^^^°^- ^"^^• AVAILABLE POWER IN THE RECESSION. From the standpoint of the recession of the falls the whole of this horse power, whether large or small, comes into opera- tion. The greatest force of the river is at the apex. This con- dition has obtained for some time, as our recent survey has found that the deepest channel is close under the American falls, and near Goat island shelf. It is seen, both from th© soundings and the present discharge over the falls, that the force at work is unequally distributed. It is only reasonable to suppose that the princijial work of new excavation is effected during periods of high water, when the rock formations are the most weakened. The apparent modern reduction in the discharge of the river has been ex- plained, but the diminished rate of recession of the falls during the last fifteen years to less than half that of the annual amount during the forty-eight preceding years (page 41) is not due to the reduction of discharge. This slower recession now in progress is largely attributable to a change of rock structure, owing to the falls now crossing the site of the ancient Falls- Chippawa valley in place of following along its course (page 166). The above brief consideration of the work of l^iagara falls shows simply the natural retreat of the great cataract ; . but now artificial conditions are beginning, and henceforth these will be greatly extended. NET MECHANICAL HORSE POWER. A change of one foot in the height of Lake Erie will in- crease or diminish the discharge by 23,205 cubic feet per second, which is equivalent to 569,000 horse power. When it comes to the question of the artificial application of power, if water be taken from the river below the First cascade, the obtainable horse power will be much less than if taken above. On the average of all the companies, the available amount will of Canada] POWER OF NIAGARA FALLS 257 be reduced by from 30 to 35 per cent,* owing to the loss in height, in the waste weir tunnels, and at the bottom of the shafts, etc. This will reduce the available horse power for a foot of fall in the lake to 400,000 or less. Where the water is taken at a considerable distance below the head of the rapids, from forty to fifty feet lower down, the available amount will be still further decreased. Accordingly, the net mean water power of the falls if entirely diverted for artificial purposes would fall to 3,200,000 horse power, and the low water discharge to 2,600,000. On the basis of 7 per cent, the available horse- power of the American falls is 350,000 for mean stages, or nearly 250,000 for low water. The preservation of Xiagara falls is now a question before both, countries. TMiere water power franchises are given they are granted in quantities irrespective of the stages of the river, or, in other words, they withdraw the water from the minimum discharge. Thus when the mean volume was reduced to the stage of February, 1902, it was about 30,000 cubic feet per second less than the mean annual discharge. This produced the lowering of the river above the rapids by over two feet below mean stage. RATIO OF THE VOLUME OF THE CANADIAN AND AMERICAN CHANNELS AND FALLS. The mean depth of the water at the American falls does not exceed 1 -5 feet, though some channels are deeper. These falls, including Luna island, have a breadth of nearly one thousand feet. The barrier whose height determines the discharge is situated near the head of Goat island, where the river is reduced to a breadth of about 340 feet. The First cas- * Power companies' heads: Canadian Niagara, 136. Electrical Develop- ment, 136. Ontario Power Company, 180. Niagara Falls Power Company, New York, 136. Niagara Falls Hydraulic Company, 210. Chicago canal, 32 feet. See Report of Queen Victoria Niagara Falls Park, by James Wilson, 1905, p. 7. These figures show the great loss of power head in transmitting the water through the tunnels. 258 FALLS OF NLVGARA ^^''°^- ^''''"■ cade between Goat island and the Canadian side forms a rim of nearly 3,400 feet, as the river cascades over the rock ledges which control the discharge of the Canadian falls, iVc- cordingly, in breadth the American chamiel is only nine per cent of the whole, xls mentioned in Chapter v., there is an almost horizontal ledge of limestone extending across the river which is covered in many parts by only a thin sheet of water, as on the eastern half. Thus for 400 feet from Goat island, during the present average water already lowered^ the sheet has a depth of less than one foot. Then for a distance beyond, beginning above the outermost Sister islands, the depth of water appears to be from two to three feet, where it descends six or seven feet. This continues for a few hundred feet, beyond wdiich tha ledge of rock is very thinly covered with water, reaching perhaps more than half way across the river. Beyond is the deep part of the iSTiagara river, extending to near the, Canadian shore, where I have estimated the depth from dis- charge values as averaging eight to nine feet, although Mr. Wilson does not think that it is so great. At the First cascade of the American channel the average depth scarcely exceeds three feet ; and being very much smaller than the deeper section of the Canadian channel, the volume of discharge is reduced considerably below nine per cent, perhaps as low as six per cent, though probably the range may be between this and eight per cent. At seven per cent the volume is about 14,000 cubic feet per second. Indeed, the shallo"\vness at the head of Goat island was shown during low water as in 1904, wdien, owing to the backing of the river due to wind, a wide expanse of rocks w^as laid bare above the Sister islands. Kalm records that four Indians forded from Goat island to the jSTew York sliore. Some idea of the character of these rapids is show^n in Plates xii. a and b, and xxxi. a and b. From Goat island to the International Boundary Line, on the First cascade, is a distance of about thirteen hundred feet, of Canada! CANADIAN rOETION 259 where the depth is shallow as just described. From this Line to the Canadian side is twenty-one hundred feet, including the deepest portion of the river, where the water is not less than three times the depth of that to the east. Accordingly, not over 13 per cent of the discharge is on the eastern side of the Bound- ary Line. This, together with, say, seven per cent in the Araercian channel, would allow 20 per cent of the volume of the river to be upon the New York side and 80 per cent on the Canadian. A mile above, the soundings give a good cross section on each side of the Boundary Line, and by Chezy's formula it is found that 75 per cent of the water flows on the Canadian side; while at the First cascade, owing to the position of the Line and the shoaler water, on the eastern side^ a smaller amount passes. At the brink of the American falls themselves the volume of the cascade, as given above, is approximately seven per cent, while the stream of water over the end of Goat island shelf, from a breadth of less than 300 feet^ is so inconsiderable that it is almost inappreciable. Above this end the whole great cataract lies within Canadian territory, FRANCHISES. The question of franchises now presents itself. These have been granted on the Canadian side of the river to: — Canadian-jSiagara Power Company... . S,(10(i ciibic feet per second ; 100,000 H.P. Ontario Power Company 11,700 ,. ,. 180,000 „ Electrical Development Company 10,750 ,i „ 125,000 ., Electrical Railway Company 400 n <• t 31,450 On the New York side : Niagara Falls Hydraulic Company. . . . 10,000 cubic feet per second ; 100,000 H.P. Niagara Falls Power Company 17,200 „ .. 200,000 „ 27,200 Chicago Drainage canal 10,000 u >■ Welland canal ( including Hamilton Cataract Company) 1,100 m m Erie canal 1,200 12,300 Total 70,950 17i 260 FALLS OF NIAGARA f^^^^- ^urv. Of the franchisee! diversion, the two companies on the iNTew York side, and the Ontario company on the Canadian, take their water from or above the First cascade, and consequently affect the river common to both countries. The intakes of the Canadian ISTiagara and the Electrical Development Companies are from the deeper part of the river, far lower than the First cascade, and they will not use the water from the shallower part of the river near Goat island side. ISTor will they lower it in the basin above the First cascade or Lake Erie to any appre- ciable extent. As a power question alone, below the First cas- cade, almost all the water passing down the Canadian falls, belongs to Canada. , It is the diversion of the water from above the rim of the First cascade which will further shrink J*Tiagara falls. Already, the crest has been curtailed by 415 feet, on the western side, owing to the artificial embankments. (See pp. 267-8.) Besides these there are ten other franchises, mostly old, some of which have been cancelled. Several are in unlimited amounts, so that- the whole volume of the river might have been given away, but none of these last are under construction. At the time of writing (winter of 1906) the Chicago drain- age canal is taking 5,000 cubic feet per second; the Canadian companies, 3,000; and the 'New York companies 9,000 cubic feet, in all 17,000 feet, besides the Welland and Erie canals. This is equal to the lowering of the water in Lake Erie by three-quarters of a foot. The figures include a liberal loss in conversion, and they should be compared with the discharge at a time of low water. Passing by the minimum daily over- flow, occurring in these later years, but adopting the mean low volume of February, 1902, which was 175,000 cubic feet per second, the above franchises, including canals, represent 40 per cent of the total low water discharge. of Canada] DIVERSION OF NIAGARA WATERS 261 LIMITATIONS OF USE. Since writing the above, tlie report of the International Waterways Commission has appeared. They recommend the limitation of the power on the Canadian side to 36,000 cubic feet per second, and on the ISTew York side to 18,500 cubic feet per second; in addition to whicli is the discharge of 10,000 cubic feet for the Chicago canal ; — in all G4,500 cubic feet, including the "Welland and Erie canals, or nearly 37 per cent of the low water discharge. This amount will lower the river, pro- ducing effects on the^ falls only a little less than the diversion of the full franchise volume mentioned. So, also, it will most seriously impair navigation on the Upper lakes as will be ex- plained in next chapter. POWER OF NIAGARA BELOW THE FALLS. The question of the power of ISTiagara does not end here. There is the same volume of water descending about another hundred feet before reaching the mouth of the gorge. Here is nearly fifty per cent of the power of Niagara falls themselves. Various schemes have been proposed to utilize it — for instance a tunnel on the Canadian side, adjacent to the Whirlpool rapids, where there is a descent of 51-5 feet. On the 'New York side it has been proposed to build a retaining wall beside the rapids which would eventually reach a height of fifty feet, and then cut down the natural slope of the side of the canon so as to build a canal between the retaining wall and the side of the gorge, now made perpendicular. On the embankment the Gorge railway would run, while the power house would be situated on the rocks at the outlet of the Whirlpool. At Foster flats a fall of twenty feet is available. Finally, the last proposal is to dam the mouth of the river to a height of 100 feet, and thus at once obtain a million and a quarter of available horse power. This would flood Foster flats, drown the Whirlpool, submerge AMiirlpool rapids, and form a great 262 FALLS OF NIAGARA ^^^''^- ^urv. mill pond. Such a proposition disregards the various considera- tions of the destruction of all these features and ignores inter- national questions. The engineering difficulties of opposing a force of 3,000,000 horse-power, which might be hurled against the works, at any time during a great flood, seem insurmount- able, and not to have been taken into account; also, the fact that the dam would require to reach more than 180 feet below the surface of the river. In this chapter I have set forth the power of the river as a natural agent, and have pointed out the capabilities for supply- ing power. Another proposition has been brought forward, namely, to divert the, water from the main river into the ISTew York channel and to save the American falls. But to pay for this expense a million more horse power is to be abstracted from the greater falls, which are Canadian. This is an insidious proposition for obtaining more water, in fact more than a third of the total available power of the river. Before making these investigations I had no appreciation of the magnitude of the changes which the proposed use of power is likely to produce. It is only by applying measurements to the falls that the comprehension of the effects of diversion can be understood. These will be shown in chapter xxi.. Part Two. The result of lowering the river in the basin above the Upper rapids by the equivalent of three-quarters of a foot in Lake HiTpnnr can be seen in Plate vii, photographed on a day when the lake level was exactly the mean of the last five years, but subsequent to the lowering of the water on the rapids, to an apparent extent. Plate XXXI. a. View of east end First Cascade at Goat Island and bared rock at Sister Island. (Sept., 1906). Plate XXXI. b. View of First Cascade above the outermost of the Sister Islam This rim determines distribution of river ilow. (Sept., 1006). 263 CHAPTEE XXI., PART TWO. SHRINKAGE OF FALLS AND LOWERING OF LAKES BY POWER DIVERSION, Effects on Niagara Falls by use of Lowering of lakes and canals, by- power, use of power at the Palls. EFFECTS ON NIAGARA FALLS BY USE OF POWER. A fall of one foot in the level of Lake Erie reduces the discharge by 23,000 cubic feet per second. This amouni divided into the franchise allowances of about 70,000 cubic feet per second represents the equivalent lowering by three feet if the water were taken directly from the lake itself. For each additional foot of the subsidence the outflow diminishes; so that the franchise amount would represent a sinking of the lake by a qualified factor. The outlet of Lake Erie at the International bridge is reduced to a breadth of 1,850 feet, and a mean depth of 22 -5 feet, with a sectional area of 42,000 square feet. As most of the water will be withdrawn from below here the effect upon the lake will be through an increased velocity, until equilibrium becomes re-established by a lowering of the lake. This question will l)e considered immediately after the present one. Below Grand island the two branches of the river unite and form a basin, which, just l)elow the creek entrance at Chippawa, is 5,050 feet across, with soundings, so that the area of the section can be determined. This is 83,000 square feet. The mean depth is sixteen feet. The discharge, is that of Lake Erie, with only a small increase from the creeks at Chippawa and Tonawanda. It took fifteen to twentv minutes for floats 265 266 FALLS OF NIAGARA [Geol. Surv. to cover the distance from the line of sonndings, where the area is measured, to the First cascade, at the end of which is situated the Forebay of the Ontario Power Company (see Plate XII. a), showing the small descent of the river. From this basin or above it the power franchises allow the withdrawal of about 41,200 cubic feet per second, which represents 20 per cent of the mean discharge, or 24 per cent of the low water flow. Ac- cordingly, the proportional diversion represents an unmodified lowering of this basin to the extent of four feet for low water stages, or 3 "2 feet for average water. To this must be added 10,000 feet for the Chicago canal, thus making a total of 51,200 cubic feet, or 25 per cent for mean water, and 30 per cent for low water. This total diversion at or above the First cascade should be modified to an undetermined amount from confining the river to a narrower channel, after the shallower rim will have been mostly drained. If the proposed limitation be carried out, then from the 64,500 cubic feet must be deducted the Canadian franchises of 19,750 cubic feet taken at points below the First cascade, leav- ing 44,750 cubic feet to be withdrawn from above it. This would make the theoretical lowering of the basin, without cor- rections, amount to more than 3 -5 feet for even mean water, or over four feet for low water. The withdrawal of the water behiiid the rim of the First cascade has the same effect as the deepening of the outlet. It increases the velocity of the river above, which for a time lessens the amount of the subsidence in the basin, until the Erie level is adjusted to the new conditions. Even now the rim of the First cascade for many hundred feet from Goat island is covered by only from one-half to one and a half feet of water. These observations were made after there had been a consider- able diversion of the river, and on a day when the lake was at its mean quinquennial level. The deeper the Avater upon the rim at present, provided it be eventually drawn off, the more would of Canada] EFFECTS OF DIVERSION OF THE FALLS 267 be the lowering of the surface behind it, dne to the diversion through artificial orifices ; while if there be less water on the rim the discharge would be diverted into the narrower and deeper channel, causing some retardation of the current in the basin. There is nothing to lead me to expect that the lowering- will be less than three feet for mean water and more for low, {See p. 272.) The diversion already at the rim of the basin has reached a foot. This has lowered the water on Goat island shelf (as shown in Plate xxxviii. a), where formerly fewer blocks ap- peared, to beyond the International Boundary Line at the cata- ract, and also near the Goat island walk {see Plate ii.). In the meanwhile the diversion is laving bare the shallow river bottom above the upper end of Goat island and about the Sister islands. The effect upon the falls is also increasing. With the lowering of even two feet the crest line from Goat island will be drained round the angle of the shelf to a distance of about 800 feet {see Plates II. and vii.), thus cutting off this magnificent sheet of the falls. This will be the first part of Niagara falls to dis- appear, as the shallowest part of the water is on the rim above. The next shrinkage will be in the American falls, which will be reduced from a sheet of about a thousand feet wide to a few narrow streams coursing down the deeper channels, as may be determined from the rim above the rapids. At the same time the main falls will have tlie depth of water reduced two or three feet. This change is arising from the diversion of the water above the rim of the First cascade ; while that from fifty feet below the First cascade, will scarcely affect the discharge at the rim above. After the loss of 20 to 25 per cent of the water, above the rim, the power dra^vu from below it will divert about one-eighth of the remainder of the discharge (for mean stage) from the deeper channel. This will further lower the water here by less than one foot. The lowering of the river above the falls in this deeper channel 268 FALLS OF NLIGARA [^eol. Surv. may contract their breadth on the Canadian side, by 200 feet more, as already 415 feet have been taken owing to the artifi- cial embankment. As as stated on the previous page, the shrink- age in progress on the western side of the falls will amount to 800 feet. Niagara falls reached their supreme magnificence by 1900, when the perimeter of the Canadian falls was 2,950 feet and the American 1,000 feet. The wdiole is now reduced by en- croachment to about 3^500 feet. With the use of the full fran- chises the entire width of the falls will be reduced to 1,500 rkx dA.aM*eWs. or 1,600 feet, and then they will lie wholly in Canadian terri- / 'i^iyo t^ tory, except small streams coursing down the ancient river T^tBt". bed over the Goat island shelf and the present route of the American channel and falls. But occasional glimpses of the ancient grandeur of the falls will be seen during exception- ally high water. LOWERING OF LAKES AND CANALS BY POWER DIVERSION AT NIAGARA RIVER. If the subject be considered at all, there seems to be an im- pression that the diversion of the water at Niagara w^ill not affect the higher lakes. If the water were all taken from below the Greens or the First cascade, at nearly fifty feet down the rapids, the affect on the upper river would be unappreciable. But most of the water under the franchises will be taken from or above the rim, which forms the barrier to the basin of the Upper rapids. In the narrow part of the outlet of Lake Erie, at the Inter- national bridge, the mean depth is twenty-two and a half feet, but the deepest part of the channel reaches to fifty-three feet. At 1-25 miles above, the rocky beds appear at seventeen to twenty-four feet. Many parts of the river below are thirty feet deep to a point just above the basin mentioned, which is nearly fourteen feet below the level of Lake Erie. The sandbars in the river are mobile and deposited by a slight diminution of the Platk XXXII. View of western end of Canadian Falls, before curtailment of 415 feet. 269 of Canada] ]j)\vekIX(; OF LAKK ERIE BY DIVERSION 2Yl current, the same would be removed, with aiT equally small increase of velocity. Thus, if the velocity be doubled, the transporting power is increased sixty-four times. This shows that the bars are as unstable as the river, the whole course of which may be considered to have a depth of between thirty and fifty-three feet. As the depth upon a great 2:>art of the rim is less than three feet, and as probably the deepest channel does not average more than eight or nine feet, the floor of the river every- where, to near Lake Erie, is practically at as low a plane as the bed of the channel across the rim. Accordingly, it is not entirely the river bed that gives the slope, but it is largely the constric- tion at the narrow outlet of the lake, whereby the surface of the river is here piled up and flow out with more rapid currents. At Fort Erie town is a broad terrace five to eight feet above the river extending to a high river bank. At the upper end, it is underlaid by rock, once a floor extending across the river. At the crib of the Buffalo Waterworks the depth of channel is seventeen feet. But immediately above and below is a large channel to twenty-four feet or more, so that the deeper part of the seventeen-foot channel may now be refilled with clay. Here the velocity of the river reaches eight miles an hour, although the width is about 1,850 feet. As the deeper channel is about 1,000 feet wide, with its limitation, by a depth of twelve feet or less, I find here the original outlet of the little Erie drainage; so that all the channel less than twelve feet deep has been only recently over- flowed (see Chapter xxvii.) by the late increase of the river discharge. At the time of the augmented volume the river flooded the rocky floor of Fort Erie to a depth of about twelve feet, since which time the lake has been lowered perhaps twenty feet by the scouring out of the clay and rocky barrier, during the recent part of the history of the river. It is thus easily seen that if the slope of the river be in- creased, its velocity at the outlet mentioned, which is now five to 272 FALLS OF NIAGARA [•^«°1- ''^"'^• eight miles an hour, must be increased also, and that the prin- cipal acceleration due to the lowering in the basin above the Upper rapids will be concentrated at the narrow outlet of the lake. At the Upper basin the velocity is less than two miles an hour, but this is two and a half times the width of the lake outlet. A change of level of Lake Huron by one foot will raise or lower Lake Erie 0-727 of a foot; also a rise of one foot in Lake Erie will cause a backwater of '346 of a foot in Lake Huron. The distance is sixty miles, and the fall of the river eight and a half feet. At Niagara I have determined the fluctua- tions in the basin from Port Day levels, and find that a change of one foot in Lake Erie produces a fluctuation of "45 of a foot in the Upper basin, about five hours later The distance is nineteen miles, with a fall of nearly fourteen feet. Under these circumstances it is easily seen that a lowering of the water in the basin above the Upper rapids must considerably increase the velocity at the outlet of Lake Erie. The result of the withdrawal of the water behind the First cascade, as has been already shovni (page 266), is to lower the water in the basin. As the water is being drawn off, it is partly compensated by the increased velocity of the river above. So long as any water runs over the thinly covered rim, the artifi- cial diversion from behind will have the same effect as if the channel were depeened. Once it shrinks below the higher part into the narrower deeper channel a new component will be in- troduced, when the result will approach that of simply a change of outlet. This, however, cannot occur until the level has fallen considerably for ordinary stages of water. If it were not for the contraction of the channel, the greatest lowering would exceed the amount mentioned (page 267). With the gradual lowering of the basin, a difference of one foot in the slope of the ri\-er should iucrease the velocity at the outlet of Lake Erie by about three per cent, which in the course of a year and a half would lower Lake Erie nearly of Canada] LOWERING OF LAKK KRIE BY DIVERSION 273 a foot," witli the {'oiii])leto effect shown sometime afterwards. While some as yet iinmeasnred factors may somewhat retard the velocity and increase the time, the ultimate effect must be the same. After that the equilibrium should be restored at a lower level of the lake. Subsequently comes in the lowering of another and still another foot, ^or more, so that I cannot see, even in this preliminary study, how that in a few years at most, from the time of the complete use of the franchise power, Lake Eric will not be lowered by three feet, or more for prevailing low water — not taking into account the effect of the further deepening of the outlet of the lake which may recur at any lime. This last statement introduces a new feature. The tendency of the river is always to scour and lower the bed of the outlet of the lake, as actually occurred to the extent of one foot immediately after 1890. Xow, with the increased velocity occasioned by the slope of the river, augmented from fourteen to seventeen feet, the deepening of the outlet must be acceler- ated at the bridge section, and on the more rocky rim at Fort Erie town. While the full amount cannot be precisely stated, it presents a serious ]:)roblem from the economic side of the question, which cannot be avoided even if overlooked. With the lowering of Lake Erie, the same effects will follow, and Lakes Huron and Michigan in their turn will be lowered a little later. So much for the physics of the river as the question appears to me. But has not the diversion of water already in use pro- duced an appreciable effect ? The five thousand cubic feet per second taken by the Chicago canal has an effect of lowering Lakes Michigan and Huron by '26 of a foot or over three inches. The volume of the Chicago canal increased to ten thousand cubic feet per second will double this amount, or lower Lake Huron by somewhat more than six inches. This Chicago diversion of five thousand cubic * Chezy formula is : v = cVR. S., where v represents velocity; c con- stant requirement of river to be determined (which in large river does not differ greatly for small changes) ; R, the radius or depth ; S, slope or fall. 18 274 FALLS OF NIAGARA t^eol. Surv. feet affects tlie level of Lake Erie by -22 of a foot,f and Lake Ontario bv "19 of a foot, which in each case will be doubled with the increased discharge bv the Chicago canal. A marked increase in the rainfall has occurred since 1900, as compared with the mean rainfall during the decade ending with that year. On examining the tables of the fluctuations of level and rainfall, and taking into account the effects of the Chicago canal upon different basins, I find from both the results that Lake Erie should have risen "46 of a foot more than it actually has done, when compared with both Lake Ontario"^ and Lake LIuron. What is the cause of this failure to raise the level of the Erie basin, where the rainfall has increased not only over the amount of the j^revious period, but also in excess of that of the other basins. Certainly ther^ is but one explanation, namely, the excess has been diverted by the power uses from above the rim of Upper rapids during the five years ending with 1905, over the mean of that of the ten preceding years of low water. Adding to this the effect of the Chicago drainage canal, the fall which the lake has experienced from the artificial diversion amounts to 0-68 of a foot in height, Ui/iJ^A. Lh Qj, about eight inches. At the close of 1905 the total diversion I^C/kfiCffii^^oJ V)f water was equivalent to the lowering of Lake Erie by about ^ nine inches. Here then is direct evidence that the diversion of water has lowered Lake Erie, to more than eight inches, while as yet not over one-quarter of the franchise power has been brought into use. But the superficial observer would not see the change on account of the increased rainfall, by which the actual present t U. S. engineers give this at 0-19 foot. * This figure is found thus : with the increased rainfall. Lake Ontario has risen 0-62 foot, to which must be added 0-19, the effect of the Chicago canal, or 0-81. As the lowering of Erie and Ontario are found to have been nearly equal, Erie should have been correspondingly raised (0-62 foot for increased rainfall and 0-22 for Chicago canal diversion, or 0-S4). This last 0-22 and the actual observed rise of 0-15, is all that can be accounted for. leaving 0-47 foot as the amount of lowering of Lake Erie by the diversion of water at Niagara falls from above the First cascade. From the changes of the rainfall, the result was found to be 0'45. of Canada] lOWERIJTG OF LAKE ERIE BY DIVERSION 275 level of Lake Erie is not lowered below that of previous years. Had not the diversion occurred, Avhile a corresponding rise actually took place in Lake Ontario and Lake Huron, Lake Erie would have been raised 0*68 of a foot (0-40 at Xiagara and 0-22 at Chicago). While it is more difficult to recognize small fluctuations than large ones, all must be taken over a considerable length of time, for neither a day, nor even a year, is sufficient, yet this demonstration of the -effect upon Lake Erie of the changes brought about by the power diversion, confirms the estimated effects on the same. A lowering of the higher lakes (LIuron and Michigan) must also occur when the change in the Erie basin is completed. This lowering of Lake Erie, during years of mean stage of water, and still more during peri^ods of unusual lowness, will produce serious effects on the canals and harbours, for Avliich provision must be made. 18i CHAPTER XXII. CHANGES OF ONTARIO LEVELS SHOWN IN IROQUOIS BEACH. Early observations concerning the Iroquois beach. Iroquois beach. Lower beach and increased descent Characteristics of Iroquois beach. of the Niagara. Tilting of the land recorded by Sudden changes of level. EARLY OBSERVATIONS COjSTCERNING THE IROQUOIS BEACH. ]\[eiitioii has been made in Chapter xv. of the higher stages of water at the month of the ISTiagara gorge shown in the now elevated beaches. The most perfectly developed, and also most carefully studied one, is the Iroquois beach, and its importance lies not merely in its height at the l^iagara, but in its now^ up- ward tilting on proceeding northeastward, so that it becomes one of the standards of late terrestrial movement. Xarrow gravel ridges traversing an otherwise clayey plain, running in directions more or less parallel to the lake shore, were made use of as Indian trails. These were adopted by the original settlers of the country as the main thoroughfare, and constitute what are called the Eidge roads. Fragments of such occur about all the lakes. Those south of Lake Ontario, from Lewiston on the E^iagara river to Rochester, were described so long ago as 1811 by DeWitt Clinton.* The character of the roads so nearly resembles that of the present lake beaches that many early writers regarded them as of such origin. But they are not horizontal. In discussing the disturbed levels of the roads between Rochester and Lewiston with Prof. Desor in 1851, f Prof. Stoddard said : ' The unequal elevation of the different parts * Journal of the New York Historical Soc, p. 68, 1811. t Proc. Boston Soc. Nat. Hist., Vol. III., p. 358, 18.51. 277 278 FALLS OF NIAGARA f^^^ol- ^urv. of the Ridge road was not incompatible with the theory of there having been a beach, if it were supposed to have been raised to its present position. The elevation of so extensive a tract to a height of 500 to 600 feet' (an incorrect reference to sea level) ' could hardly have occurred without having produced inequali- ties at different places as great as are found of some thirty or forty feet.' This is the earliest clearly expressed idea known to the writer, of the deformation of the land being recorded in the raised beaches, which was necessary to account for their origin. A fragment of the beach west of Toronto was described by Mr. (now Sir Sandford) Fleming, in 1859.'^ In 1863,f Prof. Robert Bell correlated the Davenport beach near Toronto with that at Lewiston, thereby suggesting a greater rise of land in the vicinity of Toronto than at Xiagara river. The details of this same beach through Hamilton and elseivhere at the head of Lake Ontario were described in the Geology of Canada, and later by myself in 1882.:{: At the eastern end of Lake Ontario Mr. William Dewey, in 1836,** proposed to locate the Rome and Watertown rail- way, but found that this beach rose too rapidly. Thus the atmos- phere was full of the idea of the differential rise of this old shore line. In 1842 Prof. Hall recognized that the beach was of unequal height, but did not give prominence to the deforma- tion. In 1885 Mr. G. K. Gilbert§ made further exploration from the ISTiagara river to ISTorth Adams, south of Watertown, where he terminated the beach, and gave the measurements at several points. In 1887 I continued the exploration of the beach from ISTorth Adams to Prospect farm, four miles beyond Watertown.il In that year, and subsequently, I made further * Canadian Journal, Toronto, 1859. t Geology of Canada. 1863. t ' Geology of the Region about the Western end of Lake Ontario,' J. W. Spencer. Canadian Naturalist, Vol. X., 1882. ** Quoted by Dr. Hough in History of Jefferson Co., N.Y. § Science, Vol. VI., p. 222, 18S5. II ' The Iroquois beach, a chapter in the Geological History of Lake Ontario,' Trans. Royal Soc. Can., Vol. VII., Sec. IV. pp. 121-134, 1889. Abstract in Science, Vol. XI., p. 49, 1888. of Canada] IROQUOIS BEACH 279 surveys of the delta deposits on the northern side of the Adiron- dacks, and correlated a chain of them as the plane of this beach. In 1887 I carried the survey round the Canadian side of the lake to near Belleville, about 150 miles from the head of the lake. In that region it passes round a spur of high land into the Trent valley. Additional surveys were made in 1888, 1893, and 1895. My paper on the Iroquois beach was first read before the Philosophical Society, Washington, D.C., in 1888 (see Science, vol. XI., p. 49), but the first map of the beach and this extending round both sides of the lake was published in the Transactions of the Royal Society of Canada, volume vii.^ pages 121-13-1, 1889. The paper was subsequently revised in the American Journal of Science in 1890.* The name ' Iroquois ' was given by me, upon conferring with Dr. Gilbert who had made the survey on the New York side of the lake. From the survey round the lake, with a determination of the heights at various points, those by myself being instrumen- tally measured, the amount of rise (mentioned by Stoddard) was determined as well as the direction. These could only be ascertained after the beach had been measured on both sides of the lake and afforded means of determining lines of equal rise. Interesting additional details, with the local sinuosities and an extension in the Trent valley beyond my published survey, have lately been completed by Prof. A. P. Colemanf on the Canadian side. The local details on the ]^ew York side have been further surveyed by Prof. H. L. Fairchild i, whose map marks some of the separated ridgelets at the eastern end of the lake. On this and other beaches Prof. Fairchild. has done most detailed work, as well as Prof. Coleman on the Canadian side. * • Deformation of Iroquois beach,' Am. Jour. Sci. III., Vol. XL., pp. 443-451. t Iroquois beach in Ontario, by A. P. Coleman, Rept. of Bur. of Mines for Ontario, pp. 225-244, 1894. 280 FALLS OF NIAGARA [Geol. Surv. CHAEACTEKISTICS OF IROQUOIS BEACH. At the western end of tlie lake the characteristics of this beach are shown in a sand and gravel ridge which may be 300 to 500 feet wide rising fifteen or twenty feet or even more above the sloping plain in front, which becomes more clayey on receding from the sand ridge. In front of the large old valleys great barriers have been bnilt np, snch as the Burling- ton Heights at Hamilton {see Figure 25), and Davenport Fig. 25. Map of head of Lake Ontario showing a lower modern lake beach, and the raised Iroquois beach, locall}' known as Burlington Heights. ridge west of Toronto. A common characteristic is the occur- rence of shallow depressions Avhich once formed lagoons behind the barriers. The Iroquois beach is generally found to be cut out of drift material furnishing the pebbles which have become water-worn by the waves of Iroquois lake. At certain points, as west of the ISTiagara river, the supply of gravel-making material being limited, or the waves encroaching upon the rocky shore, only cut-terraces were produced. In its western extension the beach may be occasionally seen as constituted of three separated beachlets having an amplitude of twenty to twenty- five feet. On proceeding eastward this feature of ridgelets with greater amplitude becomes more striking both on the Canadian and New York sides of the lake. In a recent paper* Prof. Fair- * Pleistocene Geology of New York, 20th Rept. State Geologist, p. 107, et seq., 1902. of Canada] IROQUOIS BEACH 281 cliilds lias marked five or six of these subridges at tlie eastern end of the hike,; bnt there are also lower shore lines. Having worked out the valley-like character of the Ontario basinf 1 was unable to account for the barrier to the lake in the St. Lawrence region until sufiicicnt tilting of the land, subse- quent to the Iroquois epoch, was found to account for the obstruction to the basin. :|: This enabled me to complete the ex- planation of the origin of the Lake Ontario basin.* {8ee Plate XXXIII.) The Iroquois beach is of further importance in showing that for a long time the water stood much higher than now at the outlet of Niagara gorge. The deformation of the beach also shows the amount of warping of the earth's crust in this region since the waters were at this height. TILTING OF LAND RECORDED BY IROQUOIS BEACH. At the head of Lake Ontario the Iroquois beach is 116 feet above the lake; at Niagara river 137 feet (Spencer); at Rochester 189 feet; at Canastota, southeast of Lake Ontario, 194 feet; and at Adams Centre 410 feet (Gilbert); at Prospect farm, four miles east of Waterto^vn, 483 feet (Spencer). The deltas which are regarded as showing the interrupted extension of this same shore plane north of the Adirondacks have a height of 582 feet at Natural bridge; near East Pitcairn 695 feet ; and at Pine 725 feet. On the northern side of the lake the beach rises from Hamilton at 116 feet f Spencer), to north of Carlton station west of Toronto at 178 feet" (Coleman). At Kingston Road, twelve miles east of t Discovery of Pre-glacial Outlet of Erie basin into that of Lake Ontario. With notes on Origin of Great lakes. Proc. Am. Phil. Soc. XIX., pp. 300-337, 188L Also Proc. Amer. Ass. Adv. Sci., Vol. XXX., pp. 131-146, 1881. % Transactions Royal Society, Canada, loc. cit. * ' Origin of Basins of Great lakes.' J. W. Spencer, Quar. Jour. Geol. Soc, Vol. XLVI., pp. 523-533, 1890. Also Am. Geologist, 1890. 282 FALLS CF NIAGARA [Oeol. Surv. or Canada] IKOQUOIS BEACH 283 Toronto, it reaches 212 feet (Spencer) ; north of Port Hope 311 feet (Coleman) ; north of Colborne 355 feet; and two miles and a half north of Trenton the beach is now elevated to 386 feet. (Spencer), The npper terrace at 436 feet, with a front shore line above, is probably the representative of the Bell terrace south of Lake Ontario, which has a height of forty feet or more above the Iroquois plain. Prof. Coleman recognized here a lower bar at 37-1 feet, and recalls attention to the fine boulder pave- ment at the rear of the terrace at 339 feet. I regarded the series from the boulder to the top of the gravel beach (forty- seven feet) as the amplitude of the Iroquois beach, which, nearer the head of the lake, is often faced with boulder pave- ments ; but the range there is only twenty or twenty-five feet. The recent publication on the extended Iroquois shore in the Trent valley, is due to Prof. Coleman, who finds that at Campbellford and near Madoc Junction the upper beach has a height of 4:75 feet. He also recognizes the smaller amount of tilting in the lower beaches. In this region the deformation of the lower shore line is four feet or less per mile, while in the upper beach it is five feet per mile. From these measurements it is found that the beach at the western end of Lake Ontario rises about two feet per mile in a direction IST. 25° E. (corrected.) This is the triangle which affects the Niagara district. The rise increases so that at the northeastern end of Lake Ontario (be- tween Trenton, Richland, and Prospect Parm, jSTew York) the average rate is five feet per mile IST. 16° E. This tilting calcu- lated for Kingston at the outlet of the lake indicates the up- lift of the lake region from the head of the lake to be 400 feet, and if computed to the first rapids of the St. Lawrence (Galops rapids), sixty-six miles directly beyond, the rise should amount to 650 feet in all. (See Chapter xxix.) In the region of * ' The Iroquois beach in Ontario,' Prof. A. P. Coleman. Rept. of Bur. of Mines of Ontario, pp. 225-244, 1904. 284 FALLS OF NIAGARA [Geol. Surv. Watertown the rise is somewhat greater, so that the total amount of differential rise might be taken to be 750 feet. These measurements show a remarkable land tilting at a very recent date, as this old shore line was horizontal when it was formed ages after the birth of jSTiagara falls. It must be re- membered that this beach line was a water level when the present barrier to Lake Ontario at Galops rapids was reduced to 500 feet or more below^ sea level, and the lake was lowered by the amount of its present elevation of 246 feet. If, how^ever, the continental region were sufficiently high, the JSTiagara river could have descended even to what is now^ the deepest part of the floor of the basin of Lake Ontario. In either case it is not surprising to find the deep Niagara River channel excavated after the withdrawal of the Ontario waters subsequent to the Iroquois episode, with the later backing of the lake to its present level. If the Iroquois plane of the Ontario basin be brought into comparison with that of Forest beach of the Erie basin, the rise, from its lowest point west of the head of Lake Erie to a point east of Sheridan, ]^.Y., is found to be 240 feet. Accord- ingly the Iroquois plane is thus proved to have been at least so much nearer sea level than its lowest point now is. Higher beaches of the Upper lakes indicate an even lower plane than the one mentioned, possibly even to that of present sea level. If the whole great lake region stood higher the relative elevations of the beaches remain as shown. However, it is not intended here to go into the question of oscillation of the earth's crust since the commencement of the beach-making period, but only to state such facts as bear upon the changes affecting Niagara river. LOW^ER BEACH AND INCREASED DESCENT OF THE NIAGARA, It was mentioned that the Iroquois beach was characterized by beachlets throughout the range of some twenty-fi.ve feet at of Canada] INCREASED DESCENT 285 the western end of the lake. At the northeastern end, on both sides, the range of beachlets is considerably more. There are lower beaches. Near Watertown is one at 288 feet above the lake, which descends nearly to water level in the vicinity of Oswego, (Gilbert). Passing under the water to the head of the lake I re- garded its position as that which determined the present for- mation of Burlington beach (Figure 25) at the end of Lake Ontario, when the lake there was nearly seventy-eight feet lower. This beach increased in height from the bottom of the lake ast the water gradually rose to the present level, upon the reflooding of the basin and drowned the deeper Niagara channel. The outer part of the drowned delta of the Niagara river owes its position to the same water line. The lowering of the lake increased the descent of the river from the Iroquois shore line to the deepest channel of the Niagara, far below the present level, thus causing the increase in the height of the falls, although later they were again lowered. SUDDEN CHANGE OF LEVEL. Other beaches are mentioned as occurring about the outlet of Niagara river. Each of these marks a period of re^ in the lowering of lake levels. Formerly I had regarded the subsid- ing of the lakes as being secularly regular, so that a general average of the time might l)e taken. The idea of general uniformity must be abandoned. The changes occurred almost per saltum, with intermittent pauses; long epochs of stability, then rapid changes, until the next water line was established, but with greater amplitude of movement in the northeastern than in the southwestern direction. This deformation was not the result of drawing off the water which had left the hori- zontal beaches, but it was due to an unequal rising of tlie land, from some internal earth movement. CHAPTER XXIII. EFFECTS OF TILTED SHORES OF WARREN WATER ON THE NIACiARA DISTRICT. Notes on the study of Lake Warren. Elevation of Forest plane above Elevation of the Forest beach. Iroquois. Tilting of land at eastern end of Warping after a long recession of Lake Erie. Niagara falls. NOTES ON THE STUDY OF LAKE WARREN. During an earlier post-glacial epoch than that of the Iro- quois beach all the Upper lakes were more or less united into one sheet of water, which I named Warren lake or water, after the late General G. K. W^arren, who had first described the shrunken Lake Winnipeg in the Eed River valley, which lake in justice should have been named after him.* The shores of Warren water were first studied on the Canadian side as well as in Michigan bj myself, but since that time Mr. F. B. Taylor has given much attention to the subject.f Earlier work had been done on fragments of this shore in Southern ]\richigan and Ohio, as the beach feature had been long- recognized. But these were not correlated into the plane of a lake shore recognisable on both sides of the lake until my work appeared. The Ridge roads about Lake Erie were old recognized fea- tures like those in jSTew York state. In 1885 Mr. G. K. Gilbert * ' The Iroquois beach ; a chapter in the Geological History of Lake Ontario.' by J. W. Spencer. Trans. Roy. Soc, Canada, Vol. VII., Sec. IV., p. 122, 1889. Advanced abstract in Science, 1888. Although Gen. Warren had described the extended Lake Winnipeg and the channel connecting it with the Upper Mississippi, Dr. Warren Upham named it Glacial Lake Agassiz, against which both Prof. J. D. Dana and myself protested, as the honours of the discovery belonged to Gen. Warren, whilst Prof. Agassiz had nothing to do with it. This was following an arbitrary rule ' that dis- coveries should not be named after living authors,' which certain people unjustly attempted to establish. However, Warren's name" attached to the Greater lakes is vindication. tThe Second Lake Algonquin, by F. B. Taylor, and Am. Geol., Vol. XV., 1895, and other papers. 287 288 FALLS or NIAGARA [Geol. Surv. of Canada] BEACHES OF WARRE:"^ WATER 289 measured the deformation of some of the beaches at the eastern end of Lake Erie.:}: The lowest of them he traced to Critten- den, !N".Y., l)nt this name he did not give to the beach. Snl)- sequentlj, I extended tliis beach to near Batavia, and again Prof. Fairchilds took up the subject and mapped the old shore line some thirty-five miles beyond. § He found it occurring as far as the Genesee valley, and fragmental equivalents of it to the head of Seneca lake. The deformation or tilting of the shore lines has an im- portant bearing on the physics of ISTiagara river. When the lake existed the sheet of water extended from Lake LIuron high over Lake Erie into the Ontario basin. There were earlier stages of the lake when it formed higher beaches, but the most important of these though not the strongest, was the last of the series. After its survey upon the Canadian side, and upon their correlation on both sides of the lake, I named the lowest shore line of Warren water, the Forest beach, as it had not been pre- viously worked out. {See map Plate xxxiv.) ELEVATIO^r OF FOREST BEACH. Southeast of Lake Huron it is strongly marked near Forest, with its lowest point standing at 715 feet above the sea. I sur- veyed the beach eastward and parallel with Lake Erie, to the meridian of the western end of Lake Ontario, where it has an elevation of 770 feet. It then turns northeastward and rises still more rapidly. It frequently cuts across the spurs of drift ridges. The IS^iagara peninsula has too low an elevation for the occurrence of this beach except at one point. From twelve to fourteen miles west of the IN^iagara river, is Font hill, an unusual ridge of drift which attains an elevation of over 300 feet above Lake Erie, rising out of a plain which is only thirty or forty feet above the lake. Around this hill is a % Science, Vol. VI., p. 222, 1885. § ' Lake Warren Shore-lines in West N.Y.,' by H. L. Fairchilds, Bull. Geol. Soc. Amer., Vol. VIII., pp. 269-286, 1897. 19 290 FALLS OF NIAGARA [Geol. Surv. strongly cut terrace witli a fine gravel floor at an altitude of 793 feet above sea level (measured). ISTortliward of Forest I surveyed the old shore line to the base of Indian peninsula where it bends eastward. In this direction it has no special bearing on Niagara falls. On the New York side this beach is 680 feet in height (bar.) at Madison, which is forty miles east of Cleveland. Here the elevation is 6Y3 feet (bar.). The beach continues to increase in height toward the east, so that at Sheridau, IST.Y., its eleva- tion is 773 feet (Gilbert) and at Crittenden 860 feet. Beyond this it rises to Indian falls and again slightly descends as it trends to the southeastward, as established by Prof. Fairchilds. Westward of Cleveland the tilting is reduced to a very small amount, as beyond Toledo its height is 653 feet. Again it rises very slowly west of St. Clair river, and on approaching the latitude of Port Huron its height is 665 feet (Spencer). However, here appears a strong eastern equivalent in the uplift, for near Forest, twenty-five miles across the former strait, its elevation is 715 feet above sea level. From this point the old shore line extends northeastward as shown on the map XXXIV. A word may be repeated with regard to the survey of the beaches. Where they form bars or ridges, with depressions or lagoons behind them, I adopted the highest line of wave action, which would bo from three to five feet above the w^ater surface. This method was pursued on account of the crest being the most constant feature. On the other hand, where the waves were cutting terraces, the most accurate point determinable at the junction of the plain with the bluif behind would bo below water level. The same is true where the l)oach is represented by sand plains. Consequently, when the shore line changes its character a personal equation appears. It was tlius on the Avestern side of the Huron strait, the measurement being low : of Canada] BEACHES OF WAKREN WATEE 291 while on the eastern, where the ridge wns strongly marked, it was slightly above the water plane.* TIT/nXc; OF LAND AT EASTEKX E]VD OF LAKE EEIE." The Forest beach has pnrticnlar importance in connexion with j^iagara, in showing that the earth movements about the western part of Lake Erie were verv small, while the rise of land at the eastern end of the lake was considerable. Between Madison, Font hill (Brown's nurseries), and Sheridan, the average rise in this large triangle is almost one foot per mile in direction X. 45° E. If the more eastern triangle between Font hill, Sheridan, and Crittenden be taken, and this em- braces the outlet of Erie, the average rise is one and a half feet per mile in direction IST. 60° E. As already noted (page 283), the deformation at the western end of Lake Ontario is two feet per mile IS'. 25° E., as recorded in the warping of the Iroquois beach. Slightly greater movements are shown in higher beaches. From these measurements on the Forest beach it is found that there has been a rise in the ISTiagara district of 135 feet more than southwestward at Madison. With this deformation straightened out. Lake Erie was reduced to a small lakelet. {8ee page 296.) This former tilting of the Niagara district would suggest the continuance of the movement so that it might be supposed that the Niagara river would b§ raised to a higher point, which would send the waters into the Mississippi. But from the present survey it is found that the stability of the region now prevails {see Chapter xxx.), without any indication of the direction which recurring movements may take. * High level stores in the region of the Great lakes, and their Deforma- tion, by J. W. Spencer. Am. Jour. Sci., Vol. XLI., pp. 201-211. 1891. 191 292 FALLS OF NIAGARA ^'^^°^- ^urv. ELEVATION OF FOREST PLANE ABOVE THE IROQUOIS. In the earth movements that have occurred the general direction is northeastward, but there are local variations as the rise is always increasing toward the northeast; so that, by adopting different triangles for the determination of the mean rise in each, variation will be found. Combining the deforma- tion recorded in the Forest beach at the eastern end of Lake Erie with the mean rise in the western triangle of Lake Ontario, a corrected determination shows that the plane of the Forest beach was between 400 and 450 feet above that of the Iroquois beach; but it is probable {see Chapter xxx.), that the deformation of both planes occurred at the same time, subse- quent to Lake Ontario falling below the Iroquois level. WARPING AFTER A LONG RECESSION OF NIAGARA FALLS. The deformation of the Forest plane did not take place to any considerable amount until the time when the Iroquois plane of Lake Ontario became warped ; consequently the warp- ing affecting the Niagara district did not occur until ISTiagara falls were very old. (See Chapter xxx.) CHAPTER XXIV. SEPARATION OF ERIE AND HURON ON THE DISMEM- BERMENT OF WARREN WATER— SHRINKAGE OF LAKE ERIE. Dismemberment of Warren water. Original discovery of the diversion Separation of Huron and Erie drain- of the Huron drainage, age Shriniiage of Lake Erie. DISMEMBERMENT OF WARREN WATER SEPARATION OF HURON AND ERIE DRAINAGE. In the last chapter different shore lines extending al)Out the Upper lakes were mentioned, showing that open water extended from Lake Huron to the Ontario basin. As the waters were lowered the successive beaches appeared — the Forest shore being the last or lowest one extending from the Huron basin to the Ontario drainage area. But at last the waters withdrew from even this lowest beach, below the level of the present higher land belt between Lake Huron and Lake Erie — the immediate result being the formation of three lakes, one including more or less of the basins of Huron, Michigan, and Superior, a small Lake Erie, and the Ontario water at substantially the same height as the shrunken Erie. The united upper three lakes I named the Algonquin. That it was separated from Lake Erie Avas shown by my survey of the Algonquin beach, which proved that the waters of this upper lake had a northeastern outlet in place of the present southern one into the St. Clair river. So far as Niagara falls is concerned it established the fact that throughout long ages they received only the Erie waters and thus were of small volume. 293 294 FALLS OF JSriAGARA [Geol. Surv. ORIGINAL DISCOVERY OF DIVERSION" OF HURON DRAINAGE. The first announcement of this diversion was made by the present writer before the Cleveland meeting of the American Association for the Advancement of Science, in 1888, with a short abstract in the proceedings for that year.* This abstract of my survey (then just made) of the raised shore lines, extending for hundreds of miles from the outlet of Lake Huron, throughout their windings to the northeast, formed the caveat of my discovery that the Xiagara river only recently received the drainage of the higher lakes ; and on account of its importance in establishing priority it is repro- duced in Appendix vii. In it this statement appears : ' Lake Warren became dismembered, and Huron, Michigan, and Superior formed one lake ; the Erie basin was lifted out of the bed of Lake Warren and became drained . . . the outlet of the upper lake was southeast of Georgian bay . . . with the continued continental uplift to the northeast . . . the waters were backed southward and overflowed into the Erie basin, thus making the Erie outlet of the Upper lakes of recent date. This it= proven by the fact that the beaoh which marked the old surface plane of the upper Great lake descends to the present water level at the southern end of Lake Huron. The Erie basin is very shallow and upon the dismemberment of Lake Warren was drained by the newly constructed Niagara river (except perhaps a small lakelet southeast of Long point). Subsequently the northeastern warping . . . eventually lifted up a rocky barrier . . . thus making the Erie the youngest of all the lakes ' (page 188-9). This announcement immediately attracted the attention of Mr. G. K. Gilbert,f with whom I discussed the matter freely in further detail, mentioning positions and heights of the Algon- quin beach at various points. * Proc. Am. Ass. Ad. Sci., Vol. XXXVII., pp. 197-8-9, 1888. t 6th .Rept. Commis. State Res. Niag., p. 71-73, for 1890. of Canada] DIVERSION OF HUKON DRAINAGE 295 Upon this data, in advance of my detailed paper and maps, he presented an essay the following year wherein my recent discoveries formed the prominent feature. Alluding to the diversion of the Huron drainage from the ISTiagara river, he begins thus : ' To illustrate the effects of the earlier system of land slopes upon the distribution of water in the region of the Oreat lakes I have constructed the map in Plate 4 ... In the ancient system of drainage Georgian bay instead of l)eing a dependency of Lake Pluron is itself the principal lake receiv- ing the overflow from Huron. Superior, Michigan, Huron, and Georgian constitute a lake system by themselves, indepen- dent of Erie and Ontario, and the channel of the Detroit river is dry. Lake Erie and Lake Ontario, both greatly reduced in size, constitute another chain, but their connecting link, the Niagara river, is comparatively a small stream.' Even in his full statement there is no mention of data for his hypothesis of the diversion of the Huron drainage from Lake Erie, but farther on he says that I stated that there was another outlet (Trent) for the Huron basin, ' demonstrating that during the existence of that outlet also the Detroit river ran dry.' Thus this subsequent reference (with ^ also ') appears as a col- lateral hypothesis, and not as the original data, which it was, of the diversion of LIuron drainage and small Erie lake. This method of presenting the main discovery of the former diver- sion of the Huron waters, the greatest by far that has been made affecting the Niagara river, could not fail to leave the impres- sion that it belonged to him, especially as he has since fre- quently referred to it without even any mention of the author. Furthermore, for his Plate 5, showing the Trent outlet, he had my own unpublished measurements without his statement of the fact, but the citation of my own abstract establishes the priority of discovery in 1888. The great importance of the diversion of the Huron drain- age lies in the fact that it enormously increased the longevity 296 FALLS OF NL\GARA tGeol. Surv. of Niagara falls, hitherto unsuspected. It also removes the possibility of an enlarged volume in the river from a melting ice sheet on the Canadian highlands, as Dr. Gilbert had formerly supposed. Other beaches were also found requiring a lower outlet than the one by the Trent valley. This was provided for by the ISTipissing trench. Along this former outlet of Lake Huron the Indians frequented it as highway, and led the early French explorers from the Ottawa river, by way of Trout lake over the low divide only three miles across to Lake ^ipissing, and thence down the French river to Lake Huron. Indeed, as early as 1615, Champlain followed this route, returning with the Hurons, by way of the Algonquin overflow past Lake Simcoe and Balsam lake and down the Trent river. While the l^ipissing outlet was mentioned theoreticallv by Dr. Gilbert, after my discovery of tlie northeastern diversion of the Huron drainage from the Erie basin. Prof. G. F. Wrighf^ was. the first to observe definite traces of the shore lines in connexion with it, though earlier explorations of Dr. R. Bell showed the occur- rence of beaches there. These have since been further explored by Mr. F. B. Taylor, who has established more fully the IS^ipis- sing outlet by direct observation. f As to the point, to which I found that the falls had receded before the addition of the LIuron drainage, see page 191. SHRi:XKAGE OF LAKE EPaE. Upon the dismemberment of Warren water a barrier existed between Lake Huron and the head of Lake Erie. As the land was not yet tilted toward the northeast, the eastern end of Lake Erie basin was much lower, so that the jiresent outlet at Buffalo s ood relatively 150 feet below the level at the Avestern end of the lake. Accordingly, Erie was drained except a small basin * Bun. GeoL Soc, Vol. IV., pp. 423-5, 1893. t Bull. Geol. Soc. Am., Vol. V., pp. 620-6, 1894. of Canada] TURNING OF HURON DRAINAGE 297 of 1,000 to 1,500 miles in area, in place of 10,000 square miles as to-day. For the relative size of Lake Erie see map p. 300, Plate XXXIV. The evaporation from the water surface would be reduced on account of the small area of the lake compared with that of the present time, as onlj a river and tributaries traversed the now filled lake basin. The area of Eric has sub- sequently been increased owing to the tilting of the land which has caused the backing of the waters. CHAPTER XXV. NORTHEASTERN OUTLET OF AEGONQLTX EAKE (THE THREE TIPPER T>AKES), Northeastern Huron drainage and Amount of tilting shown in beach. diversion from Niagara. Barrier to Algonquin lake. Northeastern rise of Algonquin beach. NORTHEASTERN HURON DRAINAGE AND DIVERSION FROM NIAGARA RIVER. The barrier between the Huron and Erie basins will be described in the next chapter. I surveyed the winding Algon- quin beach for hundreds of miles in 1887-88. It was formed after Huron had shrunk into its own basin and its level was too low for the lake to overflow the St. Clair divide. In this beach was the first suggestion that the Huron drainage had formerly been diverted by a northeastern overflow from Xia- gara falls as mentioned in Chapter xsiv. NORTHEASTERN RISE OF ALGONQUIN BEACH. At Grand Bend, on Ausable river, a few miles east of the present outlet of Lake Huron, the height of Algonquin beach is eighteen feet above the lake. Fourteen miles farther on it is thirty-six feet high, at Southampton 132 feet. Beyond it skirts the Indian peninsula, which was then a chain of islands on which the beach is preserved. This raised shore was surveyed at Ow^en Sound on Georgian bay, and has a height of 1G7 feet. At Clarksburg it is 191 feet. Four miles west of Collingw^ood its height is 187 feet; near Caldwell 170 feet. Again the beach turns northward, approaching nearer the shore of Georgian bay iit Elmsvale, 220 feet : and east of Wyebridge its elevation is 299 300 FALLS OF T^IAOAEA ^^^°^- S"^^'" 260 feet. Swinging round a bigli drift ridge southward, to Orillia, it circles the basin of Lake Simcoe, eastward of which the height attained is 293 feet above Lake Huron. Between this place and Balsam lake the old shore line indicates a former out- let with the water at one time tweutj-six feet deep. These details are taken from my paper on ' The Deformation of the Algonquin beach and Birth of Lake Huron.' "" The Algonquin beach is often broken up into ridgelets throughout a range of twenty-eight feet in height, so that the overflow of Balsam lake would not account for the occurrence of the lower beachlets. This raised shore line is shown on map Plate XXXV. There are also lower beaches and terraces much below the Algonquin plain and Balsam lake outlet, such as those at Clarksburg from eighty-one feet above the lake to lower levels. Remnants of the Algonquin beach in the broken country to the north are difficult of recognition, while the ISTipissing beach of Mr. Taylor passes below the lake level at a considerable distance from the southern end of Lake Huron. For this beach on the western side of the .ake reference must be made to Mr. Taylor's papers. ISTorthward of my farthest surveyed point Mr. Taylor carried his observations confirming the occurrence of the shore line.f The Algonquin beach lies in the region of Lake Nipissing at a theoretical altitude of 600 feet above Lake Huron. Here Mr. Taylor found the highest and best developed beach at 558 feet, overlooking the divide, between Lake ISTipissing and the Ottawa river, Avhich is 112 feet above Lake Huron. AMOUNT OF TILTING SHOWN IN THE BEACH. From this survey it was found that the Algonquin beach is now tilted and rising from the southern end of Lake Huron from one and one-third, increasing to two feet or more per mile • Amer. Jour, of Sci., cited before. t ' Second Lake Algonquin,' Amer. Geol.. Vol. XV., 1895, and other papers. of Canada] j^^^ jj_ OUTLET OF LAKE JIT RON 301 I'r.ATK XXXV. Map of Algonquin and Nipissiiig Ijeaches and nortlieastHrn outlet of the Huron drainage.* * Height of lakes here given refer to mean of 1860-1905, including that before lowering of outlets. 302 FALLS OF :v-L\GARA [G^°'- ^urv. on proceeding nortliward. Afier skirting Georgian bay the rise is found to be tliree feet per mile in a direction j^. 20° E., and farther on it is 41 feet per mile ]^. 25° E. The deformation will be further considered when discussing the JSTipissing out- let. In the lower beaches a smaller amount of tilting was found. The great valley in which Lake Xipissing lies is much lower than the temporary overflow by way of Balsam lake and Trent valley. Below the Iroquois plain fragments of the lower shore lines have been observed. In 1891 I stated that ' with the continued original uplift, the waters of the Algonquin were lowered as shown by numerous beaches, until the lake was dis- membered and Superior, Michigan, Huron, and Georgia had their birth and drained through the last at the level of Xipis- sing outlet only, by a river flow^ing through the Ottawa valley.' But the direct observations were not established until made by Prof. Wright and more fully by Mr. Taylor, who connected a lower beach with the Xipissing overflow.* It may be said that if all the warping shown in the various- lake basins were levelled ofi^, the region would be much lower than now. On the other hand, if glacial conditions were due to increased altitude the continental region would have been relatively high. BAEKIER TO ALGONQUIN LAKE. Such being the case, the question arises what held the waters at the higher level of the Algonquin plane ? Some thought there w^as a glacial dam. My w^orking hypothesis was sea level. Mr. Taylor was led to consider the Nipissing depression as * The Ancient strait at Nipissing, BulL GeoL Soc. Am., VoL, V., 1893. Of Mr. Taylor's papers see ' Ancient Strait of Nipissing,' lb. vol. 1, pp. 621-626, 1903. ' The Limit of Pre-glacial Submergence in the Highlands east of Georgian bay,' Am. Geol., vol. xiv., pp. 273-289, 1894. ' Second Algonquin and Nipissing beaches,' lb. vol. xvii., pp. 397-400,1896. ' The Nipissing and and Mattawa rivers, the Outlet of the Nipissing Great lakes,' lb. vol. xx., pp. 65-66, 1897. Prof. Wright's paper was cited in the previous chapter. Sec Bull. GeoL Am., vol. iv., pp. 423-425, 1893. of Canada] SEPARATION OF LAKES HURON AND ERIE 305 being- a strait, thoiigli later he thought that a glacial dam ob- tained at a point beyond his observations. At the southern end of Lake Huron beaches rise to twenty- five feet, which might be considered as the continuation of the Algonquin shore line, but this may include constructional planes of the rising waters of the lake. The Algonquin beach skirts the end of the lake and does not turn into the St. Clair river, which cuts directly crosswise of it. East of the river the shore line is often characterized by sand dunes, but the ridge, reaching to a height of eighteen feet above the lake, or eight feet above the ground either side, forms a spit which at one time separated a lagoon from the river. {See next chapter.) CHAPTER XXVI. HEAD OF THE ST. CLAIR TRIBUTARIES OF ALGONQUIN LAKE. Terraces about Lake St. Clair. Head of St. Clair tributaries. Depth of drift about the St. Clair Date of drowned St. Clair valley, outlet. TERRACES ABOUT LAKE ST. CLAIR. On examining the banks of the river and the shores of Lake St. Clair low terraces and sand plains, with some gravel ridges, occur to heights of ten to fifteen feet or more above Lake Huron. These are cut out of drift material prevailing every- where to a great depth. Southward of the lake, at a distance of fourteen miles, before reaching Marine city, the higher watershed of the country approaches the river with a height of forty-five or fifty feet above the lake. Here the St. Clair river has its minimum breadth and adjacent to it is a cut terrace or raised bottom thirty-five feet (bar.) above the water, with the banks rising ten or fifteen feet higher. This point is fifteen to sixteen miles from Lake Huron. A terrace on the western side is from 100 to 300 feet wide, but less conspicuous on the eastern side. It occurs for a length of over a mile. That this was a barrier when the waters of Lake Warren Vv^cre settling into the Erie and Huron basins there seems no doubt but the streams, afterwards descending rapidly to the north robbed the opposite drainage so that the real divide crept southward to the present surface of Lake St. Clair. DEPTH OF DRIFT ABOUT THE ST. CLAIR OUTLET. At the Fibre Works in Port Huron a well showed washed lake sand to a depth of seventy-eight feet, below which occurred 20 305 J06 FALLS OF NIAGARA [Geol. Surv. blue clay twenty-four feet, rounded j^Tavel five feet, clay with glaciated stones eighty feet, or 187 feet in all, reaching to a point 162 feet below lake level. This character is repeated on the eastern side of the river ; and at a point eight miles distant a well was sunk to 160 feet before reaching rock. This represents a depth of 130 feet below lake level. Fig. 2(5. Map of the head branches of the drowned cOt. Clair river, establishing a separation of the Huron and Erie drainage and former northern discharge of the Huron waters. HEAD OF ST. CLAIR TRIBUTARIES. The soundings of Lake St. Clair and the channels among the St. Clair flats proved exceedingly interesting. On observ- ing the chart the first impression produced is what appears to be a delta as the St. Clair river enters the lake. However, the river carries no sediment except the washes of its own banks. The apparent delta is low ground shading off into marshes, but among these marshes, with a depth of one to five feet through the flats barely rising above the surface of the water, the channels are found to have a great depth, reaching to forty, of Canada] KEVEKSED ST. CLAIR 307 sixty, seventy-two, and even ninety feet below lake surface. These channels appear in several branches, converging and en- larging as they approach the trunk which now forms St. Clair river, near the village of Algonac, sec Figure ^0. The occurrence of this system of drainage heading in the shallow waters of Lake St. Clair, and reaching to a depth of ninety feet, establishes conclusively that the St. Clair river lately drained to the northeast through a very deep channel, at least more than ninety feet below the level of Lake Huron, as did also the tributaries from an amphitheatre at its head. At that time the waters of Lake Huron receded somewhat be- yond the line of Kettle point, or seventeen miles north of Point Edward. Of itself this reversed drainage system would have established the former northeastern outlet of the lake, with a plane about 110 feet below the Algonquin beach, had it not been previously proved by lower beaches of the Huron basin. Within two or three miles of the head of the drowned St. Clair, the lake has nowhere a depth of more than seventeen feet, but beyond the contours show a drainage southward into the Detroit river, even before reaching the river itself. The drowned St. Clair has been partly refilled by the deposit of current- worn sand ; and it can only be supposed that the lake sand shown at the Fibre Works is the refilling of the channel as the lake waves were building up the flats on the return of the waters of Lake Huron when they overflowed into Lake Erie. DATE OF THE DROWNED ST. CLAIR VALLEY. These channels of the drowned St. Clair were a late post- glacial feature formed after the dismembering of Lake Warren, after the shrinkage of the Huron waters from the Algonquin plane, but before the tilting of the IS^ipissing outlet which turned the Huron drainage into Lake Erie and the ]*^iagara river. This new observation is interesting evidence showing 20 ,V 308 FALLS OF NIAGARA f^^^^" S"^'^' how reeentlj it is since Lake Erie was receiving the waters of the Upper lakes. In this chapter an important observation of Mr, Taylor's was almost overlooked." He had found that several streams tributary to both the river and the Lake Saint Clair showed clearly the drowning effects, and that their channels were formed while the river was a tributary of the shrunken Huron lake. I^ow they have a depth of ten feet or more and are navigable to some distance above their mouth. He concluded that the channels were cut when the base level was twenty-five to thirty feet below the present, and that they were drowned upon the diversion of the Huron waters to the south. This confirmed the idea of the northeastern outlet of the Huron basin. These tributaries observed by Mr. Taylor are not the ones brought forward by the writer, which now head beneath the surface of Lake Saint Clair coming from the amphitheatre at the Avatershed between the then Huron and Erie basins, show- ing a much greater depth than he found. The principle in- volved is the same. * Proc. Am. Ass. Ad. Sci., Vol. XLVL, p. 202, 1897. CHAPTER XXVII. AUGMENTATION OF NIAGARA DISCHARGE BY ACCESSION OF ALGONQUIN WATERS— THAT IS HURON DRAINAGE. Nipissing — Ottawa outlet. Amount of terrestrial tilting and addition of Huron waters to Lake Erie. Origin of Lake St. Clair. NIPISSING-OTTAWA OUTLET. In the surveys of the high level shores about Lake Huron, beaches at lower levels than the Algonquin were found at many points. These necessitated a lower outlet than that of the Al- gonquin shore line. In the original surveys I did not anticipate beyond the points actually observed ; nor had at that time any data been found directly connecting the beaches with the well- known ISTipissing depression. But it was self-evident that a lower outlet in that direction was needed. Trout lake on the present Ottawa drainage is only three miles from Lake Nipissing, and at its head, on the northern side, is a strongly marked terrace covered with a boulder pave- ment fifty to sixty-five feet above it (eighty-five feet above Lake ISTipissing), this being above the divide by twenty feet or more. In this direction Mr, Taylor has made the fullest explora- tions. He has connected this shore with the occurrence of beaches farther south,* and has appropriately named it the ISTipissing beach. It was only this which was needed to explain the outlet of the lake during the epoch of low shore-lines. The ISTip- issing strand is now from 162 to 175 feet (the inner margin) * Bull. Geol. Soc. cited before. 309 310 ' FALLS OF mXGARA '^^®°'- ^^'"''• above Lake Huron. Between tins and the higher shore line at 558 feet Mr. Taylorf found remains of other shore lines, but it is the Nipissing beach which is the most important feature. Only after the floor of the channel had been raised was the last drainage to the northeast entirely closed. During the time of the ISTipissing plane the waters of Lake Huron were very much withdrawn from the southern end of the lake, see Plate XXXV. In the last chapter the depths of the then tribu- taries have been shown 'to have reached ninety feet or more below the present lake surface. AMOUNT OF TEEEESTBL4.L TILTING ADDITION OF HURON WATERS TO LAKE ERIE. The floor of the Nipissing-Ottawa outlet rises to 112 feet above the lake. However, there are some narrow stream valleys crossing the divide which are only ninety-seven feet above Lake Huron (A. St. Laurent). The amount of tilting since the Algonquin waters were at the ISTiDissing plane amounts to some- what more than 112 feet and fifty feet, the height of the beach, to which must be added ninety feet, or to whatever additional depth the jSTipissing beach at the southern end of Lake Huron is submerged, — in all about 250 feet, to determine the tilting since the Nipissiug episode. But of this amount a warping of only 112 feet less about 22 feet for the now trenched St. Clair barrier was required to turn the Algonquin waters from the Huron basin into Lake Erie. At tlie time when the waters of the Huron basin were withdrawn so far fr.-m the St. Clair divide, Lake Erie, repre- sented by only a small sheet of water, was at a plane of about 150 feet be'ow the |)reseiit St. Clair level. The ])eri(»d of the northeastern drainage of the Huron basin was one of long duration, as shown both by the strength of the Nipissing shore line and the depth of the St. Clair drainage t ' The Nipisslng strait ' cited before. of Canada] ORIGIN OF LAKE ST. CLAIR 311 valleys. The story of the recession of Niagara falls confirms this, and here a compntation may be made of the date ^vhen the rim of the basin near Lake ISTipissing rose so high as to send the Huron waters over the St. Clair divide into the Erie basin. I regard the date of the warping of the earth's crust which raised the Nipissing rim to have been substantially synchronous with the rise of the land at the outlet of Lake Erie, which caus- ed the backing of the waters to the extent of the present lake. ORIGIN OF LAKE ST. CLAIR. This shallow lake excavated out of drift material may be regarded as the lowered col between the dro'svned St. Clair and Detroit rivers, where the atmospheric agents had produced a shallow amphitheatre at the head of both streams. Subse- quently, upon this district being flooded. Lake St. Clair was produced, though modified by its own lake waves acting upon the low shore. CHAPTER XXVIII. RECENT CHICAGO OVERFLOW. Old beaches about Saginaw bay show the tilting of the de- serted shore lines. West of Port Huron I have found the height of the Forest beach to be 665 feet ; northwestward at Verona Mills Mr. Taylor's measurement is 765 feet; and at Elsie southward 770 feet; and reduced to approximately 680 at Maple Eapids. This was near the Pewamo col to the southern part of the Michigan basin, and it would permit of the exist- ence of a moderate amount of warping between this point and the Chicago outlet, with the suggestion that the Algonquin plane should pass below the water surface at the southern end of Lake Michigan. But the relationship of these old tilted shore lines, and those more recent about the head of Lake Erie, has not been fully determined; so that it is not possible to state precisely what overflow has occurred into the Mississippi valley from the Algonquin basin. Some years ago I suggested a tem- porary overflow of part of the Niagara drainage. Dr. Ednmnd Andrews,f and later Mr. Frank Leverett,* have made contributions on the Chicago outlet. Omitting the height of the upper Michigan beach, which has an elevation at Homewood of seventy-four feet above the lake, and of lower ones, though of the higher series, there is a lower beach. Here the channel did not exceed seven feet in depth according to Mr. Leverett, the divide being eight feet above the lake. Beneath this level of eight feet there are lake sands superimposed upon t Trans. Chicago Acad. Sci., Vol. II., pp. 1-23, 1870. * BuU. Chicago Acad. Sci.; Geol. and Nat. Hist. Survey, No. 11, 1S97. 313 314 FALLS OF NIAGARA [Geol. Surv. drift, sliowing the higher lake stages. Again, in part, the sum- mit is marshy with beaches at ten to twelve feet. Mr. Leverett says that these shore lines ' seem to have been formed after the southeastern outlet of the lake was abandoned' (p. 74). And again he says : ' These beaches are to be referred to the action of the present lake ' (p. 78). This manifestly recent lake level of Michigan can only be accounted for by the superior height of the Saint Clair outlet. The position of the Chicago outlet is sliovai in Figure 27, from a map furnished some time ago by Dr. Andrews. Fi^. 27. Map of the low summit between Chicago creek and Des Plaines river which is only eight feet above the lake, with a trough a mile wide. A counterpart of this condition is seen about Lake St. Clair where there are sand plains cr, in one case, a gravel ridge rising to twelve or fifteen feet above Lake Huron. The St. Clair has lately been lowered 1-66 feet by the scour of the incoherent materials of its bed, so it can liardly 1)0 doubted that at a very late date, there may have been a shallow temporary overflow in the direction of the Mississippi, as previously suggested. of Canada] CHICAGO OVEKFT.OW 315 The Chicago outlet has often been associated with earlier glacial conditions. Be this a3 it may, the recent investigations point to a subsequent temporary overflow in that direction ; possibly to the extent of the channel which jNIr. Leverett thinks had a depth of seven feet. This relationship between the Chicago and St. Clair outlets now throws light on the lake history, and the contraction of the gorge, which at the Whirlpool rapids, is due to a partial diversion of the lake waters over the Chicago divide. The hypo- thesis of the future diversion of the Niagara to the Missis- sippi, with the sulanergence of Chicago, will have to be aban- doned as there has been discovered a cessation of earth move- ments (Chapter xxxi.) which would bring about such a change; and in the course of some centuries, upon I^Tiagara falls retreat- ing a half mile or a little more, the channel of the river will be rapidly deepened in the drift of the basin above the falls. This will further postpone such a western diversion, if the future tilting of the land should occur. CHAPTER XXIX. ST. LAWRENCE CHANNEL WITHOUT HURON DRAINAGE Smaller channel of St. Lawrence. Terrestrial deformation shown in the rapids. Ancient Niagara and present Ottawa rivers compared. SMALXrEK CHANNEL OF THE ST. LAWKENCE. The history of ISTiagara river is only the expression of that of the lake region in all its complicated parts. The investiga- tion of one feature imperceptibly leads us afield until we pass the boundaries of our original topic ; thus it is here. But the new observations react and more fully explain the phenomena. The tributaries of the reversed St. Clair confirm the proof of the former northern outlet of Lake Huron, with the diversion of the Upper Lake waters from the ISTiagara river. On the other side, the St. Lawrence river itself furnishes proof of its correspond- ingly small channel before Lake Huron was turned into the I^iagara river. An investigation of the St. Lawrence Avould doubtless bring to light much new interesting material, but only one or two points will here be touched upon. From the outlet of Lake Ontario to the foot of the rapids at Cornwall the direct distance is 110 miles. Then the river broadens out into an expansion from two and a half to four miles wide, extending as far as Valleyfield, a distance of thirty-three miles. This is called Lake St. Francis. There is a small descent from the foot of Lock ISTo. 15 to the lake, and again from the lake to the head of Lock ISTo. 14, at Valleyfield. These, together with the slope (?) 317 318 FALLS OF NIAGARA ^^^°^- ^"'■^■ in the lake amount to about three feet. Throughout the wider part of the lake but little of this descent can obtain, and con- sequently the current is very much reduced. Cornwall is below the foot of rapids, extending from which into the lake is a narrow deep channel, having a breadth of three-eighths of a mile in places, with a depth of sixty-seven feet. The channel is bounded by broad flats submerged to ten feet. The river-like character of this channel is unquestion- ably shown for twenty miles, beyond which the lake has been sounded at only a few points. As the older channel is filled to overflowing its size is not sufiicient for the present St. Law- rence. On the other hand, the sharply defined banks of this channel, with a great depth, show that it was made by a stream with a strong current, though now bars obstruct it at places. Two things become apparent. First, the channel was made by a smaller river; and secondly, this reach of the river has subsequently been filled to overflowing. The flooding of the river banks might occur from a rising barrier at the outlet of the lake, but the trench is too small for the river of the present capacity. The small inner channel is excavated more or less out of drift material, at the foot of rapids descending over a rocky bed. It is such a one as should be ex- pected when Lake Huron emptied by way of the ITipissing and the Ottawa in x'lace of by the Xiagara and the St. Lawrence. A few miles below Valleyfield, the river again widens out into Lake St. Louis, in which the features of the drowned channel are repeated. TJiese channels now mentioned in this connexion confirm the diversion of the upper lake waters from the J^ia- gara in a manner gratifying to one who first discovered these changes in the physics of the river. I have also found that a small channel at the outlet of Lake Erie obtained, which has recently become flooded in the same manner as the plains of Lake St. Francis. Thus there is everywhere proof of the change of drainage. of Canada! EARLY ST. LAWRENCE RIVER 319 TERRESTRIAL DEF0R:MATI0:S' SHOWN IN THE RAPIDS. The depth of the St. Francis channel, compared with that of the rapids below, indicates that it conld not have been formed after the barrier to the lake had obtained its pre- sent height, or that the valley of the lake was flooded, by merely the increased volume of the St. Lawrence, without either a raised barrier, or a diversion from the course of the pre-glacial drainage. Upon the side of the Adirondack mountains the deltas show an extension of the Iroquois plane with a mean rise of about six feet per mile, along a line nearly parallel with the St. Lawrence. How much of the rocky barrier at Valleyfield has been raised to blockade the lake has not been determined ; but a small proportion of the movement since the Iroquois epoch would be suflicient to drown this deeper channel flooded by the increased volume of the St. Lawrence river. At the Gralops rapids the rocky floor of the channel is from ten to thirteen feet below the surface of the river; yet above the rapids the river has a depth of seventy-five feet. So, also, the rocks at intermediate rapids are found to ol^struct deeper parts of the channel above them. Llere I am simply calling attention to these phenomena which are worthy of the same detailed study as has been made on Niagara river. While various questions arise from such a structure these barriers show a recent elevation of the earth's crust in this region. The proof of the elevation has been established along the Adirondacks to the south, but direct evidence adjacent to the St. Lawrence has not been previously suggested. Accord- ingly, the former assumption of the continuation of the defor- mation to the St. Lawrence river is sustained, though the pre- cise amount has not yet been actually measured at this locality. A few miles below Lake St. Francis, the Ottawa river joins the St. Lawrence, forming Lake St. Louis. It was down the Ottawa that the northern drainage of Huron came. While it is 320 FALLS OF NIAGARA [Geol. Surv. interesting to observe the characteristics of the former and in- ferior St. Lawrence river compared with those of the present, it might be equally so to observe the relationship of the present shrunken Ottawa with its recently greater channel which received the Upper Lake waters. ANCIENT NIAGARA AND PRESENT OTTAWA RIVERS COMPARED. The basin of the Ottawa river has an area of 56,700 square miles, while the Erie basin has 40,000. The Ottawa rainfall is supposed to be less than the Erie, being given at between thirty and thirty-five inches a year ; but much of this country is under- laid by rock surfaces, not to speak of the wooded character, so that the rainfall, which runs off, is relatively large. The fluctu- ations of the Ottawa are liable to great extremes* not occurring at the Niagara, yet in the mean stages of the former river may be seen to-day a stream a little larger than the Erie stage of the Niagara. On the other hand the Ottawa river, at that time was as large or somewhat larger than the present St. Lawrence above its union with the modern Ottawa. * In 1870, the extreme high water at Sainte Anne was 86,500 cubic feet per second ; ordinary high water, 41,400 cubic feet; very low water, 1906, 5,820 cubic feet. (A. St. Laurent.) CHAPTER XXX. TIME OF WARPING OR TILTING OF LAKE REGION. Post-Iroquois warping shown in Effects of post-Iroquois uplift on that beach. Niagara river. Fost-Iroquois warping in Niagara „, . „ district. Chicago overflow. Warping of .Mgonquin beach. Cause of earth movements— Fisher's Tilting of Nipissing beach affecting theory. Niagara falls. POST-IKOQUOIS WAKPING SHOWN IN THAT BEACH. The Iroquois beach is directly connected with the history of Xiagara falls. From the head of Lake Ontario to Prospect farm, east of Watertown {see map Plate xxxiii.) is a direct distance of 206 miles with a rise of 367 feet; or, to Trenton, 125 miles, the rise is 270 feet. At the eastern end of this dis- trict the rise is more than tiye feet per mile ; beyond these points deltas of riyers show the extension of thei Iroquois plane, on the northern side of the Adirondacks to a point 2-45 miles from the head of the lake, haying been uplifted by 609 feet. There are terraces in the higher yalleys which I liaye obseryed, but not written upon, and they may throw more light upon the pheno- mena in this region. Howeyer, the limitation of the Iroquois plane, whether formerly blocked by ice or not, is unknown. As shown long ago the component beachlets diyerge more towards the east and north than in the opposite direction, show- ing a greater amplitude of moyement between the intermediate pauses than at ths west. So, also, the warping is more exag- gerated in the older beaches than in the newer and lower ones. This shows a decadence in earth moyements. The great extent of the Iroquois shore line, which was a Ayater leyel, proycs that 21 321 322 FALLS OF NLVGARA '^^«°1- S"^''" it could only have been disturbed after its completion, and no glaciers could have occupied the lake basin while the beaches were being formed — an epoch of repose lasting for a long time. POST-IKOQUOIS WAKPING IN" NIAGARA DISTRICT. At the head of Lake Ontario, including the iSTiagara district adjacent to it, the tilting movement amounts to two feet per mile. On the southern side of the Niagara district the Forest plane has been warped to the extent of a foot and a half. A higher beach indicates a slightly greater uplift. But farther westward and round the head of Lake Erie the differential movements have been small. The tilting about the head of Lake Ontario is in such close accord with that at the eastern end of Lake Erie that it is seen to be one continuous feature. The conclusion, therefore, is that almost the entire amount of deformation shown in the Forest beach occurred after the construction of the Iroquois, and at the same time that the latter beach was uplifted. Had the earth movements been uniform and not intermittent these strong and distinct beach lines would not have been formed. WAKPING OF ALCfONQUIN BEACH. The Algonquin beach was formed subsequently to the Forest, but it cannot be asserted that it was synchronous with the Iroquois. As in the Ontario basin considerable time must have elapsed during the construction of the Bell and other terraces. The Algonquin beach was, however, confined to a separate lake basin, where different conditions as to outlet formerly prevailed ; but, like the Iroquois, the Algonquin plane was several hundred feet above the lowest rim of its basin in the Nipissing region. The rate of its deformation at the south- ern end of Lake Huron is only about a foot and a quarter per mile, increasing to four feet per mile, with a direction east ol north, but pointing to the foci of uplift occurring north of of Canada] TILTING OF LAKE REGION 323 the Ontario basin. Wliethei* synchronous or not with the Iroquois its bearing upon the iSIiagara river is of secondary im- portance after having established through it the fact that the Huron waters did not then overflow into Lake Erie. Stilly the lowering of the waters from the Algonquin plane was like that in the Ontario basin. TILTING OF THE NIPISSING BEACH AFFECTING NIAGARA FALLS. In the Huron basin the waters were lowered to the Nipis- sing plane, — 160 to 170 feet above the lake in the vicinity of Lake ISTipissing ; while at the southern end of Lake Huron they were more than ninety feet helow the present level. Upon the further subsidence of fifty feet the waters fell to the floor of the Nipissing-Ottawa divide. With the renewed deformation of the land the waters backed and flowed over the St. Clair divide into the Erie drainage. EFFECTS OF POST-IROQUOIS UPLIFT ON NIAGARA RIVER. This terrestrial movement was extended to and increased at the outlet of Lake Ontario, and eventually raised the barrier there so as to drown the Niagara channel by the backing of the waters into it. The warping also caused 'the overflow of the Huron waters by way of the St. Clair river. The rise of the Ontario rim so as to flood the Niagara channel was not com- pleted for a considerable time subsequent to the overflow of the Huron waters into the jSTiagara drainage. CHICAGO OVERFLOW. During the rise of the w^aters in the Huron basin a tempor- ary overflow by way of Chicago played a subordinate part, but the only effects upon the recession of Niagara falls were to partly reduce the volume of the river, when the falls were pass- ing the Whirl i30ol rapids. 211 324 FALLS OF NIAGARA ^^^^^^ ^^'-V- CAUSE OF EARTH MOVEMENTS FISHER^S THEORY. Some liave supposed that the lowering of the waters was occasioned bv the withdrawal of ice dams, still if such were the case their presence left large sheets of water forming great beaches at one level. The withdrawal of the water could lower the level but not tilt the land. The tilting is an earth movement. One stage of the earth movement carried the land unquestionalv below sea level, for at Montreal the marine deposits, with sub-arctic shells, occur at 530 feet. The minor oscillations, such as are found adjacent to the coast, are not a subject of this paper, and much is here left unwritten. As to the cause of the warping some have supposed it to be the consequence of an isostatic return to equilibrium upon the withdrawal of an ice sheet. But any ice sheet was far removed to beyond the limit of the observed shore lines, so that the withdrawal of the ice had taken place not simultaneously with the warping, but long anterior to it. In the evidence be- fore me this explanation seems to be at least largely without physical support. Even an isostatic return to equilibrium necessitates interior terrestrial movements. As to the cause of this quiet rising and sinking of the land without apparent crumbling or dislocation of the strata only a suggestion can be offered. For a working hypothesis, one enunciated by the Eev. Osmond Fisher, author of ' Physics of the Earth's Crust,' offers at least a plausible explanation. The iilu'Hduiena in the lake region indicate only one general form of earth movement, affecting, however, in a different degree, all parts of the Lake regions (Lake Superior excepted as not having been brought here under careful analytical in- vestigation as in the case of the other lakes). The foci of greatest rise are on the Laurentian highlands, a considerable distance northward of the outlet of Lake Ontario, whicli mnv b(^ ciilciilatcd from data given in this work. From of Canada] CAUSE OF EAETII MOVEME^'TS 325 these foci tlie radiatina- movements diminish to nearly zero at the head of Lake Erie. Accordingly, it is again necessary to emphasize the fact that the changes which affected Huron were the same as those operating npon Lake Ontario — a case of a rising node of the earth's interior. Mr. Fisher explains such changes on the hypothesis that there is a liquid substratum in which convexion currents exist. ' Somewhere a column of fluid Avould begin to ascend and flow away horizontally at the surface. The waste in the lower parts of the ascending column would be supplied by the inflow of the surrounding liquid, which would depress the isotherms in it, and the whole region affected Avill become cooler . . . and therefore more dense than the average of the general mass, and eventually gravity and friction will cause it to descend, when the processes of the ascending will recur elsewhere. These occur in cycles. The surface above an ascending column would be- come elevated, and above a descending one depressed (p. 216).* ' If . . . . an upward convexion current were to occur beneath the continent, the change of level would be local and would affect the superincumbent area. This would cause the kind of uplift which is found in plateau regions without much crumbling of the strata ' (page 218). * ' A suggested cause of changes of level in the earth's crust,' by Rev. Osmond Fisher, Am. Jour. Sc, Vol. XI., pp. 216-220, 1906. CHAPTER XXXI. NO PRESENT EARTH MOVEMENTS — OR STABILITY OF THE LAKE REGION. Observations about Lake Erie. Table of flucuations at Toronto Fluctuations at Port Colborne and and St. Lawrence stations. Cleveland compared. Absence of earth movements abou^ Table of fluctuations between Port outlet of Lake Huron. Colborne and Cleveland. Importance of terrestrial stability. Stability of earth's crust in Erie Former supposition of future div- basin since fifty years ago. ersion of the Niagara to the No earth movements in Ontario Mississippi. basin. OBSERVATION'S ABOUT LAKE ERIE. In preceding chapters late earth movements are shown ro have occurred, so that it was only natural to conclude that such might still be in progress. A successful effort has now been made to compare the lake levels in their relation to the present topic. In the latter part of this chapter a statement of the earlier studies and conclusions will be given. The raised shore lines westward of Cleveland indicate scarcely any changes of level, while between Cleveland and Pt. Colborne, the late rise amounts to nearly 120 feet. Therefore it is in this section that modern fluctuations should be looked for. At the entrance lock of the Welland canal, situated at Port Colborne, the daily fluctuations have been kept continuously, and those since 1849 have been investigated. The Cleveland record is continuous since 1851. These two points are located 156 miles apart. In the record personal equations appear. The readings at Port Colborne were taken by the lockmasters who recorded only the nearest inch, consequently some men might favour the larger while others the smaller readings, and others might be careless. But in the thousands of records the error will become nearly eliminated by the law of averages. jSTot so, however, if 327 328 FALLS OF NIAGARA [Geol. Surv. only levels, taken by different ol)sevvei's, during sliort periods, be compared. A very small tide exists, and with only one daily reading at a constant hour another source of error will appear during short times of observation, which would be eliminated if longer studies were made; but this is subordinate to records read- ing only to the nearest inch. The intensity and effect of the winds produce the greatest variation, but even this error is reduced when the observations extend over a long period. As the sill of the lock is taken f(jr the Port Colborne datum, there is little probability of any alteration occurring in its position at many feet below frost action. The bench at Cleveland has been changed, but with corrections as close as possible, by the Engineers of the Lake Survey. When taking long periods there is no apparent reason for going behind the datn thus furnished. The table on the following ])ages gives the fluctuations for Lake Erie at Cleveland and Port Colborne for annual and quinquennial periods, also the pi, tlicrehv ending- the life of the falls." The measure- ments were taken from a topographic map of the Lake snrvev. which was lately found to he inaccurate. Upon commencing the present survey, and in making more careful measurements, the results led to the ahandonment of the former conclusions ; hnt it was some time hefore the present stahility of the region was aihi'iucil. On the former occasion the rate of the rise of the rocky harrier was estimated to he such that Niagara river should he drained, as far as Ijuifalo, with the Niagara waters diverted into the Mississippi in 5,000 years, — a time which would have been diminished to 3,5G0 years had the measurements heen reduced to the standard of the rocky rim of the Upper rapids. About two years after the above announcement by the writer. Dr. Gilbertf adopted the same idea of eartli movements, but attempted to prove the question de novo from fluctuations of the lakes. Strange as it may seem his conclusions that Niagara falls would cease to exist in 3,500 years closely coin- cided with my own when reduced to the same datum. He selected thirty-six irregularly chosen days between August 20 and October 30, 1858, and taking the levels on them he compared the height of the water at Cleveland and Port Colborne on thirty-three days also irregularly chosen between June 28 and August 18, 1895. Therefrom he concluded that Port Colborne had risen 0-239 of a foot, or at the rate of 0-65 of a foot ])er century moi'c than Cleveh.md. From his paper I infer that he chose the days of low levels * ' Duration of Niagara falls,' by J. W. Spencer, pp. 471-2. t IStti Annual Rept., U.S. Geological Survey, Part 2, n. 622, 1S97. 22 338 FALLS OF NIAGARA ^^^o'- ^urv. wlien he thought the wind was blowing most gently. But the effect of this was that the levels taken from 1895 were only -03 of a foot below the mean of those of the three months from which he chose, while those adopted from 1858 proved to be much below the mean of the months from which they were selected. It does not appear why the averages in both cases were not taken, or why levels of autumn months, when the velocity of the wind is twenty-five per cent greater than in summer, were compared with those of months of gentler breezes. Had he compared the summer months in both cases by taking the mean levels, the difference would have been 0-04 in place of 0-239. In this case, however, had he taken the means of the whole months actually drawn from, the amount of rise would have equalled only 0-03 of a foot in thirty-seven years, or 0-08 of a foot per century; but no value can be placed on such small differences. Indeed he himself recognizes a probable error of plus or minus 0-057, which gives a range of more than a tenth of a foot. This choice of days and seasons, in- dicating an upward movement of the Niagara district, which if continued w^ould divert the Niagara waters into the Mississippi in the not distant future, must give place to the analysis of the full data covering the mean levels of thousands of days fifty years apart, which establishes the stability of the lake region during that time. The data for comparing fluctuations are given in this chapter, while the monthly variations are found in Tables 1 and 2, ApjDendix v. In the Ontario basin Dr. Gilbert compares the levels of only four days in all, taking them from August, September, and October, 1896, with thirty-four days irregularly chosen between April 17 and June 9, 1874, at Sackett harbour and Charlotte; concluding that Sackett harbour has risen 0-061 of a foot. As has been show^n in the more extended analysis, this amount is a negligible quantity as well as the observations being too few. Even if not so the records of mean fluctuations of Canada J TERRESTRIAL STABILITY 339 for many years at Galops rapids show there has not been any rise in the earth's crust during the last thirty years (this chap- ter, page 334). The same author computed the changes of Lake Michigan, comparing those for a few days in July and August, 1876, with a larger number between the end of July and the end of October, 1896. But this district lies outside the province of this paper and would not affect Niagara falls. From the fuller evidence now brought forward it may again be repeated there is not the slightest indication of present earth movements which will divert the l^iagara waters to the Missis- sippi, flooding Chicago. !N"or can we predict whether future changes will be those of upward or downward movement. Beyond the scope of this book, the question of terrestial stability established in the large lake region, of the interior of the continent, may prove of value in other problems of physical geology. 22h CHAPTER XXXII. RECESSION OF NIAGARA FALLS. Preface. Present recession. Effective height of the falls in re- cession. Former changes in the height of the falls. Variation in the volume of the water. Differential discharge from the Erie basin. Laws of erosion. Character of original river channel. Effects of Falls-Chippawa valley upon the recession. Recession of American falls. Effects of rock structure. Height and volume of falls at thi^ir birth. Increasing height of falls and establishment of second cataract. Subsidence of waters to Bell ter- race. Iroquois shore level. Union of two upper cataracts above Foster flats. Medina or third falls and their great height. Increased volume of Niagara at Foster flats. Effective ending of Medina falls — .Shallower channel below Whirl- pool. Increased effective height of falls above Foster flats. Conclusions as to stages of reces- sion. PREFACE. The many preeefliiig chapters -were a iiecessarv jireparation to this one, in ehicidating" tlie character of the recession of the Falls of iSTiagara, their changino- history, their power, and the determination of their age. This chapter seems to be separated a long way from the one on the recession lines ; but so complex has the history been that the work of the falls could not be understood until investigations had been made upon the various conditions and phenomena which have accelerated or retarded the retreat of the cataract. In a general way the recession of the falls depends upon three factors — their height, their volume, and the character of the rock formations, which are being channelled by the stream. The falls have retreated across a very level country, underlaid by almost horizontal rock formations having a great degree of uniformity, as was noted even so long ago as 1789 (Ellicott). 341 342 FALLS OF NLVGARA tGeol. Surv. Few people have supposed that the recession was not uniform; fewer still would think that the story had not all been told concerning a commonly known natural feature. The history had been partly anticipated in the discovery that the volume and descent of the water have not always been the same, and that some of the features have changed to a degree. PKESENT EECESSIOF. Although the form of the falls was sketched out in 1819, yet the actual retreat has only been measured between the years 1842 and 1905. During- these sixty- three years, seven and three-quarter acres have disappeared. The width of the gorge at the falls is 1,200 feet, the excavation of which is due directly to the cataract. The mean recession during these years is found to be 4-2 feet per annum, while during the last fifteen years it was reduced to 2-2 feet; the previous fifteen years showed an annual mean rate of 4-54 feet (page 36, Chapter iii.). It must be borne in mind that there are years of no appreciable falling away, succeeded by others wdien large areas collapse. '. EFFECTIVE HEIGHT OF FALLS IN RECESSION. The height of Niagara falls might be taken from the crest of Greens or First cascade of the Upper rapids, which is 312 feet above Lake Ontario, or 212 feet above the surface of the river in the cauldron below (as shown in longitudinal section figure 28). However, in the Upper rapids, the erosive work of the river is exceedingly slow, so that in comparison with tliat of the main cataract it is an almost negligible quan- tity. The accelerated velocity of the water adds something to the force of the descending column ; but on account of the rising knobs and blocks in the bed of the rapids, the force is impeded, especially throughout a considerable portion of the breadth where the water is not deep. Accordingly, the theo- retical gain in power by the descent of the water on the Upper of Canada] EECESSION OF NIAGARA FALLS 343 ^ '"I ILiL]i_ll_ ^i '^J2 & S =S •' ri^ o O !3 0) — ^ a; C 'SD c3 » SO I 2^ 8.° i O 05 s-rS >>^ ^'S ^ cS C rf o bo ^ 1- fi O He's s ^ tioS-g s J •c o^ g.So-;^ cs o 0'$- a o o be ^ ^-' ^^.Vi "^ oS.2 P.^l ^^ 1 1 g ;: I be ■ be '^ ^ ^ .-IS •'J-i -w_; j; 2 oj Mx:,-, ??-C SP 0) g J^ s i^^ o i £: be i Oi ^ Si, p . <4-. OJ ^ 'be Q S .2P ^ , 03 S S c s beS!S . •- 2 ^ 0)— -■'0.2 _i:> g Ij 3 M tj o _ t-. O i- cS >M ^ ' » o « ^ bOi be ; ? S 3 j=-" o '^ S3 ':3t3 '-' te o nl V n* ^ ^T^ r- ci- TS & ^ ^ f^ ^ ^ i< b OS r- -3 « 3 be 3i ^-;:3 be2 S 'Tr O cj D •S 9 m 'bcJ5 =S o:S;o -3 ° 33 h^ . ^ s ^s be a; s fullv as wide as that of the channel above and below, was nearly drained. EFFECTS OF FALLS-CIIIPPAWA VALLEY UPON THE EECESSIOX. At Hubbard point the capping thickness of limestone was increased, but south of that point it was again reduced in thick- ness, on account of the superficial depression. As the falls were receding southw^ard from Hubbard point the channel was constantly being deepened as the drift was washed away, for the Falls-Chippawa valley has a gradient becoming lower on proceeding soutliward (page 165, Chapter XIII.). Thus it would seem that the work of the falls was somr-wliat accelerated. However they are now at the point where the reopened valley is becoming buried, with the crest line turning from its recent course and commencing to ascend the side of the ancient valley, thus giving rise to the Upper rapids in ]ilace oi tlie advancing crest continuing in the pre- glacial valley. of Canada] KECESSION OF NIAGARA FxiLLS 353 Along this course the falls are constantly encountering a thicker layer of capping Niagara limestone, which at the same time, on account of its dip, is reducing the height to which the underlying shales rise. This gives a stronger and more resist- ing structure. A large part of the late reduction in the reces- sion is probahly due to this feature. So, also, in the future I should expect the rate of recession to be retarded until the falls shall have passed the line of the Greens or First cascade. After that the Upper rapids will gradually be lowered on the channel cutting through the hard rim. EECESSION OF AMERICAN" FALLS. This is very slow, amounting to about 0*6 of a foot per amium. The volume is taken at only about seven per cent of that of the whole river, but this amount subtracted from the main cataract might be expected to somewhat reduce the rate of recession of the greater falls. The American falls descend on an accumulation of dislodged blocks, so that they are not deepening the channel. EFFECTS OF THE ROCK STRUCTURE. To some extent the recession of ISTiagara falls is accom- plished by the wedging off of the upper layers along the margin of the crest (page 35). More than this the currents of the Upper rapids are producing but little effect except by dissolving out and opening up lines of joints, thus facilitating the attack of the powerful currents on the frontal masses of limestone rock. Otherwise the recession is accomplished by the undercutting in the soft shales which form the base of the walls of the gorge. As has been found from soundings, throughout a considerable portion of the breadth of the gorge, the present forces of the river are not breaking through the hard Medina sand- stones which lie nearly ninety feet below the surface of the river. The soundings under the great cataract show that their 23 354 FALLS OF NIAGARA ^^^°^- S"^^- excavating power does not exceed 80 to 100 feet. Below the Goat island shelf, the gray sandstone has been penetrated to a great depth. Where the forces of the falls were greatest they not merely pierced this hard Medina gray sandstone, but they reached to the deepest sounding, which, however, was only ac- complished when the level of the river was lower. As great masses of rock have fallen on the higher shelf since the first surveys were made it is quite probable that some of these are still remaining and protecting the higher shelf beneath the eighty feet of water. The formations have a remarkably uniform character throughout the gorge, but the Medina gray sandstone is still below the surface of the river until reaching to the mouth of the ^Vhirlpool. The depth of the river above Cantilever bridge is 186 feet or far below the zone of the sandstone. In the Wliirlpool rapids the channel reached an equal depth, though it has since been refilled by the rocks which have fallen from the sides of the gorge (page 149). The Whirlpool is also deep. Accordingly the river above this point has a channel uniformly extending to about ninety feet below the level of Lake Ontario, showing a very constant penetrating power for the falls, below the Goat island shelf, as far as the Whirlpool. All these features concern the !Kiagara river during its modern history. The rock structure beyond this point does not affect the recent history, as all the harder beds lie at a con- siderable height above the river. While the limestone founda- tion of the upper river has been variable, on account of topo- graphic features, it grows thinner at the end of the gorge. HEIGHT AND VOLUME OF THE FALLS AT THEIR BIRTH. Upon Lake Ontario subsiding below the level of Erie there was a sudden drop of about thirty-five feet. Then a terrace was formed which now has a length of 900 feet, extending up- ward along the river. The northern projecting end (450 to 500 of Canada] RECESSION OF NIAGARA FALLS 355 feet), beyond the line of the ' monntain/ is covered by a delta deposit formed by the river accumulations covering the under- lying beds, while the more southern 450 feet are cut out of rock formation as shown in Plate xxviii. The falls then, at their birth, had a height of thirty-five feet, and the episode lasted sufficiently long, so that the incipient canon commenced to be formed. Of the distance of 450 feet which might be assigned to it, a portion seems due to the older topographical depression in the edge of the escarpment. As the river was falling into the lake, wave action as well as river currents prevailed; accordingly the excavation due to the falls does ont equal the above amount, while the delta deposit extended the length of the terrace. Here the capping of the ^Niagara limestone was entirely removed, and the floor of the terrace was superimposed on jSTiagara shales. This water level was of considerable dura- tion as the well-defined shore line west of the gorge proves (page 19Y). Upon the withdrawal of the lake w^aters, there seems to have been a short pause allowing for the structure of another subordinate terrace. The breadth of the river may have reached 1,400 feet, though this width might have to be reduced before the subse- quent falling of the walls on the eastern bank. The average depth of the water at the modern American falls is taken at 1-5 feet, and its volume does not exceed seven per cent of that of the whole river of to-day. The original river having a greater breadth than that of the American falls, and larger volume of water (fifteen per cent) the Niagara falls at their birth should have had a somewhat greater depth of water than the smaller American cataract of the present day. The terrace showing height of ISTiagara falls at their birth is seen in Plate xxviii., page 199. This being densely wooded, the terrace form is not seen so well as when examined on the ground. This is the first time that the original falls have been des- 23* 356 FALLS OF NIAGARA l^^^"'- ^"'■''• cribed, although an uncorrected estimate of the volume of the river was announced in 1894.* INCEEASING HEIGHT OF NIAGARA FALLS AND ESTABLISHMENT OF A SECOND CATARACT. The waters of the Ontario basin suddenly sunk to a level 200 feet above the present lake surface. This exposed above the river level the thick band of Clinton limestone, the height of which on the western side of the river is 240 f 0) 0) T1 ^ C f/j o o s o U . if! II _bC-5 377 CHAPTER XXXIV. FUTURE RECESSION OF THE FALLS. The survey of Niagara falls would be uufiuislied if the facts which bear upon the future recession as they finally appear were not brought together. There are two sides to this ques- tion: one the natural changes in progress, and the other those affected by the artificial modifications. The falls have been undergoing changes tending to modify the recession. The rocky rim at the First cascade is lower on the Canadian side than on the Goat island, and the current naturally sweeps against the banks of stony clays above the Dufferin islands. These have been constantly washed away, tending to lower the river as the bedrock sinks in that direction. The rate of recession of the falls during the last fifteen years has been greatly reduced, this being caused by changing features, with effects due to the diversion of the water for power purposes. The change is mostly due to the turn in the course of the river, with the apex of the falls now begin- ning to creep up the side of the Falls-Chippawa valley, where the thickness of the limestones forms a more resisting arch, — retarding the recession. About 2,500 feet above the apex, the rim of the First cascade, should be reached in about 600 years, if the mean rate of recession (during the last sixty-three years, or that of the 227 years since the falls were first seen by Father Hennepin) were to continue. In the meanwhile, as the falls retreat the hard capping stratum is not merely increased by the thickness of the rising floor of the rapids, but also by the dip of the limestone formation. On this account the 379 380 FALLS OF NLVGARA ^Geol. Surv. rim at the First cascade might not be reached for much more than 600 years. At the rate of recession since 1890, the time required would be 1,500 years. The falls, having reached this point and gaining a height of fifty-five feet with the extinction of the Umier rapids, would commence to trench the rim whereby the surface of the upper river would soon be lowered to the floor of the basin in the soft Salina beds south of the rim mentioned. The height of the falls would be then reduced to less than that of the present day. A new Upper rapids just below the outlet of Lake Erie would be established to a height of sixty feet or more, eventually becom- ing another cataract as the river passed over the Corniferous rock between Fort Erie and Black Hock, thus lowering the surface of Lake Erie. Before the present investigations it was supposed that there were upward earth movements here which would counteract this lowering of the lake level, but these are found to have ceased, at least for the present. Such is the natural immediate future of the LTpper rapids, while the falls themselves will be considerably reduced on account of the lower floor of the Chippawa valley. "WTiile the re-excavation of the superficial drift deposits in the basins, underlaid by the Salina formation, will be much more rapid than that of the rock, yet it might require a long time to complete these changes of the new Upper rapids. Accordingly there is still a long natural future for Xiagara falls. Turning to the artificial modifications of the falls. The protection of the l)anks of the Upper rapids is already com- pleted by the railway company. The full diversion of the water under the present active franchise will reach -tO per cent of the low water discharge. This must greatly reduce the rate of recession of what is left of the falls, though to what extent only future measurements compared with those already made, can determine. Plate XXXVITI. a. View of G(_)at Kland shflf, adjacent to tlie Falls, which is heiiii; . Ii.i ,iiii|, w ith rock.s appearing at surface. (Sept. 1906.) Electrical DeveUjimient Coiupany Power House in Victoria Park, shown in background. Plate XXXVIII. b View of Carter Cove, Upper Arch Bridge, and factories on the N. Y. bank of the gorge beyond. 381 of Canada] FUTURE OF NIAGARA FALLS 383 In Plate xxxvit. is a picture of the falls diirina; high water (October 25, 1906), where thin streams are seen flowing under the retaining wall, made in curtailing the breadth of the falls on the Canadian side, thus in a degree restoring for a day nearly their former width. Plate xxxviii. a shows the present draining of Goat island itself. Plates xxxix. a and b are views of the Greens or First cascade over rocks which will be the first laid bare. Plate XXXIX. a. Piotile ^'iuu• of First Cascade from (ioat Island. Plate XXXIX. b. ^'iew of First Cascade above Middle Sistt-r Island, and n\<-r lird ticmi w hirli water has lately been drawn off. 25 385 T H K OUTLETS OF ERIE BASIN AND ORIGIN OF LAKE BASINS 25^ 389 CHAPTER XXXV. ORIGIN OF THE LAKE BASINS— DROWNED AND BURIED VALLEYS. Preface. Buried valley between Georgian Topography of the Lake basins. bay and Lake Ontario, etc. Features of the Ontario basin. Former higher elevation of the con- Features of the Erie basin. „ tinent. . ■i7'oot„^„^ f +h^ TT„„ i^ocjr. Southern tributaries of the ancient Features of the Huron basin. ^ ,. „ii„„ //->i,;„ „„^ „ ^ . ^> ,,. , . ^ . Laurentian valley (Ohio and Features of the Michigan basin. others) PREFACE. As lias now been seen ISTiagara falls primarily owe their origin to tlie changing levels of the lakes. Their history has been modified by the buried channels which were tributaries of the ancient valleys now converted into the Great lakes of America. Again, these ancient valleys were more or less obstructed by drift deposits, and since the glacial period the region has been greatly modified by the tilting of the land sur- face, the evidence of which is seen in the now raised beaches, causing changes in the course of the lake drainage. In reality the history of Niagara began long posterior to the for- mation of these valleys, and if the history were told in chrono- logical order the ancient valleys and their blockades should Have been described before commencing the study of the falls. However, their origin has really been discovered by the obser- vation of phenomena which of themselves were not obtained in their natural order, and it required many years to collect sufficient evidence to explain them. The exploration of the Whirlpool and Whirlpool rapids has brought to light an ancient gorge which of itself was unimportant ; but owing to its effects 391 392 FALLS OF NIAGARA [Geol. Surv. upon the recession of the falls, giving rise to the Whirlpool, it assumes a prominence much greater than many other similar buried channels of larger size. The discovery of the Falls-Chip- pawa channel explains a still more remarkable feature, trend- ing in a direction opposite to that of modern drainage and demanding the existence of a further buried valley sweeping round and entering the Ontario basin by some otlie:- route, while upon the surface of the country no such feature appears. TOPOGEAPIIY OF THE LAKE BASINS, J 11 front of the highlands of l^ew York and Pennsylvania are the great plains of the Lake basins, broadly speaking, though really this term is not quite applicable. The Erie plains, north of the central part of the lake, rise slowly into hills and a high rolling country covered with drift, reaching to 1,700 feet above the sea. In front of the higher slopes of l^ew York is the jSTiagara escarpment facing Lake Ontario, from the foot of which are plains extending to the lake. On the northern side of Lake Ontario the flat country passes into high drift ridges. While the iSTiagara escarpment has an elevation of 330 to 340 feet at ISTiagara river it rises somewhat higher at the head of the lake^ and forms the northeastward face of the high country just mentioned beyond the plains of Lake Erie. This escarp- ment is the characteristic feature of the peninsula between Georgian bay and Lake Huron, of the islands of Lake Huron, across the upper peninsula of Michigan, and in the peninsula separating the lake of that name from Green bay. Between Lake Ontario and Georgian bay the face of the Niagara escarp- ment is so obscured by hills of drift that its true character is not always apparent in crossing the rugged, drift-covered country. Beyond Lake Ontario, the St. Lawrence valley extends from the foot of the Adirondack mountains of New York northward for thirty miles or more to other ridges, character- of Canada] ONTARIO BASIN 393 izcil l)y (li-it> hills. In tlit' diroction of the lower Ottawa river the undiihiting plains become more extended as the coun- try is much lower than farther west. In fact the St. Lawrence valley may he considered as the extension of that of Lake Ontario, with the river now traversing it near the surface and not in a deep vallev; althnngh terraces occnr to a height of 100 feet or more. These terraces are much below the level of the great deserted and much tilted beaches of the Ontario basin which have hitherto been surveyed. Xot only is Lake Ontario excavated out of almost level Paltpozoic formations, but this is equally true of Erie, Huron, and Micbigan basins, except on the northern side of Huron, where the formations are crystalline rocks. FEATURES OF THE ONTARIO BASIN. The special feature of the basin of Lake Ontario is the dro-\\med escarpment on the southern side, showing the valley- like character of the lake as was first pointed out many years ago.* At that time there was an exaggerated idea of glacial erosion, which was a plausible Avay of explaining unknown features. The late Prof. J. S. Xewberry was the leader of this school, and yet he thought that an ancient channel be- tween Lake Erie and Lake Ontario might be found. Such a channel was described by the writer which is now known to be only one of two or three depressions across the is^iagara peninsula. Lake Ontario is simply a deep trough reaching to 738 feet in depth, while it has been only 246 feet above sea level. The deepest point is nearly north of Seneca lake which reaches 165 feet below sea level. Farther toward the outlet of Lake On- tario, the floor rises without any form of a channel such as * ' Discovery of the Pre-glaclal Outlet of the Basin of Lake Erie.' By J. W. Spencer. Proc. Am. Phil. Soc, Phila. Vol. xix., pp 300-337. Also in Report of Geological Survey of Pennsylvania. Vol. qqqq. 394 FALLS OF NIAGARA [Geol. Surv. prevails to the eastward. This feature made the investigation of the origin of the basin very difficult, for undoubtedly there was a river-like valley throughout the lake without any mani- festation of its continuing to the present outlet. It was to account for the rising floor, as well as the great depth of the lake below sea level, that recourse was had to glacial erosion, although the striations upon the- rock surfaces were strongly marked in a direction always oblique to the face of the escarp- ment. This would disprove such an origin for the basins. The more precise study of the details showed that at a point about twenty miles east of Toronto, and some half dozen miles from the northern shore, there was a channel reaching to 474 feet, while the soundings on either side of it have a depth of only about 200 feet.* At the head of the lake there is a broad valley which is really a refilled caiion. This is the Dundas valley with the rocky walls rising nearly 500 feet above the lake. On the side of this valley, extending to the city of Hamilton, a boring (at the Royal Hotel) showed that the buried channel' reached 292 feet below Lake Ontario and cut into the Niagara plateau to a depth of more than 800 feet. The depth of the lake even as far west as south of Toronto is 400 feet. This buried Dundas valley completely penetrates the zone of I^iagara limestone, and connects with a basin in the Salina formation beyond. Returning to the soundings of Lake Ontario it is found that the northeastern extension of the lake is shallow and contains many islands. These are of limestone formation (Trenton and other allied series), lying in almost horizontal strata. There are undoubtedly deeper channels ; for among the Thousand islands, a few miles below the outlet of the lake, the water reaches to a depth of 240 feet. Here the river crosses a narrow zone of crystalline rock of the Laurentian system. But the first rocky barrier across the river occurs at the Galops rapids * On the B. A., but not on the U. S. chart. of Canada] ONTAKTO BASIN 39i sixty-six miles in a direct line from Kingston, at the outlet of the lake. Here the depth is reduced to twelve or thirteen feet at a point one-third of a mile below the old lock Xo. 27. This barrier is composed of limestones of the same character as those at the ontlet of the lake. It was several years after the observations of the fluviatile character of the lake basins before any rational explanation could be given of the barrier just mentioned. I was able subse- quently to apply the w^arping recorded in the post-glacial fea- tures. From the deepest point at the outlet a line may be drawn somewhat parallel to that from Canastota to Cape Rut- land (Prospect farm four miles east of Watertown), at the eastern end of the lake. Between these two points the Iro- quois beach rises 282 feet. (See map Plate xxxiii.) As has been mentioned in Chapter xxii., page 284, the post-glacial rise between the outlet of Lake Ontario and the Galops rapids amounts to 500 feet or more. Here is found a differential warp- ing sufficient to acount for the barriers across the outlet' of Lake Ontario, for if straightened out it would not be suffi- )cient to obstruct the flow of the river through the present lake basin. This result is obtained without having determined the last remnant of the differential change of level. Under such conditions the outlet of Ontario over the rocky rim would be 500 feet below sea level. At the same time it shows the con- tinuity of the character of the Laurentian valley with that of the basin of Lake Ontario. These features have been some- what fully explained in a general way in ' Origin of the Basins of the Great Lakes of America,'* and earlier papers. A full investigation of the St. Lawrence valley has never been made, and possibly somewhat deeper channels adjacent to the rapids may be found, as recent observations suggest. Under any circumstances, however, such a deep valley below * Quar. Jour. Geol. Soc. London, vol. xlvi., pp. 523-'533, 1890. 396 FALLS OF iS-IAGAEA ^Geol. Surv. sea level demands that it should have been formed during a period of high continental elevation. FEATURES OF THE EEIE BASIN. Lake Erie is situated in a broad plain. Much more than half of it has a depth not exceeding eightv-four feet; not over one-sixth of its basin has any considerable depth — the most profound soundings reached to only 204 feet, as shown in maps, Plates xxxiv. and xl. jSTmnerous tributaries of the ancient basin have been discovered by borings. For in- stance, at Cleveland one such was found by Dr. ]N^ewberryf- with a depth of 228 feet below lake surface, although adjacent portions of the lake are nowhere over eighty- four feet deep. Later, Dr. Warren ljpham:{: mentions the occurrence of another at this locality with a depth of over 400 feet. Long ago Dr. T. Sterry Hunt" recorded channels at Port Stanley and Vienna, reaching below lake level to a depth of 152 to 200 feet, and at Detroit another to 212 feet. These are sufficient to show that channels everywhere traversed the now shallow basin of the lake, refilled during the glacial period. At the eastern end of Lake Erie, Dr. Julius Pohlman called attention to the buried river channel at Buffalo creek, reaching to 120 feet,ff and Irving P. Bishop** reports that in this locality, at Blackwell canal, that bridge piles were driven to 120 feet without i-eaching rock, although the present floor of Lake Erie is found tea^ more than forty feet below the surface, so that the deep channel of Buffalo creek is buried beneath the much shallower lake. One of the piers of the International bridge across the Niagara river rests upon glaciated rock sur- faces to a depth of forty-five feet, while at other piers the depth is greater. At a point, little more than a mile above Fort Erie, t Geology of Ohio. t BulL GeoL Soc. Am., vol. vii., p. .328, 1S96, lb., vol. viii., p. 8. 1S97. * Geology of Canada. 1863. tt Life History of Niagara, Trans. Am. Inst. Min. Eng., vol. xvii., 1889. ** Geology of Erie county, N.Y., Rept. State Geologist for 1897. of Canada] HURON BASIN 397 the head of the low terrace (five to eight feet) has a rock floor, with the river reduced in depth to seventeen feet. Bnt the chan- nels on either side of this col is thirty and twenty-four feet re- spectively, suggesting a buried valley between, but there was no passage for the ancient Buffalo creek. Other buried channels are also known, but these illustrate the features of buried ones to a considerable depth about the Erie basin. FEATURES OF THE HURON BASIN, The southern half of Lake Huron is a plain traversed by valleys submerged to form a shallow lake. jSForthward of this basin and extending obliquely across the lake for ninety miles is a submerged escarpment, facing the northeast with a height of from 300 to 450 feet (Map Plate xl.). The extreme depth of the lake reaches to 750 feet, while the deepest channel between Lake Huron and Georgian bay is unknown on account of drift filling, but soundings show 306 feet. On the western gide of Georgian bay is a deep channel reaching to 510 feet in front of the Xiagara escarpment. These channels in front of the submerged escarpment of Lake Huron and of the Niagara escarpment on the western side of Georgian bay are fragments recording the ancient drainage of the lake basin. On both sides of the Huron-Georgian lake our knowledge of the drainage sys- tem is further extended by deep borings, FEATURES OF THE MICHIGAN BASIN. The lake is divided into two basins, the more northerly with a maximum depth of 864 feet, in part bounded by vertical escarpments now submerged. One of these abrupt features shows a descent of 500 feet, with the deepest sounding in the outlet of the lake at only 252 feet. A fiord tributary is seen in Grand Traverse bay to a depth of over 600 feet. This and shallower fiords indicate the existence somewhere of a deep channel connecting Michigan with Huron, like the river valley 398 FALLS OF JSTIAGARA [Geol. Surv. buried beneath the drift material of the modern floor of Lake Erie. They establish the existence of channels though not shown by the soundings. Lake Michigan as a whole is carved out of Palaeozoic formations, the newest of which are the Coal Measures. There is a smaller basin in the southern part of Lake ]\Iichi- gan with a depth of 576 feet, and a ridge between the two basins submerged from 300 to 312 feet. Across the southern part of the peninsula of ]Michigan there is a valley dissecting the high- lands of that State, the western portion of which is occupied by Grand river, and the eastern portion by a stream emptying into Saginaw bay. Well borings in the western section show the absence of rock to a depth of from 100 to 200 feet below lake level without reaching the greatest depth, while farther east there are several wells, one of which in the drift is 500 feet. This is equal to 350 feet below the surface of Lake Huron. The location is not, however, in the middle of the old valley. Other buried valleys and channels submerged demonstrate that the lake basins are simply the valleys of a great river and its tributaries of high antiquity before the glacial period. BUEIED VALLEY BETWEEN GEOEGLiN BAY AND LAKE ONTARIO. The distance between Georgian bay and Lake Ontario is about seventy miles, in a direct line. The country is character- ized by rising plains, crossed by high transverse ridges of drift trending eastward and westward. Lake Simcoe, with a diameter of twenty miles, is situated between such ridges. These rise from 200 to 550 feet above Lake LIuron. From Georgian bay to Lake Simcoe in a direct line is a dis- tance of twenty-five miles. It is a low, flat country. Rock is known to be absent far below the level of the bay. x\t Barrie, on the shore of Lake Simcoe, 140 feet above Georgian bay, a well was bored to a depth of 280 feet without penetrating drift. Twenty miles farther inland at Xewmarket, and elsewhere, of Canada] SOUTH OF GEORGIAX BAY 399 wells were also found rcaeliino- below the level of Georgian bay, watliout finding rock. Southward of iSTewmarket is a transverse drift ridge beyond which at Richmond hill (217 feet above Georgian bay) a w^ell was sunk 400 feet without penetrating the drift. This proves that the channel is at least 700 feet under the drift ridge and here is a boring reaching to 183 feet below Georgian bay without meeting with rock. Southward of Rich- mond hill the country falls away to Lake Ontario, and this is channelled by deep ravines showing the absence of rock much below the level of Lake LIuron. Upon the eastern side of Lake Simcoe the country is covered with limestone rising 150 feet above the lake, while on the west- ern side of the broad buried valley are the JSTiagara limestones. Thus a trough is found which is the continuation of Geor- gian bay extending southeastward in the direction of Lake Ontario, and parallel to tliG jSTiagara escarpment, although the deepest borings, to nearly 200 feet below Georgian bay, have not revealed the floor as they have not been sunk to the underlying rock. This trends toward the most interesting deep channel mentioned page 391, as occurring about twenty miles east of Toronto, to a depth of 474 feet below^ the surface of Lake Ontario. The evidence now shows that with the drift removed the Iluron-^Iichigan lakes would be very much lowered below the present level, even if the outlets mentioned had no greater depth than has been ascertained by direct observations. In- deed this is the only line along which channels so deep as this have been found. South of Lake Michigan deep borings also occur, which at one time Avere thought to be indicative of a Michigan outlet toward the south. More recently Mr. Frank Leverett collected evidence on this subject, and finds the lowest rock floor near the border of Lake Michigan is only about 230 feet below the lake, with the floor having an altitude of 350 feet above sea level. This is not the slightest indication 400 FALLS OF NLIGARA ^*^^°'- S"""^- that higher rock does not obtain between it and the Mississippi drainage. Even this level is hnndreds of feet above the floor of Lake Michigan. The fiord-like valleys of Green and Grand Traverse bays show by their direction that the ancient drainage of Michigan basin was towards the northeast and not southwest. While borings to the fnll depth of the bnried channel have not been made every^vhere, yet they have reached far below that of the present lake surface, and where this method of investigation has been carried sufficiently far the chain of evidence is com- plete. As in the case of the Erie outlet, newly discovered and described in Chapter 'xxxvii., such a chain is not suggested in the surface features. FORMER HIGH ELEVATIOX OF THE CONTINENT. In glancing at the features of the Ontario basin the post- glacial deformation has been pointed out. So also deformation of Lake Huron is strongly marked by the raised shore lines, which are nearly horizontal at the southern end of the lake; but they increase in height so that they rise at the rate of nearly four feet per mile to the northeast (Chapter xxv.). With the post-glacial deformation or warping of the earth's crust removed, by a depression of the raised parts of the plane of the earlier water levels, a considerable difference will be found in the topographic relief on proceeding to the northeast. Such a depression would admit the St. Lawrence estuary with a depth of 500 feet below sea level, where now is found the rocky rim over which the river passes by rapids on its way to the sea. As all the features of the lake basins are those of atmospheric action, such a trough becomes evidence of late high continental elevation of the land. This hypothesis is sustained by the deep channel of the Gulf of St. Lawrence.* Here Avas a plain now forming the floor beneath sea level. This, being a continua- tion of the Laurentian valley of fluviatile origin, necessitated * ' High Continental Elevation preceding the Pleistocene period,' by J. W. Spencer. Bull. Geol. Soc. Am., vol. I., pp. 65-70, 1890. of Canada] ELEVATION OF COXTIXENT 401 a liig'h elevation for the lake reo-ion, vliicli lasted for long ages. That the continent was raised still higher at a later date is shown by partly explored canons at the mouth of the Gulf of St. Lawrence, at the mouth of the Gulf of Maine, and many others ; but especially that of the Hudson, which I have recently been able to explore in detail. Here the caiion of the Hudson, with a length of fifty miles, has a depth reaching to 4,000 feet below the floor of the continental shelf, and is traceable to a depth of over 7,500 feet below sea level. According to my working hypothesis, requiring an elevation of the land to a height of 2,000 feet or more, the period of great elevation was very long, during the making of the Laurentian channel through the Gulf and the forming of the valleys of the Great lakes. Such was the condition of the continent when the lake valleys were completed, before the glacial period. Whatever abrasion and rounding of the abrupt surfaces occurred during that time, the valleys were not ploughed out by glaciers, as shown by the direction of the glaciation on the rock surfaces, which is everywhere oblique to the face of the escarpment. As before mentioned the lake basins were formerly supposed to have been excavated by glaciers, but since the appearance of my earlier papers upon the origin of the basins this sub- ject has been one that has received very little attention, con- troversy having ceased, interest in the subject has declined. While several geologists have added to our knowledge of the features south of the lake, the revival of the subjects of the basins themselves is seen in a later contribution by Prof, A. W. Grabau.* In attempting to give a full history of the lake basins from the Palneozoic days he has continued the ancient drainage of the Archasan highlands of Canada, as he supposed it to be, to the * Bull. N. Y. State Museum, No. 15, 1904. Also 18th Kept. Com. Reserv., Niag., for 1901, pp, 42-54. 402 FALLS OF NIAGARA [Geol. Surv. latest times preceding the Ice age. In doing this he has drawn some remarkable conclnsions. Assuming the country to have been 2,000 feet higher than now, l)y a process of tilting, he sees a valley running down from Georgian bay to Lake Ontario and joining the main valley of the lake. The course from Georgian bay to Lake Ontario is along that which I have described ; then through Lake Ontario, producing a channel valley rising from a depth of about 500 feet below sea level, to more than 700 feet above it. His channel passes up the Dundas valley and beyond. From the Dundas valley he assmnes that there must have been a continuation of the river southwestward to the Erie basin, between exposures of the Devonian and I^iagara lime- stone systems, now 700 or 800 feet above the sea; but it is not certain whether it continued up thq Cuyahoga valley at Cleve- land or the Maumee and across to the Wabash. The post-glacial tilting is everywhere to the northeast, con- sequently this must be disposed of, and in so doing it increaseJ; the necessary rise from the deepest part of Lake Ontario to the Erie basin by a considerable amount. Beyond Dundas valley, in the direction which he indicates, there is no channel, but the buried valleys found are those transverse to it. It is now know^n that the deepest channel between Lake Erie and Lake Ontario is some distance eastward of that observed at Dundas. The topography of the country is entirely against this drainage. The post-glacial warping is a constantly increasing movement, but his channels would require deformation amounting to ten feet per mile if uniform (in excess of that obtaining since the glacial period) of which there is not the slightest evidence ; nor would any uniformly progressive deforma- tion give rise to the features which he supposes. In short, Prof. Grabau was not familiar Avith the geology or topography of the country he attempted to describe. Even if he had supposed the lake basins to have been excavated by glaciers, or if bv some internal earth movement the lake basins alone had of Canada] ELEVATION OF CONTINENT 403 sunk, the topographic features of this region wouhl not have supported his conclusions. Against all of this is the fact that the channel in the Ontario basin would have to convey waters upward to a height of more than 1,000 feet to the top of the N^iagara escarpment, and proportionally would indicate thou- sands of feet of pre-glacial uplift to the northeast, while \yv know that the relative elevation there was much loAver than now. At the same time this Niagara escarpment everywhere faces a broader and more open country to the northeast, without any suffffestion that at the time of the formation of the lakes the country was so high as even the escarpment, but quite the con- trary. The work of Prof. Grabau, to which reference has been made, was intended as a guide to Niagara, and he has assembled into it much valuable information. Whatever were the earlier con- ditions in the Palaeozoic days it is unfortunate that he should have attempted to carry his uniform hypothesis over such vast geological periods down to the time of the completion of the lake plateaus, such as obtained just before the glacial period. SOUTHERN TRIBUTARIES OF THE ANCIENT LAITRENTIAN VALLEYS NOW REVERSED THE OHIO AND OTHERS. As found many years ago by Mr. J. F. Carll, the upper branches of the Allegheny formerly emptied into Lake Erie. Later, from the evidence of borings, I suggested that the Upper Ohio, with its tributary the Monongahela, M-as for- merly reversed, draining by way of the Beaver, Mahoning, and Grand Iliver valleys into the Erie basin. This subject was fur- ther investigated by Dr. P. Max Foshay,"' who named the reversed drainage after the present writer. This named river was again called into question by Prof. I. C. White, but upon further observation by ]\Ir. Leverett the plane of the river was found at a higher level than that which had been questioned, so * Pre-glacial Drainage, Western Penn., Am. Jour. Sc. iii, vol. xl., pp 397-403, 1890. 26 404 FALLS OF NIAGARA [Geol. Surv. that Prof. White reversed his opinion, stating that — ' It is now pretty surely established through the work of Carll, Spencer, Foshay, Ilice, Chamberlin, Leverett and others that the Monongahela, Lower Allegheny and Upper Ohio waters drained northward into the Lake Erie basin in pre-glacial times.'f Accordingly, the pre-glacial drainage area of the Erie basin had an extended watershed reaching for a hundred miles into AVest Virginia, thus the drainage and the whole upper basin of what is now the Ohio river then flowed into the Erie basin. The Genesee of jSTew York and Pennsylvania was a pre- glacial tributary of Lake Ontario, but its course was blocked at one point Avhich it left, making a new channel for itself before reaching the lake. The modern river passes along a course forming three waterfalls at Rochester. Seneca (605 feet deep or 359 feet below the level of Lake Ontario) and Cayuga lakes Avere tributary valleys. To these I gave considerable attention when studying the origin of the basin of Lake Ontario, but more recently Prof. R. S. Tarr:]: has called attention to a buried channel at the head of Cayuga lake (435 feet deep) reaching to a depth of 430^feet below its surface, or 274 feet below the level of Lake Ontario. This lake is nearly due south of the deepest point in Lake Ontario (738 feet). The conclusions as to the origin of the lake basins are tliatl they were the valleys of the ancient Laurentian river and its tributaries, obstructed by deposits of drift and by subsequent warping. These valleys were formed when the continent stood much higher than ' at present, and the broad valley of the St. Lawrence river extended continuously from the lake region to the plains between New England and the Laurentian moun- tains ; and thence on to the plains now forming the floor of the Gulf of St. Lawrence, traversed by a channel 2,000 feet below t Am. Geol. vol. xviii., p. 368. 1806. JJour. Geol. vol. xii., pp. 70-73, 1904. of Canada] UPPER OHIO REVERSED 405 sea level. This deep channel receives tributaries from several directions as lias been described.* The character of the lake basins being now described it will be easier to understand the relationship of the ancient drainage of the iSTiagara district. The outlets from the Erie valley to the Ontario basin have been deferred in this work, on account of their forming an additional subject beyond their immediate bearing upon Niagara. * ' High Continental Ellevation preceding the Pleistocene period,' by J. W. Spencer. Bull. Geol. Soc. Am., vol. i., pp. 65-70, 1890. Also, ' Evidences as to Changes of Sea Level,' Bull. Geol. Soc. Am., vol. vi., pp. 141-166, 1895. 26* CHAPTER XXXVI. NIAGARA PENINSULA. Preface. ' Short Hills ' and the sand ridges. Features of Niagara peninsula. Grand river-Dundas valley — ancient Geology of the peninsula. drainage. Elevation of the Erie plains. PREFAcr-:. The discovery of the Falls-Chippawa valley was the result of investigating the nature of the basin at jSJ"iagara falls. On account of its depth it demanded a search for its continuation with an outlet into the lower Ontario basin, if such could be found, for no other suggested itself except by the Dundas valley far away. Without finding such outlet, the basin at Niagara falls would have been unintelligible, but would pro- bably have been attributed to glacial erosion. On the other hand the discovery which has been made by following theoreti- cal lines completes a chaY)ter in the Lake History. FEATURES OF THE NIAGARA PENTNSUEA. The Xiagara peninsula which separates Lake Erie from Lake Ontario shows a serrated shore line along the northern side of Lake Erie, but if the points be connected by a line this is found to be comparatively straight. On the other hand, while none of the sharp points project into Lake Ontario, yet there is a sweeping indentation in the shore line of the latter, which is deepest between Port Dalhousie and Jordan harbour. Here the embayment indents the shore line by several miles in- w^ard of the mouth of the Niagara river. Inside of this shore 407 408 FALLS OF NIAGARA t^eol. Surv. line is a gently rising plain to the foot of the l^iagara escarp- ment, which at the river is abont seven miles from the lake, while at Grimsby, twenty-seven miles to the w^est, the escarp- ment is only a mile or less from the shores of Ontario. The foot of the escarpment limited by the plain in front of the Iro- quois beach has a height of less than 130 feet at Niagara river. This slightly diminishes on proceeding westward. I^ear its brow the escarpment has a general elevation of 320 to 340 feet above Lake Ontario, with the watershed generally distant not more than from two to four miles. Southwest of St. Catharines — that is behind the deepest part of the indentation in the shore of Lake Ontario — there is a large embayment in the IJ^iagara escarpment. (See Chapter XXXVII.) GEOLOGY OF THE PENINSULA. The Niagara escarpment is formed by the same series of limestones and shales as are described along the gorge of the Niagara river. The limestones give rise to the bold feature. \\Tierever the small streams cross it picturesque waterfalls occur, with small cahons sometimes a half a mile in length, as along; the Beaver Dam creek at De Cou falls, and Twentv-mile creek, south of Jordan. After passing the brow of the escarp- ment the superficial limestones are usually covered by a con- siderable depth of drift, with the widest exposure of the rock along the Niagara river. On the Erie side of the peninsula there is another ridge of limestone buried beneath the plains. This occurs along the Niagara river to a distance of two miles from the lake. At the surface it is generally covered by a thin layer of drift, but at a depth of a few feet it is found. Still the surface of the country south of Lake Erie seldom rises higher than from twelve to twenty feet above the lake, though near Fort Erie it is higher as also near the watershed back of the Niagara escarpment of Canada] NIAGARA PENINSULA 409 Upon reaching the surface of the country above the Niagara escarpment the general appearance is that of a remarkably level plain, and to tlie eye the features are the same all the way from it to the northern side of the low ridge of Cofniferous limestone near Lake Erie, although there is a broad basin between these limestone zones underlaid by the soft Salina formation. This is composed of light, or dark or black shales, with or without gypsum nodules, or layers of gypsum and impure limestones, all easily destructible materials. This has given rise to a de- pression in the rock surface below the country, recognized for more than sixty years. Still it is covered by drift so as to make a continuous plain only a few feet above Lake Erie without showing any depression upon the surface due to the buried features. ELEVATION OF THE ERIE PLAINS. These plains are traversed in a western and eastern direction by a creek, which at Chippawa village is ten feet below the level of Lake Erie ; at Welland, where it passes under the Welland canal, it is nine feet below the surface of the lake ; and at Wel- land Port, as shown on the map, the creek is still below lake level. The southern part of the peninsula from Dunnville to Welland is traversed by the Feeder canal with the surface of the water eight feet above the lake. Extensive marshes have also continued until recent times, and, although now drained, the surface at many places can not be more than five feet above the lake. This gives a general idea of the flatness of the Niagara peninsula. Some exceptions must be made to this statement. ' SHORT HILLS ' AND THE SAND RIDGES. Sand and gravel ridges occur to a height of 443 feet above Lake Ontario south of St. David (page 215) ; a similar hill at Lundy Lane, Drummondville (page 212) rises to 465 feet; extending as a ridge to the vicinity of Niagara falls. Another 410 FALLS OF NL^GAKA "^^^o^- ^urv. eminence southeast of C]iij)pawa village occurs. Tlie most re- makable ridge rises west of Font hill, to 320 feet above Lake Erie. It tapers south west ward for three miles, to near Fen- wick, with a breadth of nearly a mile and a quarter, though the higher part does not exceed half a mile. This is entirely com- posed of drift, which in the deep gullies upon its northern side is shown to be largely composed of stratified clayey sand and gravel, although th? higher part of the summit is a sandy clay containing a few boulders and small stones of drift origin. Round the hill occurs a fragment of the Forest beach, formed by wave action upon the material of the hill. Its height is 221 feet above the lake. The giillies on the northern side of this hill (called the 'mountain') descend and unite with deeper ones which have partly excavated the buried valley. The country is broken up into sharp hills and deep ravines, which are designated the ' Short hills ' and so known in history. The main stream traversing the Niagara plateau has an old name which it received from the Indians, and throughout the country it is everywhere known as Chippawa creek, or ' the creek/ while ' the river ' refers to Xiagara. By an Order in Council in 1820 the name was officially declared changed to Welland river, yet this has never been adopted locally, though it appears upon late official maps. Away from the sand ridges mentioned the surface soil is that of clay or sandy clay, with only a few places where bands of surface sand occur. At a few feet below the clay, sands or gravels are found. Where the sands surmount the clays, they are of the nature of beaches. ^Nowhere on the surface of this country soutli of Font hill is there any suggestion of buried valleys. This hill, rising higher above the plains, would seem to be opposed to the occurrence of such. (For characteristics of the upper ISTiagara river district, see Chapter ix.) of Canada] GKAXDKIVKK COUKTKY -ill gka:nd kivkk-duxdas valley, axciext deaixage. Westward of the meridian of Dunnville the land rises to a greater height with more varied features, though underlaid by the Salina formations, over which the Grand I'iver generally flows. About Brantford, there is a deep, bi-diul ro-exravated valley forming a strong surface feature. The northern watershed of the Grand river approaches the margin of high country faced by the jS^iagara escarpment, as south of Hamilton^ where it is 492 feet above Lake Ontario. l\ear the southern margin of the valley, at Onondaga, is a buried valley to 110 feet, or to a level twenty feet below Lake Erie. To the northward the rocks are absent for a greater depth, as at Jorseyville (about twelve miles from Lake Ontario) Avliere wells are 150 feet deep to rock, or to the level of Lake Erie. Thence to the buried Dundas valley the depth is known to be very great. I^Tear Ancaster the rocky wall of the southern side of the Dundas valley occurs under the drift, which also forms hills in the upper part of the valley. Its breadth, cutting through the limestone escarpment, is two and a half miles, but it partly expands so as to include the plains of Hamilton in an enlargement at the head of Lake Ontario. At Hamilton, the buried Dundas vallev reaches 292 feet below the lake level. Llere is found to be a deep buried depression through the Niagara formation extending to the buried Salina valley and Lake Erie, with a tributary from the upper part of the Grand river district, also joining it. The datum plane of this partly explored depression, below that of the present surface of Lake Erie, was further lowered in olden days, before the post-glacial warping which raised the dis- trict south of Dundas to a considerable height above the Erie plane. Before the recent explorations the Imried Grand Biver- Dundas valley was the only known depression between Lake Erie and Lake Ontario capable of lowering the upper lakes pro- 412 FALLS OF NIAGARA IGeol. Surv. vided the drift de^DOsits were removed. But the studies here were among the first investigations made, and they have never since been extended. Later it became apparent that there was also a shallow de- pression near the Welland canal sufficient to lower Lake Erie if drift were removed. When the Falls-Chippawa valley was found to reach far below the surface of Lake Erie, the question of the pre-glacial outlet of the Erie basin required further investigation. (See next chapter). CHAPTER XXXVII. DISCOVERY OF THE ERIE OUTLET. Deep channels about Lake Erie. Erigan canon. Salina basin. De Cou and Swaze falls. Well-borings. Erigan channel and buried Erie Southern boundary of Corniferous outlet. limestone. Slope of the Erigan outlet of the Northern boundary of Niagara lime- Erie basin. stone, and Thorold depression. Erigan tributaries. Features of the ' Short Hills ' dis- Crossing of the lake depressions. trict. DEEP CHANNELS ABOUT LAKE ERIE. Profound channels buried to depths below the floor of Lake Erie have been mentioned on page 396. Thej occur to depths of 120 feet or more near Buffalo, and 228 to 400 feet in the vicinity of Cleveland, as illustrating the pre-glacial drain- age that formerly obtained beneath the present floor of the lake, which has since been covered with drift deposits. These demanded the discovery of an outlet for what is novv'^ the Erie basin. SALINA BASIN. It was ascertained that on Grand island (at Sour Grove) wells in drift reached sixty feet, though surface rock occurs on both the northern and southern ends of the island. At Tona- wanda the buried rocks are fifty feet below the surface, while at Gatzville, five and a half miles east, a rock floor was shown as ranging from twelve to ninety-six feet deep (near the creek). The Falls-Chippawa valley Avas found to reach ninety feet and more below the Erie level. This was deeper than any channel at the Welland canal. The Dundas valley was a long way off, so that further borings were sought for in order to ascertain what was the outlet of the Falls-Chippawa valley. 413 414 FALLS OF I^IAGAEA ^^^''^- S"^^" WELL BORINGS. Scattered well-records showed the great depth of the buried valley or basin. Others were found indicating that the deeper parts of the basin did not extend nearly so far as the Dundas valley, although deep wells occurred over the whole distance. Fortunately, at this time, a great deal of information was obtainable from the recently sunk gas wells, and water wells. The officers of the gas well companies and the water well borers and others were most courteous in placing informa- tion at my disposal. Many of the locations of these wells were visited by me in order to ascertain the surface topography. The positions and depths of the Wells are shown in the map Plate XLT. Their heights above lake level are not given on the map, although many have been determined. Approximately, however, their elevation is known. Thus, the Feeder canal, which is eight feet above the lake, is at many places higher than the level of the country. The main creeks to beyond Welland Port are below lake level, with the banks rising twelve, or in some cases to twenty feet. The country is generally a plain, except at Font hill ridge, on the flank of which is Fenwick well (forty-eight feet above the lake). At the Quaker church the well is at fifty-one feet above the lake. This is just south of Ridgeville. Where not specified the general level south of Ridgeville and Fenwick does not exceed ten to twenty feet. The same may be said to be characteristic everywhere away from the mentioned ridge which rises to over 300 feet above Lake Erie. This hill is shown on the map by contour lines. Upon the depths of the wells being plotted it became mani- fest that buried channels could be traced across the floor of the Saliiia basin. These ranged themselves into a system — a feature that was hardly to be expected without many more borings thai) were obtained. of Canada] BURIED ERIE OUTLET 415 l'l..STK XLI. 416 FALLS OF ?riAGAEA ^'^^°^- ^urv. SOUTHEEN BOUXDAEY OF COEXIFEEOUS LIMESTOiS^E. This has a general breadth o£. about two miles, though some- what wider on approaching the ^N^iagara river and narrower towards the west. It is two miles in width where crossed by the Welland canal. Its surface is thinly covered with earthy deposits. The northern side of this belt is sharply defined on account of its forming a now buried escarpment of eighty feet or more in height. This feature, all of which is buried, has been brought out by the well borings which reveal the abrupt disappearance of the limestone. The same rocks are occasion- ally seen upon the lake shore and pass under the shallow waters. Xear Lowbank post office there is an embayment in the lake shore about two miles across. Here the shore is faced by sand dunes ten to twenty feet high, with a flat country behind not more than five or six feet above the water. ISTear the lake shore, east of Lowbank village (on the farm of Cyrenus Barrick) the limestone is entirely wanting, while the well is in drift to a depth of 157 feet or 150 feet below lake level. Another well a little farther inland shows the same absence of rock to a depth of 155 feet. Other wells to the east also indicate that there is uo rock to a depth of 100 feet, with the Corniferous limestone ridge just beyond coming to near the surface. In short, the embajmient of the shore of Lake Erie owes its form to the present buried valley crossing the belt of limestone. This valley has now been traced across Moulton and Wainfleet townships, but it is situated east of Lowbanlv village, which is on its western margin. Everj^vhere beyond Lowbanlc the limestone approaches the surface and gives character to the lake shore. NORTHERN -BOUNDARY OF NLVGAEA, LIMESTONE, AND THE TIIOEOLD DEPRESSION. At the ^Niagara river the Niagara limestone belt is nearly seven miles across, and its southern boundary, as determined by well borings, is found to continue nearly due west from Chip- of Canada] BURIED ERIE OUTLET 417 pawa village and south of Allenbnrg village, so that in the vicinity of the ' Short hills ' it is reduced to an approximate breadth of four miles. This restriction in width, however, is due to an indentation in the brow of the escarpment on the surface of which the rocks appear. But, in going southward, they soon become buried to a moderate depth by drift. The surface of the mountain facing Lake Ontario is no higher than the level of Lake Erie, or it may be twenty-five feet or more lower. The country behind seldom rises more than twenty-five feet higher. The Font hill ' mountain ' already mentioned is only a ridge of drift forming the highest portion of the ' Short hills,' shown on map by contours. Approaching the southern side of the Niagara limestone zone, two miles south of Allenburg, the wells become deeper and deeper as they pass into the Salina basin. The deep cut of the Welland canal (located south of Allenburg) is thirty-six feet above Lake Erie. At Thorold the rocks are seen adjacent to the canal. However, here is an ancient pre-glacial valley, indenting f omewhat the face of the escarpment as if there might once have been a channel drawing waters from the Erie basin. It was formed by a little stream, which made a slight topographic feature in the face of the escarpment, possibly somewhat greater than the Whirlpool-St. David channel already described. FEATURES OF THE ' SHORT HILLS ' DISTRICT. Westward of Thorold the Niagara escarpment sets still far- ther southward. For a distance of some twelve miles the trend of the ' mountain ' forms an embayment, receding to a depth of three miles in the face of the escarpment. It reaches its farthest point in front of the ' Short hills.' So, also, in front of the indentation just mentioned is a corresponding but much deeper embayment of the shores of Lake Ontario, carved out of soft Medina shales, while the escarpment itself is capped by hard limestone. This indentation is a topographic feature 418 FALLS OF NIAGARA [Geol. Surv. of pre-glacial date clue to erosion and it is not a bend in tlie zone of limestone, but it is cut out of them. Moreover, this em- bayment begins at a conspicuous promontory east of Mer- ritton, and extends to another one on the western side, nine miles distant, at Jordan. Xor is this incision shown alone upon the face of the escarpment and shore of the lake, but it is seen in the depression on the surface of the country, in that the alti- tude is reduced twenty-five feet or possibly fifty feet, thus bringing the surface of the country below the level of Lake Erie, in conformity with the slopes of ancient times. At the head of this embayment, beginning at De Cou falls, is a great buried canon, trending westward of south, but now re-occupied by drift deposits rising from the Bell terrace which crosses the outlet of the gorge. These plains have been extensively eroded and carved into great valleys 100 feet or more in depth, thus forming the ' Short hills,' which on their southern margin rise into the great drift ridge of Font hill already mentioned, which blocks the heads of the valleys. ERIGAN CAJSrON. Eric'an was the name I gave the buried vallev"^ which tra- versed the Erie basin in pre-glacial times. It is formed from the Indian word Eriga (Erie). As its continuation across the Xiagara peninsula has been discovered the name must be ai> plied here as elsewhere. The eastern end of the canon begins near De Cou falls^ and lies beneath the broken country just described, along Twelve- mile creek. Even at De Cou falls, known in the early settle- ment of the country as Beaver Dam falls, the higher rocks are removed to sixty-three feet Ixdow the level of Lake Erie, so that the outer margin of the canon is somewhat farther east beneath the drift covering, as may be found in the shallower wells. * Proc. A.A.A.S. for 1888, cited before. of Canada] BURIED ERIE OUTLET 410 At St. Johns West, two and a half miles within the ancient gorge, but upon its risinti' eastern side, the rocks occur at 140 feet below the Erie level. Beyond this point the valley trends more to the east, and depths of 100 and 68 feet, below Lake Erie, to rock were found at wells, showing that the zone of the ISTiagara limestone as a surface feature had been penetrated One mile east of St. Johns West the rocks on the right side of the valley have risen 115 feet (or twenty-five feet below the Erie level). {See Plate xlii.) J^earer the middle of the buried gorge, at a point two miles inward from Do Con falls, the channel is open to a depth of even 216 feet below the Erie level, with the depth of drift not determined. On the western side the rock walls are better shown. At the outer point is a conspicuous headland cape. Ilere also is a step in the old topography, so that there is also an inner wall, as the canon appears to have been double. The top of the inner gorge is here shown by the rock surface, being 232 feet above Lake Ontario, while the outer at half a mile farther has a height of 280 feet. The country rises higher beyond. Tn ascending these valleys the lower bench becomes covered with drift, except where lateral streams cross it ; while the higher one can easily be traced as an escarpment two and a half miles to the south- west, at Effingham, or three miles and a half from the mouth at De Cou falls. At the Effingham mills the lower bench of rock is seen at 105 feet below the Erie level, while the gully leads down to the deeper valley extending from 140 to 190 feet below the Erie level. Here Medina sandstone of the western slope of the caiion is exposed. Above the mill the limestone rocks of the outer wall of the gorge are slightly exposed, but generally lie buried in drift. This is on the western side of the ancient valley, which has been reopened by streams for a further dis- tance of about a mile towards the southwest, where the surface rises to the level of Lake Erie. From Effingham onward to ths southwest, the wells show that the southern edge of the !N"iagara 27 420 FALLS OF NIAGARA ^^^°^- S^^^" limestone zone to a depth of fifty feet or more, below the Erie level, has been passed. As a like turning has been found on the eastern side it is evident that the buried outer caiion dissects the hard beds of limestone for a breadth of between two and three miles, while the inner gorge is narrower. (See Plates SLi. and XLii.) The hill of deep drift makes exploration diffi- cult, but towards the eastern side a ridge of rock rises to fifty feet below the Erie level, with a deeper channel beyond, where the Electric railroad crosses the great ravine. The borings in ravine to thirty feet do not reach rock. There seems to have been here an ancient tributary. The trend of the principal valley is to the southwest under the western part of the mass of Font hill. With the establishment of the fact that the buried valley crossed the Niagara zone, it was desirable to ascertain if rock formations occurred near the surface upon the southern side of the hill. Accordingly, borings were made at the Quaker church south of Ridgeville, through surface clay of eight feet, and then through quicksand to a depth of 132 feet without reaching rock. Operatior.s were then stopped. It is a great pity that the full depth of the drift was not here definitely ascertained, but th? operations were carried to a point eighty feet below the Erie level. As has been shown, there is an indentation of three miles in the face of Niagara escarpment — a feature which suggests that here was the broad outlet of a great river, in ancient times. At the head of the embay ment is now found a canon more or less buried, with the creek bed within the gorge now reopened to a depth of 270 feet below the Erie level without reaching the rocky floor. For three miles farther in the caiion is the valley opened to 140 feet below Lake Erie, and lesser depths are exposed beyond, until blocked by the great drift ridge. Font hill, beyond which the valley has again been found by borings. This is the canon section, through the Niagara formation. (Ses Plate xlil) of Canada] BURIED ERIE OUTLET 421 271 422 FALT.S OF NIAGARA [Geol. Surv. The trend of the caiiion is slightly oblique to the line across its month westward from De Con falls. As shown in Figure 29, the outer caiiou is two miles wide, and the inner a mile and a quarter. Fig. 29. Section across mouth of Erigan canon. D., De Con falls ; Pp., points of outlets of outer and inner gorge on western side ; L. O., level of Lake Ontario ; b. , bed of Twelve- mile creek ; refilling of gorge to above Bell terrace or 165 feet above Lake Ontario. There is no similar incision in the Niagara escarpment be- yond this point until reaching the Dundas valley. The one at Thorold is comparatively small. N'o apparent stream could have made the great canon, or accompanying features. The country is away from any railway, and being broken with difficult roads it is avoided^ although beautiful and fer- tile. The available maps do not show any topographic features. Accordingly this district has escaped the attention of explorers, or the railway geologists. But the indentation, which might have been suggestive, is further obstructed by Fon;t hill, be- yond which upon the surface of the low plains there is not the slightest topographic relief, or geological feature even until reaching Lake Erie, which suggested any buried valley. DE COU AND SWAZE FALLS. De Con falls. — De Con falls is at the angle which the valley of Twelve-mile creek makes at its exit through the escarpment. The falls are on the Beaver Dam creek, at the outlet of Erigan caiion. After descending from the Power Company's dam, about forty-two feet to the edge of the rock, they fall in three cascades Avith iiilcrvciiiiig i'aj)ids, the full height of 224 feet. of Canada] BURIED ERIE OUTLET 423 The canon wliicli llicy have made is .nearly half a mile long. The upper ealaracl over the lower Niagara limestones and shales, now caseades about seventy feet ; the second from the Clinton beds has a height of forty-five feet ; and the third descends over the Medina gray sandstones. The whole is a post-glacial feature of the same age as I*[iagara falls. Swaze falls is in a pretty little ravine on a western branch of Twelve-mile creek, three-quarters of a mile ncn-th of Effing- ham. It is partly excavated out of drift material and partly out of the Niagara limestone, with the underlying shales ex- posed. A little farther down the ravine the thick Clinton band of limestone, overlying a thin layer of Clinton shales, forms a second small fall or rapid. Farther on, some of the upper beds of Medina shale and red sandstone are also seen. The ravine is a tributary to the reopened Erigaii canon. ERIGAlsr CHANNEL^ OR BURIED ERIE OUTLET. Turning now to map, Plate xli., and studying the borings, I was surprised that the evidence, so far obtained, showed not merely a general basin, in the Salina formation, l)ut also chan- nels crossing it, now buried under even as much as 192 feet of drift, where the to]i of the well is no more than ten or twelve feet above Lake Erie. Accordingly, these borings bring to light a channel 180 feet l)elow the lake level, which has been ex- cavated out of the soft shale rock, while the general depth of the drift filling of the Salina l)asin is perhaps 100 feet or a little less. The deep trough seems to have a breadth of about two miles. Lowbank is on the western side of the channel, but here the rock approaches the surface. At no considerable distance to the east a boring in the channel shows a depth to 150 feet below the lake. This also is west of the middle of buried trench. ISTear Winger many borings have been made, and these are often so close together that they can not be given on the map. 424 FAT.I.S OF NIAGxiRA ^^^^'- ^"''^• Here the clianiiel reaches to 180 feet below the hike. Farther on, evidently without quite reaching the deepest part, the liorings show an absence of rock to a depth of 160 feet. At Fenwick, on the western side, the buried valley Avas found by two wells (192 feet deep) to reach 134 feet below lake level. From Lake Erie to this point the valley has been located by borings at almost every mile, or at shorter distances. (See map, Plate xli.) The greatest depth should be found be- tween Fenwick and Ridgeville. Following the trend of the buried valley, from Lake Erie to Fenwick, and extending it along the same course, across a gap of only four miles, without direct observation, beneath the great drift mass of Font hill to Effingham, it is now found to be an open caiion whose course is in the same direction. This caiion has already been described. Accordingly the buried channel from the Erie basin across the ISTiagara peninsula and through, the escarpment has now been found. SLOPE OF THE EEIGAIST OUTLET OF THE ERIE BASIN. The deepest part of Lake Erie is twenty-seven miles south of Lowbank (as shown in the longitudinal section, fig. 30). The post-glacial rise of the region is about a foot and a half per mile in a northeastward direction (determined from the elevation of Forest beach at Brown's nursery. Font hill, and at Sheridan and Crittenden, X.Y.). The post-glacial rise at the western end of Lake Ontario is two feet per mile N". 25° E. Making corrections for post-glacial deformation it reduced the bed of the known channel at Low^bank, on its western side, to 190 feet; at Winger to 228 feet; at Fenwick on the w^estern side, to 194 feet ; and even at the Quaker church on the eastern side, where the Avell was not carried to rock, the depth is now known to more than 140 feet below the level of Lake Erie. As has been * Proc. Am. Assoc. Ad. Sc, Vol. xxxii., p. 199, for 1888. of Canada] BURIED ERIE OUTEET 425 h4-aJi-^ S brii COS ^oiig r^E' 426 FAT>LS OF NIAGARA [Geol. Surv. shown the depth of Lake Erie is only 204 feet. This channel is thns fonnd to reach below the floor of the lake. Again in the creek bottom at Effingham (three miles within the gorge from De Con falls) the open valley is now 255 feet below the Erie level, withont reaching its rocky floor. Accordingly I have been able to obtain data which conclu- sively establishes the existence of a buried channel from one to two miles wide crossing the jSTiagara j)lateau and dissecting the escarpment to a great depth. As there has been a post-glacial uplift in Forest beach, of 121 feet between Cleveland and Font hill, of which only sixty feet pertains to the region between Cleveland and the deepest part of the lake, the other sixty feet belonged to the stretch between the 204 foot sounding and Font hill. By this amount the original outlet of the Erie basin was lowered in pre-glacial days below that of the present level. Accordingly, the floor of the open vallev, without reaching l)ed rock, now within the limits of the buried gorge, represents a level of 305 feet below deepest sounding in Lake Erie. At the same time, there was a somewhat greater post-glacial change in the levels than that which has been measured. From these measurements the buried Erigan channel is found to reach to a depth more than sufficient to have drained the Erie l)asin in pre-glacial days. ERIGAjST TKII3UTAEIES. The Erigan channel is found to be not the only one upon the surface of the Salina formations. A considerable number of channels have now been located. There is one from the region to the west of the Erigan, approximately along the present drainage creek, which at Welland port is 100 feet below the Erie level. Another tributary valley is adjacent to Grand river, with an equal depth, which extends eastward and joins the Erigan, while the modern Grand river leaves its ancient course and passes over a rocky zone below Dunnville. (^See map, Plate XLI.) of Canada] BURIED ERIE OUTLET 427 Again, from tlie ca^itward, a luiricMJ tributary has been found along tlie northern foot of the buried Corniferous escarpment and another still fartlicr north and ]iai'allel to it; also one in !N"ew York adjacent to the Tonawanda creek. 'I'liis crosses Grand island and extends westward to the Erigan channel already described ; but on its way the Falls-Chippawa valley joins it. This was the feature which led to the discovery of the buried valleys, which gave the true explanation of the Upper rapids of JSTiagara river. Another tributary from the southeast is traced through Welland (town) joining the one just mentioned, which as a broad valley enters the southeastern side of Erigaii canon, south of St. Johns West; and at the same time rounding off the corner of the Xiagara limestone belt. (See map, Plate xlii.) The buried Buffalo creek joined the Erigan, or some tribu- tary to it now buried under the floor of Lake Erie. The ancient Carll or Allegheny crossed the present lake basin directly to the debouchure at Lowbank, while the Spencer or reversed Ohio en- tered the Erie farther west. These rivers are shown on Map, Plate XL. The depth of the Buft'alo buried valley is much greater than that near Welland. The trench through the Cor- niferous limestone, now occupied by the Xiagara river just be- low the present outlet of Lake Erie, Avas a low col between a small valley tributary to liuffalo creek channel, and another joining the valley passing by Welland. CROSSING OF THE LAKE DEPRESSIONS. In pre-glacial times, the Erie formed a long trough, as to- day, with another one parallel to it in the Salina formation, but separated by the ridge of C\)rniferous limestone on the southern side of the present lake shore. The central Salina basin was separated by a prominent rit^ge protected by the ISTiagara lime- stone from the parallel Ontario basin, which last is excavated out of softer shaly rocks. The Corniferous range was dissected 428 FALLS OF NIAGARA f^^°^- ^"^^• at Lowbank and at Niagara river. The Niagara highlands were incised bj the trough or valley at Diindas, another little one near Welland canal at Thorold, and the great transverse Erigan channel and canon, here for the first time described. The Erigan channel crossing and connecting three parallel basins is a repetition of the feature of the Delaware river dis- secting mountain ridges and also crossing the intervening valleys, of the Appalachian mountain system. All of tliese features of the pre-glacial topography of the resion described have been so concealed with drift and terres- trial warping as to have only slowly come to light. The Wainfleet Marsh. — The great marshes of AVainfleet, and other lowlands in Niagara peninsula south of the escarp- ment, are of the latest origin — having been produced by the flooding of the lake region owing to the tilting of the land which raised the barrier at the outlet of Lake Erie, so that the waters rose more than 100 feet in the Erie basin. These at a recent date flooded the marshes mentioned, though now again the plains have emerged owing to the cutting down of the barrier at the outlet of Lake Erie, where may be seen the terrace of the town of Eort Erie, the height of which, with steep bluff behind it, is from five to eight feet above the river. This terrace is even cut out of limestone which occur at the southern end of the town, as well as out of drift deposits. APPENDICES 429 Plate XLIII. 3^ Niagara Falls as repnisented by Father lleniibpin who saw them in IGZS. Date of print 1(597. 430 APPENDIX I. EARLY DESCRIPTIONS OF NIAGARA FALLS. (For t-aily discoveries and name, sec Appendix vm.) (A) Hennepin's description of Niagara fai.ls. (seen 1678)" ' A description of the Falls of the river Niagara, whicli is to be seen betwixt the Lake Ontario and that of Erie.' ' Betwixt the Lake Ontario and Erie there is a vast and pro- digious column of water, which falls down after a surprising and astonishing mamier, inasmuch that the Universe does not afford a parallel. ' Tis true Italy and Swedenland boast of some such things ; but we luay well say that they are sorry pat- terns when compared to this of which we now speak. At the foot of the horrible precipice we meet with the Nia- gara river, which is not above a quarter of a league broad, but is wonderfully deep in some places. It is so rapid above the descent that it violently hurries down the wild beasts while endeavouring to pass it to feed on the other side, they not being able to withstand the force of its current, which invariably casts them headlong about six hundred foot high. This wonderful downfall is composed of two great cross streams of water, and two falls, with an isle sloping along the middle of it. The waters Avhich fall from this horrible preci- pice, do foam and boil after the most hideous manner imagin- able, making an outrageous noise more terrible than that of thunder; for when the wind blows out of the south, their dismal roaring may be heard more than fifteen leagues off. The river having thrown itself do\^Ti the incredible preci- pice continues its impetuous course for two leagues together to the Great Rock above mentioned, with inexpressible vapidity: But having past that, its impetuosity relents, gliding along more *LondoQ Edition of 1698. (Chapter VII.) 431 432 FALLS OF NIAGARA [Geol. Surv. gently for other two leagues, till it arrives at Lake Ontario, or i'rontenac. Any boat or greater vessel may pass from the Fort to the foot of the high Kock above mentioned. This Eock lies to the westward, and is cut off from the land by the river ISTiagara, about two leagues farther down than the great Fall: for which two leagues the people are obliged to transport their goods over- land by the way very good: and the trees are but few, chiefly hrs and oaks. From the Great Fall unto this Rock, w^hich is to the west of the river, the two brinlvs of it are so prodigious high that it would make one tremble to look steadily upon the water, rolling along with a rapidity not to be imagined. Were it not for this great cataract which interrupts navigation they might sail with barks or greater vessels more than four hundred and fifty leagues, crossing the Lake of Huron, and reaching the farther end of Lake Illinois ; which two lakes we may easily say are little seas of fresh water.' (Reprinted from a photographic plate of the original in Report of New York State Survey, 1879.) (B) A LETTER FROM MR. KALM^ A GENTLEMAN OF SWEDEN, NOW ON HIS TRAVELS IN AMERICA^ TO HIS FRIEND IN PHILADELPHIA; CONTAINING A PARTICULAR ACCOUNT OF THE GREAT FALL OF NIAGARA.* {This is flic earliest account of Niagara falls turitten in English.) Albany, September 2, 1750. ' Sir, — After a pretty long iourney made in a short time, I am come back to this town. You may remember, that when I took my leave of you, I told you, I would this summer, if time permitted, take a view of Niagara Fall, esteemed one of the greatest curiosities in the World. When I came last year from Quebec, you enquir'd of me several particulars concerning this *This is not published in the Travels of the author, but appears in Appendix to Travels of John Bartram, from Pensilvania to Onondaga, Oswego and Lake Ontario. (J. Whiston and B. White, Fleet St., London, 1851). of Canada] KALM S DESCRIPTION 43^ fall ; and I told you what I heard of it in Canada, from several French gentlemen who had been there; but this was still all hearsay ; I could not assure you of the truth of it, because I had not then seen it myself, and so it could not satisfy my own, much less your curiosity. K'ow, since I have been on the spot, it is in my power to give you a more perfect and satisfactory descrip- tion of it. After a fatiguing travel, first on horseback thro' the country of the Six Indian Nations, to Oswego, and from thence in a Canoe upon lake Ontario, I came on the 12th of August in the evening to Niagara Fort. The French there seemed much per- plexed at my first coming, imagining I was an English ofiicer, who under pretext of seeing Niagara Falls, came with some other view; but as soon as I shew'd them my passports, they changed their behaviour, and received me with the greatest civility. Niagara Fall is six French leagues from Niagara Fort. You first go three leagues by w^ater up Niagara river, and then three leagues over the carrying place. As it was late wdien I arriv'd at the Fort, I could not the same day go to the Fall, but I prepar'd myself to do it the next morning. The commandant of the Fort, Monsr. Beaujon, invited all the officers and gentlemen there to supper with him. I had read formerly almost all the authors that have wrote any thing about this Fall ; and the last year in Canada, had made so many enquiries about it, that I thought I had a pretty good idea of it, and now at sup- per requested the gentlemen to tell me all they knew and thought worth notice relating to it, which they accordingly did. I ob- served that in many things they all agreed, in some things they were of different opinions, of all w^hich I took particular notice. "When they had told me all they knew, I made several queries to them concerning wdiat I had read and heard of it, w^hether such and such a thing w^^s true or not and had their answ^ers on every circumstance. But as I have found by experience in mv other travels, that very few observe nature's works with accuracy, or report the truth precisely, I cannot now be entirely satisfied without seeing with my own eyes whenever 'tis in my power. Accordingly the next morning, being the 13th of August, at break of day, I set out for the Fall. The commandant had given orders to two of the Officers of the Fort to go with me and show me every thing, and also sent by them an order to Mons. Jonqiieire, who had liv'd ten years by the carrying- place, and knew every thing worth notice of the Fall, better than any other person, to go wnth me, and show and tell me whatever he knew\ 434 FALLS OF NIAGARA tGeol. Surv. A little before we came to the carrying-place, the water of the Niagara river grew so rapid that four men in a light birch canoe had much difficulty to get up thither. Canoes can go half a league above the beginning of the carrying- place, tho' they must work against a water extremely rapid ; but higher up it is quite impossible, the whole course of the water for two leagues and a half up to the great Fall, being a series of smaller Falls, one under another, in which the greatest Canoe or Battoe would in a moment be turn'd upside down. We went ashore, therefore, and walk'd over the carrying-place, having besides the high and steep side of the river, two great hills to ascend, one above the other. Here on the carrying-place I saw above 200 Indians, most of them belonging to the Six Nations, busy in carrying packs of furs, chiefly of deer and bear, over the carrying-place. You would be surpriz'd to see what abundance of these things are brought every day over this place. An Indian gets 20 pence for every pack he carries over, the distance being three leagues. Half an hour past 10 in the morning we came to the great Fall, which I found as follows: To the river (or rather strait), runs here from S.S.E. to iST.i^.W. and the rocks of the great Fall cross it, not in a right line; but forming almost the figure of a semicircle or horseshoe. Above the Fall, in the middle of the river is an island, lying also S.S.E. and jST.^.W. or parallel with the sides of the river; its leng-th is about seven or eight french arpents (an arpent being 180 feet). The lower end of this Island is just at the per- pendicular edge of the FaU, On both sides of this island runs all the water that comes from the lakes of Canada, viz. : Lake Supe- rior, Lake Michigan, Lake Huron, and Lake Erie, which you know are rather small seas than lakes, and have besides a great many large rivers that empty their water in them, of which the greatest part come do-s\Ti this Niagara fall. Before the water comes to this island, it runs but slowly, compared with its motion when it approaches the island, where it grows the most rapid water in the World, running with a surprising swiftness before it comes to the Fall ; it is quite white, and in many places is thrown high up into the air ! The greatest and strongest battoes would here in a moment be turned over and over. The water that goes down on the west side of the island is more ranid, in greater abundance, whiter, and seems almost to outdo an arrow in swiftness. AMien you are at the Fall, and look up the river, you may see that the river above the Fall is every where exceeding steep, almost as the side of a hill. "When of Canada] KALm's DESCKIPTION 435 all this water comes to the very Fall, there it throws itself down perpendicular ! It is beyond all belief the surprise when you see this ! I cannot Avith words ex- press how amazing it is ! You cannot see it without being quite terrified ; to behold so vast a quantity of water falling from a surprising height ! I doubt not but you have, a desire to learn the exact height of this great Fall. Father Hennepin, supposes it 600 Feet perpendicular; but he has gained little credit in Canada; the name of honour they give him there, is un grand Mcniew, or The great Liar ; he writes of what he saw in places where he never was. 'Tis true he saw this Fall : but as it is the way of some travellers to magnify every thing, so has he done with regard to the fall of Niagara. This humour of travellers, has occasioned me many disappointments in my travels, having seldom been so happy as to find the wonderful things that had been related by others. For my part, who am not fond of the Marvellous, I like to see things just as they are, and so to relate them. Since Father Hennepin's time, this Fall, by all accounts that have been given of it, has gTown less and less ; and those who have measured it with mathematical instruments find the perpendicular fall of the water to be exactly 137 feet. Monsr. JlLorandrier, the king's engineer in Canada, assured me, and gave it me also under his hand, that 137 Feet was precisely the height of it; and all the French Gentlemen that were present with me at the Fall, did agree with him, without the least contradiction* it is true those who have tried to measure it with a line, find it sometimes 110, some- times 150 feet, and sometimes more; but the reason is, it can- not that way be measured with any certainty, the water carry- ing away the Line. When the water is come down to the bottom of the rock of the Fall, it jumps back to a very great length in the air ; in other places it is white as milk or snow ; and all in motion like a boiling chaldron. You may remember to what a great distance Hennepin says the noise of this great Fall may be heard. All the gentlemen who were with me, agreed, that the farthest one can hear it, is 15 leagues, and that very seldom. When the air is quite calm, you can hear it to Niagara Fort; but seldom at other times, because when the wind blows, the waves of Lake Ontario make too much noise there against thfe Shore. They informed me, that when they hear at the Fort the noise of the Fall, louder than ordinary, they are sure a l^orth East Wind will follow .which never fails: this seems wonder- ful, as the Fall is South W^est from the Fort: and one would 28 436 FA-LLS OF NIAGARA ^^^°'- S"^^'' imagine it to he rather a sign of a contrary wind. Sometimes, 'tis said, the Fall makes a much greater noise than at other times ; and this is looked upon as a certain mark of approaching bad weather, or rain; the Indians here hold it always for a sure sign. When 1 was there it did not make an extraordinary- great noise* just by the Eali we could easily hear what each other said, without speaking much louder than common when conversing in otlier places. I do not know how others have found so great a noise here, perhaps it was at certain times, as above-mentioned. From the Place where the water falls, there rise abundance of vapours, like the greatest and thickest smoak, sometimes more, sometimes less : these vapours rise high in the air when it is calm, but are dispersed by the wind when it blows hard, if you go nigh to this vapour or fog, or if the wind blows it on you, it is so penetrating that in a few minutes you Avill be as wet as if you had been under water. I got two young Frenchmen to go down, to bring me from the side of the Fall at the bottom, some of each of the several kinds of herbs, stones and shells they should find there ; they returned in a few minutes, and I really thought they had fallen into the water: they were obliged to strip themselves quite naked, and hang their clothes in the sun to dry. When you are on the other East side of the Lake Ontario, a great many leagues from the Fall, you may, every clear and calm morning, see the vapours of the Fall rising in the air ; you would think all the woods thereabouts were set on fire by the Indians, so great is the apparent smoak. In the same manner you may see it on the West side of the Lake Erie, a great many leagues off. Several of the French gentlemen told me that when birds come flying into this fog or smoak of the fall, they fall down and perish in the Water ; either because their wings are become wet, or that the noise of the fall astonishes them, and they know not where to go in the Dark ; but others were of opinion, that seldom or never any bird perishes there in that manner; be- cause, as they all agreed, among the abimdaiice of birds found dead below the fall, there are no other sorts than such as live and swim frequently in the water; as swans, geese, ducks, water-hens, teal, and the like. And very often great flocks of them are seen going to destruction in this manner: they swim in the river above the fall, and so are carried down lower and lower by the water, and as water-fowl commonly take great delight in being carried with the stream^ so here they indulge themselves in enjoying this pleasure so long, till the swiftness of Canada] KAL.m's DESCIJIPTION 437 of llie water becomes so great, that 'tis no longer possible for them to rise, but they are driven down the precipice, and perish. They are observed when they draw nigh to the fall, to endeavour with all their might to take wing and leave the water, but they cannot. In the months of September and October j such abim- dant quantities of dead waterfowl are found every morning below the Fall, on the shore, that the garrison of the fort for a long time live chiefly uf)on them ; besides the fowl they find also several sorts of dead fish, also deer, bears, and other ani- mals which have tried to cross the water above the fall ; the larger animals are generally found broken to pieces. Just below the fall of the water is not rapid, but goes all in circles, and whiter, like a boiling pot ; which however doth not hinder the Indians going upon it in small canoes a fishing; but a little lower begins the smaller fall. When you are above the fall, and look down, your head begins to turn; the French who have been here 100 times, will seldom venture to look down, without at the same time keeping fast hold of some tree with one hand. It was formerly thought impossible for any body living to come at the island that is in the middle of the fall : but an acci- dent that happened twelve years ago, or thereabouts, made it appear otherwise; the history is this. Two Indians of the Six Nations went out from Niagara Fort, to hunt upon an island that is in the middle of the river, or strait, above the great fall, on which there used to be abundance of deer. They took some French brandy with them, from the fort, which they tasted several times as they were going over the carrying place; and when they were in the canoe, they took now and then a dram, and so went along up the strait towards the island where they proposed to hunt ; but growing sleepy, they laid themselves down in the canoe, which getting loose drove back with the stream, farther and farther down till it came nigh that island that is in the middle of the fall. Here one of them, awakened by the noise of the fall, cries out to the other, that they were gone! yet they tried if possible to save life. This island was nighest, and with much working they got on shore there. At first they w^ere glad ; but wdien they had considered every thing, they thought themselves hardly in a better state than if they had gone down the fall, since they had now no other choice, than either to throw themselves down the same, or to perish with hunger. IJut hard necessity put them on invention. At the lower end of the island the rock is perpendicular, and no water is running there. This island has plenty of w^ood, they 28i 438 FALLS OF NIAGARA [Geol. Surv. went to work directlv and made a ladder or shrouds of the bark of lindentree, (which is very tough and strong,) so long 'till they could with it reach the water below ; one end of this bark ladder they tied fast to a great tree that grew at the side of the rock above the fall, and let the other end do^vn to the water. So they went down along their new-invented stairs, and when they came to the bottom in the middle of the fall, they rested a little; and as the water next below the fall is not rapid, as before mentioned, they threw themselves into it, thinking to swim on shore. I have said before, that one part of the fall is on one side of the island, the other on the other side. Hence it is, that the waters of the two cataracts running against each other, turn back against the rock that is just under tlie island. Therefore, hardly had the Indians began to swim, before the waves of the eddy threw them with violence against the rock from whence they came. They tried it several times, but at last grew weary; and being often thrown against the rock they were much bruis'd and the skin of their bodies torn in many places. So they were obliged to climb up their stairs again to the island, not knowing what to do. After some .time they per- ceived Indians on the shore, to whom they cried out. These saw and pity'd them, but gave them little hopes of help ; yet they made haste down to the fort, and told the commander where two of their brethern were. He persuaded them to try all possible means of relieving the two poor Indians ; and it was done in this manner. The water that runs on the east side of this island is shallow, especially a little above the island towards the eastern shore. The commandant caused poles to be made and pointed Avith iron; two Indians determined to walk to this island by the lielp of these poles, to save the other poor creatures, or perish themselves. They took leave of all their friends as if they were going to death. Each had two such poles in his hands, to set against the bottom of the stream, to keep them steady. So they went and got to the island, and having given poles to the two poor Indians there, they all returned safely to the main. Those two Indians who in the above mentioned manner Avere first brought to this island, are yet alive. They were nine days on the island, and almost starved to death.* N^ow since the way to this island has been found, the Indians * These Indians had better fortune than ten or twelve Utowawas, who attempting to escape here the pursuit of their enemies of the Six Nations, were carried down the cataract by the violence of the stream and every one perished. . . . No part even of their canoe being ever seen again. of Canada] kalm's description 439 go there often to kill deer, wliicli having tried to cross the river above the fall were driven upon the island by the stream; but if the King of France would give me all Canada, I would not venture to go to this island ; and were you to see it, Sir, I am; sure you would have the same sentiment. On the west side of this island are some small islands or rocks of no consequence. The east side of the river is nearly perpendicular, the west side more sloping. In former times a part of the rock at the Fall which is on the west side of the island, hung over in such a man- ner, that the water which fell perpendicularly from it, left a vacancy below, so that people could go under between the rock and the water ; but the prominent part some years since broke off and fell down; so that there is now no possibility of going between the falling water and the rock, as the water now runs close to it all the way down. . . . The breadth of the Fall, as it runs into a semicircle, is reckon'd to be about 6 Arpents. The island is in the middle of the Fall, and from it to each side is almost the same breadth ; the breadth of the island at its lower end is two-thirds of an Arpent, or thereabouts. . . Below the Fall in the holes of the rocks, are great plenty of Eels, which the Indians and the French catch with their hands without other means ; I sent down two Indian boys, who directly came up with about twenty fine ones. . . Every day, when the Sun shines, you see here from 10 o'clock in the morning to 2 in the after- noon, below the Fall, and under you, when you stand at the side over the Fall, a glorious rainbow and sometimes two rainbows, one within the other. I was so happy to be at the Fall on a fine clear day, and it' was with great delight I view'd this rainbow, which had almost all the colours you see in the rainbow in the air. The more vapours, the brighter and clearer is the rainbow. I saw it on the East side of the Fall in the bottom under the place where I stood, but above the water. When the wind carries the vapours from that place, the rainbow is gone, but appears again as soon as new vapours come. From the Fall to the landing above the I'all, where the canoes from Lake Erie put on shore, (or from the Fall to the upper end of the carrying- place) is half a mile. Tjower the canoes dare not come, lest they should be obliged to try the fate of the two htdians, and perhaps with less success. . They have often found below the Fall pieces of human bodies, perhaps of drunken Indians, that have unliappily come down the Fall. I was told at Oswego, that in October, or thereabouts, such plenty of feathers are to be found here below the Fall, that 440 FALLS OF NIAGARA tGeol. Suvv. a man in a day's time can gather enough of them for several beds, which feathers they said came off the birds kill'd at the FalL I asked the French, if this was true ? They told me they had never seen any such thing, but that if the feathers werfe pick'd off the dead birds, there might be such a quantity. The French told me they had often thrown whole great trees into the water above, to see them tumble down the Fall. They went down with surprising swiftness, but could never be seen after- wards; whence it was thought there was a bottomless deep or abyss just under the Fall. I am also of Opinion that there must be a vast deep here ; yet I think if they had watched very well, they might have found the trees at some distance below the Fall. The rock of the Fall consists of a grey limestone. Here you have, Sir, a short but exact description of this famous Niagara cataract : you may depend on the truth of what 1 write. You must excuse me if you find in my account no ex- travagant wonders. I cannot make nature otherwise than I find it. I had rather it should be said of me in time to come, that I related things as they were, and that all is found to agree with my Description ; than to be esteem'd a false Relator. I have seen some other things in this my journey, an account of which I know would gratify your curiosity; but time at present will not permit me to write more ; and I hope shortly to see you. T ara, &c. PETER KALM. DESCRIPTIOX OF THE FALLS OF NIAGARA BY ANDREW ELLICOTT IN 1789*. Amono" the manv natural curiosities which this count rv affords the cataract of jSTiagara is infinitely the greatest. In order to have a tolerable idea of this stupendous fall of water it will be necessary to conceive that part of the country in which * In a letter from Andrew Ellicott, Esq., to Dr. Rush, of Philadelphia. Illustrated by a well engraved view of the Falls, delineated by Mr. Elli- cott. Massachusetts Magazine, July, 1790, pp. 387-388. (Note.— The illustra- tion can scarcely be considered as good.) of Canada] ELLICOTT S DESCKIPTION 441 Lake Erie is situated to be elevated above tliat which contain? Lake Ontario about three hundred feet. The slope which separates the upper and lower country is generally very steep and in many j^laccs almost perpendicular. It is formed by hori- zontal strata of stone, great part of which is what we commonly call limestone. The slope may be traced from the north side of Lake Ontario, near the bay of Toronto, round the w^est end of the lake ; thence its direction is generally east, between Lake Ontario and Lake Erie. It crosses the strait of Niagara, and the Cheneseco river, after which it becomes lost in the country towards the Seneca lake. It is to this slope that the country is indebted, both for tlie cataract of !N"iagara and the great falls of the Cheneseco. The cataract of iSTiagara was formerly down at the northern side of the sloj)e, near to that place which is now known by the name of the Landing; but from the great lengtli of time, added to the great quantity of water and distance wdiich it falls, the solid stone is worn away for about seven miles up towards Lake Erie, and a chasm is formed which no person can approach M'ithout horror. Do\vn this chasm the water rushes wdtli a most astonishing velocity, after it makes the great pitch. In going up the road near this chasm the fancy is constantly engaged in the contemplation of the most romantic and awful prospects imaginable ,until, at length the eye catches the falls — the ima- gination is instantly arrested, and you admire in silence ! The river is about one hundred and thirty poles wide, at the falls, and the perpendicular pitch one hundred and fifty feet. The fall of this vast body of water produces a sound which is fi'e- quently heard at the distance of tw^enty miles, and a sensible tremulous motion in the earth for some poles round*. A heavy fog, or cloud, is constantly ascending from the falls, in which rainbows may always be seen w^hen the sun shines. This fog, or spray, in the winter season falls upon the neighbouring trees where it congeals, and produces a most beautiful crystalline appearance. This remark is equally appli- cable to the falls of the Cheneseco. The difficulty which would * It is said by those who have visited the stupendous cataract that the descent into the chasm is exceedingly difficult, because of the great height of the banks. A person, having descended, however, may go up to the bottom of the falls, and take shelter behind the torrent, between the falling water and the precipice, where there is a space sufficient to contain a number of people, in perfect safety; and where conversation may be carried on, without much interruption from the noise which is less here than at a considerable distance. This is not unworthy the at- tention of the philosophic reader. 4:4:2 FALLS OF NIAGARA [Geol. Surv. attend leveling the rapids in the chasm prevented my attempt- ing it; bnt I conjecture the water mnst descend at least sixty- iive feet. The perpendicular pitch at the cataract is one hun- dred and fifty feet; to these add fifty-eight feet, which the water falls in the last half mile, immediately above the falls, and we have two hundred and seventy-three feet, which the water falls in a distance of about seven miles and a half. If either ducks or geese inadvertently alight in the raj^ids above the cataract, they are incapable of getting on the wing again, and are instantly hurried on to destruction. There is one appearance at this cataract worthy of some attention and which I do not remember to have seen noted by any writer. Just below the great pitch the water and foam may be seen puffed up in spherical figures, nearly as large as com- mon cocks of hay; they burst at the top and project a column of spray to a prodigious height; they then subside and are suc- ceeded by others, Avhicli burst in like manner. This appearance is most conspicuous about half way between the island that divided the falls, and the west side of the strait, where the- largest column of water descends. I am, etc., Andrew Ellicott. ]^iagara, December 10, 1789. Mr. Ellicott's estimate of the age of Niagara is not given here, but it is found in the ' Journal of William Maclay,' (Appleton's, 1890.) This memorandum of Mr. Maclay, ex- pressing astonishment, was made in 1789. (See page 20.) ot Canada] RECESSTON NOTES 443 APPENDIX II SURVEY NOTES OF THE RECESSION OF THE FALLS. Prof. Hall's survey, ^vliicli was the first made for determin- ing the recession, must remain the starting point; and, accord- ingly, his map is reproduced in Plate v. He and subsequent surveyors have left certain permanent monuments, to which I have added others. Only a few of these need be preserved, and the following is a list of them : — stations. T. P. No. 6, 1842 (Hall).— A square-topped stone monument about six inches across, rising seven inches above ground, in the path along the bluff on the southern side of Goat island, opposite the apex of the Canadian falls. A rude figure " 6 " appears on the northern side M. 1890 (Kibbe).— A stone monument in edge of gravel walk at the head ot the path leading to Terrapin rock on the southwestern side of Goat island ; now covered with ten inches of gravel, but at present accessible through a small tile pipe. I used a station " S " 493.5 feet from T. P. 6, (which is an excellent point). " S " is almost identical with " M." Loretto. 1886 (Woodward).— A brass screw one-half inch in diameter set into tin deck of cupola on Convent, directly under centre of cross. This station, which has an altitude of 192.5 feet above the bench mark at Table Rock House, is one of the best for observation. The top of the screw is marked by a cross. G. 1890 (Kibbe). — Brass bolt one inch in diameter near edge of cliff 262.5 feet from the southeast corner of Table Rock House, and thir- teen feet from crest line. The ground at this point has been covered by several feet, with a stone wall in front. However, a cast iron pipe about ten inches in diameter surrounds it and rises above the rocks near the pump house. D. 1904-1905 (Spencer).— This monument of stone is six inches square with brass bolt. It is marked G. S. D. This is in the centre of the sidewalk about sixty-four feet south of the northeast corner of Loretto grounds. It was originally used by the Electrical De- velopment Company, and is 123.0 feet above the bench mark at Table Rock House. C. 1904-1905 (Spencer).— Brass bolt in stone monument six inches square marked G.S.C. on the brow of the hill above the Michigan Rail- way terrace (Falls View) 930 feet northward of Station D, between which points base line was taken. It was also an Elec- trical Development Company station. It is 138 feet above bench mark Table Rock House. 444 FALLS OF NIAGARA [Geol. Surv. F. 1890 (Kibbe). — Brass screw with cross on top in platform round cupola of Table Rock House. It is 23.3 feet from northeastern edge of main body of the more eastern of the two southern chimneys, and 28.2 feet from northwestern edge of the more westerly one. The distance from Loretto to D 226.5 feet. C 1,143 " ' G 1,842 T P. 6 2,916 G C 1,036 G D 1,620.5 " G T. P. 6 1,547 The base of the flagstaff of the jSTatural Food ComjDany, K'iagara Falls, X.Y., is a very conspicuous feature, and may be found convenient in measuring angles from stations along the brow of the Canadian highland. From Loretto Convent bolt the angle between the flagstaff and T.P. 6 is 6° 12' 30", and between flagstaff and M it is 15° 43' 30". Other temporary positions were used which were easily connected with some of the points mentioned. For the survey of the greater part of the crest, stations on the hill from 120 to nearly 200 feet (Loretto) above the datum of Table Rock House on the Canadian side, were found very much more valuable than the low ones on the Goat Island side. From these high positions the sharp line of the water was de- fined along the crest on the western side ; while with good light the edge of the rock was distinctly shown beneath the thin sheet of water on the eastern side. Indeed from the Goat island side, it would be iinpossible to make a survey of the crest on account of the low sloping rock surfaces, and inability there- from to determine the exact edges. However, a good tangent line at the head of the apex, and some of the rock features below the surface farther toward the Canadian side, could be well seen. There were many points of rock that could be well distin- guished from two or more stations. Again little channels in the river had more or less permanent features and could be used. Such was the foundation of the survey of 1904. In addition enlarged photographs were used in 1905, and it was found that many small rapids had points that could be recog- nized when using the photographs. I had hoped to correct and fill in additional data from the use of vertical angles after taking levels of the river; but in the irregular surface it was found to be impossible to get levels sufficiently accurate, as the horizontal error would amount to from five or twelve times that 01 the vertical. As may be seen from a comparison of the successive sur- of CanadaJ EECESSION NOTES 445 veys, some crest lines show more advanced positions than older ones, indicating an error of judgment. ISTo two surveys would show exactly the same results in the minute detail owing to the personal equation. In my survey of 1904 all positions were agreed upon by Mr. Goodwin and myself. Our object was to find the position of the edge of the rock. In throwing the transit lines across the crescent, those striking the water at high angles were unquestionably good ; but in the lines crossing the direction of the current, especially where the water was deep, the question of judgment was involved, — as to what point to take, whether to adojit that where the water first broke, or where it was actuallv f-illing, or an intermediate point. There is absolutely no means of eliminating this personal element. In order to establish doubtful points it was said that Mr. Hall of the U. S. Geological Survey used searchlights, and worked at night, in order to eliminate the difficulties arising from want of proper pickets, as has been done by the Japanese. At first glance this method appeared to solve the difiiculty ; but after the experience with the question of personal element in fixing the points it remained precisely where I found it — judg- ment as to what part of the curve one should take. This could not be eliminated even by cross lights. But I lived with the falls and learned to recognize minute features. The appear- ance of the apex varies from different positions. ^fiTear station C one can look directly along the axis of the little central chan- nel which is being formed, {see Plate 6) but nowhere is there a large V or a channel reaching down through the rocks. It is only a superficial incision. From T. P. 6 on Goat island, two or three small superficial crevices are apparent ; from M on Goat island, owing to the low position and configuration of the floor of the river so that the actual edge of the rocks is not well seen, one might be led to suppose that a trough of some size existed. This idea is dissipated on examination from the Canadian side. My survey of 1904 was the fifth made for the purpose of determining the recession, and the resurvey in October and j^ovember of 1905 makes it the last. Mr. Hall's survey came in between these two, but it was not published until after my second edition was printed. The first edition was made to accompany Summary Report of the Geological Survey of Canada, 1905. The accuracy of apex in my survey is sho\^Ti in photograph Plate vi. Besides determining the rate of recession during the last 446 FALLS OF NIAGARA [Geol. Surv. fifteen years, my survey establishes the fact that there has been no central recession. This confi-rms the earlier views of nearly twenty years ago — that the crest is alternately one of flattened form, succeeded by another with a sharp apex-retreat. The form of the crest line in 1819 has not been previously studied, but it establishes this view. The fall of the rock during Xovember (1905) mentioned on page 35 is sketched on map (Plate ii.) from the appearance, such as is shown in a photograph. I had expected such a fall as this, and should look for a very considerable widening of the crescent, as the river beneath Goat island shelf is being under- mined along the side of the deepest part of the channel. If reference be made to Plate xiv. there will be observed a transverse zone of comparatively smooth water extending from the apex diagonally across the course of the rapids above the falls. This is a most important feature as here is a fragment of a transverse valley of pre-glacial date. of Canada] WELLS 447 APPENDIX III. WELLS IN NIAGARA TOWNSHIP BELOW ESCARPMENT. Feet. Lot 90 N. W. to Medina ledge 30 M 89 S.W. on Bell terrace in sand and gravel to water 7o „ 96 to water 90 „ 97 S. of road (dug) 60 „ 133 S. on terrace 65 (?) M 99 W. below terrace rock 12 , 131 E. rock 20 „ 25 E 40 „ 86 W. dug 35 „ 84 W. „ 31 ,,129 S. E. (Falty) rock. 55 ,,127 E. ' „ 32 „ 74 S. E. „ 70 „ 64 S. „ 84 ,1 125 S. E. no rock 35 „ 58 E. well no data 105 „ 80 N. W. (Soules) no rock 100 „ 81 S. W. dug 35 „ 83 S. „ 35 „ 103 W. „ 36 „ 155 N. rock 14 „ 120 N. „ 30 „ 112 N. W. rock 40 „ 110 N. centre rock 42 „ 109 N. dug 40 „ 77 W. rock 40 „ 61 dug 40 ,, 55 no rock 35 „ 58 rock 50 „ 64 S. middle rock. 90. „ 66 E. ,- 60 „ 6G W. no rock 60 „ 73 S.W. no rock 72 „ 39 M. red shale 13 „ 38 E. „ 13 ,,72 no rock 33 „ 71 N. „ 47 „ 35 E. ,- 45 „ 34 rock 44 „ 31 no rock 53 Niag, on Lake S. W. rock 25 to 30 „ 19 E. no rock 80 (60 ft. above ri v.) „ 20 E 75 „ 80 S. W. dug 68 448 FALLS OF NIAGARA [Geol. Surv. APPENDIX IV. METEOROLOGICAL TABLES. TABLE I. -SHOWING IN INCHES THE MEAN MONTHLY RAINFALL IN THE BASIN OF LAKE SUPERIOR * Years . ® 0-99 1-75 0-61 1-75 1-24 112 1-39 2-34 1-62 103 200 0-64 o < i' ^ 3 5o < 1 o > o 6 CD Q Is s 1882,. 1883 . 1-49 0-83 1-32 1-71 0-89 1-88 1-87 2-30 1 03 1-33 1 58 1-70 2-27 176 2-5S 113 1-61 1-29 1-60 1-74 0-82 0-90 1-50 9. 68 14 36 27 50 85 56 02 70 11 55 28 67 23 89 55 11 30 16 79 19 90 10 i 2 1 1 1 2 3 1 1 2 2 1 3 25 57 48 76 45 01 35 67 87 64 66 08 82 2 3 1 1 1 2 2 2 2 2 3 2 4 I 1 1 2 3 3 3 48 4 32 3 95 74 04 50 20 85 35 19 17 75 62 02 22 ()2 (iS 42 20 32 48 10 i 5 2 2 1 2 3 3 3 3 2 2 2 2 2 3 4 2 i 16 08 02 84 74 02 12 53 74 05 28 62 55 4.S 28 99 22 is 3U 60 40 2 5 2 3 1 2 3 2 o 1 2 2 4 1 2 2 2 6 2 2 5 4 4 23 25 04 (56 18 72 61 45 70 79 58 75 68 59 03 43 97 22 69 23 35 50 50 2 4 1 2 2 •1 i 2 2 •7 3 4 1 3 2 2 2 2 2 2 3 3 2 72 52 63 84 58 03 09 03 63 18 03 51 81 08 51 76 62 46 75 49 03 60 90 2 1 2 T 2 1 1 2 2 2 1 4 1 1 1 1 T 2 93 ?. i'> 26 33 23 23 20 25 28 26 22 23 28 27 26 27 29 2() 28 30 27 25 33 29 32 93 1884 221 98'3 43^ 212 52 Ls 711 93 5 791 104 094 46 3 82 2 22 2 88 2 42 3 65 1 04 1 49 1 62 89 05 41 86 47 84 03 39 74 63 88 66 58 40 29 29 36 14 30 00 20 49 1885 68 1886 . . 80 1887 . 98 67 22 36 58 49 13 78 87 60 50 37 78 36 84 62 76 40 00 I 3 2 3 1 1 3 1 3 1 3 4 4 2 4 2 1 3 6 87 1888.. 00 22 69 68 01 03 39 63 2 2 1 2 I 1 62 1889 97 1890 00 1891.. 1892., 1893.. 1894.. 59 07 88 71 1895 1030 0-961 1-351 1 77 1 105:2 21 2 37 1896 78 16 28 28 04 55 88 49 40 30 1 5 2 2 4 4 3 5 3 4 •^3 51 1897.. 1 60 77 74 73 59 80 70 84 SO 1 1 2 1 1 2 T 17 1898 09 1899 2 I 1 1 2 1 1 16 G7 01 76 45 30 40 94 1900,, 1901 , . 1902 1-55 0-80 1-09 111 100 0-60 1 2 2 1 1 29 16 86 1903 49 4 18 1901,. 1905.. 40 10 3 3 60 40 2 1 60 40 In table given by the Meteorological Bureau, years run from November to Octo- ber. In present tables, calendar years are taken. Means will thus differ slightly. *For years 1882-1898 as adapted from meteorological report in report of Chief of Engineers upon Survey of Northern and Northwestern Lakes, 1903, pp. 2878-2879 ; for subsequent years data obtained from Meteorological Bureau. of Canada] METEOKOLOGICAL TABLES 449 TABLE II.-SHOWING IN INCHES THE MEAN MONTHLY RAINFALL IN THE BASIN OF LAKES HURON AND MICHIGAN. Years . 1882 . . 1883 . . 1884.. 1885.. 1886.. 1887. 1888.. 1889.. ISDO . . 1891 . , 1892. 1893 . 1894. 1895. 2-31 267 3 03 4 00 326 2-20 2-42 1 2-44 2-77 2-78 2-57 3-79 3 1896 il-90 1897. 1898* 1899. 1900. 1901 1902 1903. 1904 1905. 3-70 2-79 i-5:> 1-36 !l-78 ,0-71 1-46 I- 50 Moil 3-281 10 313 2 28 1-741-30 2-203-21 3-86 110 1-81 2-84 205 0-65 2-30 200 2-92 2-83 1 951 15 2-222 16 1-522 60l 1-64 1 19 1-531-57 1-33 2 98 ,2-2H|3-12 1-342 68 '2-851-67 1-54 316 1-462 57 11-89 2-21 jl -80 3-30 thlyl rec lake s. . . 1-651 2 02'; 2 ' 25 : 2 -23 1-51 2 20 1 58 2 73 2 04 1-89 .3 -881 1-82, 1-45 ^2-84 i3-18 11-79 !l-53 |l-92 0-95 il-72 !3-25 j2-10 ord 3315-8616-11 .57-2-733 in 5S 3 61, 2 -87 S(i2 67 143 310 1-77 632-11 11 1-63 59 4-46 2-73 62 3-94|3-07i SO, 1-9712-34: 17 501305 2-832-143-S2 4-822-70 1-45 290 1-49 1-21 3-0(i'2-34 2 26 3-06 2-60 3 47 2-58 3-7711-59 3 70 3-2013-79 2-31 2 51 5 61 2-25 2-83|4-10 4-72 50615-23 12-86 2-28:4-25 ,2 10 3 10 3-10 omi ttedlin 5§ 64 3-66 47! 3 -.50 69|3 04 9414-62 73 2-52 59 3-11 52 3- 18 17 1-97 911-80 ■16,2 81 •012-84 16 3-72 -96|3 0C)| -49 4-451 111-34 -3812-46 -37 2 87 -78 3-38 -07 2-56 -27i3-50 -793-65 -90 3 90 uniting 13-19 3-63 l-87i 4-442 164-54 3 09 306 3-24 12-67 2-66 207: 3-221-78 3-78 l2-43 2-64"2-00' 1-22 2 84 3-23 3 -6812 01 1-58, 1-745-102-49 2-13 2-64 2-39 3 65 3114-24 3 26 2-56 2-52 1-40 2-97 3-84 ,1 713 72 1-33 2 36 2-60 2-381 ,2-34;2 291-58 2-58 114 2-35 3 29 3-36 75 3-29 134 2-21 1-99 2-52 2-33 1-94 1.50 1-90 2 90 0-301 80 records ofth 38-63 35-61 34-50 33-56 29-60 29-33 29-47 33 78 30-38 33-12 35-48 30 -70 27-90 30-20 3111 29-95 28 05 32-79 28 08 33-08 31-98 28-80 e two 35-30 TABLE IIL -SHOWING IN INCHES THE MEAN MONTHLY RAIN- FALL IN THE BASIN OF LAKES ST. CLAIR AND ERIE.f Years. 1882. . 1883. . 1884. . 1885. 1886. <^'§ 1 66 3 215 3 2-70 1 3-391 1887 19716 1888 2-54 1 1889 |2-99 1 1890 14113 1891 214 4 1892. 1893. . 1894. . 1895. . 1896. . 1897. . 1898. 2-26 ;2-79 1-92 3-15 |l-73 2 -.51 3-75 -941- •93 2 •38' • -92 2 111 -762 -661 10 2 20 3 -76 2 -58 2 751 -711 -60 2 -66 3 -47|4 452 611 962 52 2 941 88'2 60 2 89.3 13|2 42 2 13|4 812 431 60 2 62 2 11,1 03 6 03 512-72 643-91 612 84 73 2-49 25 2-85 15 4-51 315-33 02 1-61 55 7-87 75 4-11 08 5-52 66I2-29 2-57 3-83 3-21 4-965- 2-593- 4-78 2' 2-47 1 303 1 2-36 2 3-432 3-8811' 2 922 5-98 3 2-892 202 1-79 5-34 2-97 2 -.33 90 2-55 59 2-62 1-50 3'- 79 2-77 2-91 711 -15 .38 3-44 76 3-21 43 2-88 42207 6310-67 09i2-74 58i3-54 0412-82 9912-91 2-5713- 2-3712- 2-56I3- 4-45 1- 2-71 2- 2-2212' 2-77jl- 3-49I4 1-962 3 ■48^1 1-503 401 3 2 321 4-50 1 ,0-78 1 13-312 213012 31 46|l-96 2-81 7913-212-50 49 3 77 312 253-I22-80I 87 3-481-85 443-413-42 89 2-981-43 07,5-55 2-20 10' 3 08 1 63 80 316 3-23 2l!l-88 2-28 27 4 62 4-63 33 2-51 1 65 215-28 209 40 3-37 2-72 38 -.32 30-51 36-66 .33-46 31 96 30-51 31-24 40-23 33-77 39-44 37 43 30 -.38 28-41 35-38 .33-20 36 -02 * Michigan only. +ror yeai's 1882-1898 as adapted from Meteorological reports in Rept. of Chief of Engineers upon Svirvey of Northern antl Northwestern Lakes, 1903, pp. 2878-79 ; for subsequent years data obtained from Meteorological Bureau direct. 450 FALLS OF NIAGARA TABLE III.— Continued— OF ERIE ONLY. [Geol. Surv. j= 4J Years. •' fi4 _2 198 C3 4-19 a, <^ 117 4-17 3 ^-5 1-98 m 3.10 1-89 o 2.62 2-23 > o — 1-67 3-91 o P. 3 39 = 1 1899. 1 203 3-76 32 00 1900 9 10 2 04 2 39 2 11 2 313 06 5 31 3 18 1 00 31 53 190L. 92 1 67 o 74 >7 68 4 083 20 3 11 3 47 2-54 1-49 207 3 81 32 78 1902 29 94 3 96 79 •> 9 79 •-> 00 58 3 2 79 7 24 3 315 95 5 98 06 1 47 23 5 47 2 13 2-55 2-71 2 03 1 ilO 3 11 20 38 38 75 1903 42 4 15 1904 50 3 00 4 90 3 00 2 10 4 20 4 20 3 30 2-60 T80 0-40 2 20 30 20 1905 , 1-90 1-60 1-80 2-80 4 70 4 -50 4 00 3 40 2-90 2 90 2-80 200 35 30 TABLE IV. -SHOWING IN INCHES THE MEAN MONTHLY RAIN- FALL IN THE BASIN OF LAKE ONTARIO. Years. a 4 ^ < ^ § 6 a >-3 bo o O o >■ o 6 1882. . . . 1883 2 00 o 94 1 84 2 01 60 4 90 4 59 2 37 2 64 2 30 2 18 1 83 35 20 1884. . 3 67 3 0(i 3 11 1 13 3 64 3 46 3 45 •7 59 2 22 2 43 2 28 3 08 34 12 1885. . 2 82 1 79 89 2 34 2 74 4 06 3 18 5 52 3 12 4 09 2 39 9 86 35 80 1886. . 3 79 2 02 3 05 3 20 2 31 2 33 3 35 2 / < 3 77 1 93 4 74 1 97 35 23 1887. , 2 75 4 20 1 83 1 82 1 64 77 3 78 o 77 1 74 2 30 2 00 2 73 30 33 1888. . 2 47 1 84 2 75 2 85 2 35 3 39 1 88 4 13 3 40 4 2913 59 2 70 35 64 1889.. 3 91 2 21 1 1)2 3 16 3 39 6 79 17 2 02 3 10 3 47 ;5 20 3 41 43 45 1890. . 3 98 3 29 3 09 2 80 6 26 4 24 2 76 4 81 6 68 4 9012 93 2 98 48 72 1891. . 3 06 3 65 3 25 1 85 1 19 2 78 3 32 4 24 1 62 2 96 i 2 86 4 03 35 81 1892. . 3 72 2 28 2 93 1 09 5 61 5 52 4 28 5 70 o 16 1 8913 65 1 46 40 29 1893. . 2 03 3 56 2 13 3 53 5 81 2 50 3 33 5 65 3 24 2 .31 1 2 05 3 62 39 76 1894. . 2 !)4 2 61 1 65 3 55 6 77 3 23 •7 62 1 43 4 97 4 O-l 9 07 2 45 38 51 1895. . 2 84 i 77 1 47 1 86 2 58 2 57 2 60 3 89 2 29 1 4113 54 3 94 30 76 1896. 2 16 4 47 3 96 1 07 2 59 2 89 4 86 2 50 4 36 2 66 '2 ()8 1 41 35 61 1897. 2 27 1 53 3 07 2 45 3 35 2 96 13 2 35 2 03 80i4 26 3 04 33 24 1898. 3 71 2 02 2 19 2 51 3 58 3 00 2 30 5 57 3 20 4 92i2 90 2 84 38 74 1899. . 2 31 1 70 3 81 i 64 3 41 1 85 2 82 1 39 3 57 2 861 86 3 98 31 20 1900. . 3 13 3 97 3 84 1 41 1 71 2 27 4 40 2 10 9 38 3 08 5 54 9 21 36 04 1901. 2 51 1 94 3 05 4 16 4 06 3 31 3 83 3 56 2 92 1 73 3 04 4 36 38 47 1902. . 2 49 2 50 2 93 2 59 2 69 4 78 6 52 2 63 2 71 3 28 1 71 3 47 38 30 190.3. . 3 05 2 65 li 8S 2 CS 46 .) 75 4 33 4 82 1 50 4 88 2 00 3 10 39 10 1904. 3 40 2 9(1 :i .-.(» 3 2( 3 m 4 4(1 4 40 3 40 3 90 2 40 70 2 ()0 38 10 1905. . 300 i-90 2 00 2 00 ..■•* 5-30 3-80 430 2-40 3-80 2-20 2-90 36-50 of Canada] IiIETEOROLOGICAL TABLES 451 TABLE v.- SHOWING IN PER CENT THE MEAN MONTHLY RELA- TIVE HUMIDITY IN THE BASIN OF LAKES ST. CLAIR AND ERIE.* Years. c J3 i & o 72-4 €8-5 78-5 73-5 -3 . s c 1882. 1883 78-2 77-3 70 65-6 68 70-670 5 67-3l71-9J72-8 1884 77-8 80-7 75-2 68-7 67-3 70-l;68 2 68-9:71 71-9 76 781 1885 7G1 76-5 75-2 72-8 73-2 72-5 74-6 79-8!73-3|78-5 80 79-3 1886 81-8 77-8 78-6 74 3 72-3 72-3 69-6 72-8170-672-8 74 79-8 1887 78-1 82 76-5 68-8 68 72-6 68-1 66-372-3 69-8 71-5 77-7 1888 81 80 79-8 82-5 77-5 80-5 79-8 77 1 75] 74-3 75-6 64-7 676 70-3 65-5 68173-3 68-8 76 68-6 71 -8 70-674-6 67] 74 70-877-5 70-1171-6 74-5 70-8 80-6 69-5 75-3;75-8 1889 80-5 74-3 73-6 77-1 76-8 72-5 1890 72-5 65-8 71 667-6 1891 77-672-665-6 1892 791 82-3 77 168-5 77-6 77. 7:68 -8 72-3 74-8 71-7 78-8 80-3 1893 827 82-3 76 ,74-871-8 71-5 66 66 69-6 73-3 72-679-8 1894 76-8 78 70-8;70 73 1 67-8 61-3 65 72-3 74-5 74-177-0 1895 83 80-3 74-7169-3 661 65-8.62-7 67-168-3 65 -1 76-3811 1896 82-7 81 781i75-l!68-5 70-6 73 5 73-3176 6 72-1 76 |80-1 1897 82-5 80-6 77-5 73-671-8 67-8,71 -1 7ll'65-5 69 1 77 -8179 -8 1898 79-5 801 181 9 79-8 75-5 65 5;71-8 70-567-8 75-372-575-3 Mean 79-6175 4 1 808:77-8 70-1 73-7 70-1 711 68-5 70-2 72-3 72-6 75 1 77-9 73-60 Superior. . 68-9 73-1 73-2 75-0 75-0 76-5 800 81-7 76 40 Huron and Vlichig'n 81 8 81-2 177-5 72-8(71-5 74 72-] 74-3 75-8 76-5 78-8 81-2 76.46 Ontario :80-9 80-2 76-5 70 71-2 72-8 72 72-4 73-7 75-6 75-7 78-9 74-99 TABLE VI. -SHOWING IN DEGREES F., THE MEAN MONTHLY TEM- I'ERATURE IN THE BASIN OF LAKES ST. CLAIR AND ERIE. Years. 1882. 1883. 1884. 1885. 1886. 1887. 1888. 1889. 1890. 1891. 1892. 1893. 1894. 1895. 1896. 1897. 1898. Mean 245 259 I I 26 29 16 24 28' 5 25 8120' 34 4 32 1|29 223 2'24 3 18 9 26 5 28 7 27 28-5 33 1 24-6 9:33-2 4 29-8 2-2S-2 42 3;jr)-8:45' J30-6i45 |31-3 47' 7|30 !44 6 33-144 5'40-S,46' 2 29 146 2 2S ■ 4 51 2 35-8 45 42 44 531 66 57 i68 55-5 64 56-8:65 61-667 7 54 3 67 4 57 4 63 .54 -2169 53-9 67 54-7 69 3 54-3 68 5 56 5 68 9 58 -7 69 1 63 8 66 3 54-4 63 3 57-9 68 670 5 68 572 :69 8 75 (i9 971 971 4 66 171 32145-3 56-5 67-3 71-1 67-2 67-3 65-6 »;8-7 68-5 68 1 67-5 62 66 -7160 69 l67 70-5 63 69- 162 68-7 66 70-5 66 69 -9 60 67 -2 65 71-6 66 68-6 63-2 40 50 7 43 39 30 49 (40 53 137 48 1.38 45-8 40 46 5 40 51-6 46 51 |38 51 -6 36 52-9 38 53 1 36 45-6 39 17 8 42 55-8 40 531 507 4 28 2 32 3 30 7 30 623 30 S31 6 39 7 28 337 9 27 9 29 3 33 6 31 6 30 2 30 40 30-9 480 -5g * Authority same as proceding tables. 29 452 FAI.T.S OF NIAGARA [Geol. Surv. TABLE, SHOWING IN DEGREES F., THE MEAN MONTHLY TEMPER- ATURE IN THE BASIN OF LAKE SUPERIOR. From the year 1882 to 1898 inclusive. i 6-2 4 83 i 18-4 ^ g s 62-3 < 601 a m -s o > o 25-8 6 c lr7 Annual Mean. Mean 35-7 46-6 57-2 1 53-9[ 41 1 35-98 HURON AND MICHIGAN. Mean 16-8 18-1 25-6 402^ 50-4 1 i 1 61-6 66 1 63-8 58-3 45-8 33 9 1 24-5 42 08 LAKE ONTARIO. Mean 20-5 21-3 274 41-6 530 63-8 67-4 05-2 58-9 46-6 36-4 26-5 44 01 Authority cited for preceding tables. TABLE VIL— SHOWING IN MILES PER HOUR. THE MEAN MONTHLY VELOCITY OF WIND IN THE BASIN OF LAKES ST. CLAIR AND ERIE.* Date. < a . . . ST > 10 13 2 11-6 11-4 12-7 12-1 11 10-2 10-8 12-6 12-2 131 14 1 11 14-5 14 12-1 9-5 11-4 12 5 d P 11 CM 18'^2 11-5 11-3 12 6 12-9 10 4 10 6 12 3 12 11-8 13-1 11-1 14 1 13 2 12-6 11-9 12-7 12 1 9-4 11 3 12-9 1883 ii-6ii-7 11-5 10 9 10 10 11 9 9 10 9 12 14 9 10 11 12 11 8 9 3 2 8 6 2 9 2 2 8 2 9 () 5 10 3 8-8 9 1 8-3 7-3 9-3 9-7 8-6 8-5 10-7 10-6 10-2 9-7 11-3 10 2 9 4 9 5 9-3 10-4 9-4 7 6 7-4 8 9 7-7 7-1 8 6 8-9 6 8 7-5 9-2 9-3 8-5 8-1 8-7 91 9-2 8 3 8-2 8-5 8-2 7-5 81 8-3 7 8 7-5 7-3 8-2 8-7 7-9 9-2 9 8-5 9-6 8-9 8-7 8-3 7-5 8-3 8-3 8-1 8-7 7-9 9 8-5 9 7-7| 9 7 8-2 8 5 8-6 7-7| 9 3 7-8 8 7-7^ 7-9 8-710 9-7 10-4 9-4 9 1 10-7 10-7 9-8 8-8 10-4 11 11-3 10-3 12.8 10-5 10-8 12 10-5 9-5 10-4 10 6 1884. .. 11-4, 10-2 !V8 13-2 10-2 11-2 11-2 12-8 10 8 14-2 13 |l0-9 10 710 412 12 12-11 9-6 12-811-6;ll-7 9 1 12-411-6 11 810 310 4 11 7 13 4 12-7 121 13-2112 1835. . . 1886. . 1887. . , 1888. . . . 1889. 1890. . . 1891. . . 1892. . . . 1893. . . . 8-4 7-8 8-1 9-2 9-2 8-7 8-2 7-8 8-3 8-1 11 9-8 10 10 3 8-8 10-4 9-2 9-3 9-4 90 1894. . . . 1895. . 1896.. . 1897.. . 1898. .. 13 2 10 3 13 9 13-4 12 9 11-2 13-2 13 1 13-8 12 1 12-5 12 9-4 11-5 ^?,■9. 12-5 13 4 13-S 11-5 11-6 9-8 11 4 12 3 10-9 10 11 5 10-5 Superior Huron ai Ontario. 10-4 9-05 idMichig'n 10 3 10-7 Authority as cited in preceding tables. of Canada] T.AKE FLUCTUATIONS 453 APPENDIX V. TABLE I.— FLUCTUATIONS OF LAKE ERIE AT PORT COLBORNE.* Years. 1850.. 1851.. 1852.. 1853.. 1854.. 1855. 1856.. 1857 . . 1858. . 18511.. I860.. 1861.. 1862. . 1863.. 1864.. 1865.. 1866. . 1867 . . 1868.. 1869.. 1870. . 1871.. 1872. 1873. . 1874.. 1875. . 1876. . 1877. 1878.. 187^.. 1880. . 1881.. 1882.. 1883.. 1884.. 1885. 1886. . 1887.. 1888.. 1889. 1890. . 1891 . . 1892. 1893.. 1894.. 1895. 1896.. 1897.. 1898.. 1899.. 1900. . 1901.. 1902.. 1903.. 1901.. 1905. 1906.. 94' 2 84|l 742 24il 352 563 042 43 12 293 50 3 391 341 98 09 26 20 81 52 46 16 06 33 75 50 64 05 49 16 16 41 75 00 08 83 08 66 41 16 50 91 16 251 08|1 910 5810 11 2 30 L' 86 { 2 1313 3 3 4 3 3 3 3 22i2 47 12 622 09 2 21 1 27 26 73 82 8712 38 1 28 3 93 2 52 2 04 2 90 2 68^2 923 15 2 74 3 75 2 08 331 33,0 160 1611 5012 33|l 25 !l 16 33 1 58 2 33!2 4110 0811 21 2s'2 S9 2 40 2 oS 28 93 si 01 97 2 68 2 85,2 79'2 02 2 90:3 372 312 16 2 03 3 17 3 16 3 16 75 25 33 83 75 16 08 58 50 58 75 58 08 58 1 41 1 50 25 1 25 2 66,2 91 2 50 1 74 3 62,2 9014 30 3 _12[2 57 2 97 3 32 4 194 59 4 96 3 114 51 3 -143 52 2 61 2 39 2 26 2 19 3 79; 3 55! 1 5512 612 112 964 .'^4 2 183 212 502 27 12 14 3 514 58 2 913 33 3 413 33 2 08 2 16 3 001 83 2 50 2 91 2 911 1 2 2 2 1 1 25|1 1612 7512 002 751 05 2 17 2 82,2 16 1 42,2 57 !3 1514 103 49 3 05 4 45 3 45 3 73 2 85,3 ..'3 96 2 84 3 27 3 06 3 851 65 2 042 372 004 55 2 083 38 2 70 2 512 63 3 30 3 58 3 502 50 2 .50 3 91 58 66 41 2 00 1 1611 00;2 1612 00 1 58 il 251 .58 2 33 2 3 4 3 3 3 3 33 3 17|3 56^2 (•1 2 06 2 88 2 23 2 46 3 012 961 83:2 84'2 452 02 3 76 2 02 2 32 2 74 2 69 2 50 3 69 3 413 58 3 41 3 413 83 2 W^ 2 25 2 081 83 2' .50 2 25 1 110 331 00 2 00 1 911 911 331 33 2 412 912 58 2 001 98 1 8(12 25 3 44 3 64 2 3912 702 42,3 53 3 92 3 55 3 713 87 3 50 3 05 2 32 2 S2 2 77 2 62 2 99 2 27 2 7512 7811 50 2 59 2 39 2 66 3 63 2 90 2 141 30 2 27 1 29 2 36 2 33 2 .50 3 08 [3 oo;2 752 412 75,2 661 50 '2 ooi CO 1 910 08 1 33 2 08,2 66 2 412 92,1 1 2 2 3 2 3 1 2 3 3 2 3 60 3 12 3 86 2 43 2 39 3 28 2 27 11 702 732 .6 2 80 37 27 38 26 43 83 91 34|2 951 83!2 87'2 83 2 33j3 252 66^2 33 2 08 1 •),- ■> 41 jl 0S|1 661 .50 1 910 08 50 1 66 1 410 50 1 16 00,2 001 33 68 3 00 83 58 25 66 83 001 411 91 1 08 1 91 10 45 1 62 2 9S3 472 4711 14 3 70 1 19 3 46,3 34|3 29 2 70 3 92 3 66 2 83^2 22 2 713 74il 68 1 3l!2 68 '2 ()6 2 84 1 20 2 94 1 942 9513 28|2 30 3 60 1 OOil 32:2 26,2 00 2 42 2 33 3 83J2 25^2 91 2 50 1 41 2 OOjl 501 1 3311 .5011 250 75[0 25 1 25' 1 00, 1 16 1 26 1 7'.t 2 21 2 673 89 '2 36'2 98 2 22 2 56:3 16'3 87 '3 62 3 03 3 94 3 70 2 05 2 17 2 42 2 8712 462 69 2 00,2 22 1 %\ 2 59 02 04 36 19 55l2 90 2 2711 33'2 80 2 ;!.•{ 2 50 2 .50 2 66|2 08 2 6(; 1 08 2 331 16 1 75!0 33 1 33I1 501 501 971 S: 1 i a a < c 1855 2-39i 2 14 1 2-95! 3 73 3-95 3 •3013 54 3 613-89 3 10 1856 3 25 2-85 2 52 3-01 2 97 3-35 3 38 3 232 9^2 322 202 49 2 88 1857 1 74 L-77 3 50 ^■88 3 97 3 93 3 683 22 3 76 3 76 3 32 1858 3 85, i-45 3 48 3-59 3 85. 5-21 5 16 5 074 514 413 99 4 09 4 22 1859 4 08 . i-86 4 27 4-84 4 72 1-69 4 75 4 45 3 851 063 88 1 3 68 4 26 1860 3 26' 2-90 3 30 400 4 21^ 1-18 3 92 3 76 3 42,3 123 03 2 87 3 49 1861 2 6i; i-33 2 77 3-81 4 24 I 314 06 4 103 923 693 67 3 44 3 58 1862 3 43i. ?]4 3 28 4 IS 4 42^ 1424 39 4 013 703 32 2 98 3 01 3 69 1863 3 46; J-75 3 69 3 -81 3 99; 5-85 3 73 3 65 3 26 2 82,2 41:2 38 3 40 1864 2 09 1 3 24 2 45 2 9.J 3 65: 5-60 3 34 p 07 2 85:2 54 2 3712 44 2 79 1805 2 01 ] L-43 1 75 2-47 3 05l: 5032 99l2 91 12 872 57 2 19 2 05 2 44 1866 1 78] L-62 2 01 2-59 2 81 1; 5-07 3 182 9352 87 2 86 2 62 2 63 2 58 1867 2 34: } 02 2 42 274 3 26: 5-57 3 .38 '3 073 68 1 2 34 1 84 T 62 2 60 1868 1 42] L-04 1 63 2-46 2 91 ': 5 -.30 3 27(2 75 2 48 2 03 1 87 1 66 2 23 1869 1 65 L-58 2 06 2-36 2 91 ;: 5-30 3 58 3 48 3 21, 2 76 3 30 2 65 2 65 1870 2 89: 5 12 2 89 3 54 3 75: 5-72 3 76 3 7113 463 08 2 78 2 66 3 28 1871 2 45 i 212 2 573-05 3 .32: 5 35 3 33 3 12|2 95 2 28 2 101 66 2 69 1872 1 58] L 34 1 251 4P 1 891 2-26 2 25 22il 991 82 1 49 1 26 1 73 1873 1 16!] L-17 1 24 2-52 3 19: 5-27 3 25 3 192 79 2 49 2 29'2 66 2 43 1874 3 05:: 510 3 13 3 30 3 39i: 5-46 3 49!3 33 2 872 43 2 01 1 80 2 94 1875 1 57 1] L-40 1 541 94 2 4lll J-84 2 97 2 96 2 82:2 33 2 i8;2 40 2 28 1876 2 36 il 292 3 57 4-09 4 41^ 1-52 4 41 4 11 3 94 3 41 3 493 15 3 69 1877 2 75 ' J-59 2 36 2-79 3 04: 5-12 3 36 3 22 3 142 74 2 66 2 74 2 87 1878 2 X2'. 2-96 3 09 3-51 3 75; 5-753 75 3 53 3 40 3 05 2 85 2 93 3 28 1879 2 511 2 37 2 40 2 76 2 91: 500 3 03 2 81 2 48,2 25 1 :8i2 04 2 52 1880 2 541 2-58 2 72 2-88 3 15: 5 26 3 35 3 11 2 8812 44 2 36 2 02 2 77 1881 1 61] L-72 2 04 2-74 3 14: 5-38 3 33 3 Oi 2 66 2 61 2 432 64 2 61 1882 3 11 : j-n 3 56 3-78 3 98^ I 13 4 06 3 92 3 6513 20 2 882 37 3 48 1883 2 281 J-49 2 68 2-80 3 26: 5 96 4 16 4 10 3 7913 47 3 09 3 12 3 26 1884 2 79: 505 3 24 3-79 4 06- 1 14 3 ;»2 3 76 3 33 i3 00 2 52 2 45 3 33 1885 2 271 206 1 92 2 74 3 47: 5-98 3 94 3 95 3 80 i3 70 3 5813 .53 3 24 1886 3 551 2-82 2 63 3 51 3 81 : 5-91 3 89 3 69 3 44 3 21 2 92 12 83 3 35 1887 2 61 : ? 04 3 82 3-87 4 05- I 07 3 S4 3 52 3 29 2 70 2 43:2 45 3 29 1888 2 271 200 2 10 2-7;^ •J its ; 5 11 3 2(5 3 16 2 72 2 35 2 41 2 29 2 60 1889 2 31- 215 1 99 2-34 2 r>'2 ' 1 ■ 95 3 15 •) S4 2 45 2 03 1 76 2 02 2 37 1«90 2 381 2 (,7 2 79 3-28 3 62|: 5-99 3 61 3 15,2 98 2 79 2 76 2 53 3 05 1891 31' 2-29 2 75 2-62 2 44il 2 -.58 2 48 2 2112 03 1 651 21 28 2 15 1892 31 LIO 1 14 1 70 2 50: 5-26 3 3S 3 03,2 71 2 15 1 82 .55 2 13 1893 17 La5 1 47 2-20 3 04: 5-23 2 95 2 6112 23|1 88 1 48 56 2 08 1894 84 1-72 1 75 2 15 2 541 2-75 2 73 2 36 2 19 1 87 1 63 56 2 09 1895 23 LOO 1 01 1-26 1 48i^ L-57 1 46 1 381 28 8o;o 70 86 17 1896 96 )-88 831-28 1 66 •93 1 81 2 021 70 46 1 09 12 1 39 1897 . 09 1-29 1 6612-21 2 .54 1 2-64 2 63 2 47;2 19 70,1 57 54 86 1898 .59 I 79 2 05,2-63 2 78' 2-81 2 59 2 39 2 01 811 69 52 2 13 1899 67 1-46 1 83:2-13 2 44 2 .56 2 28 2 091 85 Blil 62 34 90 1900 36 L-57 1 92 2 23 2 39 2 47 2 34 2 311 00 751 49 45 94 1901 35 100 88 1-29 1 31 L-72 1 91 1 781 71 331 16 19 .38 1902 08 D-63 94 1-49 1 86 212 2 74 2 72j2 38 2 292 02 82 84 1903 72 1-70 2 283-05 3 09 305 2 98 2 76 2 59 2 251 77 31 2 37 1904 19 1-28 1 86 2-91 3 17; 3 32 3 41 3 102 84 2 492 12 77 2 45 1905 52 1-31 1 18 1-83 2 46 2-98 " 07 2 88 2 59 2 :30 1 93 92 2 16 Rept. Engineers Northern and Northwestern lakes. of Canada] LAKE FLUCTUATIONS 455 TABLE III. -FLUCTUATIONS OF LAKE HURON* MONTHLY MEAN HEIGHT OF WATER SURFACE AT SAND BEACH f ABOVE MEAN TIDE AT NEW YORK. Add 580 feet (except when whole number is 9 when add 570).+ Date. 1854. 1855. 1856. 1857. 1858., 1859. 1860. 1861. 1862., 18(W. 1864. 1865., 1866.. 1867. 1868., 1869 , 1870. , 1871., 1872. 1873. 1874. 1875. 1876. 1877. 1878., 1879. 1880. 1881. 1882 , 1883.. 1884. 1885.. 1886. , 1887.. 1888.. 1889.. 1890. 1891.. 1892.. 1893. . 1894.. 1895.. 1896. 1897.. 1898. 1899. 1900.. 1901. . 1902.. 1903. . 1904.. 1905.. 103 1.71 1-58 2 23 2-70 2-78 03 2 43 2-21 1-75 0-81 63 0-39 05! 1 10 0-57 0-48 1-37 1-65 079 0-57 1-86 161 1- 10 741-72 46 2-45 1-89 1-29 0-99 1-5511 1-721 1-681 2-41 2 2-38 2 2-74 3 2-26 2-45,2 34 1-25 251117 78 0-1)6 53 42 93 9 -87; 9 84 9-73,9 23 0-24 03,9.919 9-?89 1 1 1 2 2 3 2 2 2 2 1 1 28! 1 25 1 22 6712 252 29 641 001 141 852 382 06 1 251 98 (I 56 1 86:2 72,1 45 1 2 38 '2 !i:^ 3 66 2 421 151 64 39 9-43:9 9-790 9-61 9 9 90 9 9 -81 ,9 9 79 9 90|9-920 38l0-30l0 02 76 9 93 26(1 73 9 92 09 29 lo 80 75 29 522 15 3 3 23 2 95 2 471- 881- 571- 151- 59 0- 09 '2 45'2- 711- 05 1- 75 1- 331- 13 2- 46 2 99 2- 28 1- 99 1 5S 1 - 08 2- 73,2 76'2- 49 2 22 3- 57 2- 562- 081 • 781- 72 0- 01 0- 25 62 1- 020- 2] 9- 910- 51,0 93 0- 05 ."..•) ■ 910 27 0- 610- 66i0- 92 3 552 38 2 90 68 73 63 39 41 39 90 ll 25 1 55 3 772 00 2 221 09ll 960 14|0 88,1 02 1 170 61 '9 4810 630 470 20'0 ,si> KliO 07 08 22 82 57 55 18 33 04 3 (is 2 39 2 2-39J2-35 2 2 3 3 4 3 511 2 2 95 24 99 2 22 3 59 2 56 2 551 S,S2 07 2 48 2 723 043 18 3 64 3 89 2 30 2 551 52 1 910 67 (J31 32 1 26 98 72 79 861 37:0 911 48|0 56 41,1 4011 2-28 2 10 3-44 3-92 3-903 3193 3-56 3 3 07 2 •J '6(1 2 -.2-23 1 10 2-12 2 1 1 1 1 2 2 1 1 2 2 3 2 2 T (19 1 22 2 on 1 62l2-68j2 20|3-40 3 12 3 OOI2 243-393 483-3713 (^2 i-(;8 25 2-lt- 67 1-33 83 2-09 68 2-59 87 2-56 61 1-58 15 2-16 4{ 2-29 If- -.'-15 6( 3-60 77 2-67 60 2-50 59 1-46 19' 2-762 2-372 1-751 1 711 0-840 1040 1-21:0 1-201 0140 08 9 0-9110 (V78 1-06,0 0-710 1110 0-840 0-730 5511-531 5ll-56ll 79 88 520 800 361 451 671 2412 79,1 36 '2 37l3 33 2 62 2 26.2 132 20 2 86,0 62 1 33 1 74 1 001 09 26 22 14 61 14 33 75 81 06 02 19 21 1 1- 0- 23 1 28 9 (36,1- 710- 77 0- 7219 ■ 70,9- 30 0- 2810- 45|0- 74 0- 66 0- 27 0- 83 26 0- 22 0. 401-22 04 2 14 73 1-50 55 2 -59 04 3 06 83 2-90 50 2 20 95 2-82 812-52 16 1 93 21 1 08 20 89 331-(J7 120-77 7910-51 501-22 931-58 53 1 16 06 0-77 86 1-70 30 1 45 89 1-62 94!2-75 21 2 16 031-83 02 1 00 491 29 27 2 10 08 1-93 86 2-75 46 2 20 87 2-67 75 '2 -43 741-452 60 1-391 8710-7111 03 0.731 97 9 910 33 09 48 0-28 58 0-28:0 33 9 (19*9 .-169 44 9 18 9-.SS0 20 (JllO 310-12 73|0-53'0 43 012 14 9 93 41012,0 95 0-50 93,0-67i0 *From Rept. of Engineers, U. S. A. 1904. fFliictuations from 1860 64 at Port Aux Barqtie, Mich.; 1865-70 Milwaukee, Wis.; 1871- '4 Port Austin, Mich.; 1875-1900 Sand Beach, Mich. JThe above are new corrected elevations. 507 ft. above old which were based on G.T.R. gauge 80 miles below Lake Hui'on and 0'66 ft. below lake at Ft. Gratiot. 456 FALLS OF NIAGARA IGeol. Surv. o - ^ • = o ^ ^ I > O ^ •< sis 2 bb I-:) o a> i-l.-lr-l(Me' c S5S?fe£§?5S^^:i^??^^c^^??Sgg??§SgS?5c^i^ti§ I-l J— 1 fl r-l 1-1 I— 1 1— 1 i-H I-l r-H ■ ^^ 1 1 1 o O tH(Mi-nM(Mi-ii-l(Mi-li-li-l rt r-li-i —1 ^ ^ i-iC-lrHr-(rHCqCi 1-^ cc ^ i-H :c ^ ^ im c^ t-i i-h .-h i-^ .-i jm s CM c; t^ t~ o 00 c: ^ t ^ ;2 c; Tj- 1^ i~ - CO r^ l^ C^ h^ O O "^ O CO 'I' CO CO g ^ S ^ i^O c^ O O CO QD 05 CO ^ CO O T. •* O CO .O O CO I I eo^cCCOiOiOr-OOCOCOjHt-.CCiO^t-C^gj^OOr^ Cq(NC0S^1.-I.^C1--l'^'-"^'|"|' C>0,-(IM(Mr-l I I OOCOIMl-CStNlMCOCOt-tOCOCOOSD'Jt-OjDOp'-O^ oco'S?^?^-*o"-ccoi--i^cooinaiO»aj-*-;*'i.^«P7f' CO (N CO CO 1-1 1^ li-lT3COCOC5;-J03;Te SoS^l0010©©.-ll-CO©Ct-OinTt<,-l© -■^-' ■--■, I '77 I + + '**??"" SS§tScqb^©t-^-5'©iN-*©(Nco9t-'r''-'r':^ ^'r^i^'i— IrH T— l^■ lO O o o o © o © o S to ^■'^ o aO S tpi o %f^ ^^ ^ . tT . n ■jj ^'Z c3 <«> r^ u> 1 i 1 w D - oj n ^ t! go S"1^ ^ g^ 0) flT- a2 ,3j 4S Ph S-< ^ Pm* c ^*aow»'XxXx>-XXxxxx^02=^=^25S 458 FALLS OF NIAGAEA [Geol. Surv. TABLE v.— FLUCTATIONS OF LAKE ONTARIO AT TORONTO, CHAR- LOTTE AND OSWEGO. Charlotte Oswego Years. Toronto. Charlotte. above Toronto. Oswego. above Toronto. 1850 245-74 245 49 1851 245 47 ' 245 72 1852 246 34 246 48 1853 246 91 246 96 1854 246 68 246 21 --47 246 46 --22 1855 246 28 246 02 -■26 245 62 -■66 1856 246 247 56 07 246 246 33 74 --23 -•33 245 246 78 72 - -78 1857 --35 1858 V47 40 248 33 + ■93 248 22 + -82 185lt. 247 246 14 31 247 246 02 83 + -78 + -52 248 247 10 05 + •96 1860 + •74 1861. 247 246 05 92 247 247 61 60 + -56 + ■68 247 247 52 72 + -47 1862 + •80 ISG:^ . ... 246 246 246 245 246 245 50 29 03 62 44 17 247 247 246 246 247 245 20 02 76 42 37 95 + •70 + ■73 + -73 + -80 + -93 + -78 247 247 246 246 246 245 27 03 98 22 92 64 + -77 1864 + -74 1865 + -95 1866 + -60 1867 + -48 1868 + •47 186'.l 24(; 06 246 80 + -74 246 62 + •56 1870 247 245 20 84 247 246 88 51 + -68 + -67 247 246 67 27 + -47 1871 + -43 1872 244 41 244 98 + -57 244 91 + •50 1873 245 246 244 53 28 96 246 246 245 05 80 51 + ■52 + -52 + ■55 245 246 245 95 66 36 + -42 1874 + -38 1875 + -40 1876 246 76 247 37 + ■61 247 19 + -43 1877 245 58 246 18 + ■60 246 01 + -43 1878 246 245 10 70 246 246 78 43 + •68 + •73 246 246 59 30 + -49 1879 + •60 1880 245 51 246 03 + ■52 245 90 + •39 1881 245 246 14 13 245 246 72 79 + ■58 + ■66 245 246 60 6t» + •46 1882 + ■52 1883 246 246 31 96 246 247 82 34 + ■51 + ■38 246 247 76 37 + •45 1884 + •41 1885 246 59 246 80 + -21 246 92 + •33 1886 247 246 31 77 247 247 52 10 + -21 + -33 247 247 70 14 + •39 1887 + •37 1888 245 56 245 88 + -32 245 99 + •43 1889 245 67 246 11 + ■44 246 11 + •44 1890 246 73 247 08 + •35 247 24 + •51 1891 245 77 246 17 + •40 246 20 + •43 1892 244 93 245 45 + ■52 245 47 + •54 1893 245 49 246 09 + ■60 246 07 + •58 1894 245 31 245 89 + -58 215 87 + .56 1895 243 244 81 06 244 244 44 70 + ■63 + ■64 244 244 37 71 + •56 1896 + •65 1897 244 44 244 94 + ■50 244 90 + •46 1898 245 03 245 40 + 37 245 45 + ■42 1899 244 97 245 33 + ■36 245 28 + 31 244 244 91 73 245 245 35 24 + ■44 + -51 245 3? + •41 1901 245-27 + •54 1902 245 01 245 46 + ■45 1903 245 54 245 99 + •45 1904 246 29 246 63 + ■34 1905.. 245-91 246-31 + -40 of Canada] DISCHARGE OF NIAGARA RIVER 459 APPENDIX VI. TABLE I. -MEAN MONTHLY DISCHARGE OF NIAGARA RIVER IN UNITS OF A THOUSAND CUBIC FEET.* Years. 1860. 1861. 1862. 1863. 1864. 1865. 1866 1867. 1868. 1869. 1870. 1871. 1872. 1873. 1874. 1875. 1876. 1877. 1878. 1879. 1880. 1881. 1882. 1883. 1884. 1885. 1886. 1887. 1888. 1889. 1890. 1891. 1892. 1893. 1894. 1895, 1896. 1897. 1898. 1890. 1900. 1901. 1902. 1903. 1904. 1905. 234 219 238 23K 207 206 201 213 192 198 225 216 196 187 229 196 214 222 224 217 218 197 230 212 233 212 240 21!i 212 212 214 212' 190 187 202 188 182 185 196 198 191 191 185 225 213 231 245 211 l".i;! 197 204 184 196 231 208 191 187 230 192 226 219 227 214 218 199 230 21(; 229 207 224 22! t •20b 209 i 220 212 1S."> 189 199 183 180 189 201 193 196 1.S3 175 235 250 223 246 234 254 243 24G 21(1 227 200 2l(i 206 219 215 222 197 216 207 21 4 1 225 240 255 256 260 250 243 229 254 257 250 247 241 229 2491 245 237 2.52 2.53 249 259 251 244 244 243' 234 236! 229 225 228' 226 225 218 189 188 231 195 241 214 230 214 222 206] 240 229 193 217 235 204 252 223 239 223 225 222 245 221 223 233 246 204 220 246 207 222 247' oool 205 1 213, 223! 234 1 222 21!li ]'.)9| 210 209 189' 189 210 220 208 211 189 194 1S( 194 200 183 179 198 20 202 204 180 182 224. 229 234 241 226 235 226 235 245 244 235 236 203 211 232 234 237 i 238 215 224 260 262 229 231 245 245 226 228 231: 234 231 236 250! 253 224; 249i 252 254 238 250| 246: 248; 251' 252 227 230 218; 227 242 1 250 2151 218 217 234 229 233 232! 226 236 i 229 234 222 241 239; 233 245 241 238 225 221 216' 2311 229 I 243 243' 235 227 224 215 21s 214 1 218 210 225 1 2191 213 206 223 202 202 212 235 211 234 239 227 260 236 245 229 236 235 2271 2051 223 230 210 232 235 i 225 227 224 253 249 233 231 240 237 224 21(i 230: 225 228 220 252' 2491 243 254j 253| 245 249 1 245 235 249 i 249 246 248| 243 238 247 240 234 2341 232 231 I 224 242 232 222 216 227 21() 210 206 218 194 198 218 223 215 222 196 2(4 220 224 218 214 216 199' 203 202 208 236 229 227 219 222 214 193 192 201 206 220 216 219; 214 212! 207 213 212 203 201 222 222 221 211 210 189 199 210 206 202 205 199 214 230 223 212 208 201 194 216 212 215 201 i 213 20!) 237 239 222 220 229 225 211 201 215 214 219 215 232 225 238 230 228 218 1 244 241 233 226 221 215 213 215 206 200 223 223' 198 188 209 201 203 194 203, 197 178 176' 193 185 199 196 201 198 1971 197 1 200 194 190 187 212, 206 225 238 228 214 215 207 220 19 198 220 220 11 189 220 201 214 231 222 226 206 206 220 214 231 216 240 224 216 212 206 217 189 195 196 196 180 239 241 243 237 223 215 218 219 211 220 234 221 199 215 227 212 244 225 234 217 223 219 239 234 236 233 236 235 219 214 229 209 209 207 207 187 186 192 195 205 195 209 191 203 193 204 187 193 201 202 213 215 208 From Rept. of Chief of Engineers, U.S.A., pp. 2875-76, 1903. 4G0 FALLS OF NIAGARA [Geol. Surv. TABLE II.— MEAN MONTHLY DISCHARGE OF ST. CLAIR RIVER IN UNITS OF A THOUSAND FEET.* 1 Years. c 4 ti a ^ 259 ^ ^ >* 1 c5 a 198 178, 1851 201 g 1 230! 254 2571 254 O 245 242' Q 231 ^ 1860 228 1861 182 163 171 191 231 257 263 2661 265 2601 251' 225 227 1862 190 1711 174 194 230 247 249 252 2481 259i 252 213 223 1863 -189 167, 171 188 223 241 239 241 239 235 237 197 214 186t 178 158 1 163 183 220 231 235 2341 226 213 208 193 203 1865 157 140 148 174 205 217 231 2321 229 904 210 196l 197 1866 156 1321 138 163 195 208 215 218l 214 211 210 191 187 1867 163 155 145 154 135 152 176 168 208 201 226 214 212 215 212 209 227 204 220 201 210 200 186 198 1868 200 188 1869 151 133 134 157 192 209 218 226 225 218 218 207! 191 1870 168 151 1611 186 221 239 242 240 216 237 227 19(5 210 1871 182 161 178 198 233 250 251 243 235 225 221 191 214 1872 162 139 143 160 195 219 225 221 220' 216 213 209 194 1873 155 135 142 166 204 225 230 231 228 232 231 203 198 1874 17(! 160 168 181 210 235 238 235 232; 226 211 207 207 1875 165 145 151 171 206 230 233 233 238 233 233 229 206 1876 177 157 165 186 226 248 262 263 2671 258 254 218 223 1877 190 :7i 175 193 224 246 249 247 2411 237 233 212 218 1878 183 173 160 149 169 153 184 170 220 201 239 19(5 238 215 2.36 231 213 213 235 208 234 208 196 201 210 1870 192 188f* 164 166 143 154 148 159 165 176 201 210 222 225 230 230 229 228 219 217 240 217 213 199 18S1 225 224 237 205 1882 181 157 165 185 217 223 22S 2.30 228 236' 230 229 209 1883 176 156 162 179 218 238 213 2:)2 219 243 255 219 216 1884 192 170 176 198 231 245 247 24S 241 251, 245 227 223 1885 190 170 175 193 229 254 258 21 -.2 261 2581 -251 223 227 1886 194 177 185 207 242 272 266 262 259, 2571 253 204 232 1887 187 171 180 195 227 248 251 246 23^1 236 1 22(; 203 217 1888 169 148 157 175 212 240 236 239 232 227' 222 190 204 188!) 167 158 147 152 16G 161 198 195 214 206 222 214 222 1 219', 213, 205 191 193 1890 137 142 2185 213 209 202 175 186 1891 154 132 137 1.59 193 202 201 200 195 187 180 178 176 1892 142 123 128 146 177 185 193 199 1951 195 190 182 171 1893 140l 120; 127 150 191 219 208 205 200 i 196 193 L82 178 1894 148 129 137 158 194 210 216 208 204, 199 197 159 180 1895 144 123 128 146 178 189 187 184 182 178 Kfe 166 164 1896 127 134 110 113 114 121 131 144 167 183 177 195 177 199 175 200 174 195 169 171 : 174 196 1.56 1897 188 186 171 1898 138 120 130 1.55 186 194 19(, 195 192 186 183 179 171 1899.. 140 141 117 122 125 128 144 147 183 178 199 187 207 192 204 j 194 202 200 192 190 200 202 186 179 174 1900 173 1901 149 127 132 157 190 207 210 212 201. 200, 19( 17; 180 1902 144 121 124 145 178 192 19U 194 190 18C 182 169 167 1903 . 1904 tl70 1905 t179 *From Rept. of Chief of Engineers, U.S.A., p[). 2874 -'75, 1903. t AiJproxiniately. of Canada] DISCHARGE OF ST. LAWRENCE RIVER 461 TABLE III.-.MEAX ANNUAL DISCHARGES OF ST. LAWRENCE AND ST. MARY RIVERS.* Years. 1860 1861 1862 18fi3 1864 186.J 1866 1867 . 1868. 1869 . 1870 1871. 1872 1873. 1874 . 1875 . 1876 1877. 1878 1879 1880 . 1881 . 1882. 1883 . 1884 1885 . 1886. 1887 1888 1889. 1890 1891. 1892 1893 . 1894 1895 1896. 1897 . 1898. 1899 1900. 1901 . 1902 . 1903 . 1904. 1905. Saint Lawrence. ( Saint Mary's. 271,510 91,429 287,921 92,223 288,625 86,540 277,158 79,934 271.063 73.733 269,369 81,569 250,559 81,294 . 268,441 85,763 2.36,011 81,974 260,929 88,4.34 287,209 82.460 251,596 74,5'.tl 217,423 77,959 244,173 82,462 261,439 82, .397 228,842 85,245 275,449 91,235 244,376 82,768 200,138 73,475 251.966 60,635 242,402 69,637 234,732 78,347 261,593 77,508 264,699 73,252 279,607 69,912 268,382 75,173 287,753 71,539 273,812 70,296 242,381 66,956 247,784 67,532 274,235 65,174 249,782 58,199 231.594 57.427 246,915 63,048 241,539 74,143 203,625 7C.,796 212.293 77,978 217,155 78,879 231,123 72,231 226,661 85,655 227,714 87,211 226,499 87.220 232,100 245,800 262,300 255,300 '^Report of Chief Engineer,?, LT.S.A., pp. 2872-73, and 2877, 1903. 462 FALLS OF NIAGARA [Geol. Surv. APPENDIX VII. The following is a reprint of the original abstracts which formed the careat of the discovery of the diversion of the Huron waters from Lake Erie^ which became shrunken to a pond, and the subsequent overflow of the Huron waters into the Erie drainage. The channel of the buried Laurentian valley between Georgian bay and Lake Ontario, as also other features connected with the lake history, were described for the first time in the abstract reprinted here. ISTOTES ON THE OeIGIN OF THE GrEAT LAKES OF ^ORTH America. by j. \v. spencer. (Eead before the Cleveland meeting of the American Asso- ciation for the Advancement of Science, 1888.) Proceedings of American Association for the Advancement of Science, volume xxxii, pp. 197-199. abstract. Discovery of the ancient course of the Saint Lawrence river. — Previous investigations by the author showed that there was a former river draining the Erie basin and flowing into the extreme western end of Lake Ontario, and thence to the east of Oswego, but no further traceable, as the lake bottom rose to Ihe northeast. Upon the southern side there was a series of escarpments (some now submerged) with vertical cliifs facing the old channel. By recent studies of the elevated beaches it is demonstrated that the disappearance of this valley of the Saint Lawrence was one with that of Lake Ontario. Recent discoveries of a deep channel, upon the northern side of Lake Ontario (a few miles east of Toronto), and of the absence of rocks to a great depth under the drift, far beneath the surface of Lake Huron, between Lake Ontario and the Georgian bay, — and in front of the Kiagara escarpment, between these lakes, — of Canada] PKE-GLACIAL DRAINAGE 463 of a channel in Georgian bay at the foot of the escari> nient, and of the channel across Lake Huron, also at the foot of a high submerged escarpment, show that the ancient Saint Lawrence during a period of high continental elevation rose in Lake Michigan basin, flowed across Lake Huron basin, and down Georgian bay and a channel, now filled with drift, to Lake Ontario ; thence by the present St. Lawrence valley to the sea — receiving on its way the ancient drainage of the Erie basin and other valleys. OEIGIN OF TilE BASINS OF THE GREAT LAKES. The two questions involved are the ' origin of the valleys ' and the ' cause of their being closed into water basins.' The basins of Lakes Ontario and Huron are taken for consideration. The previous paper upon the course of the ancient St. Lawrence shows that the Huron and Ontario basins are sections of the former great St. Lawrence valley, which was bounded, espe- cially upon the southern side, by high precipitous escarpments, some of which are submerged. Upon its northern side there were lesser vertical escarpments, now submerged, with walls facing the old valley. The valley was excavated when the con- tinent was at a high altitude, for the eastern portion stood at least 1,200 feet higher than at present, as shown by the chan- nels in the Lower St. Lawrence, in Hudson straits and off the ISTew York and Chesapeake bays. The valley was obs- tructed in part by drift and in part by a north and northeast- ward differential" elevation of the earth's surface, due to terres- trial movements. Tlie measurable amount of warping defied investigation until recently, but it is now measured by the up- lift of the beaches and sea cliffs. Only one other explanation of the origin of the basins need be considered — that of the • Erosion by Glaciers,' (a) because the lake basins occur in glaciated region; (b) glaciers are considered (by some) to erode; (c) supposed necessity, as the terrestrial warping was not known. In reply : living glaciers abrade but do not erode hard rocks, and both modern and extinct glaciers are known to have flowed over even loose moraines and gravels. Again, even though gla- ciers were capable of great plowing action, they did not affect the lake valleys, as the glaciation of the surface rocks shows the movement "^ to have been at angles (from 15° to 90°) to the trend of the vertical escarpments against which the movement 464 FALLS OF NIAGARA f"^*^^" S"'"^" occurred. Also the vertical faces of the escarpments are not smoothed off as are the faces of the Alpine valleys, down which the glaciers have passed. Lastly, the warping of the earth's surface in the lake region, since the beach episode, after the deposit of the drift proper, is nearly enough to account for all reeky barriers which obstruct the old valley and form lake bapins. ESTABLISHMENT AND D1S]ME:MBERMENT OF LAKE WARREN. This is the first chapter in the history of the Great lakes and is subsequent to the deposit of the upper boulder clay, and therefore the lakes are all very new^ in point of geological time. By the warping movements of the earth's crust, as shown in the beaches — after the deposit of the later boulder clay — the lake region was reduced to sea level and there were no Canadian highlands northward of the Great lakes. During the subsequent elevations of the continent beaches were made around the rising islands. . . With the rising of the continent. Lake (or per- haps Gulf of) Warren — a name given to the sheet of water cov- ering the basin of all the great lakes — was formed. A succes- sion of beaches of this lake have been worked out in Canada, and from Lake Michigan to iSTew York, extending over many hundreds — almost thousands — of miles. Everywhere the dif- ferential uplift has increased from almost zero about the w^est- ern end of the Erie basin, to three, five, and, in the higher beaches, more feet per mile. With the successive elevations of the land this lake became dismembered, as described in succeed- ing papers — and the present lakes had their birth. DISCOVERY OF THE OUTLET OF HURON-MICHIGAN-SUPERIOR LAKE INTO ONTARIO^ BY THE TRENT VALLEY. With the continental elevation described in the last paper — - owing to the land rising more rapidly to the northeast — Lake Warren became dismembered and Huron, Michigan, and Su- perior formed one lake ; the Erie basin was lifted out of the bed of Lake Warren and became drained, and Ontario remained a lake at a lower level The outlet of the upper lake was south- east of Georgian bay by way of the Trent valley into Lake Ontario, at about sixty miles w^est of the present outlet of this lake. The outlet of this upper lake was twenty-six feet deep where it connected with the Trent valley, and tlie channel was from one to two miles wide. This, for a few miles, is cut across of Canada] CHANGE OF HURON OUTT.ET 465 a drift ridg-e to a depth of 500 feet. With the continued con- tinental uplift to the northeast (which has raised the old beach at the outlet, into the Trent valley, about 300 feet above the present surface of Lake Huron), the waters were backed south- ward and overflowed into Erie basin, thus making the Erie out- let of the upper lakes to be of recent date. Tliis is proven by the fact that the beach, which marked the old surface plain of the upper great lake, descends to the present water level at the southern end of Lake Huron. ERIE TJIE YOUNGEST OF ALL TJIE GREAT LAKES. The Erie basin is very shallow, and upon the dismember- ment of Lake Warren was drained by the newly constructed Xiagara river (except perhaps a small lakelet southeast of Long point.) Subsequently the northeastward w\arping (very much less in amount than farther northward at tlie Trent out- let) eventually lifted up a rocky barrier and formed Erie into a lake in recent times, thus making Erie the youngest of all the lakes. The beaches about Cleveland are not those of separated Lake Erie, but belong to the older and original Lake Warren. Note. — To distinguish from the modern, the ancient valley of the St. Lawrence, above described, is named the ' Lauren- tian ' ; the ancient river from the Erie basin, the Erigan ; the Huron-Michigan-Superior lake, the Algonquin, as also the beach which marked its shores, and the river which discharged its waters by the Trent valley. The expanded, but separate, Lake Ontario is named the Iroquois, as also its principal beach, now at 116 feet above its modern surface, at the extreme west- ern end of the lake, while at about 135 miles northeastw^ard (near Trenton) its elevation is 386 feet. 46G FALLS OF NIAGARA LGeol. Surv. APPENDIX VIII. ON THE DISCOVERY OF NL\GARA FALLS AND THE NAME. OjSt the discovery. Cartier. — After the discovery of the northern continent by Cabot in 1497, it is said that French fishing vessels visited the Banks of ISTewfoundland, but concealed the knowledge of their hunting ground, so that not until the voyages of Jacques Cartier, in 1534 and 1535, were the Gulf and River St. Law- rence discovered. On October 3, 1535, Cartier ascended Mount Royal (at Montreal), and from there he saw the impetuous Lachine rapids, with a smoother stretch of the great river of Canada (St. Lawrence) extending far above. He was also shown another river (Ottawa) coming from the west. The guides indicated three rapids above the Lachine, beyond which one could navigate westward for three months Avithout obstruc- tions.* Accordingly Cartier was the first European to dis- cover the beginning of the route to Niagara. Champlaix. — The Lachine rapids were not again visited until Champlain reached them in 1603. These he tried to ascend, but failed. However, he inquired of the Indians about the head of the ' Great River of Canada,' Their accounts are given in ' Histoire de ISTouvelle France ' by Marc Lescarbot, published in 1609, and were taken from Champlain's 'Des Sauvages' (1604) f. These descriptions closely agree, and two of them, taken from an original copy of "fe^'Ntarbot (pp. 381-384), may together be briefly repeated. A third account, by other Indians, gives more details as to the St. Lawrence, differs somewhat in distances, and refers to the Thousand Islands. After passing the Lachine rapids, at Montreal, which they could see, at a distance of two or three leagues, there is a river leading to the Algoumequin country (this reference is to the Ottawa river). Continuing up the river, they pass five rapids in a distance of eight or nine leagues, where each rapid is a * ' Jacques Cartier and His Four Voyages to Canada.' By Hiram B. Stephens. Montreal, p. 66. t See Quebec edition (1870), Vol. II., and "Champlain not Cartier" by Peter A. Porter, Niagara Falls, 1899. of Canada] DISCOVERY AND NAME 467 quarter of a league. The rapids are most difficult to pass. Then the J enter a river which is like a lake, fifteen leagues long according to one, and six or seven as given by the other account. (These are the rapids between the expanded por- tions of the St. Lawrence, forming Lakes St. Louis and St. Francis.) Beyond, they pass five other rapids in a distance of twenty or twenty-five leagues, above which they enter a ' very great lake,' 150 leagues in length according to one, and 300 leagues according to the other. At the end of this great lake (Ontario) are other falls, a league in breadth, descending with a very great current of water into the lake. Passing these falls, and having to carry their canoes, they enter another great lake, where they see no land on either side, as the lake is as large as the first. At the end of this second lake there is a sea, but beyond this second lake they have not been. In this combined narrative is a very clear account of Lake Ontario, where they also mention what is now called the Trent river and Lay of Quinte, down which the Algoumequins came to war with the Iriquois, who descended other rivers on the south side of Lake Ontario. Here is the first reference to Lake Erie, which is clearly set forth, and also to the Niagara river and the great Falls mentioned— with even an estimate of their breadth. But all these distances are largely divisible. The location of Niagara river some forty miles out of position is here only an error of detail. .Such a clear account of neces- sity largely by diagram obtained from the Indians, who spoke a tongue that could not have been very well understood, is most remarkable; and had the book been published at a later date, one might have doubted the antiquity of the narra- tive. This first announcement, to the world of the existence of the celebrated Falls, must be credited to Champlain, al- though he never saw them, even in his later travels. Traders and missionaries were in the region of Niagara as early as 1626 (perhaps Brule in 1611) ; but they have left no account of the Falls. On Champlain's map of 1632 (of which I have seen an original copy) Lake Erie is represented rather as a strait or expanded river, with islands non-existent, or exaggerated, connecting the ' Fresh water sea of the Hurons ' (Lake Huron) with the end of Lake Ontario, where Champlain locates the great falls. In the explanatory note (No. 90) in the volume he says : ' Sault d'eau au bout du Sault Saint-Louis fort haut, ou plusieurs sortes du poissons descendans s'estour- dissent.' Translated — "A verv high fall of water, at the end 30 468 FALLS OF NIAGARA [Geol. Surv. of tlie rapids of St. Louis, where many kind, of fisli in descend- ing, are stunned. (Lac St. Louis is the same as Lake Ontario). Champlain's map is inferior to Indians' account of Lake Erie, given a quarter of a century before, and shows no advance in the knowledge of ISTiagara falls. Lalement and Ragueneatj. — In 16-41, Father Lalement came from the Ste. Marie mission, in the Huron country, seem- ingly across the peninsula of southwestern Ontario, into the country of the ' ^Neutral ISTation,' and thence ' four days going to the entrance of the so celebrated river of that nation, into the Ontario or Lake St. Louys.' ' The stream or river is that through which our Great Lake of the LIurons, a fresh-water sea, empties ; it flows first into the Lake of Erie, or of the l^ation of the Cat, and at the end of the lake, it enters into the territory of the ISTeutral Nation, and takes the name of On- guaahra, until it enters into Ontario.'* In 1648, Father Ragueneau wrote : ' Almost due south from the country of the Neutral IS'ation we find a great lake 200 leagues in circumference called Erie. It is formed by the discharge of the fresh water sea (that is, Lake Huron) and throws itself over a waterfall of dreadful height into a third lake named Ontario. 'f While these accounts more fully describe Lake Erie and correctly locate the N^iagara river, they add no information as to the character of the falls, beyond that narrated by the In- dians to Champlain, although one gives the name of the river and the other mentions the falls. Who was the first European to see the falls will never be known, as he was probably among the coureurs cle hois, or perhaps some missionary, who did not leave a written record. But he or they had made them known as 'so celebrated' when Lalement visited the river, and later, Rague- neau the falls, the names of whom thus become the pioneers them among the visitors to the region. The hostile Iroquois had turned the tide of early exploration from this region, and sent it to Lake Huron and beyond, by another route, before Lake Erie was known except by report. La Salle, accompanied by Dollier de Casson and Rene de Gallinee, was in this region in 1669. Gallinee speaks of l^assing near the mouth of Niagara river and hearing the roar of the falls. :|: On his map of 1670, he says that the falls * Jesuit Relations. 1641, Thwalte's edition, pp. 191, 193. t lb. 1648, (French and English edition, by Reuhen Gold Thwaites, 1988), p. 63. t O. H. Marshall's writings, p. 219, where he cites Galinee's Journal. of Canada] DISCOVERY AND NAME 469 descend, according to the report of the Indians, more than the height of 200 feet.* Another map, unnamed, was made three or four years later, on which N'iagara is described as ' Chute haute de 120 toises par ou le lac Erie tombe dans le lac Fron- tenac ' ( Ontario). f Thus, the estimated breadth and height of Xiagara falls were mentioned before the visit of La Salle and Hennepin in 1678. Hennepin's celebrated visit was soon followed by a glimpse of the cataract by La Houtan in 1688 (who assigned a height of 800 feet to them). Chaklevoix:}: was the first careful observer who visited the falls (in 1721). He estimates the crest line at 400 paces. He says that exactly in the middle, the cataract was divided in two, by a very narrow island half a mile long, which comes to a point, but that the two falls soon reunite. He tried to measure their height, and gives it at 140 feet (French). He said that the one falls had several points which jutted out, but that the other appeared very smootho At that time the cross- fall of LTennepin had disappeared. The narrowness of the end of the islantl was confirmed by Kalni in 1750, and by Pierie's picture in 1768. Cavagnac, the son of the Governor General of Canada, was here in 1722, and gives the height of the falls at 26 fathoms (166 English feet). The fuller account of Kalm (1750) is reprinted in Appen- dix L This short account covers most of our information, regarding the great Falls of JSTiagara, which was handed down during more than a century after they were first mentioned by Cham- plain. ORIGIN OF THE NAME NIAGARA. Lalement tells us that the name given to the river, by the people through whose territory it flowed, was Onguaahra, but he gives no meaning. In the Mohawk language the name Oneagerah-Onyara as given by one, and Oh-nya-ga-ra by * Parkman's La Salle. Notes at end of volume. t Parkman also mentions the second man. t ' Voyage to North America,' by Charlevoix. Dublin Edition, 1766, p. 206. 30i 470 FALLS OF NIAGARA tGeol. Sur7. another,! meaning the ' neck/ in allusion to the river cutting off the Niagara peninsula between the two lakes. The Seneca Indians called them Nya-geah. The Iroquois called it 0-nj-a- ka-ra, and also Oienkwara, which last meant tobacco smoke. On Sanson's map (1656) the name is given as Ongiara. They are also given as Unghiara. Hennepin gives the spelling ISTiagara in the map of 1683, accompanying his volume on Louisiana. In 1686, the form Oneigra was given in a docu- ment. On Coronelli's map of 1688 the word i^iagara is used. In 1701, the Senecas deeded to the English a tract of land, 'including likewise the Great Falls of Okinagaro,' and in the treaty of 1726, the word used is Oniagara. The Indians pronounced the word IvTee-aug-ara. In the latter part of the Eighteenth Century it was given as ISTiagara, and in the early ISTineteenth Century it appeared as Niagara. P.S. — This appendix should have been placed as the first article of Appendix I, but it was only prepared while the work was awaiting the press, and the page references in index could not be disarranged. t See O. H. Marshall, for many forms of the word ; also ' Brief History of Old Fort Niagara,' by Peter A. Porter. Niagara Falls, 1896. INDEX Note. — After primary references to features have been made, subsequent mention of tliem, when introduced in an explanatory manner, is not always indexed. Page. Accession of Hm-on waters to Erie drainage, Date of 370,376 Adams, Alton D 14 Adams Cejitre located on Iroquois beach 281 Age of Niagara Falls, conjectures 371-373 determined from upper cataract 360 during Erie stage 368 during modern stage 368 total 370 Age of Whirlpool 370 Alleghany river tributary to Erie basin 403 Allenburg and Thorold depression 417 Algonquin beach about Lake Nipissing 300 map of 301 rise or tilting of 300, 302 lake 299 barrier to 302 defined 293 American falls, depth of water on 257 profile view of 37 recession of 353 view of 113 volume of 258 'American Falls,' The term 33 Ancient valley at Narrows, Whirlpool rapids 147,152 Andrews, Dr. Edmtmd, on beaches at Chicago 313 Apex at Falls 41, 86, 112 Area of Basin above Upper rapids 265 lakes and their drainage fields 207 outlet of Lake Erie ." 265 recession. See Recession of Niagara Falls. Augmentation of Niagara river 311 Bakewell, E., on name ' Canadian Falls ' 13 recession of 372 Balsam lake outlet 296 Banks. See Original Banks of Niagara river. of lower Niagara river 137,138 view of 205 in Victoria Park, view.. 113 Barclay, Anth., of International Boundary Survey 14 471 472 FAI^LS OF NIAGARA ^*^^°'- S"^^" Page. Barrier to Ontario basin 283,284,395 Erie " 293 Basin above Upper rapids, section of 265 . . affected by Erie level 272 See also First Cascade. Beaches. See Algonquin, Bell, Forest, Iroquois, Nipissing-, &c. of southern end of Lake Huron 303 Beaver dam creek .... 408,418 Beauharnois canal 244 Bell, Dr. Eobert, on Davenport ridge 97,278 Nipissing outlet 296 See Preface. Bel] terrace, description, and showing height of Falls. ..137,201,202 Berryman's sand hill and Tiew^ 127,219 Bird, Wm. A., of International Boundary Survey 16 Birth of Niagara falls, and vievv^ of Roy terrace at 191, 199, 354 Bishop, Irving P., buried valleys beneath Lake Erie 396 Borings at Barrie, (Lake Simcoe)) 398 along Erigan channel on Niagara peninsula 423 in Falls-Chippawa valley 161,166 at Fenwick 414 south of Fonthill (at Quaker church) 414 at Hamilton, in buried Dundas valley 411 Cantilever bridge (Narrow^s of Whirlpool rapids), 59, 147, 149 Eichmond Hill 399 about Saginaw bay 398 in Whirlpool-St. David gorge 132,133 Boat, view of, in Whirlj)ool 67 below Whirlpool (in sounding) 67 Boundary line at brink of Canadian falls 259 off Goat island 16 off Grand island 16 map, 15 ; and opposite 19. at Niagara Falls 13 Bowman ravine 127,134 Breathing well 135 Brewery old, buried channel boundary 129 Brown's nurseries. Forest beach at Fonthill 424 Brock's monument above Eoy terrace 197 Buffalo creek buried valley 396 Buried valleys in Lake basins 391 of Dundas 394 Falls-Chippawa 161 between Georgian bay and Lake Ontario.. .. 398 under Lake Erie 396 of Whirlpool-St. David channel 128,130,131,132 Burwell, Keating and Hawkins, surveys of Niagara Falls.. .. 22 Burlington beach and Heights 280 of Canada] ll