FOR THE INTERNATIONAL GEOLOGICAL CONGRESS 
 
 OUTLINE OF THE EVOLUTION OF THE 
 FALLS OF NIAGARA 
 
 Contrast uith the Falls of the Zamhesi 
 
 By J. W. SPENCER, A. M., Ph.D., LL.D. 
 
 (Author of "The Falls of Niagara; Their Evolution," Etc. ) 
 
 Special Commissioner of the Geological Survey of Canada, undoi 
 
 the Directorates of Doctors R. Bell and A. P. Low, for 
 
 the Scientific Investigation of the Falls of Niagara 
 
 w 
 
 Washington, D. C:. 
 
 Press of Judd &: Detweiler, Iiu . 
 
 1913. 
 

 
 U.S.A. 
 
 2 ^ 
 
 3 
 
 I 
 
 3 V/. SPE^'CE^ 
 
 ©CU-^^ 
 
 04'-^^^ 
 
 ^/(.Z 
 
 ^ 
 

For the International Geological Congress 
 
 OUTLINE OF THE EVOLUTION OF THE FALLS OF NIAGARA: 
 CONTRAST WITH THE FALLS OF ZAMBESI 
 
 By J. W. Spencer, A. M., Ph. D., LL. D. 
 
 (Author of "The Falls of Niagara; Their Evolution." etc.*) 
 
 The magnitude of Niagara Falls has popularly made them "Amer- 
 ica's unique and greatest natural feature." Their evolution is more 
 wonderful. In geological chronology they are most important. Their 
 development doubtless may seem complex to and bewilder the superficial 
 observer, but the student can see that the changing history is due to two 
 or three primary causes. 
 
 The ordinary waterfall only requires a stream to descend somewhat 
 abruptly from a higher to a lower country, having begun its descent at a 
 not distant date in the past. The character of the rooks determines 
 whether the form is a cataract or a series of rapids. Its age should be 
 found by dividing the length of the gorge by the rate of recession of the 
 fall, if such can be determined. These factors are known at Niagara, 
 where, however, the results have been modified by great variations in the 
 volume of water and in the height of the fall. 
 
 The variation of the volume of Niagara Eiver has been due to earth- 
 movements affecting the whole lake region, whereby the discharge of 
 the river was increased. The Falls have also varied in height in both 
 directions. The increasing height was caused by the lowering of the 
 waters of the Ontario basin, due to the withdrawal of the ice-sheet, and 
 the subsequent reduction in height was due to the raising of the lake 
 level, owing to the recent tilting of the region. The complexity lies in the 
 student being compelled to take into consideration many measured phys- 
 ical data which are not generally presented to the geologist. 
 
 The student of physical geography must refer to the author's full 
 works, where the original discoveries of the great changes occurring in 
 Niagara Eiver have been described. Some of the features briefiv men- 
 
 *"The Falls of Niagara; Their Evolution aud varyiug Relations to tlie 
 Great Lakes ; Characteristics of the Power and the Effects of its Diversion." 
 By Joseph William Winthrop Spencer, pp. xxxi -f 490, with many maps and 
 illustrations, 1907, Geological Survey of Canada. 
 
tioned will make the casual visitor better understand the exceptional 
 cliaracter of the Falls of Niagara. 
 
 The crest-line of the double cataract (including (Joat Island) was a 
 mile long before it was shortened by 415 feet (1901-1903) at the Cana- 
 dian end, due to the installation of power plants, leaving the "Canadian 
 Falls," or greater cataract, 2,500 feet broad, and the smaller one — 
 "American Falls" — 1,000 feet wide. The height of the main Falls is 
 160 feet, but this begins at the foot of the upper rapids, which descend 
 55 feet. These rapids, however, have not been made by the river pi-oj)er, 
 but are primarily due to the water flowing down the side of a ])resjlacial 
 valley which trended to the southwest and joined the "Erigan River," 
 or ancient outlet of the Erie Valley, situated 12 to 1-1 miles west of the 
 Falls. The grandeur of the Upper Kapids, as seen from the Canadian 
 Park before the construction of the power-houses, was not inferior to 
 that of the Falls themselves. The mean volume of the water descending 
 the main falls was 194,000 cubic feet per second, with 10,000 cubic feet 
 additional flowing over the American Falls. Thus 95 per cent of all the 
 water passes over the Canadian Falls. 
 
 Only a few centuries have elapsed since the two cataracts were nnitcil 
 below Goat Island, and only a few more will be required before the main 
 falls will have receded above the island, thus causing the diversion of the 
 water from the smaller cataract. 
 
 The water in the cauldron below the great falls has a depth of 72 feet 
 to the loose blocks of limestone, which have fallen, owing to the removal 
 of the soft underlying shales, etcetera. The depth of the water is here 
 about 100 feet; but a quarter of a mile below the apex the deeper inner 
 channel reaches to 192 feet, or to 94 feet below the level of Lake Ontario. 
 For a distance of two miles below there is a navigable stretch of river 
 having approximately the same depth. This is in the wide portion of 
 the gorge, wliich iu some places exceeds its mean breadth of 1,200 to 
 1,300 feet. Immediately beyond, the gorge becomes narrow, shrinking 
 to 700 feet in width. Here the waters dash over the great blocks qf 
 limestone which give rise to the Whirlpool Rapids, with a descent of 52 
 feet, and reach the celebrated Whirlpool. Its measured depth is 126 
 feet; but as the points sounded were not quite in the middle of the outlet, 
 this may be increased to 140 feet, or 94 feet below the level of Lake 
 Ontario, which is the depth of the river above the rapids (see mapV 
 
 The Whirlpool is situated at the head of a deep preglacial valley 
 formed by an ancient small stream flowing northwestward (as di.«<covered 
 by Lyell, who, however, erroneously thought that it was the course of 
 the ancient Xingara River, hi ISSl, Spencer first showed that the en- 
 
tire river is modern). This channel was refilled by the drift of differ- 
 ent Glacial epochs, which has been penetrated to a depth of 2-26 feet 
 without reaching the bottom. This buried gorge has been reopened by 
 the modern river with the formation of the Whirlpool itself. The nar- 
 rows of the Whirlpool Eapids represent the shallow upper extension of 
 the same buried valley, deepened by the modern falls (but also at a time 
 when the volume of the river was temporarily reduced (Taylor)) to 
 conform with the slope of its bed above and below this section. Here 
 the river-made channel has been partly refilled by the masses of rocks 
 which have since fallen into it (from the sides of the gorge), thus pro- 
 ducing the Whirlpool Eapids. 
 
 A quarter of a mile below the Whirlpool the maximum depth of the 
 river is reduced to 100 feet. In the direct line beyond, the whole natural 
 breadth of the gorge is occupied by Foster's Flats, around which the 
 narrow river channel curves at Foster's Eapids. 
 
 Foster's Flats preserve the most wonderful records in the history of 
 Niagara Eiver. Here are terraces, covered with hard rock strata, which 
 once formed the floors of three separate cataracts, thus registering the 
 heights of the extinct series of waterfalls. (The upper two were each 
 about 120 feet high at this point; the lowest had a variable history of 
 having once a great height reduced again.) Xot only this, but the upper 
 two cataracts were here united into one. The lowest floor (composed of 
 hard sandstone resting on soft shale) was broken through at the head of 
 the Flats when the volume of water was increased by the additional dis- 
 charge from Lake Huron and the upper lakes; for until the falls had 
 receded to this point, Xiagara Eiver drained only the waters of Lake 
 Erie. Henceforth Xiagara Eiver was increased to nearly seven times its 
 original volume. Heretofore the drainage of the three uppermost lakes 
 had been to the eastward ; but a differential tilting of the earth's crust 
 had turned the drainage southward into Lake Erie and the Xiagara 
 (3,500 years ago). The lowest fall had been at the foot of the Flats; 
 but later it channeled through the Flats. All the cataracts were then 
 united. 
 
 While the head of Foster's Hats is situated about four miles below 
 the Falls, the distance from this point to the foot of the gorge is scarcelv 
 three miles. In this distance the most important feature is a peculiar 
 lateral chasm (Smeaton Eavine), where no existing stream could have 
 formed it. However, it was due to a small stream flowing from the 
 river itself around a little island, after the falls had receded above this 
 point (see map). 
 
The end of the canon of Xiagara is characterized by a series of ter- 
 races or other deserted shorelines which mark the lowering of the lake 
 waters with the increasing height of the falls. At their birth the falls 
 were only 35 feet high. As they were intennittently increased in height 
 a series of cascades M'as formed, due to the harder layers of rock. After 
 the waters had fallen below the level of the Iroquois Beach, the lowest 
 of the great cataracts came into existence, descending from the Medina 
 sandstone, now forming a splendid terrace at the mouth of the gorge. 
 But the level of Lake Ontario continued to recede until it was twelve 
 miles distant, subsiding nearly 180 feet below the present level (as found 
 by soundings within and without the end of the caiion). This last cata- 
 ract eventually readied a height of over 300 feet ; but the volume of the 
 water was only that from Lake Erie (15 per cent of the present). The 
 strata beyond, all the way to the lake, were soft shales or drift materials 
 and easily removable, not being protected by layers of hard rocks. 
 
 But the same tilting of the earth's crust which turned the waters of 
 the upper lakes into the Xiagara also raised the outlet of Lake Ontario 
 and lowered the height of the lowest cataract by nearly 180 feet, which 
 was further reduced when it had reached Foster's Flats, owing to a 
 slight tilting of the rock strata. 
 
 The nearly horizontal strata are remarkably uniform in their charac- 
 ter. The uppermost beds consist of hard Xiagara limestone resting 
 upon Xiagara shales. Beneath these are the hard bands of Clinton lime- 
 stone, underlaid by variable, easily yielding strata. Xext comes the 
 durable Medina sandstone, with hundreds of feet of shales beneath. 
 
 The buried valleys produced relatively little effect upon the whole 
 recession of the falls, the greatest amount being at the Whirlpool, which 
 was equivalent to the omission of only about 700 feet of solid rock in 
 the direct course of the river. 
 
 Under the laws of erosion the recession in the same kind of strata 
 varies according to the volume of the river and the height of the falls. 
 The Falls of Xiagara are receding at the mean rate of 4.2 feet a year. 
 Taking the variations at each point of the excavation of the canon, it is 
 found that the recession of the falls for the upper four miles has re- 
 quired (approximately only) 3.500 years. The falls were located at the 
 AMiirlpool 3,000 years ago. Before soundings were made under the 
 Canadian Falls, the required time was calculated to have been 5,000 
 years. But the soundings revealed the fact that the falls throughout 
 the .section above the Whirlpool were higher than now. which fact short- 
 ened the first calculation of the time. Indeed, tiic Wiiirlpool Uapids 
 were completed less than 300 years ago. 
 
The uppermost of the cataracts below Foster's Flats presented features 
 so uniform that the mean height could be taken at 105 feet. The volume 
 of water, as stated, was only 15 per cent of that of the whole modern dis- 
 charge. The height of the Falls at their birth (35 feet)' was also con- 
 sidered. The calculations based upon the changing physics give 35,500 
 years as the time required for the excavation of this lower and older sec- 
 tion of the canon. Accordingly, the whole life of the Falls of Xiagara 
 is calculated at approximately 39,000 years. There does not appear to be 
 any feature of great importance affecting the age which has been over- 
 looked. Eecasting the use of the data (determined by measurements) 
 and minor variations would not indicate a greater variation than ten 
 per cent, or 4,000 years, unless the measurements be replaced by specu- 
 lations based upon a priori deductions. If new discoveries should be 
 made, the computations might have to be amended. 
 
 The question of the age of the Falls has always awakened tlie greatest 
 interest. Ellicott (in 1789) gave then 55,410 years. Bakewell (in 
 1829) reduced the estimate to 12,000 years. Lyell (in 1841) raised 
 the conjecture to 35,000 years, which figure became popular. All of 
 these were based upon the length of the gorge divided by a conjectural 
 rate of recession. Pohlman was the first to use a measured rate. Gil- 
 bert (1886) adopted the maximum measured rate of recession, reducing 
 the time thereby to 7,000 years or less. Had any of these writers used 
 the mean rate, the result would have been 9,000 years. Upham and 
 Wright followed with 7,000 years, but did not take into consideration the 
 changing physics. Spencer (in 1894) was the first to apply the varia- 
 tion in volume, height, and measurements of the different sections, and 
 then provisionally computed the age to be 32,000 years. Gilbert after- 
 wards withdrew his previous figures, but has never given others. Taylor 
 (in 1898) gave the age at 50,000 years, but added that it might be as 
 low as the figures of Lyell or Spencer. His figures were partly based 
 upon the changing physics and partly conjectural. Taylor is the only 
 student of Niagara, other than Spencer, who has used in any degi'ee the 
 changing conditions in estimating the age of Xiagara Falls. Finally 
 Spencer's revision (in 1907, in which soundings and borings have been 
 most important) places their age at 39,000 ± 4,000 years, more likelv to 
 be in excess of than below the principal figure. 
 
 The age of the Falls does not g\\e us the date of the close of the Ice 
 Age. From the region to the south of Xiagara the glaciers had with- 
 drawn some thousands of years before the birth of the Falls. The 
 glaciers left the Saint Lawrence Yalley a considerable time after the 
 beginning of the history of the Falls, but no sufficient measurements 
 
have been made for determining the exact date, which, based upon par- 
 tial data, would seem to be more like 25,000 or more years ago than 
 only a few thousand. 
 
 The Falls of Niagara reached their greatest perimeter or breadth about 
 1890. Then commenced a period of low water; but when high water 
 stages returned after 1903, the breadth of Falls had been curtailed bv 
 man, and since then much water has also been diverted for power pur- 
 poses ; so that the present high water has not brought back the former 
 conditions. 
 
 The mean gross horsepower of the Falls is almost 5,000,000, but 
 much must be lost in its application, which would reduce it for mean 
 water to 3,200,000, or for low water to 2,600,000. About 93 per cent 
 of all the water passes down the Canadian side of the International 
 Boundary. When the total amount of water now allowed to be diverted 
 shall have been used, the volume of the discharge over the Falls will be 
 reduced to 68 per cent of the natural amount for mean stages, or to 60 
 per cent for very low water. The effects of diversion may now be re- 
 peatedly seen during low water and prevailing ice conditions in winter, 
 when the American Falls are broken into fragments, and the eastern side 
 of the Canadian Falls runs dry. When the total amount of water now 
 permitted to be used shall be diverted, the crest-line of both falls will 
 have been reduced from 3,950 feet (in 1900) to 2,100 feet; but in the 
 future the Falls are destined to be destroyed by man. 
 
 The investigations of the evolution of the Falls of Niagara have been 
 closely connected with those of the Great Lake history, which should also 
 be consulted. 
 
 The Victoria Falls of the Zambesi (discovered by Livingstone) 
 had been popularly so lauded as to make the great African falls almost 
 outrival those of Niagara. The writer turned to the lengthy papers of 
 Mr. Lamplugh for reliable data, by which to compare the two cataracts. 
 In these were found the description of local features, which are incon- 
 sequential at a distance. The features which are consequential relate to 
 the character of the rocks, their age, the peculiar configuration of the 
 falls and chasm and causes therefor, the size and volume of the falls, 
 etcetera; but concerning them Lamplugh's papers were so deficient and 
 speculative that they throw no new light which had not been shown in 
 the ..hort hut .leap and comprehensive paper of M..|vMeux.* Indeed. 
 
 •••I'h.vsleal IIIs,„rv of Vi,t,.ria Falls." l,v .V. .F. C. .Molynoux. <;e.«. .Unu: 
 London, vol. xxv, |)p. .J()-."»4. ino't. 
 
Lamplugh leaves himself open to the suspicion of mistaking hurried re- 
 vision for research.! But from Molyneux ma}' be gathered the data for 
 comparing the Falls of the Zambesi and N^iagara. 
 
 The Grand Canon of the Zambesi is more than 40 miles long, excavated 
 out of jointed basaltic rocks covered with a superficial formation of sands. 
 The basalts are supposed to be of Tertiary age (Mennell and Molyneux). 
 The breadth of the river above and at the falls is 5,580 feet, or 300 feet 
 wider than the combined falls at Xiagara. In both cases the islands are 
 included. Deducting the breadth of these, the water-line of the Vic- 
 toria is 3,200 feet long, and that of X'iagara was about 4,000 feet. As 
 one looks at the African cataract, its height toward the left side is 256 
 feet, while it increases to 343 feet toward the right side. Xiagara Falls, 
 including the rapids immediately above, are 215 feet high. The Vic- 
 toria Falls cascade over the left wall of the gorge into a chasm only from 
 80 to 240 feet wide. At Xiagara the gorge is 1,200 to 1,500 feet wide. 
 Below the outlet of the Zambesi chasm, which is 400 feet deep, the river 
 turns sharply into a similar chasm a mile long, stretching diagonally 
 across the old course of the river. This zigzag course is repeated with 
 the amplitude of the breadth of the river before it sank into the gorge. 
 
 This peculiar feature is plainly shown on Molyneux's map as being 
 due to the opening of joints in the rock in two sets oppositely diagonal 
 to the main course of the river. These are opened in tlie bed of the 
 
 t Although the gorge of the Zambesi below the Victoria Falls had been 
 discovered by Livingstone so long ago, and had been specifically and appro- 
 priately designated the Grand Canon of the Zambesi l>y Molyneux (there 
 being a smaller one far up the river), yet Mr. Lamplugh offers us a new 
 name — "The Batoka Gorge" — as if remiming constitutes the discovery, the 
 more so as he thus entitles his paper. With such a beginning, one needs not 
 be surprised at hyixjtheses in place of research. He discredits the Tertiary 
 age of the basalts (Mennell and Molyneux). He says that the evidence of 
 their being Mesozolc. as supposed by Passarge. is on a "highly speculative 
 basis," but himself gives no evidence of their age. Yet in his summary Lam- 
 plugh says that they are "probably Mesozoic." Again, in describing the super- 
 ficial sands. Lamplugh adds : "The hypothesis that they are wind-blown under 
 conditions different from those which now prevail agrees best with the gen- 
 eral characteristics," without telling us how the wind acted differently. Such 
 other .speculations resulted from his very brief visit to the Zambesi. \Mnle 
 his map indicates a jointed structure of the rocks, it gives no idea whatever 
 of the manner of development of the Victoria Falls alternately from one side 
 of the chasm to the other in place of at the end of the gorge, which is its 
 most wonderful feature. (Brit. Assoc. Ad. Sc. Rept. for 1905, pp. 292-.304; 
 Quar. Jour. Geol. Soc. Loudon, vol. Ixiii, pp. 162-216 ; Geog. Jour. London, vol. 
 xxxi, pp. 133-152, 287-303, 1908.) 
 
river before it descends into the chasm. In course of time these sub- 
 merged channels become so deep as to take the whole volume of water 
 and thus withdraw it from the recent long crest -line. Xow the stream 
 runs around the newly formed spur, with the water cascading in the new 
 chasm, but down the opposite side from that of the previous one. This 
 alternating process is repeated, thus giving the Falls of the Zambesi 
 their unique character. It would be difficult to conjecture the rate of 
 recession, although obviously faster than if the erosion were at the head 
 of the gorge, as in the case of other cataracts. 
 
 The rainfall in this African country occurs mostly in February and 
 almost entirely within five months, so that the volume of the Zambesi 
 is sometimes very large and at others very small. While the measured 
 mean discharge of the river could not be obtained from the Chartered 
 Company wliich has obtained control of the power, it is ai)parently much 
 less than at Niagara. Under all of these conditions, the Victoria Falls 
 may be considered a grand rival of Niagara and also unique, but not as 
 eclipsing our famous American cataract or giving any measurements of 
 geological time. 
 
 , WlUH'U'i- (f. s. 
 
 Note.- In the in 
 
 vestigial 
 
 ;ion of till' |,li, 
 
 vsi.al i-rol.le 
 
 nis of 
 
 fnndanientiil resfiii 
 
 •clios <tl 
 
 • Messrs. Tlioi 
 
 nas Unssel ;: 
 
 iiul K. 
 
 Coi-ps 1)1' Kii^'inetM-s 
 
 ) are ii 
 
 lulispensahle. 
 
 
 
Ij 
 
 J 
 
LIBRARY OF CONGRESS 
 
 029 708 335 1 
 
 A