\^ X^ ENGINEER DEPARTMENT, UNITED STATES ARMY. 
 
 F 606 
 .U28 
 Copy 2 
 
 A.lSr ESS-A.Y 
 
 COXCERNIXG 
 
 IMPORTANT PHYSICAL FEATURES EXHIBITED IN THE VALLEY 
 OF THE MINNESOTA RIVER, 
 
 UPON THEIR SIGNIFICATION, 
 
 G. K. WARREN, 
 
 lIAJOr. OF ENGIXEEKS AND BVT. MAJ. GEXEEAL, XI. S. A. 
 
 [rAET II OF REI'OliT OX THE MIXXESOTA lUVER, SUBMITTED TO BRIG. GEX. A. A. 
 HUJirilREVS, CHIEF OF EXGIXEERS, OCTOBER 31, 1874.1 
 
 WASHINGTON: 
 
 GOVERNIMENT PRINTING OFFICE 
 
 1874. 
 

WASiimaTON, D. C, December 7, 1874. 
 Gexeral : I liave the honor to request that Part II of mj' recent 
 report ou the Miunesota liiver be printed with the small maps, (each 
 an octavo page,) for special distribution. 
 
 The title of this part is, "An essay concerning important physical 
 features exhibited in the valley of the Minnesota Eiver, and upon their 
 signification." 
 
 The conclusion is that these features are a part of extended physical 
 phenomena on this continent, which are of much jiractical importance 
 to engineers, and possessed of special scientific interest. The object of 
 this distribution is to submit the matter to the consideration of scientific 
 men, so that the truth of the conclusions may be tested. 
 V^ry respectfully, 
 
 G. K. Warren, 
 Major of Engineers^ Bvt. Maj. Genl., U. 8. A. 
 
 Brig. Gen. A. A. HuMrHREYS, 
 
 Chief of Engineers. 
 
 [ludorscments.l 
 
 Office of the Chief op Engineers, 
 
 December 8, 1874. 
 Eespectfully submitted to the honorable the Secretary of War, and 
 recommended that the report and maps be jirinted for distribution as 
 requested. 
 
 A. A. Humphreys, 
 Brig. Genl. and Chief of Engineers. 
 
 Approved : 
 
 By order of the Secretary of War. 
 
 H. T. Crosby, 
 
 Chief Cleric. 
 December 8, 1874. 
 
AN ESSAY CONCERNING IMPORTANT PHYSICAL FEATURES 
 EXHIBITED IN THE VALLEY OF THE MINNESOTA RIYERj 
 AND UPON THEIR SIGNIFICATION. 
 
 BY G. K. WAEREX. 
 
 Introductory — Minnesota Valley formerly the course of a great river — General view of 
 the Mississippi Basiu — Valley of Minnesota formerly drained the basin of Lake Winni- 
 peg — Evidence of former great extension of Lake Winnipeg — Professor Hind's descrip- 
 tion of Winnipeg Basin — Hypothesis to account for the former drainage of the Winniiieg 
 Basin along the valley of the Mississip])i, and for the change to the present outlet 
 by Nelson's River — Hypothesis confirmed, by Lake Michigan and Illinois River ; by 
 Lake Winnebago and Fox River; by Saint Joseph's and Saint Mary's Rivers chang- 
 ing from Wabash River to Maumee River ; by subsidence along the North Atlantic 
 shore ; by effects along the Saint Lawrence and Niagara Falls ; by results on the 
 shores of the Great Lakes ; bj'' the canons of the Rio Colorado of the West and Rio 
 Grande ; by the basin of Great Salt Lake ; by the peninsula of Florida ; by the re- 
 cent extension of the Gulf of Mexico northward in the valley of the Mississippi — 
 Reference to a former continental change of level — Local changes of elevation of 
 earth's surface inadmissible — Great permanence of continental structure since recent 
 Tertiary period — Same structure in Glacial period — Shore-lines of continent indicate 
 northern subsidence and southern elevation — Map restoration of ancient basin of the 
 Mississippi. 
 
 The name first given to the Minnesota River by travelers and explorers 
 was Saint Peter's, believed by Mr. Nicollet to have been 
 derived from a Frenchman by that name, who, before Le iiitroductory. 
 Seuer visited it in 1795, had located himself at its mouth. This 
 name it retained on all the maps and works of subsequent explorers, 
 down to the organization of a territorial government. Mr. Nicollet 
 especially urged the preservation of the name on account of its early 
 adoption and long historical use, and had it been a question simply 
 among scientific men it probably would not have been changed. 
 
 It has, however, been a common occurrence for our settlers, when the 
 country came to be occupied by them, to change names given by early 
 French travelers. Sometimes they were merely grotesque changes in 
 pronunciation and spelling by which the origin and significance of the 
 original name was nearly or quite lost, and sometimes it was to adopt 
 a name more easily spoken or one whose meaning was more appropriate 
 and pleasing to them. 
 
 Not unfrequently the name of the river was changed to that of the 
 native tribe found there, and at other times, as in the present case, the 
 name which these aborigines gave to the stream was adopted. These 
 new names given by the settlers afterward became incorporated in the 
 laws, and thus acquired a precedence over those assigned by the first 
 explorers.* 
 
 * In some noted instances scientific explorers have inexcusably changed the names 
 given by previous explorers and writers of the highest standing. Frdmont changed 
 " James's Peak "to " Pike's Peak," " Lake Bonneville" to " Great Salt Lake," " Ogden'a 
 River " to "Humboldt River," and thus men who had fairly won the houQr of so naming 
 such important places were unfairly deprived of it. 
 
It would be fortunate if the new uaine was always as appropriate as 
 to this ease. The river flowing- through the laud of the Dakotas, whose 
 language abounds iu pleasant sounds, and whose names of natural ob- 
 jects are often expressive of characteristic features, pleasing or other- 
 wise to the senses or to the imagination, was named by them Minnesota. 
 Sota in their language means nearly clear or clouded. The whitish 
 water of the Minnesota makes it appear very distinct from that of the 
 Mississippi where the rivers join, the latter having an amber-tint and 
 appearing quite dark where it is several feet deep. 
 
 The line where the two waters join and mingle is marked by little 
 whirls and eddies, and by ascending and descending currents, imitative 
 of gentle ebullition. 
 
 Here the whitish water rising through the amber-colored has the pleas- 
 ing effect of thin, e^■er-varying clouds or curling smoke. It is altogether 
 probable that this optical effect gave origin to the name Minnesota, 
 Cloudy-water. 
 
 This pleasing effect is only seen during the low-water stages of both 
 rivers, the amber-tint of the Mississippi being derived from the drain- 
 age of forests and lakes coutaining decaying vegetation, and the whitish 
 tinge of the Minnesota probably i'rom minute particles of clay obtained 
 from the Cretaceous or Tertiary deposits along its southern branches. 
 
 The Minnesota Kiver valley is most elevated at Lake Traverse, lying 
 in the valley, latitude 45° 40', longitude 90° 55', about 
 ie^former\y\1ie 3,000 feet above the occan-lcvel. Its general course is 
 course of a great southeast to Maukato, latitude 44° 09', longitude 94° 05', 
 ^^^®'^' where the elevation is above 705 feet above the ocean. 
 
 From Mankatf> downward, the general course is northeast to the junction 
 with the Mississippi, where the elevation is 095 feet above the ocean; 
 latitude 44° 42', longitude 93° 35'. The valley at the upper end is as 
 wide as at its lower end. Intervening parts are sometimes much wider, 
 and rarely narrower. The whole is sunk from 130 feet to 250 feet be- 
 low the general level of the country. 
 
 The distance from Lake Traverse to Mankato, by the valley of the 
 river, is one hundred and seventy-seven miles ; thence to the mouth, 
 seventy-nine miles; total, two hundred and fifty-six miles. 
 
 The valley of the Mississippi below the junction, and of the Minue- 
 sotaabove it, is wide and beautiful, and is continuous in direction and of 
 nearly the same breadth, varying from about one mile to two miles. [See 
 map.] In marked contrast is the valley of the Mississippi above their 
 junction, it beingonly about one-quarter mile wide and nearly at right an- 
 gles with the other, it is a mere gorge, whose bottom is almost completely 
 tilled by the river, and evidently has its origin iu the waterfall now at 
 Saint Anthony. This fall in quite recent times must have been where the 
 river now joins the main valley, and since receded to its present posi- 
 tion, seven miles above the junction. 
 
 The valley of the Minnesota above, and of the Mississippi below, is 
 much wider than the existing streams require. It could not have been 
 formed by the action of existing forces, as the Mississij^pi above has 
 been. It must have been excavated by some force that is no longer in 
 operation, and if this was a river it must have been of much larger size 
 than the Mississippi below the juuctiou with the Minnesota. It will be 
 the object of this essay to show that this force was a river ; that it 
 drained, in times subsequent to the last glacial drift formations, all 
 the Winnipeg I>asin. I shall entleavor to explain the cause of its dis- 
 appearance, and show that this cause has produced the same effects 
 •elsewhere ; and I shall brietly trace certain other effects, which an 
 

 207^. 90 ft 
 
 JJhch —JMUe- 
 TerticaZ Scale 
 
 ^^p f7ton'l7i^ t/it' ^-fi^e/' ycnY^&y wy>^^'^ 7^7^^ J/l7^7^^.^ofa fr7u7 
 777ejzi-?zc7io7i, t/7i^ a7i^er/c7i. <*Zf^ve ^/7^j7/7ie77o7/. 
 
 PETERS. PHOTO-LITHOQR*PHEfl, WASHINGTON, D C, 
 
Jin -U r^tiZe^ 
 
 //•o7// ^nt/'y^v ^}^€<r/c^^ //7ic^ff <:7^7^ec^</?t (//^C</Z.K/^^^Iaco7^i/> 7.:SSnys. 
 
 * PETERS. PHOTO-LITHOGRAPHER. WASHINGTON. D C. 
 
operatiou of nature so extended most probably would produce and has 
 Ijrodnced. 
 
 It is a general truth that the valleys of nearly all our rivers show 
 indications of volumes of water in former times greatly exceeding those 
 now tloAving. I say nearly all, because some of our rivers in southern 
 latitudes and rainy regions are an exception. But generally it may be 
 said that our rivers might have their volumes restored to what they 
 were by a change of meteorological conditions, so as to increase the 
 rain-fall to the required extent without changing the areas of their 
 drainage basin. If this were practicable without invoking other changes 
 on the earth's surface of the greatest magnitude, it is evident that it 
 would affect rivers like the Minnesota and Mississippi above their junc- 
 tion (lying contiguous, and having basins of limited areas) proportion- 
 ately', and the valleys would always be proportionate to the areas of the 
 drainage basins and volumes of discharge. This in reality is so generally 
 the case, that a marked exception like that presented here, (where the 
 much larger stream occupies the much smaller valley,) requires us to 
 admit that if the great valley is a valley formed by a river, it must have 
 been by a river draining a basin of vastly greater area than now drained 
 by the Minnesota. 
 
 The disproportionate size of the Minnesota Valley at the mouth 
 of the river increases as we ascend. For whereas the valley nearly 
 maintains its width on the average all the way to the source, the 
 stream gradually dwindles away to nothing, so that at the upper 
 end of the valley where it is a mile wide, the only water there is, is 
 received from small streams coming through the ravines on each 
 side, which have formed dams by their torrent-like deposits, and the 
 intermediate spaces are occupied by lakes of considerable depth. That 
 these small affluents could not have formed the great valley is ob- 
 vious enough. Their own small ravines show their power to erode, 
 and this is proportioned to their volumes; but, on their waters reach- 
 ing the great valley, their power is lost, and instead of eroding 
 there, they are making deposits and filliug up. This feature of the 
 affluents of the Minnesota Valley I believe I was the first to note and 
 attribute proper importance to. I have in my report of January, 18G7, 
 shown how this effect Avas taking place in the Mississippi, and hy it 
 explained the formation of Lake Pepin and Lake Saint Croix. 
 
 The Qn'Appelle branch of the Assinniboine River in British America, 
 closely resembles the Minnesota Eiver. There are similar lakes along 
 its upper course, described by Prof. H. Youle Hind, in his report of 
 explorations of Assinniboine and Saskatchawan rivers, &c., concerning 
 which I make the following extract in relation to the Qn'Appelle Eiver: 
 
 The narrowest breadth of the bottom of the Qn'Appelle Valley is half a mile; its 
 greatest breadth about one mile and a half. Its shallowest part is 120 feet below the 
 level of the prairie, and its greatest depth is between 350 and 400 feet. * * The high- 
 est part of the bottom of the Qn'Appelle Valley is only 85 feet above the south branch 
 (of the Saskatchawan,) at its summer-level, and from 75 to 78 feet above it during the 
 spring elevation of its waters. This occurs at a point distant IH miles from the junc- 
 tion, (of the two valleys,) where a lake is found which discharges itself both into the Sas- 
 katchawan and Assinniboine. Before connecting with the Assinniboine it falls about 
 "280 feet in 25(i miles, or 1 foot 1 inch per mile. The difference of level between the South 
 Branch on one end and the Assiuniboiue at the other, does not exceed, according to our 
 estimate, 200 feet. 
 
 In its long, deep, and narrow course there are eight lakes, having an aggregate length 
 of tifty-three miles. * * * Numerous soundings of the Qn'Appelle Lakes showed 
 them to hold from 40 to GO feet of water, which depths are maintained with great reg- 
 ularity. 
 
 The construction of a dam 85 feet high and 800 yards long would send the waters of 
 the South Branch down the Qn'Appelle. 
 
In another place Professor Hind says : 
 
 How Avere the deep lakes hollowed ont? Lakes fillin<i the breadth of the valley,, 
 hut during the lajiHO of ages uot having increased its breadth, preserving, too, for 
 many miles such remarkable depths ; and although in some instances far removed from 
 one another, yet maintaining those depths with striking uniformity. What could be 
 the nature of the eroding force which dug out narrow basins 54 to 66 feet deep at the 
 bottom of a A^alley already 300 feet below the slightly undulating prairies, and rarely 
 exceeding one mile in breadth ! * * * They seem to point to the former existence of a 
 much deeper valley now broken into detached lakes by the partial filling up of inter- 
 vening distances. 
 
 The valley of the Minnesota was not formed b}- glacial action. There 
 is nothing in the configuration of the country to have caused glaciers to 
 act along snch a line. Had this valley existed at the time the glaciers 
 covered the country, they would have obliterated it, as we know they 
 have many former river-valleys. Moreover, this valley, at its upper part 
 and all along its course, is through the glacial drift itself, and therefore 
 must have originated since the deposit of that material. A 'priori we 
 may assign to water alone the formation of the valley, for we know of 
 no other suflicient cause. But the evidence that it was done by the 
 action of water is direct, and exists, in deposits and terraces, such 
 as are made by streams; in isolated high islands in the valley, such as 
 are left behind by cataracts; in i^ot-holes and water-marks high above 
 any present level of the stream ; and in the fact that there is nothing in 
 the valley to interrupt the continued flow of water. 
 
 Step by step along the valley the indications of the former great 
 river since the last Glacial period are apparent, and I will here intro- 
 duce a specimen of this evidence, taken from the report of a scientific 
 explorer who makes his statement without reference to any hypothesis. 
 It is from Owen's Geological Eeport. In that Dr. Shumard says, j). 492 : 
 
 Two or three miles below the mouth of this river (Red Wood) is one of the most in- 
 teresting exposures of granite, on the left bank of the river. It rises in irregular 
 smooth knobs to the height of 125 feet. At an elevation of 40 to 50 feet above the 
 present channel an ancient bed of the river is distinctly recognizable. The bottom of 
 the bed, as well as the sides, are worn into polished grooves 5 to 12 inches deep; there 
 are besides smoothly-worn pot-holes, uot only in the former bed, but also at various 
 heights even to the top of the rock, and most of the granite surface is rounded and 
 almost polished, all giving evidence of having been laved for a very long period of time 
 by a swift current, and corroborating the conclusions heretofore drawn from the exist- 
 ence of level terraces of alluvial land far above the present highest-water mark, and 
 1 from the position of strata containing fresh-water shells in elevated positions, that the 
 Saint Peter's once flowed at a higher level, or rather that the land has been elevated at a 
 very recent period. 
 
 There is sufficient evidence in the manner in which this valley has 
 been eroded to show that it was done by the action of a river which 
 formed water-falls at certain points, which gradually receded up stream. 
 An instance of the high islands of rock left in the valley, such as occur 
 where a receding cataract divides into two channels, is noted by Dr. 
 Shumard in Owen's Geological Ileport, and a sketch of it shown on page 
 486. A better re]iresentation of it is gi^■en on sheet No. 13 of my map 
 of the ]\Iinnesota A'allej', on scale two inches to a mile. 
 
 There is a much finer exhibition of this effect in the vicinity of Little 
 Rapids, thirty miles from the mouth. It is shown on sheet No. 18 of 
 my map of the valley. There the falls on the former river must have 
 divided into at least three channels. A map of this locality is given 
 Avith the text. 
 
 We have evidence that since the di^apjiearance of the ancient Minne- 
 sota I\iver, the Mississi})[)i has filled Tip the valley at the mouth not less 
 than sixty feet. There is only very little slope to the present river from 
 
J c cite or 3£il&s 
 
 12 3 
 
 N. METERS. PHOTO-LITHOGHAPHE 
 
9 
 
 the mouth up thirty miles to the Little Rapids. This space in the valley 
 is like a lake that has been nearly tilled with alluvium. At these rapids 
 the hard limestoue rock is from forty to fifty feet above the river-sur- 
 face, aud underneath it is soft sand-rock, easily worn away. The high 
 detached islands in the valley show there must have been a cataract at 
 this point, in the course of the ancient river. The u[)per level was that 
 of the litne-rock and the lower level was that of the ancient bed, which, 
 through the action of the Mississippi, has since filled up as stated fully 
 sixty feet. A sudden descent, then, must have occurred here, as great as 
 one hundred feet. Such was the ppwer at work to form this valley. 
 
 The valleys of the tributaries of the Minnesota are sensibly propor- 
 tioned to the present volumes, with the exception of the Chippeway, aud 
 there is reason to believe it formerly received the drainage of the region 
 of Otter-Tail Lake, which now goes to Lake Winnipeg. 
 
 The topographical and hy drographical features of the Minnesota Valley 
 are shown in detail as far as known on the map, on scale of two inches 
 to the mile, accompanying the report of which this is a part. 
 
 I introduce here in the text a reduction of the portion exhibiting the 
 summit-level between Lakes Big Stone and Traverse. 
 
 I consider it well established, that a river of large volume greatly 
 exceeding the present stream formerly flowed throughout 
 the Minnesota River Valley ; that this river existed since of the Missis- 
 the dei^osit of the glacial drift, as shown by the valley, at ^^pp^ Basin, 
 its upper end being excavated in that material ; and that it did not dis- 
 appear merely from a diminished supply of rain in the basin, now drained, 
 since other adjacent basins are not thus affected. 
 
 I will now endeavor to show what was the source of this former great 
 river, and in order to make this more readily perceived I will present 
 some general observations concerning the present Mississippi Basin.* 
 
 The area drained by the Mississippi, exclusive of the mountain-regions, 
 is formed mainly by two great plains inclining toward each other, and 
 whose intersection forms the lowest longitudinal line of the basin. The 
 Minnesota River occupies this lowest line from Lake Traverse to the Mis- 
 sissippi ; thence southward it is the Mississippi that is so situated. The 
 river- valley along this line is generally, and as a whole, much wider and 
 indicative of greater eroding action than that of any of the valleys of the 
 tributary streams. The slope of this main valley is however generally 
 less than that of the tributaries, and the volume of water flowing in it 
 is, in several places, less above a tributary than in the tributary- itself. 
 With but few exceptions, the valleys of the tributaries are relatively 
 proportional to their present volumes of water, so that we must look 
 beyond the present Mississippi Basin to explain the anomaly presented 
 by its axial valley, when compared with the like things in all the rest 
 of this region. 
 
 If now we extend our consideration to the basin north of the Missis- 
 sippi Basin, we find its lowest line to be the prolongation of vaiiey of the 
 the Minnesota Valley, declining with the gentlest of slopes Minnesota tor- 
 in the opposite direction to the north, the axial valley soon ^e^basin oTiake 
 widening out into a great expanse of smooth, flat land. Winnipeg. 
 This must have been the bed of a former great lake, whose waters flowed 
 down the Minnesota Valley ; and if we follow this ancient bed north- 
 ward, we find it passing, by an almost imperceptible slope, beneath the 
 
 * I shall use the word " basin" to denote the whole area drained by a river, and em- 
 ploy the word " valley" to designate the space between the main bluffs, whether occu- 
 pied by water or by bottom lands or by alluvial terraces. 
 
10 
 
 waters of Lake Winuipe^'. The present level of this lake is determined 
 l)y the ontlet to Hudson's Bay, known as Sea Eiver or Kelson's Kiver. 
 Its elevation above the sea, as estimated by Major Long, and adopted 
 by Professor Hind, is about G'M) feet, but there is considerable limit of 
 error in this determination. The summit-elevation along the lowest line 
 between this lake and the Mississippi is in the Minnesota Valley at Lake 
 Traverse, which is elevated about nine hundred and ninety-five feet above 
 the sea, so that if the lake-surface were to rise three hundred and sixty- 
 five feet, it would again flow to the Mississippi. 
 
 This most probably was the real condition at a time when the outlet 
 by Sea River to Hudson's Bay did not exist. 
 
 Abundant evidence exists as to the former great extension of Lake 
 Winnipeg ; the work of Prof. Hind is full of it. The great 
 mfJ^great^ex/en- lake-bed, which has been left dry, has excited the wonder 
 sion of Lake Win- of all who liavc Seen it, and I i^reseut the following quota- 
 ^^^^^' tions to show how it impressed itself upon different ex- 
 
 plorers as viewed in regions many miles apart. 
 
 The first is that from Long's Expedition,* vol. 2, p. 3, describing the 
 appearance, just after leaving the drift-formation along Lake Traverse: 
 
 The plain upon whicli we were traveling was apparently boundless. It was covered, 
 with a short grass of a pale yellowish-green hue. The eye of the mineralogist could not 
 detect a single stone within a mile's travel, and the few that were detected during the 
 day were rolled and uninteresting. In some places pebbles were as abundant as if we 
 had been traveling ui^on the bed of some former river or lake. The mind endeavors in 
 vain to establish limits to the vast expanse of water, which certainly, at some former 
 day, overflowed the whole of that country. 
 
 Near the junction of the Red Lake River these travelers note, vol. 2, 
 p. 32 : 
 
 This completed a journey of two hundred and fifty-six miles, [that is, since leaving 
 the head of Big Stone Lake.] * * * -phe dull monotony of a journey upon 
 prairie-land never appeared to us so fatiguing. * * * j/,e comdril '"« ^'^'V/ 
 Jiat and remarkably deficient in water. There are no valleys, and but few brooks and 
 even springs. 
 
 I italicise this quotation where it specially indicates lake-bed. The 
 season of the year was the last of July. On reaching Lake Winnipeg 
 it is stated, vol. 2, ]). 70 : 
 
 Among the remarkable features of Red Eiver may be enumerated its total want of 
 islands, excepting near its mouth, and the circumstance that it has no bottom or val- 
 ley, properly speaking. It runs in a mere treucli in the prairie. Toward the mouth 
 of the river the country becomes an imjjeuetrable swamp. 
 
 So far, then, as Long's expedition shows the character of this region, 
 it is just what the bed of a receding lake would be with a stream, "a 
 mere trench" along the lowest line, cut in the surface of a plain which 
 passed insensibly beneath the existing water. 
 
 Major Long in his topographical report, p. 224, vol. 2, says : 
 
 The innuediate valley of Eed River is not bounded by parallel ranges of blutis or 
 banks like that of the Saint Peter and other tributaries of the Mississippi, but expands 
 to a great width. ****#«* 
 
 The fiatness of the surface that almost uuiformly prevails throughout the valley of 
 Red River may be regarded as a defect in its natural character that cannot easily be 
 remedied. 
 
 Mr. Nicollet saw this ancient lake-bed for the first time at a point 
 
 * 8vo. Printed by G. B. ^Vhittaker, Ave-Maria Lane, London, 1828. 
 
11 
 
 about 90 miles from wbere JMajor Long's expedition entered upon it. 
 He says, [Sen. Doe. 237, 2Gth Cong., 2d sess., p. o2 :J 
 
 We reached a spot which my harometer suflicicutly informed me was on the dividing 
 ridge, aud uot far from its eastern verge; so that we had actually traveled over the 
 great swell of the plateau that separates the Upper Shayeu from Devil's Lake aud the 
 Ked Eiver; but we were uot yet iu sight of the great valley of the latter, though every 
 moment expecting to come withiu it. While proceeding ouward we saw Dixou, who 
 always preceded us at a distance to indicate our best route, and with his eagle eye to 
 recounoiter the country before us, suddenly come to a stop, light his pipe, aiul qui- 
 etly sit himself down upon a small knoll with his back turned toward us. This was 
 contrary to his usages, for he had always before, when he suggested a halt for any pur- 
 pose, been iu the habit of turning round to wait our comiug. On this occasion he was 
 immovable, and I could not guess the cause. Wheu we reached him, w^e found him iu 
 the most ecstatic contemplation before the vast and magnificent valley of the Red 
 River, which there displ.ayed itself before us, spreading itself iu an almost insensible 
 slope to the east, to the north, and to the south, aud bounded only by the horizon. 
 
 jMr. Nicollet then begs leave " to introduce a few reflections npou 
 the magical influences of the prairies," aud he concludes by saying, "I 
 pity the man whose soul could remain unmoved." 
 
 The following is from the "Report of a Geological Survey, &c., by 
 David Dale Owen, U. S. Geologist," (p. 175.*) He was then just west 
 of Otter-tail Lake : 
 
 Nothing but personal obserYation cau convey to the mind the singular effect pro- 
 duced by this dead level plain. The line of the horizon is so perfectly straight that it 
 might almost serve the purpose of astronomical observations for determining the alti- 
 tude of the heavenly bodies. While standing ou this great savauna, strainiug the 
 eyes in quest of some object more prominent than a blade of grass, it occurred to me 
 that there is probably no spot ou the globe more suitable than this on which to meas- 
 ure a degree of latitude. 
 
 A few remarks will suffice to record our observations from this point to the settle- 
 inenta of Red River, since these plains extend the whole distance # * » 
 aud give a monotonous sameness to the whole face of the country. The descent of 
 streams watering these jilaius is effected more by an excavation into the alluvial de- 
 posits than by any depressions of the country. 
 
 When near the 49tli parallel, at Pembina, Mr. Owen reports thus : 
 
 After a hot and fatiguing ride over the plains we arrived an hour after sunset at 
 the foot of Pembina Mountain. In the twilight, as we stood, at our encampment on 
 the plains, it looked as if it might be 300 feet or more in height ; but in the morning 
 by broad daylight it seemed less. When I came to measure it I was somewhat sur- 
 prised th.at it did uot exceed 210 feet. I observed on this, as on many other occasions, 
 that a hill rising out of a level plain appears higher than it really is, especially when, 
 as in this case, the trees on its Hanks and summit are of small growth. Pembina 
 Mountain is, iu fact, no mountain at all — uor yet a hill. It is a terrace of table-land, 
 the ancient shore of a great body of water that once filled the whole of the Red River 
 Valley. 
 
 Ou its summit it is quite level and extends so for about iive miles westward to 
 another terrace, the summit of which, I was told, is level with the great buffalo jilains 
 that stretch away toward the Missouri. * * » * « 
 
 Instead of being composed of ledges of rock, as I was led to suppose, it is a mass of 
 incoherent sand, gravel, aud shingle, so entirely destitute of cement, that with the 
 hand alone a hole several feet deep may be excavated in a few minutes. The Pembina 
 River has cut through this material a deep, narrow valley, but little elevated above 
 the adjacent plain. 
 
 The height of this lake-terrace above the ancient bed west of Pembina, 
 as given by Mr. Owen, is 210 feet. The lake-bed here is about 50 feet 
 above Red River, which at this point is about 100 feet above Lake 
 Winnipeg, making this lake-terrace 360 feet above the level of Lake 
 Winnipeg, a height sufficient, as I have shown already, to extend the 
 present Lake Winmj)eg to Lake Traverse. The uncertainty that exists 
 
 *4to. Printed by Lippincott, Grauibo &, Co., Philadelphia, 1852. 
 
12 
 
 as to tlie coirectuess of the elevations, requires me to state that too 
 mncli reliance mast not be placed on them. The hypothesis of changing 
 continental slopes, which appears farther on, does not require that we 
 should fiud an ancient lake shore in the middle or northern part, high 
 enough above the sea at present to be on a level with Lake Traverse. 
 On the contrary, it rather requires that the shoreline that formerly ex- 
 isted should be at a lower level now. 
 
 I Avill introduce here an extract from Professor Hiud's report upon 
 the Winnipeg Basin, but any one interested in the questions of physi- 
 cal geography should consult his report. 
 
 Extracts from the Karrativc of the Canadian Red River Exploring Ex- 
 Xjedition of 1857, and of the Assinnihoine and Saslcatcheican Explorinf/ Expe- 
 dition o/"iS58, by Henry Youle Hind, M. A., F. R. G. *S'., Professor of Chem- 
 istry, (£•<?., tOe. * * In two rohimes. Published by Longman, 
 Green, Longman and Roberts, 18G0. Vol. 2, p. 230 et seq. 
 
 Chap, xxxvi. Surface Geology. 
 
 GEOGKAl'IIICAL BOUNDAlilES. 
 
 The basin of Lake Winnipeg, including under tbis term tlae country drained by the 
 
 rivers flowing into it, extends from the 90th to the 118th meridian, 
 
 serfraticm^of Win- ^^^ i^'ost easterly point being the lake and swamps from which the 
 
 nipe^ Basin. Savaune Eiver takes its rise, in longitude 90*2 14', latitude 48*^ 53', 
 
 and the most westerly limit from wbich it draws contributions is 
 
 probably the Glacier, near Howse Pass, in longitude 117^ 35', latitude 51*^ 52'. 
 
 The southern extension of its boundary is at Lake Traverse, iu Dakota Territory. * 
 
 It stretches north as far as Frog Portage, longitude 103° 30', latitude 55° 26'. 
 
 This basin couse(iuently extends over 28 degrees of longitude and 10 degrees of lat- 
 itude. The elevation of its eastern boundary is 1,485 feet above the ocean, and the 
 heiii'ht of land near the sources of the tributary which rises farthest to the west is 
 6/347 feet above the same level. 
 
 Its northern boundary is separated from the valley of the Missinnippi by a low port- 
 age, over which waters flow during floods, while toward the south, Lake Traverse, 
 "which also sends watbr into the Mississippi during spring freshets, is only 820 feet 
 above the sea.* 
 
 The outlet of Lake Winnipeg is through the contracted and rocky channel of Nelson 
 Kiver, which flows into Hudson's Bay. 
 
 The mean breadth of this great inland basin is about three hundred and eighty English 
 miles, and its mean length nine hundred and twenty miles, hence its area is approxi- 
 mately three hundred and sixty thousand square miles. Its shape assimilates to that 
 of a truncated section of a circle, lying in a position the reverse of that produced by the 
 meridians and parallels distinguishing its boundary. 
 
 The eastern rim of the basin of Lalie Winnipeg is formed by part of the Laurentide 
 range of mountains, which consist of gneiss interstratified in some localities with bands 
 of crystalline limestone, and much embossed by domes of iutrusive granite, syenite, 
 and occasionally with elevations of trap. 
 
 The highest summit of the Laurentide Mountains, as far as known, does not exceed 
 1,950 feet above the sea, or 1,350 feet above Lake Superior. 
 
 Striking off in a southwesterly direction from the Savaune Lake the height of land 
 extends to Lake Traverse, and is composed of drift-hills covering Laurentian, Silurian, 
 and Devonian rocks, with low granite ranges and exposures of trap. From the divid- 
 ing ridge, at the source of the Pigeon Piver, southwesterly to Veiniilion Lake, the 
 Height of Land ridge is called by the Indians Mis-sahe-Wa-chu ; in its continuation to 
 the low water-shed, between Rainy River and the Mississippi, it has the name Ish-ko- 
 )ia-bi Ma-chii. Continued soxithwesterly from the Falls of Pokegama, this ridge would 
 pass south of Leech Lake, and strike the Red River of the North near the Great Bend, 
 at which point Doctor Owen found silurian rocks iu 1848. 
 
 From this point on Red River the rim of the basin assumes a northwesterly direc- 
 tion as far as the elbow of the South Branch of the Saskatchewan, along the denuded 
 
 * NoTK. — This last elevation is too little. Mr. Nicollet's barometrical determinations 
 made it about 975 feet elevation, and this is conlirined by the le%'^els of various surveys 
 made since, although iiaving a probable error of 25 feet.— G. K. W. 
 
fliinks of the Grand Coteau dii Missouri. From tlie elbow it turns southwest as ftir as 
 Chief Mountain Lake, under the ll4th meridian, hxtitude 49^, where it meets with the 
 flanks of the Rocky Mountains, trendinsf first nortli, and then northwesterly. The 
 North Branch and the main Saskatchewan indicate approximately its low northern 
 boundary. 
 
 SURFACE FEATURES. 
 
 Lake Winniiteg, at an altitu de of 028 feet above the sea, occupies the lowest depres- 
 sion of this great central basin, covering, with its associated lakes, Manitoba, Winni- 
 pegosis, Dauphin, and Saint Martin, an area slightly exceeding thirteen thousand 
 square miles. *♦**#** 
 
 The country, possessing a mean elevation of one hundred feet above Lake Winni- 
 peg, is very closely represented by the outline of Pembina Mountain, forming part of 
 the eastern limit of the Cretaceous series. *<«#*# 
 
 The area occupied by this low country, which includes a large part of the valley 
 of Red River, the Assinniboiue, and the main Saskatchewan, may be estimated at 70,000 
 square miles, of which nine-tenths are lake, marsh, or surface rock of Silurian or 
 Devonian age, and generally so thinly covered with soil as to be unfit for cultivation 
 except in small isolated areas. 
 
 Succeeding this low region, there are the narrow terraces of the Pembina Mountain, 
 which rise in abrupt steps, except in the valleys of the Assinniboino, Valley River, Swan 
 River, and Red Deer's River, to the level of a higher plateau, whose eastern limit is 
 formed by the precipitous escarpments of the Riding, Duck, and Porcupine Mountains, 
 with the detached outliers, Turtle, Thunder, and Pasqnia Mountains. This is the great 
 Prairie Plateau. * * *■ * area in the l>asin of Lake Winnipeg is abour 
 1'2U, 000 square miles : it possesses a mean elevation of 1,100 feet above the sea. * * 
 * * * * If * *f 
 
 HEACIIES AND TERRACES. 
 
 In the valley of Lake Winnipeg the first prominent ancient beach is the Big Ridge. 
 Commencing east of Red River, a few miles from Lake Winnipeg, this ridge pursues a 
 southwesterly course until it ai^proaches Red River, within four miles of the Middle 
 Settlement ; here it was ascertained by leveling to be 67^ feet above the prairie. On 
 the opposite side of the river, a beach on Stony Mountain corresponds with the Big 
 Ridge, and three or four miles farther west it is observed marking the limit of a former 
 extension of the valley of Lake Winnipeg. On the east side of Red River the Big 
 Ridge is traced nearly due south from the Middle Settlement, to where it crosses the 
 Roseau, forty-six miles from the mouth of that stream, and on or next the 49th parallel. 
 It is next met with at Pine or Tamarac Creek, in the State of Minnesota, and from this 
 point it may be said to form a continuous and horizontal gravel-road beautifully arched, 
 and about 100 feet broad the whole distance to the shores of Lake Winnipeg, or more 
 than 120 mijes. 
 
 On the west side of Red River, and north of the Assiuniboine, I traced the Big Ridge 
 from a i)oint about three miles west of Stony Mountain to near Prairie Portage. Here 
 it appears to have been removed by the agency of the Prairie Portage River and the 
 waters of the Assinniboiue, which are said to pass from the valley of that river into 
 Lake Manitoba during very high floods. 
 
 Another and higher ridge was observed on White Mud River, about twenty miles 
 west of Lake Manitoba. It resembled in every particular the ridge on the east side 
 of Red River, being about 100 to 120 feet broad, and about 25 feet above the level 
 prairie. It was again noticed in the rear of Manitoba House, where the same char- 
 acteristics were preserved. It probably crosses the Assinniboiue three or four miles 
 west of Prairie Portage, and is perhaps identical with the lowest ridge or step of the 
 Pembina Mountain. 
 
 Ill the rear of Dauphin Lake, the next ridge in the ascending series occurs; it forms 
 an excellent pitching track for Indians on the east bank of Riding Mountain. Proba- 
 blj' these ridges are found close together at the foot of Pembina Mountain, where no 
 less than four distinct steps occur close together, near the sources of Scratching River. 
 The summit of these steps may be the plateau, whose altitude was ascertained by 
 Dr. Owen to be 210 feet above the i^rairie-level, and the first steps may be continuous 
 with the Big Ridge, limiting the level prairies of Red River and the Assinniboiue. 
 
 The pi'airics inclosed by the Big Ridge are everywhere intersected by small subor- 
 dinate ridges, which often die out, and are evidently the remains of shoals formed in 
 the shallow bed of Lake Winnipeg when its waters were limited by the Big Ridge. 
 Many opportunities for observing the present formation of similar shoals occurred iu 
 Lake Manitoba, St. Martin's Lake, Lake Winnipeg, and Dauphin Lake. These, when 
 the lakes become drained, will have the form of ridges in the level country then ex- 
 posed. Indeed, it may be said that the region between Dauphin Mountain and Lake 
 Manitoba, in the direction of Ebb-aud-flow Lake, and south of that body of water, is 
 
14 
 
 but recently drained, or still in process of draining^, being removed from tbe snrface of 
 Ebb-and-liow Lake by a very few feet, and covered with water to a lar<i;e extent in the 
 sprint^. At present it consists of marsh, bojif, and ridj^e, in continued succession. When 
 completely drained the country will resemble the present prairies of the Assinuiboiue, 
 with the o-ontle rich depressions, and the low dry, gravelly ridges. 
 
 The Pembina ]Mountain is par excellence the ancient boajsh in the valley of Lake 
 Winnipeg. Dr. Owen thus describes it as it presents itself a few miles south of the 4Uth 
 parallel. 
 
 I have already quoted tliis from Dr. Owen in a preceding part of the 
 essay, to show that his determination of the elevation of this ancient 
 beach makes it sufficiently high to restore the outlet southward by the 
 Minnesota Valley, so I do not repeat it here. But this testimony of 
 Professorllindtothegenuinenessof this ancient beach, after his thorough 
 explorations in the Winnipeg Basiu, leaves no room to question that the 
 level of Lake Winnipeg was formerly so much above what it now is as 
 to overflow all the intervening country, and create the river that eroded 
 the Minnesota Valley. 
 
 I have ventured to provisionally restore this aucieut lake, on the sup- 
 position that its surface be raised about 351) feet above its present level. 
 It is not here claimed that this would, if accurately done, give the 
 shores of the former lake, for there has probably been much change of 
 the slopes of the surface since the lake had its southern outlet, as tl^ere 
 seems no way to account for its disappearance but by a change of the 
 inclination of the longitudinal valley of the continent. 
 
 The consideration of the probable cause of the former extension and 
 subsequent shrinking of this great lake nest demands attention. < 
 
 Having in the iireceding discussion shown that Lake Winnipeg form- 
 erly was continuous to the Minnesota Valley, with au 
 Hypothesis to outlet along the valley to the Mississip])i, we must at 
 
 3>CC0llIlt lOV til6 ' I i- / 
 
 former drainage ouce infer that the existing outlet by iS^elson's Kiver to 
 Vaifey^S^fof *'*<^ Hudson's Bay did not then exist. We have no exain- 
 thechangetothe pie in nature of any great lake having two outlets at the 
 NeYson*s Kiver.^ same time in operation. It is not readily conceivable how 
 two such could ever have formed. Our general idea is that 
 a lake is a depression of the earth's surface, whicii becomes partly or 
 entirely tilled with water, and in the latter case it flows out at the 
 lowest point in the margin or rim of the basin. Thereafter this out- 
 let prevents further rise of the water, by draining it off from the lake 
 as fast as it enters, and thus it cannot overflow at some other point. 
 Now, inasmuch as the closing of the Nelson's River outlet at the present 
 time would cause Lake Winnipeg to rise till it would run out at the 
 Minnesota Valley, it at first seemed plausible to suppose that if the 
 Glacial period tempered off gradually into the present geological epoch, 
 there might have been a long time when tlie glaciers had still sufficient 
 extension southward to close this outlet to Hudson's Bay. Then, on the 
 further recession of the glaciers northward, the present outlet would be 
 Ijresented and the lake drained off. Although something like this may 
 have occurred, it is, so far as 1 now know, an unsupported hypothesis, 
 and barren of any fruit. It will not aid us in explaining any phenomena 
 presented by other lake-basins and water-courses of North America, nor 
 enable us to predict what probable results we shall find in other regions, 
 and thus intelligently direct further investigations. 
 
 The hypothesis which I have fouud to account satisfactorily for this 
 change of outlet from what it formerly was down the Minnesota and 
 Mississippi to the Gulf of Mexico, to what it now is by way of Nelson's 
 or Sea lliver to Hudson's Bay, is to regard it as a result of a gradual 
 
15 
 
 change of inclination of the surfiice of the low interior portion of the 
 continent, caused by a slow elevation of the southern part and subsi- 
 dence of the northern part, extending through a vast period of time, 
 and probably still going on. The elevating force appears to come from 
 the part of the earth-surface occupied by the Pacific Ocean, and the 
 line of greatest depression is somewhere near Greenland, or between it 
 andthe continent. At some intermediate region there would be no change 
 in the elevation, but the change of slope would be going on in the same 
 direction throughout. With these assumed conditions we can mentally 
 go back in time to a period when all of the Winnipeg Lake Basin — the 
 lake being shallow — was at a higher level than the lowest axial line of 
 the continental basin south of it of which it would form a part. In that 
 case no lake would exist in the basin, it being completely drained south- 
 ward. At such a time the surface of Hudson's Bay and of the Arctic 
 Ocean woiild be much less than now, and there would be a much greater 
 extension northward of the waters of the Gulf of Mexico. With such 
 conditions the cold at the north would be intensified and extend farther 
 south than now, while the northern extension of the Gulf would carry 
 the warm, moist winds of the south farther north than now. This effect 
 may have gone to considerable extent, and would greatly favor the for- 
 mation of glaciers. The former existence of glaciers is therefore favor- 
 able to this hypothesis. 
 
 Let us now consider what would happen along the course of the then 
 great southerly-tiowing river when the northern snbsidence set in. First, 
 there must have been a decreasing river-slope in the northern portions, 
 and a diminishing power to erode. Wherever some formation of hard 
 rock was met, if the erosion failed to keep pace with the decreasing 
 slope, a lake would begin to form above it. This lake would thereafter 
 catch and hold all the hard abrading material formerly washed along 
 by the stream, and thus farther decrease its eroding power. The lake 
 must gradually expand above the barrier, and its limit would be reached 
 when a new outlet formed. In the present case the outlet was on its 
 northern rim. 8uch I take to be the history of Lake Winnipeg. Its 
 southern outlet, the Minnesota, met, in the ledges of granitic formation 
 which extend from Big Stone Lake southward along its valley for 110 
 miles, a material which its eroding power was too small to remove fast 
 enough to prevent the formation and expansion of a lake above it. This 
 growing lake finally found a new outlet by overflowing near Nelson's 
 Elver. The first material of the bed of the new outlet was probably 
 loose drift, so that it was easily removed, and the outlet widened and 
 deepened rapidly. When the hard rocks in the bed of Nelson's Eiver 
 were reached abrasion proceeded slowly, causing, along with the gradu- 
 ally changing slope, a slow farther recession of the southern shores, to 
 be hastened occasionally by more rapid lowering of the level, as some 
 long-resisting barrier was finally removed. The records of this action 
 in producing ancient shores and beaches are shown in the quotations I 
 have made from the writings of others. 
 
 The Nelson's Eiver has every indication of being of recent origin. 
 Professor Hind says of it: "It is characterized by having falls and 
 rapids which effectually oppose communication even by canoes." 
 
 The direction of the tributaries of the Winnipeg Bawin all indicate that 
 they belonged to the same system of rivers as those of the Mississippi 
 Basin. The course of the Eed Eiver as it comes from Otter-Tail Lake, 
 of the Wild Eice Eiver, Shayen Oju Eiver, and other smaller tributa- 
 ries, all point to a southerly outlet, until in their course they reat^h the 
 ancient lake-bed, and then they turn in regular curves through an arc 
 
16 
 
 of more than 9(P and take an opposite direction. This change of coarse 
 is just what we shonhl expect from a change in the outlet of Lake Win - 
 uipeg, and its gradually draining off. Professor Hind's examinations 
 show that the South Branch of the Saskatchewan formerly continued its 
 southerly course through the valley of the Assinniboine Kiver. I think 
 the changed direction resulted from the formation of a lake on its upper 
 course, which opened a new outlet, and drained off under the action of 
 the same cause which changed the course of the drainage of the Winni- 
 peg Basin. The Little Souris is said to have formerly had a southern 
 course through the valley of the Pembina Eiver, and there is evi- 
 dence of the former great lake through whose agency the change was 
 effected. It seems to me probable that the Upper Little Souris formerly 
 connected with the James or Dakota River of Dakota. 
 
 I will note here that, so far as I know, I was the first to point out the 
 effect of lakes originating along a water-course undergoing a diminution 
 of its slope, which, by expanding, overflow and form new outlets, thus 
 reversing the courses of rivers and changing their channels. 
 
 A reference to the map (at the end of this essay) of the ancient valley 
 when Lake Winnipeg had its greatest expansion, and still had its out- 
 let by way of the Mississippi, will show the harmony there is in the 
 directions of all the rivers as members of the Mississippi Basin. 
 
 There is no doubt an immense alluvial deposit in the ancient bed of 
 Lake Winnipeg. Where this is cut through by the Hed River it is 40 
 to 80 feet deep. All the tributaries from the west flowed through the 
 soft cretaceous and tertiary strata, and carried to the lake silt, which, 
 spread out over the bottom, gives it such a wonderfully level appear- 
 ance. These tributaries still carry a great deal of clayey material to 
 the lake, so that it is now, and always has been, as its name signifies in 
 the Ojibbeway tongue, Wi-ni2n — muddy water. 
 
 The following facts and reasons confirm tlie truth of the hypothesis 
 of northern subsidence : 
 
 An exactly similar case to that of the Minnesota Valley and ancient 
 Lake Winnipeg is presented by the valley of the Illinois 
 gan and ilii- River and Lake Michigan. The Illinois Valley is through- 
 nois River. q^^ broad and deep, far beyond that of rivers generally of 
 the same volume. At its upper end its width is great, even where there 
 is no stream, and it widens out suddenly into what was, undoubtedly, 
 the ancient shore of Lake Michigan, whose waters must then have flowed 
 down the valley of the Illinois. So little have the waters receded from 
 this ancient outlet that the city of Chicago has cut down the barrier 
 sufficiently to serve as a canal for navigation, and a drain for its sewer- 
 age, by a direct flow of the water. 
 
 The hopothesis given to explain the shrinking of Lake Winnipeg 
 answers equally well for this case. The much less change in the sur- 
 face-elevation of Lake Michigan, since the southern outlet closed, may 
 be readily accounted for in the greater length of Lake Michigan's uew^ 
 uorthern'^outlet than that of Lake ^Vinnipeg. The hard, resisting gran- 
 ite, which, in the Minnesota Valley, prevented erosion, had its counter- 
 part at the head of the Illinois Valley in the continuous, nearly hori- 
 zoutly, stratified, compact, magnesian limestone. 
 
 Another similar case is presented by Lake Winnebago and the Upper 
 
 Fox River, in Wisconsin. All the drainage of the present 
 
 bago aM Up- Winnebago Basin was formerly southward through the 
 
 per Fox River, valley of the Wisconsin River to the Mississippi. The 
 
 Wolf River, whose course is southwesterly, formerly flowed direct to 
 
17 
 
 the Wisconsin, bnt near its Junction it traversed i;ranitic ridges, (tLie 
 same as those met by the ancient Minnesota ;) its erosions coukl not 
 keep pace with the changing continental slope, and a large lake was 
 formed, much larger than the existing Lake AVinnebago. This over- 
 tiowed, finally, into the depression forming Green Bay, and the lake 
 was thereafter greatly reduced. The Little Upper Fox and the Wolf 
 rivers, and others, turned back in their course on reaching this former 
 lake-bed, ]>resenting the anonnilous appearance which it has on the 
 maps, so like those of the Red River of the North. 
 
 The features of this line between Green Bay and the Mississippi Val- 
 ley early facilitated communication, and excited wonder, as those at 
 the source of the Minnesota did. There are the same kind of lakes on 
 the Upper Fox as on the Minnesota, and the outlet of Lake WinnebngOj 
 like that of Lake Winnipeg, is full of rapids. 
 
 These four streams are all nearly united at Fort Wayne, Ind. By 
 the course of the rivers Saint Joseph and Saint Mary's, 
 (on the map,) we should expect them to How southward st. iiiai^v'^s. Mau'i 
 to the Wabash. An examination which I have made gf^g^^,"^ Wabash 
 shows that they once did so. The Wabash Valley is con- 
 tinued up, so as to unite with the valleys of these streams, and there are 
 deposits in the W^abash Valley which could only have come from these 
 streams. A little way down the Wabash Valley, below where these 
 streams join it, is found horizontally stratified hard magnesian limestone, 
 resisting erosion, the same as that at the ancient Illinois outlet to Lake 
 Michigan. The anomalous present course of the Saint Joseph and Saint 
 Mary's rivers also indicates that they formerly flowed down the Wabash 
 Valley, and the great width of this valley throughout its length, compared 
 with its present volume of water, contrasts strongly with the relation 
 between neighboring streams and valleys. We cannot believe that Lake 
 Erie ever had an outlet down this valley. Its present level is 190 feet beloAv 
 the summit of the valley, and even if Niagara did not exist, an elevation 
 of 50 feet would have flowed it back into Lake Michigan, and thence 
 down the Illinois River. What seems most reasonable, is to consider 
 this change in the direction of the flow of the Saint Joseph's and Saint 
 Mary's rivers to be due to the formation of a lake at a higher level than 
 Lake Erie, (in the same way as we have already noted having taken 
 place on the Fox River of Wisconsin,) which lake formed a new outlet 
 along the present course of the Mauraee, the same as Lake Winnebago 
 has along the lower Fox River, but, unlike that lake, this one has been 
 entirely drained off. I have never had time to examine the course of 
 the Maumee to verify this hypothesis. There should exist an ancient 
 lake shore, and a gorge somewhere along theMaumee, if rock is encoun- 
 tered in its course. It, therefore, furnishes a test by which further ex- 
 amination may verify or reject this partial hypothesis. 
 
 So far the confirmation to my hypothesis has been mainly tested by 
 its explanation of similar phenomena in not very distant ^^ ^^^^.^ 
 regions, and largely dependent upon my own observa- tainid by-effects 
 tions, but the hyj)othesis itself is in no way startling. ^0"°^*^^'^°^^ '^®- 
 Professor Dana, in his Manual of Geolog}', says : 
 
 Altlioujrli the eartli, iu its last stage, has readied a state of comparative stability, 
 eh luges of level iu the laud still take place. The movemeuts are of two kiuds : 
 
 1. Secular, or movemeuts pro^ressiug slowly by the ceutury. 2. Paroxysuial, tak- 
 ing;' placL^ suddeuly, iu couuectiou usually with earthfjuakes. 
 
18 
 
 It is to the Oist of tliose, of course, that my liypolliesis belongs. 
 Professor Dana says : 
 
 The secular movemeuts which huve been observed are coufined to the middle aud 
 hif^her temperate latitudes, and are evidently a continuation of the series which char- 
 acterized the post-Tertiary period. In this and other dynamical changes the post-Ter- 
 tiary aud the age of man have intimate relations. 
 
 As I have stated that the cliauge we are considering is one of the 
 present time, I liave some difficulty iu making quotations from author- 
 ities to sustain that view, for such similar effects belong to an antece- 
 dent time. I therefore take that evidence which appefirs unquestiona- 
 bly to belong to the existing period. Professor Dana's manual says : 
 
 In Greenland a slow subsidence is taking place. For 600 miles from Drisco Bay, 
 near 69^ N., to the Firth of Igalika, 60^ 43', the coast has been sinking 
 
 Subsidence f^y. fo,^^ centuries past. Old buildings and islands have been sub- 
 f^ntic shore. merged, aud the Moravian settlers have had to put down new poles 
 ^ for their bolts, and the old ones stand, as Lyell observes, " as silent 
 
 witnesses of the change." 
 
 On the North American coast south of Greenland, along the coasts of Labrador to 
 New Jersey, it is supposed thatsimilar changes are going on ; though more investigation 
 is required to establish the fact. G. H. Cook concludes, from his observations, that a 
 ■slow subsidence is iu progress along the coast of New Jersey, Long Island, and Martha's 
 'Vineyard. 
 
 I have, myself, examined a considerable portion of the coast betweeu 
 IS'ew Jersey and Maine, and I think that no hypothesis other than a 
 o-radual submergence going on at the present time will account for the 
 .effects observed. 
 
 If there is a subsidence along the Northern Atlantic shores, and along 
 
 the interior axial valley, it is not unreasonable to infer 
 
 -th^^^aSt Law? that all the intermediate northern region partakes in it, 
 
 rence River. and thus it may be that the Saint Lawrence liiver is a 
 
 northern outlet, aud of recent existence as such. 
 
 This river is a vast arm of the sea, from the Gulf of Saint Lawrence, 
 for a long way, such as might result from a partially sunken river-basin. 
 The great depth in the Saguenay Eiver, over that in the Saint Law- 
 rence at their junction, indicates that a corresponding depth which must 
 once have existed in the latter, has been filled up by detritus brought 
 down by the upper Saint Lawrence aud deposited in the lower part, 
 after .the course of the stream became reversed. 
 
 The cataract of Niagara, a northern outlet, is of recent formation, 
 and it should be noticed as a significant fact that all the 
 
 NiagaraFaiis. jjQj,^|jgj,u outlets abound in cataracts and rapids, while the 
 streams flowing south are comparatively free from them, indicating the 
 more recent origin of the former. A few rapids in the course of the main 
 Mississippi A' alley can be shown to have a much more recent origin 
 than the rest of the valley. 
 
 Ao-ain, if we consider the lakes by themselves, we see that if the con- 
 tinental slopes are gradually inclining more and more 
 
 Shores of the toward the northeast, these bodies of Avater must have a 
 great lakes. tendency to move iu that direction, which they would obey 
 until stopped by some rocky barrier ; these southwestern shores would 
 have smooth rounded outlines like land rising from the water, and 
 the opposite ones would be sharp aud ragged iu outline from abrasion 
 of the waves and from the submergencies of land long furrowed by 
 atmospheric iufluences. Lakes Winnipeg, Winnebago, Superior, Huron, 
 Ontario, and Champlaiu, all are bounded on the east aud north by 
 rocky shores where the water is deep, and the reverse condition exists 
 
19 
 
 on tlie opposite shores, which are more or less shoal. It cannot be that this 
 difference is due to the character of the rock in place, for this could not 
 determine the direction of the action to be the same in places so distant 
 and differing" so much from each other. Were the tendency of the chang- 
 ing continental slope to be in the opposite direction from what it is now, 
 there can be little doubt the lakes would be found at the base of the 
 Eocky Mountains as they formerly w^ere in the Tertiary period of geology. 
 
 The evidence of recent elevation of southern areas pari passu with 
 the northern depression comes from parts of the continent 
 far removed from the valleyof the Minnesota where this sou'tiFen^^^con-^ 
 investigation was begun, and as this is not the place to tinentai eieva- 
 
 * tion 
 
 treat of the subject fully, I intend to present only certain 
 general considerations concerning this feature of the subject. 
 
 The region. of vast canons embraced in the drainage-basin of this 
 river exhibits effects which would result from its gradual 
 elevation above the sea by a force which acted witli maxi- o/\\°e West**^^ 
 mum intensity at the part of the basin nearest the sea. 
 The sources of this river originating in the high mountains of the most 
 elevated central region could not have the slopes at these places mate- 
 rially diminished by the changing level of the continent, while there 
 would be an appreciable increase of the slopes in the parts near the 
 ocean, where steep shores became more elevated. The surface strata 
 in the basin of the Colorado were mainly made up of soft creta- 
 ceous and more recent geological formations, into which the streams 
 readily cut their way. Abundance of hard rocks, such as bowlders, con- 
 cretions, &c., fell into these streams, and, by their abrading action in 
 moving along, cut the channel deep, as a saw supplied with sand and 
 water works its way through a block of marble. The clashing sound of 
 these rocks in the bed of these streams has been heard by persons stand- 
 ing on the banks. Thus the abrading action in the Colorado Basin has 
 ke])t pace with the elevating action, and the great depth of these caiions 
 in a manner measures the extent of the elevation. Had such soft rocks 
 existed along the Minnesota, under like conditions, so that the drainage 
 could have always remained southward, we should have had escarp- 
 ments there 500 or GOO feet deep. 
 
 Ko serious water-fall exists along the canons of the Colorado, which 
 is a proof of the powerful and long-continued eroding action, which in 
 some places has sawed into granite rocks. A result similar to this is 
 displayed along the Rio Grande del Norte. 
 
 The peculiarities of the Great Salt Lake Basin admit of an explana- 
 tion on the hypothesis of a recent elevation of the south- e^eat Salt Lake 
 western part of the continent. The ancient beaches or 
 shore-marks of a former extension of this lake by an elevation of the 
 waters above the present level as much as 1,000 feet are seen on the 
 west slope of the Wasatch and east slope of the Eumboldt Mountains. 
 This depth of waters spread the lake over an immense area and far into 
 the present basin of the Columbia River. Before this elevation began 
 we can suppose the ocean to have extended far inland to the south of the 
 Salt Lake Basin, and that the drainageof the latter was toward the Gulf of 
 California. As the elevation proceeded this lake would expand north- 
 ward until the new outlet was formed through the Cascade Range, by 
 way of the valley of the Columbia. A rapid lowering of the surface 
 and contraction of the area would soon follow, to be continued indefi- 
 nitely as the new outlet deepened and extended. At some period, as 
 
20 
 
 this elevation went on, the Sierra Nevada and Coast ranges would reach 
 a height snfficient to arrest the moisture coming from the Pacific Ocean 
 so that the evaporation in the interior basin would exceed the precipi- 
 tation when the lake would begin to dry up, and finally become as it 
 now is. 
 
 A similar result in the basin of the Humboldt Eiver and Mud Lakes 
 can be similarly accounted for. 
 
 I could bring many observations in support of this view, and may do 
 so on some other occasion. 
 There seems to be every evidence of this being very recently elevated, 
 
 but, as it is not in the drift region, I do not know whether 
 ^Peninsula of t|je evidences there can be distinguished from those which 
 
 an elevation preceding that time would show. 
 In my report on the Mississippi River I think I shall establish the fact 
 
 that since the Drift period the Crulf of Mexico has extended 
 sionXtiie^Giaf <is high up the valley at least as the neighborhood of 
 of Mexico. Dubuque, Iowa. This will mainly depend upon the obser- 
 
 vations and measurements, which show an interruption in the general 
 width of the main valley at the rapids near Kock Island, near Keo- 
 kuk, and at the Grand Tower, and at the Grand Chain. At all these 
 places the river now flows through exceptionally narrow and obstructed 
 channels, while a more extensive and unobstructed ancient valley is 
 found alongside of tliem. The reasoning is this, that this uniformly 
 wide, ancient river-valley existed before this extension of the Gulf 
 northward. That when it afterward became submerged these old 
 channels became so filled with silt, that on the region undergoing a 
 more recent elevation the river at these exceptional localities did not 
 regain its ancient valley, but has since excavated a new one, recent 
 in origin, as its incompleteness shows. It wnll require much space to 
 fully elaborate this point, and I reserve it for my report on the Missis- 
 sippi Eiver. The questions involved concern also similar effects on the 
 Ohio and Missouri liivers, for all of which the same explanation applies. 
 
 It will be observed that, in speaking of the effects at the rapids of the 
 Mississippi, &c., I have introduced a change of level pre- 
 a ^rmer^conU- ceding the Continental movement going on now, by which 
 nentai change a part that is now undergoing elevation experienced a 
 ot level. reverse action, and from being above the sea was de- 
 
 pressed beneath it. This last action, too, was post-Glacial, or, at least, 
 was since the most southern glacial drift was formed. I acknowledge 
 two such oscillations as being made out, and have been as careful as 
 possible not to confound their effects, in which they much resemble each 
 other. 
 
 In my discussion I have held that the movement of theearth's sur- 
 face we have been considering is secular, and affects large 
 of^eievaUon^ot areas at the same time, and that local changes could not 
 
 earth's surface be admitted to explain the effects to which attention 
 inadmissible. ii-ir-i,Vii i • i- 
 
 was called. Such local changes only occur m volcanic 
 
 regions. The thickness of the earth's strata at present precludes such 
 
 limited movements elsewhere. 
 
 The tertiary strata, where I examined them in Nebraska, were found to 
 
 have a gentle inclination from the liocky Mountains, ex- 
 
 nence of struct- tending over hundreds of miles, yet everywhere appearing 
 
 ures^nce recent level and undisturbed, except by erosion. Viewing them 
 
 '^ at places where they are margined by the rocks, which 
 
21 
 
 formed the shore of the ancient hilce into whicli tlese tertiary deposits 
 were made, it seemed like viewing that ancient shore itself. 
 
 To show that this permanence of stracture in the earth's crust exists, 
 I will present a few points in regard to it as deduced 
 from phenomena of the Glacial period which have come tinentai^"^truct- 
 under my observation. The line of distinct iiumodilied ure exists i^ the 
 glacial drift, as I have observed it, extends somewhat ^°^^ ^^"° 
 south of Warsaw, on the Mississippi ; thence in an undefined line west- 
 ward to the Missouri ; thence along that river, mostly on the east side of 
 it, as far north as about latitude 46° ; thence westwardly till it joins the 
 moraines made by the ancient ]locky Mountain glaciers, which extend 
 eastwardly from the mountains 50 or 100 miles. 
 
 A large intervening portion of the plains shows no signs of glacial 
 action, and the tertiary and cretaceous fossils almost everywhere lie 
 exposed at the surface. The great glacier, then, must have had its south- 
 western terminus on these plains, very nearly along the line of the Mis- 
 souri, south of the parallel of 46° north. This line is very nearly i^aral- 
 lel with the summer isothermal of the present time. That this was the 
 line of terminus of the glacier seems confirmed by the glacial scratchings 
 in Wisconsin and Minnesota which are to the southwest, and this being 
 the direction of the shortest line to the terminus, would be the line of 
 least resistance to the glacial movement. This condition evidently 
 implies a disposition of the continental masses of mountain and plain, 
 much as now — an arid region on the i)lains which limited the formation 
 of glaciers for want of moisture, which presupposes a mountain-mass to 
 the southwest to intercept it then as now.* 
 
 Since the relative structure of the continent in its arrangement of 
 mountain and plain appears to have been permanent for ^ xj, ■ i- 
 long period, I will introduce the indications afforded by of the continent 
 the ocean shores as to the changing level of the continent, ern^'^^subsfdence 
 
 The existence of fiords and deep, narrow indentations and southern eie- 
 has been noted as belonging to northern latitudes and '^^^^°^- 
 glacial regions. Glaciers must do their work mainly above the sea-level, 
 and it must require subsequent submergence to convert valleys thus 
 eroded into fiords. Fiords may, however, be independent of glacial 
 action, for there can be no doubt that a submergence of the caflou 
 regions of Arizona and Kew Mexico would form fiords out of valleys 
 eroded simply by the atmosphere and by streams. Moreover, the forma- 
 tion of eroded valleys must have been an exceptional form of glacial 
 action. Its most common eiiect is to grind down all but the large masses 
 composed of the hardest rock, filling, and obliterating, all former ine- 
 qualities of the surface. 
 
 The termination of the great glacier on the Atlantic seems to have 
 been marked all alongour shore by an immense terminal moraine, much of 
 which is still visible at Long Island, Block Island, and Martha's Vineyard 
 and Nantucket Islands. The wearing away which these shores are 
 undergoing I can account for only on the supposition of their gradual 
 submergence. AVherever valleys were filled by glacial deposits the sub- 
 sequent action of rains, if above water, would wash them out, and then, 
 on submergence, fiords would appear. 
 
 The material which the waves remove from the shores is not swept 
 out into deep water. The lighter clays float inland, to aid in forming 
 flats. The smaller stones and gravel are slowly moved along the shore, 
 
 * The wonderful labyrinth of erosions forming the Bad Lands of White River appear 
 to have entirely escaped glacial action and submergence by water, and as they exist 
 now are a monument of atmospheric action, including even the Glacial period. 
 
22 
 
 rolling up and dowu with the waves imtil reduced to sand aud deposited 
 in littoral cordons or in slioals like Nantucket and Great South Shoals. 
 The large rocks remain distributed over the bed of the ocean. 
 
 Looking at the map of the North American continent we see on its 
 northern shores the sharp and ragged lines similar to the contours of 
 the contiguous dry land, while along the southern shores we see the smooth . 
 rounded lines such as form the contours of laud beneath the water. 
 
 Note. — What is said here iu this essay might, perhaps, more properly have beeu said 
 in 1868, at which time the substance of it was given iu tlie annual report. But a some- 
 what larger scope was there given to my remarks, and time did not suffice to reduce 
 them to definite order. Even now my time is too much occupied to enable me to prop- 
 erly compare my own observations with otliers. The map of tlie ancient valley, and 
 the diagram giving a profile of the contiuent along the lowest line, from the Gulf oi 
 Mexico to the Arctic Ocean, aud along the lowest line from the head of Chesapeake Bay 
 to San Francisco, were prepared and submitted in l.-<3-^, but now again repeated here, 
 having never been published. 
 
-.C^ 
 
 -^ 
 
 Ci 
 
 
 - \. 
 
 ^1 
 
 GULF OF 
 M E X I C b 
 
 TROPIC OF 
 
 "A 
 
 CAM 
 
 CE^J 
 
 
 N PETERS. PHOTO-UrHOOMPHEfl. 
 
•I 
 

 
 .7fft7iso^c7.JB!i 
 
 
^ T 7Q6 t^ee 9X0 
 
 . IHHii