Geological Survey Professional Paper 296 Geology of Glacier National Park And the Flathead Region, Northwestern Montana

STRUCTURE (continued)

GEOMORPHOLOGY

GENERAL FEATURES OF THE GEOMORPHOLOGY

The deformation that reached a climax in the Lewis overthrust was followed by uplift and accelerated erosion early in Tertiary time. The major structural features trend northwest and had a marked effect on the development of the initial drainage pattern. Erosion by the streams thus started has produced ranges and groups of ridges with trends that are influenced by structure. The steep faults that border several of the ranges made great contributions to outlining the mountain masses. Irrespective of whether these faults are normal or reverse, they constitute large, persistent fractures that, relatively speaking, dropped blocks of the earth's crust along the sites of certain major valleys. The depressions thus formed became sites of deposition for soft rocks of Tertiary and later age, whose relatively easy excavation by streams has tended to perpetuate the topographic forms. Subsidiary fractures of northeasterly trends produced weak zones in hard rocks that also facilitated the work of the streams. The present reticulate drainage pattern is, to a degree, a survival of the pattern that originated along structural lines, modified in various ways by incidents that accompanied the removal by erosion of thousands of feet of rock. On the plains the rocks above the Lewis thrust have been almost completely removed by erosion. The vulnerability to erosion that permitted this is still operative and is resulting in the relatively rapid retreat of the mountain front, with consequent steep slopes and cliffs.

No summit peneplain is thought to exist, but the gradual rise of the mountainous region was checked sufficiently to permit formation of a surface of mature relief, termed the "Blackfoot surface." This surface reached its greatest degree of maturity late in the Tertiary. It was still little modified when the earliest of the glaciers of the Pleistocene epoch formed. After one or more glacial episodes, renewed erosion cut deep valleys into the modified Blackfoot surface. Glaciers of Wisconsin age occupied and reshaped these valleys. Later, revived stream erosion cut gorges ranging in depth from a few score to over 100 feet. Within the last few thousand years new glaciers have appeared in the cirques left by the far larger glaciers of Wisconsin age. During the present century these have declined, but it seems quite possible that fluctuation of the climate will revive them before they are all gone.

EFFECTS OF THE EARLY TERTIARY UPLIFT

The Lewis overthrust left most of the two regions here described at altitudes much above those before the diastrophism. Carving of the present mountain masses began at that time. Erosion was guided by structure, and in a general way this has continued to the present. Most ranges and major components thereof correspond rather closely to structural trends, and their crests coincide reasonably well with anticlines and upthrown fault blocks. The major syncline in Glacier National Park lies in part between the Lewis and Livingstone Ranges, and there are many similar correspondences between structural and topographic depressions. Fault scarps have retreated enough to obscure the character of the fractures, but there has been little of the tendency to that inversion, as a result of which, in many other regions, mountains are underlain by synclines and valleys by anticlines as a result of long-continued erosion. This situation has persisted in spite of the removal from the overthrust block of nearly all the Paleozoic and Mesozoic strata. Because of the large amount of rock that has been eroded away, present topographic features probably should be regarded as resequent rather than consequent, resulting from adjustments to structure at levels far below the original surface.

One probable major modification in the topography results from the retreat of the mountain front, especially in the latitude of the park. In that latitude a block of resistant, but in part extensively fractured, rock originally had been shoved at low angles over far less resistant rocks to a distance far east of the present mountains. As soon as the leading edge of the block was exposed to erosion, it must have been vulnerable to sapping because of the great difference in the character of the rocks above and below the thrust zone. Lateral corrasion by streams would have progressed faster in the yielding Cretaceous rocks beneath the fracture zone than in the hard rocks of the Belt series. Landslides similar to those that now interfere with road maintenance in the vicinity of the thrust trace must have been frequent and large. They would have aided stream erosion in the removal of the overthrust block and, hence, in promoting the retreat of the mountain front. Wherever streams succeeded in cutting through the hard rocks of the overthrust block into the soft, clayey rocks beneath, new areas subject to sapping were exposed, and destruction of the outlying part of the overthrust block was correspondingly hastened.

Some of the largest valleys are related in origin to faults that are, at least in part, later than the Lewis overthrust. These are the valleys cut by the principal forks of the Flathead River. Each of these valleys is, in varying degree, associated with large faults which appear to have moved at times subsequent to the major thrusting. The faults tended to produce grabenlike depressions of northwesterly trend that were filled with sediments, the oldest of which may be of Eocene or Oligocene age. The South Fork of the Flathead and the part of the Flathead River locally called the North Fork still occupy faulted and filled depressions. The same is true of the Middle Fork, but the relation is less obvious because hills of hard rock intervene between the faults and the river channel. The fact that the faulted depressions were filled to depths of thousands of feet by clastic material shows that the faulting disturbed the then-existing drainage. It seems a logical assumption that streams corresponding fairly closely to the three forks of the Flathead River were present after the thrusting but were interfered with sufficiently by the later faulting, so that they could not keep their valleys cleaned out. Perhaps the streams were merely checked enough so that marshes and ponds developed in places along the faulted parts of their valleys. Perhaps at intervals during the Tertiary period and the early part of the Pleistocene epoch, the streams were blocked entirely, and lakes filled their valleys. Such interference with drainage could be brought about in various ways, such as renewal of movement along the faults, landslides, or glacial activity.

Some transverse valleys within mountain masses are believed to have formed along zones of weakness related to faults. Even where faults of marked displacement cannot be detected, shearing parallel to the regional system of transverse faults may have facilitated erosion. Many valleys originated before the post-Belt rocks were removed and may have been controlled by structural features in those rocks not detectable in the resistant rocks that remain.

CONCEPTS AS TO A SUMMIT PENEPLAIN

If a peneplain or similar feature once extended over the summits of the mountains, no evidence of it remains. Such relatively flat erosion remnants as exist are far below the peaks. Both Daly (1912, p. 599-642) and MacKenzie (1922, p. 105-106, 118-119) oppose the concept of a high-level peneplain in the mountains near the international boundary. Even though some of their arguments may be open to doubt, it is abundantly clear that these two independent observers of the mountains just north of Glacier National Park found no topographic features that they regarded as remnants of a peneplain at or near present summits. Much of Daly's discussion is devoted to showing that many factors in the history of a mountain range tend toward accordance of summit levels. In and south of Glacier National Park, however, the peaks have such a wide range in altitude that any surface passed through them would have a relief of the order of fully 3,000 feet. Daly's discussion suggests that the actual relief on the original surface from which the peaks were carved could have been much greater because normal mountain erosion has some tendency to produce equality of summit levels.

W. C. Alden (1932, p. 4-10) has described remnants of an extensive surface of Oligocene age, called the Cypress Plain, east and northeast of Glacier National Park but identified nothing correlative with that plain in the immediate vicinity of the Park. He concluded that there are no known "* * * remnants of land forms of Oligocene time in the mountains of the Glacier National Park region, unless they are the peaks of the mountains themselves, and even these have probably been greatly modified by erosion in subsequent ages."

A. C. Lawson (1925), using data presented by Alden (1932, p. 4-17) as to erosion surfaces in eastern Montana has attacked the problem from the standpoint of the theory of isostasy. Using assumed data for the specific gravity of the rocks involved, he derives figures for conditions in Canada just north of Glacier National Park, though he emphasizes they can be approximations only. His calculations indicate a removal of a prism of rocks 9,075 feet thick from the mountains since Oligocene time, with a resulting uplift of 5,675 feet. This would give a net reduction in altitude of the mountains of 3,400 feet. If anything like 9,000 feet of rock has been removed from the mountains since the Cypress plain was formed, it is obvious that no traces of that surface would be preserved in the present mountains. Alden's estimates are not as large as Lawson's and might permit remnants of a surface correlative with the Cypress plain to lie below the present mountain peaks, but he recognized no such surface.

The various lines of argument outlined above appear to justify the assertion that since the Lewis overthrust and related phenomena disturbed the region, plains topography has at no time succeeded in extending itself over the site of the present mountains. In other words, there is no summit peneplain.

BLACKFOOT EROSION SURFACE

Although the mountains have been actively eroded since near the beginning of the Tertiary period, the process has not been continuous. Flat-topped ridges on the plains just east of the mountains have been interpreted by various observers, notably Willis (1902, p. 310, 336) and Alden (1932, p. 13-17) as remains of old erosion surfaces. Their counterparts must be present in the mountains, although the evidence is not everywhere as clear as could be desired. Willis called the oldest erosion surface he recognized the Blackfoot surface, whereas Alden spoke of the same surface as equivalent to his Flaxville plain or No. 1 bench and proposed to drop the name "Blackfoot surface." As the present report concerns areas in and adjacent to that for which Willis originated that name and areas far from the type locality of the Flaxville plain, it is proposed here to revive the use of the term "Blackfoot surface" for the general vicinity of Glacier National Park. It is not intended, thereby, to question the validity of the use of the term "Flaxville plain" in its type locality, which is in northeastern Montana, or in areas sufficiently near that locality so that the plain can be traced fairly continuously. The erosional remnants close to the mountains that are tentatively postulated by Alden to represent the Flaxville plain are isolated features. Their correlation with the Flaxville plain is based on assumed gradients projected many miles westward from definitely recognizable parts of that plain. Alden recognizes erosional remnants at one or more levels below his No. 1 bench near the mountains, which adds to the uncertainty as to the correlation of any set of benches with the Flaxville plain of eastern Montana. Alden is in a better position than anyone else to make broad correlations of erosion surfaces in the general region, and the term "Blackfoot surface" is revived in the present report solely for the purpose of being cautious, a caution which is shown to some extent by Alden in his latest published paper (1932, p. 15). As the name "Blackfoot" is taken from the name of the Indian Reservation east of the park and the tribe living there, it might be more appropriate to spell it "Blackfeet". This would be in accord with local usage and apparently with the usage favored by F. W. Hodge (1910, p. 570-571); but as usage is not strictly standardized, the change would serve no useful purpose.

Whether one chooses to speak of the Blackfoot surface, the No. 1 bench, or the Flaxville plain, the oldest erosion remnants immediately east of the park described by Willis and by Alden are the same topographic features. Concepts as to the age and origin of the old surface of which these features are modified remnants have varied since Willis proposed the name Blackfoot surface, but his name nevertheless has priority for the region close to the park. Willis (1902, p. 340), in line with his concept of the structural history of the region, thought that the Blackfoot surface was warped and that correlatives of it within the mountains have been elevated out of harmony with the surface in the plains area as a result of the Lewis overthrust. Present concepts as to the regional structure require the feature Willis called the Blackfoot peneplain, or surface, to be younger instead of older than the thrusting, so that his assignment of an early Tertiary age requires revision. Alden's discussion (1932, p. 4-17) indicates an age of Miocene or Pliocene. Apparently the plain had been essentially completed by the end of Miocene time, but streams did not dissect it extensively until much later.

The remnants of the Blackfoot surface and other similar surfaces on the Great Plains are so flat and so nearly at the same altitude that the term "peneplain" has been applied to them. However, the remnants at some distance from the mountains are mantled by gravel interpreted as deposits made by swift streams in the course of eroding the plains. They originated far from the sea and at altitudes thousands of feet above sea level. It would appear that the Blackfoot, Flaxville and similar surfaces were made up mainly of stream valleys that had widened and coalesced as a result of lateral planation by active streams. Wasting away of interstream areas by soil creep and similar processes undoubtedly occurred, but there is no evidence that stream erosion had become so feeble that such wastage processes were dominant. Hence, none of these surfaces seems to fit strictly the concept of a peneplain as advanced by W. M. Davis (1899, p. 207-239; 1902). They are more nearly panplains (Crickmay, 1933, p. 337-347) or, close to the mountains, pediments (Bryan, 1925, p. 93-97). A significant difference in the present connection is that the production of panplains or mountain pediments by active corrasion would require much less time than the wastage that would culminate in a peneplain (Cotton, 1948, p. 273-275). Perhaps even more significant is the fact that panplains and pediments are more local, restrictive features than peneplains, strictly defined. The Blackfoot surface, and associated surfaces, may have been broad expanses in the region of the present Great Plains, but they came into abrupt contact with the ancestral Rocky Mountains. Areas of subdued topography are believed to have anastomosed into the mountains, but there is no evidence that the mountains were obliterated or even extensively subdued as they would have been by peneplanation.

To a degree, the hesitation to apply the term "peneplain" to surfaces in the Glacier National Park and Flathead regions may result from inadequacies in present general concepts as to how peneplains may originate and as to the expectable land forms related to them. Difficulties in accepting conventional views as to peneplains have been pointed out by L. C. King (1953). This is not the place to discuss his paper in detail, but it may be remarked that his proposal to substitute "pediplain" for "peneplain" is of fundamental interest. If regional surfaces of low relief can be thought of as resulting from the coalescence of many pediments, difficulties in comprehending, correlating, and interpreting features currently spoken of as peneplains might be lessened.

Both Willis and Alden regarded the numerous relatively flat-topped ridges just east of the mountains as remnants of the Blackfoot surface (No. 1 bench or Flaxville plain). These include Kennedy Ridge, Swiftcurrent Ridge, St. Mary Ridge, Milk River Ridge, Two Medicine Ridge, and numerous similar ridges farther south in the Flathead region. The ridge tops are mantled by gravel which Willis (1902, p. 310, 315, 328-330) interpreted as debris carried from the mountains by streams. Alden (1912, 1914; Alden and Stebinger, 1914), in part in company with Stebinger and others, has reexamined the high-level benches east of the park mentioned by Willis and accumulated data on similar benches to the north and south. He presents much evidence in support of his contention that the gravel on these benches is glacial drift of early Pleistocene age, rather than the alluvial deposits that Willis envisaged. Nevertheless Alden's most recent publication on the subject (Alden, 1932, p. 14, 15, pl. 1) mentions and maps similar benches a little farther east capped by gravel of nonglacial origin and regarded by him as the probable equivalent of the Flaxville gravel (Miocene or Pliocene). Thus, in spite of Alden's discovery that some of the high-level gravel is of glacial origin, it seems clear that benches whose tops represent the Blackfoot surface and which are in part still mantled by alluvial gravel of suitable age remain on the plains close to the mountain border.

If the Blackfoot surface east of the mountains is the product of lateral stream corrasion coupled with the deposition of alluvial fans, then that surface must have tended to extend itself back into the mountains, and remnants may be expected there. M. P. Billings (1938) has disproved Willis' thesis that such a surface within the mountains would be out of harmony with the part in the plains area. Remnants of the Blackfoot surface are believed to be present within and west of the mountains in and south of Glacier National Park although the evidence in many places is obscure. The topographic feature that most attracts attention in this connection is the top of Flattop Mountain, between the Lewis and Livingstone Ranges. This feature has been vividly described by F. E. Matthes (1904), who. however, did not speculate as to its origin. Its possible relationship with the Blackfoot surface has been pointed out in the papers by Willis and Billings above cited. In the unpublished report on the work of Campbell's parties in Glacier National Park, Alden expresses his opinion that the top of Flattop Mountain, some other relatively flat surfaces within the mountains, and the spurs east of the Flathead River (North Fork) north of West Glacier are all to be correlated in age and origin with the Blackfoot surface (his Flaxville plain). He speaks of them as having been affected by early Pleistocene glaciation although he cites no drift of comparable age remaining on them.

Observations made during the present investigation are in agreement with those of Alden as to preservation of erosion remnants of Blackfoot age in and near the mountains. In the descriptions that follow the data obtained during the fieldwork under Campbell (assembled and interpreted by Alden) are freely drawn upon. The general conclusion is that the highest flat-topped ridges on the plains to the east, the prominent bench-tops in the principal valleys of the Flathead drainage system, and numerous high-level benches and related features within the mountains at high altitudes are all remnants of the Blackfoot surface. That surface was fairly level on the plains and in the largest valleys, but it spread through the mountains as a flood-plain system with many and prominent unreduced elevations between the streamways. That is, the area now occupied by mountains was hilly but contained broad, gently sloping valleys. The first of the three views in figures 30-32 represent a landscape near the crest of the Lewis Range in Blackfoot time. The illustration is based on an oblique aerial photograph by the U. S. Army Air Corps taken from the northeast and showing the head of Red Eagle Creek with Blackfoot Mountain in the background. The third of the views in figures 30-32 represents modern conditions and is essentially a copy of the photograph. For the purpose of bringing out pertinent features, all three views show more of the foreground than was included in the original photograph.

FIGURE 30.—View showing topography after the Blackfoot surface had formed in the area of Glacier National Park. Based on an oblique aerial photograph by the U. S. Army Air Corps; taken from the northeast and showing the crest of the Lewis Range in the vicinity of the head of Red Eagle Creek.

FIGURE 31.—View showing topography in the area of Glacier National Park at about the middle of Pleistocene time, during the stage of accelerated erosion that preceded the main Wisconsin glacial advance. Based on a oblique aerial photograph by the U. S. Army Air Corps; taken from the northeast and showing the crest of the Lewis Range in the vicinity of the head of Red Eagle Creek.

FIGURE 32.—View showing the topography in Glacier National Park at the present time. Based on an oblique aerial photograph by the U. S. Army Air Corps; taken from the northeast and showing the crest of the Lewis Range in the vicinity of the head of Red Eagle Creek.

Probably the summits of Flattop and West Flattop Mountains together represent the largest single intramountain plain of Blackfoot time. It formed and has been preserved in large part because of the aid to erosion afforded by the structure of the underlying rocks. The two mountain tops lie along the trough of the major syncline in the region, which is here exceptionally wide and gently dipping. Presumably the undulating plain preserved on their summits originally merged with the relatively flat areas now known as Granite Park, the bench north of Glacier Wall, the Hanging Gardens, and others. It may well have extended northward over the present Waterton Valley, remnants being preserved in the flat south of Kootenai Peak, the nearly flat crest of Porcupine Ridge, the bench south of Campbell Mountain, and similar features. The tops of Flattop and West Flattop Mountains have benches that may record incomplete attempts at leveling through lateral corrasion by streams that have now vanished. Few and generally feable streams now traverse them. Undrained depressions, in part occupied by ephemeral ponds, are common. Kip Creek, which flows southeast through marshes not recorded on the map and which swings abruptly east down the cliffs on the flank of Flattop Mountain into the canyon of Mineral Creek, is obviously abnormal to the present drainage pattern. Perhaps its ancestor flowed northwest over the Blackfoot surface through the present Kootenai Pass. There are many benches in nearby parts of the Lewis Range such as the one utilized by the trail from Logan Pass to Granite Park; these benches may be remnants of the old Blackfoot surface. Some of these have been modified by glaciation and are now occupied by cirque lakes or even by small glaciers such as Chaney Glacier. Many of the wind gaps that now serve as passes are at altitudes sufficiently comparable to the benches to suggest that they originated in Blackfoot time, or approximately so. Erosion by streams and glaciers has been so active in Pleistocene and later time that many details are obliterated.

Figure 33 is intended to give an idea of the topography in Blackfoot time. It is drawn on the assumption that erosional remnants already cited and others to be mentioned below are parts of the Blackfoot surface, which was lowered and modified by later events. The map shows the principal areas thought to be correlatable with the Blackfoot surface—the contours correspond to altitudes greater than the present altitudes of the erosional remnants. This is to allow for the fact that most or all the remnants have been lowered by weathering, glaciation, and other processes that have occurred since the Blackfoot surface was first formed. The altitudes assigned to the diagrammatic contours are referred to present sea level without regard to any oscillations of the land that may have taken place between Blackfoot time and the present. Figure 33 is intended to represent the surface at the time of its maximum extent. The various erosional episodes from the end of the disturbances related to the Lewis overthrust to the time when gravel of late Miocene age was deposited on the plains to the east contributed to the carving of the Blackfoot surface. According to Alden's concepts of erosional history east of the mountains (Alden, 1932, p. 4-17), that surface may have persisted with only minor modification until the close of the Tertiary period.

FIGURE 33.—Diagrammatic contour map of the Blackfoot surface in the region now occupied by Glacier National Park, with a section diagonally across Glacier National Park illustrating the difference in character between the topography of Blackfoot time and that of the present day. (click on image for a PDF version)

Figure 33 contrasts the topography of the Blackfoot surface with that of the present day (taken from pl. 1). Relative to present sea level, the Blackfoot surface was at greater altitudes than comparable parts of the present surface. It was, however, far less irregularly and sharply dissected.

The long smooth-topped spurs on the southwest side of the Livingstone Range are so different from the rugged mountains immediately east of them as to demand an explanation. They are composed mainly of the sediments of early Tertiary age that fill the ancient fault-outlined valley of the northern part of the Flathead River. The old fill is deeply dissected by streams from the mountains. The resultant spurs are now mantled by unconsolidated debris that has lost its original topographic forms to such an extent that the details of origin are obscure. Data collected by Campbell and his men lead Alden (written communication) to the opinion that, in various localities unconsolidated material includes stream gravel on the spur crests and glacial drift, probably of two or more ages. The topographic form induced Daly (1912, p. 538-584, pl. 2) to suggest that the spurs are moraines, but this idea is entirely at variance with the facts since discovered. Alden's ideas accord with the concept gained during the present study that the spur crests are modified remnants of an erosion surface carved on the old fill, itself somewhat deformed, before the rejuvenation that gave rise to the present incised valleys.

Alden in his unpublished manuscript on the park postulates that certain high-level cirques and bench remnants at the heads of the stream valleys now cut into the old fill east of the Flathead River mark the approximate positions of the upper reaches of these streams before incision. One of the best preserved of the upland benches is at the head of Logging Creek. This bench has about a dozen small rock-bound lakes on it, part of which drain north into Waterton River. It is thus a relatively level area that straddles the continental divide and is out of harmony with the present topography. Any sediments or soil that may have formed on it in a previous erosion cycle have been removed by glacial action, but in other respects the form may be little modified. Alden, largely on the basis of the field notes of E. M. Parks and C. S. Corbett, cites also a remnant of a smooth high-level bench at the head of Quartz Creek. This bench, which is about 1,000 feet above the one at the head of Logging Creek, has 9 small lakes on it and small glaciers above it on the northeast flank of Vulture Peak. The bench is transected by a narrow gulch whose V-shape indicates that it was cut by a stream but whose rock walls have been smoothed and striated by glacial ice. The benches at the heads of Logging and Quartz Creeks and similar benches elsewhere along the crest of the Livingstone Range, like benches and wind gaps farther east, are regarded as erosion remnants broadly correlatable with the Blackfoot surface.

Precise correlation among these widely scattered erosion remnants is obviously impractical, especially, as all have been modified in varying degree by post-Tertiary events—notably glaciation. With due allowance for this, it is believed that the Blackfoot surface on the plains to the east extended headward into the mountains in some such way as is indicated by the diagrammatic contours on figure 33. One of the larger intramountain valleys was at the site of Flattop Mountain. Another broad valley was on the site of the present valley of the Flathead River, at altitudes corresponding to present altitude in the neighborhood of 5,000 feet. From this valley the surface rose within the mountains to above a present altitude of 6,500 feet, locally over 7,000 feet, and sloped off to merge with ridge tops in the plains area east of the Lewis Range at present altitudes of 5,500 to somewhat over 6,000 feet. Hills or small mountains near the crests of the present mountain ranges maintained themselves well above the broader expanses of the Blackfoot surface, which had gradients sufficient to permit transport of gravel (figs. 30-33). The present altitudes of a few upland remnants, such as East Flattop Mountain, are too great to fit the concept of a single erosional surface rising smoothly from the plains into the mountains. The explanations that might be thought of to account for apparent discrepancies of this kind are too numerous to warrant discussion on the basis of the incomplete data at hand.

In much of the Flathead region, the record of events of Blackfoot time is even more obscure than it is farther north. One reason for this is the much greater diversity in rocks and in structure. The part of the valley of the Middle Fork between West Glacier and the mouth of Bear Creek is essentially a continuation of the valley of the main Flathead River. Where the valley of the main Flathead River now swings southwest in a gorge at the north end of the Apgar Mountains, the major topographic depression continues southeast in line with the valley of the main stream farther north and with the valley of the Middle Fork above West Glacier. The part of the depression not now occupied by a master stream is underlain by Tertiary and later deposits like those in the valley of the main river farther north. An observer whose first impression of the country was gained from such vantage points as Garry Lookout or Double Mountain might easily get the idea that the Middle Fork of the Flathead maintained a northwesterly course over McGees Meadow between Howe Ridge and the Apgar Mountains and thence into Canada through the valley of the southeasternward-flowing Flathead River. Such a concept would, of course, now be a mistaken one, but it fits the general topography so well as to suggest strongly that in the geological rather recent past it would have been correct. Perhaps the ancestral Flathead River flowing on the Blackfoot surface had its head in the present valley of Bear Creek and discharged its waters into Canada.

Other features of the present topography indicate drainage changes of the first magnitude, which is in accord with the concept just expressed. The narrow gorge of the Flathead River that passes around the north and west sides of the Apgar Mountains into the open valley to the south is abnormal. Similarly, the Flathead River, west of the region mapped during the present investigation, cuts squarely across rock of the Belt series at the northern end of the Swan Range in the narrow rock-walled gorge called Bad Rock Canyon (pl. 3). There is a striking difference in the gradients of the different forks of the Flathead above and below this place (U. S. Geol. Survey, 1937, 1939, 1947, 1950). In Flathead Valley the river meanders over a nearly level plain. Its rise for the first 21 miles above the north end of the lake is almost imperceptible. Even at Bad Rock Canyon, a little over 45 miles along the river from the lake, the rise is only 135 feet. Above the canyon the gradients of all forks of the Flathead are much steeper. In a little less than 66 miles from the canyon to the international boundary, the main river rises 950 feet. In the 44.5 miles from its mouth to the confluence with Bear Creek, the Middle Fork rises 760 feet, and above Bear Creek the gradient is even steeper. The South Fork rises 1,700 feet in 105 miles. While these gradients are far less than those attained by mountain torrents, they are in strong contrast to the almost stagnant condition of the river for most of the distance from Bad Rock Canyon to Flathead Lake.

Conditions at Bad Rock Canyon have been mapped and described by C. E. Erdmann (1947, p. 124-128, 147, 160). He appreciated fully the complexities of the drainage pattern and offered explanations which differ from those given here principally in that Erdmann did not suggest a possible reversal of flow in the upper valley of the Flathead River and did suggest a series of drainage changes related to recurrent movements along the major faults. If the faults renewed their activity at intervals as he suggests, they would, doubtless, have influenced the drainage. Perhaps the tilted beds near the mouth of Kintla Creek reflect such movements. The earthquakes still felt in northwestern Montana occasionally have been attributed (Pardee, 1926; Scott, 1936) to renewed disturbances along old faults in modern times. If, as Erdmann's data indicate, Bad Rock Canyon contains no deposits of Tertiary age, it was probably cut after the Blackfoot surface was formed. He reports unconsolidated deposits, including material of glacial origin to a depth of several hundred feet below river level in the gorge, which is in accord with the idea that the gorge originated some time ago, quite possibly before the Pleistocene epoch. Anyone interested in possible drainage changes should explore the hills north of Bad Rock Canyon more than has yet been done. These hills have not been mapped geologically or topographically and may contain clues to some of the problems under discussion.

The idea that the Flathead River may once have flowed northwest requires testing in Canada; this has not yet been done so far as the writer knows. The present Flathead River heads a little south of latitude 49°30'. Maps show enough apparent abnormalities in drainage pattern to suggest that in southern Alberta and adjacent parts of British Columbia the erosional history has been fully as complex as it has been south of the international boundary.

Returning to consideration of the Flathead region, it may be said with assurance that the dissected erosion surface marked by spur crests close to the main Flathead River continues to the south in the valley of the Middle Fork. There the surface is carved largely on rocks of the Belt series and has more irregularities than remain on the soft materials of the spurs farther north. Even so, flat spur crests such as those topped by Loneman Mountain, Double Mountain, and others are striking features of the topography. Some of the spur crests exceed 6,500 feet in present altitude, but others, such as the one above Garry Lookout and that north of lower Dickey Creek, are lower. Along the spectacular gorge of the upper Middle Fork above the mouth of Bear Creek, the topography is more irregular and is dominated by the steep valleys of actively eroding streams. Even here, relatively level ridge tops, such as the one on which Spruce Lookout stands, may represent modified fragments of old erosion surfaces. Farther south, parts of the ridges topped by Lodgepole Mountain, Gable Peaks, and Union Peak could be easily traversed by a jeep, if such a vehicle could be gotten up to the crests. Most of the ridge crests in this part of the Flathead region are near or somewhat above an altitude of 7,000 feet.

The valley of the South Fork of the Flathead is more open than that of the Middle Fork. Like the latter it has relatively flat-topped spurs, which are well below the range crests. The altitudes of the more distinct spur tops are somewhat lower than in the vicinity of the Middle Fork. Pioneer Ridge, one of the largest, culminates somewhat above 6,000 feet above sea level. Kah Mountain, 6,485 feet, is a knoll surmounting a gently rolling upland just south of the boundary of the Flathead region. On the opposite side of the South Fork, corresponding ridge crests, such as those above Felix and Silver Basins, reach altitudes near 6,500 feet. Some of the benches near the crest of the Swan Range, such as Jewel Basin and those near Mount Orvis Evans, are probably analogous to the upland benches of the Livingstone Range already referred to. Benches of this kind, together with relatively flat-topped spurs, led Davis (1916, p. 271-276; 1921, p. 80-97) to suppose, on the basis of distant observations, that the Swan and Mission Ranges displayed remnants of a high-level surface, "* * * perhaps deserving to be called a peneplain." They are far below the summits and are here interpreted as corresponding essentially to the Blackfoot surface.

Throughout the valley of the South Fork, the soft Tertiary and later deposits wrap around the spur ends and are plastered against the slopes of the spur ends to altitudes that have not been fixed closely because of poor exposures and dense forest cover. They have been noted at altitudes above 4,500 feet but certainly do not reach the tops of the more prominent ridge crests. It would appear, therefore, that here, as well as along the Middle Fork, these deposits have been more extensively eroded than on the west side of Glacier National Park. If the spur crests there, where they are underlain by the Tertiary deposits, are correctly correlated with the spur crests along the Middle and South Forks as remnants of the same erosional surface, the Tertiary fill, which antedates that surface, must have reached comparable altitudes in all three valleys when originally laid down.

The old fill along the South Fork has in places a strathlike top that could represent a partial erosion surface carved on it at some time after the spur crests were formed. The old, largely abandoned trail that extends along the east side of the valley from near the former Riverside Ranger station to Devil's Corkscrew Creek took full advantage of this strath, and the proposed new road will, in part, do so also.

In the eastern part of the Flathead region, there are relatively flat uplands at various altitudes. Some of these doubtless are remnants of old erosion surfaces, but present data do not permit correlations that seem valid and significant. There is little to tie most of these level patches either to the spurs along the Middle and South Forks or to deposits of Tertiary age. For the present it is sufficient to say that the topography in the mountains of the eastern part of the Flathead region is not incompatible with the idea that the Blackfoot surface was once present there—much as it is supposed to have been farther north.

In summary, an erosion surface that may have been mature but was certainly far from being a peneplain extended throughout the Glacier Park and Flathead regions at some time late in the Tertiary. The higher mountains had so successfully resisted attack that even today correspondence between surface forms and rock structure can be discerned, but the principal stream valleys had been widened sufficiently, so that broad expanses of fairly level land ramified throughout the mountainous area.

Together these are spoken of as the Blackfoot surface. Eastward, this surface joined that represented by the highest of the erosional remnants on the western part of the Great Plains—the No. 1 bench of Alden. The Blackfoot surface must be much younger than the old alluvium that filled valleys in the Flathead drainage basin in the Oligocene epoch and later and was well developed at the time that the earliest of the Pleistocene glaciers occupied the mountains. Much of the Tertiary period was occupied in forming this surface, and conditions did not change enough to seriously deface it until some time in the Pleistocene. Hence the Blackfoot surface is broadly correlatable with the Flaxville surface of Alden. How nearly the two merge and can be considered identical is left for future workers to decide.

POST-BLACKFOOT EROSIONAL FEATURES

Although the Blackfoot surface may have reached its culmination as recently as the early part of the Pleistocene epoch, it has since been attacked so vigorously that in many places its remnants have not been identified with assurance. Only the major features of the somewhat complex post-Blackfoot history are known. Those features related primarily to stream erosion are described here; features of glacial origin being touched upon in later pages.

After the Blackfoot surface was formed, rejuvenation took place, and streams cut deep valleys in the mountains, reducing the nearby plains to a level close to the present one and well below the ridge tops regarded as remnants of the Blackfoot surface. As some of the ridges are capped by early Pleistocene drift, much of the rejuvenation that largely destroyed the Blackfoot surface took place after the first Pleistocene glaciation. The mountain valleys produced by this rejuvenation were modified by and hence are older than the glaciers of the Wisconsin stage. This indicates that much of the downcutting took place near the middle of the Pleistocene.

The larger mountain valleys contain terraces that record pauses in their formation. Most of the higher terraces are poorly preserved, and many are concealed or obscured by forest growth and landslides. Probably there are three and possibly more terraces between the flood plains and the Blackfoot surface. On the plains likewise, there are bevelled ridge tops that correspond to one or more pauses in down-cutting later than the Blackfoot cycle (Alden and Stebinger, 1913, p. 532-542; Alden 1932, p. 14-15).

The fact that the downcutting did not extend to the heads of present streams is shown by the absence of incised gorges in most of the higher remnants of the Blackfoot surface. This is illustrated diagrammatically by figure 31, which shows a gorge in the foreground that dies out before reaching the higher part of the broad valley carved in Blackfoot time. In a few places the gorges of middle Pleistocene age must have reached almost to the range crests, for Alden mentions one such gorge. His unpublished manuscript on Glacier National Park calls attention to a place at the head of Quartz Creek in the northern part of the Livingstone Range. Here a valley whose shape testifies to its fluviatile origin is incised into relatively level terrain interpreted as a remnant of the headwater part of a surface of Blackfoot age. Although surrounded by cirques, the V-shaped stream valley bears no evidence of glaciation other than smoothed and striated surfaces on its rock walls.

GLACIAL FEATURES

Much of the spectacular character (figs. 3, 16, 27, 28, 30-32) of the present topography of the Glacier Park region results from glaciation. The Flathead region (figs. 1, 2, 22, 23) has also been glaciated, but with less spectacular results, which is largely because much of it is lower and some of the rocks are not adapted to the production and preservation of cliffed forms. While the principal glacial features of the two regions have long been known, the summaries in the present report can do no more than point out salient features. Detailed description and interpretation require much more information than is available.

Alden's descriptions prove that glacial deposits of early Pleistocene age are plentiful on ridge tops in a rather small area in the plains east of Glacier National Park that escaped later glaciation. Evidently when glaciers of the Wisconsin stage spread over the landscape, they removed, or at least rendered unidentifiable, the products of the early glaciers. The known remnants of the earlier deposits range up to more than 200 feet in thickness and reach altitudes of 7,000 feet; so the mountain glaciers that dumped these deposits on the plains may have been extensive. Most of the old drift now remaining lies on flat-topped remnants of the Blackfoot surface (the No. 1 bench of Alden), but some appears to be on surfaces which are over 100 feet lower and belong to Alden's No. 2 bench. Glaciation may have begun after erosion had reached the level of No. 2 bench, or there may have been two early glacial episodes, in one of which ice moved over the Blackfoot (No. 1) surface and the second of which did not take place until after that surface had begun to be dissected to depths of one or two hundred feet. In either case, the glaciers presumably extended headward to gathering grounds high on the peaks of that day and covered much of the mountainous region with ice. That region had topography like that shown in figure 30—much less rugged and less deeply incised than it was during Wisconsin time. This, coupled with the fact that all early Pleistocene drift seems to have been removed from the mountains, suggests that the early glaciers were not as thick as the later ones.

Information as to early glaciation west of the Continental Divide is even more scanty than it is to the east. It is a fair assumption that the western valleys contained glaciers similar in magnitude to those to the east. Presumably ice streams from the mountains poured into the valleys of the ancestral Flathead River and its major tributaries.

After the early glaciers retreated vigorous stream erosion, as already noted, incised the mountain valleys to close to their present depths—locally as much as 3,000 feet below the Blackfoot surface. Figure 31 illustrates the character of the topography that resulted close to the range crests at an early stage in the rejuvenation. Before the return of glaciation, the incisions were somewhat deeper than figure 31 indicates. The deepened valleys were filled with glacial ice, which widened and somewhat deepened them. Most of the spectacular topographic features for which Glacier National Park is noted were formed at that time, and the higher parts of the country came to resemble the scenery shown in figure 32. The glaciers poured far out into the lowlands beyond the mountains. The largest glaciers on the eastern slope were about 50 miles long. On the west the mountain glaciers of the Glacier National Park and Flathead regions joined ice streams from the north that extended as far as the present outlet of Flathead Lake, a distance along the course followed of more than 50 miles from the border of the park. All of the mountains except the higher peaks and ridge crests were buried under the ice streams. Probably few peaks rose more than 1,000 feet above the ice.

A general idea of conditions during the Wisconsin stage of glaciation can be obtained from Alden's map (1932, pl. 37), which shows much of the eastern part of the Glacier Park region and the country toward the east past the site of the town of Cut Bank. Glaciers extended into Canada along the valleys of Belly and Waterton Rivers. Another, fed largely from gathering grounds in the headwaters of Kennedy Creek and St. Mary River, reached the edge of the continental ice sheet near the international boundary. South and east of the site of St. Mary, a large area of the plains does not appear to have been covered by ice during Wisconsin time. The glacier that issued from the valley of Cut Bank Creek was able to push only a few miles east of the right of way of the present highway that skirts the mountain front. The glaciers emerging from the valleys of forks of Two Medicine Creek and from valleys farther south coalesced to form a large glacier that stretched more than 35 miles east of the mountain front and almost reached the lobe of the continental glacier that passed east of the site of Cut Bank. According to Alden, that town and much of the country north and west of it are within the limits of a lake that formed west of the continental ice sheet. It is somewhat strange that the glaciers that emerged from the southern tip of the present park and areas farther south were able to maintain themselves farther across the plains than those from the more rugged, intensely glaciated mountains farther north. The reason may be in part that numerous individual mountain glaciers were able to unite on the plains drained by Two Medicine Creek. Also, the glacial divide, according to Alden's unpublished data, was farther southwest in the general vicinity of Marias Pass than the present Continental Divide.

West of the Continental Divide the effects of mountain glaciation of the Wisconsin stage are somewhat less spectacular than they are to the east, where already steep slopes on hard rocks lent themselves to the formation of cliffs. Nevertheless, all of the valleys were filled almost to their tops, and the ice streams in each united to form large glaciers in the valleys of the principal forks of the Flathead River. Most, if not all, of the spurs whose crests mark the approximate position of the old Blackfoot surface were inundated by the ice and cloaked with morainal material. Data gathered by members of Campbell's parties and by Alden in later visits show that the ice covered the Apgar Mountains completely and left striae on their crests. Thus the glaciers must have attained thicknesses in excess of 3,000 feet. The ice streams joined and passed westward through Bad Rock Canyon and other gaps into Flathead Valley (pl. 3). There they joined large glaciers moving down from Canada. Together the ice streams pushed their way southward to the site of Polson. The terminal moraine near that town forms the southern shore of Flathead Lake and marks the most southerly point reached by the ice during the Pleistocene. According to J. T. Pardee (1942) the ice there met and dammed glacial Lake Missoula, a large water body that was fed by melting glaciers and in part contained by walls of glacial ice. Lakebeds supposed to have been deposited in Lake Missoula are plentiful south of Polson and extend somewhat north of that town on both sides of Flathead Lake. According to unpublished maps made in 1911 by Eugene Stebinger, the moraine at Polson is the younger of two terminal moraines deposited during Wisconsin time. The older moraine has its center about at Charlo, 18 miles farther south and is now surrounded by the deposits (largely silt) laid down by glacial Lake Missoula. The glacier that occupied Flathead Valley must have been one of the principal ones in the Rocky Mountains region. The great masses of ice from the vicinity of Glacier National Park were merely tributaries. Probably the largest single contribution to the ice in Flathead Valley came from Canada through the depression that has been termed "the Rocky Mountain Trench." Other substantial contributions came from the west. The glacier in Flathead Valley was in places more than 15 miles wide and presumably over 3,000 feet in maximum thickness; and according to W. M. Davis (1916), it was powerful enough to shear the ends off the spurs of the Swan Range.

In the preceding summary the glaciation of Wisconsin age has been treated essentially as a unit. This is because of lack of information on which to base division. Most of the glacial sculpture manifest in the present mountains was performed during the Wisconsin. Two or more major fluctuations must have occurred, but adequate discussion cannot be made until detailed studies with this objective in mind have been undertaken.

It appears that the glacial activity of the Pleistocene epoch came to an end roughly 10,000 years ago and was followed by a period of increasing warmth in which most of the glaciers in the Western United States, probably including all in the vicinity of Glacier National Park, melted (Matthes, 1942, p. 190-215; Dyson, 1949b, p. 7-11). About 4,000 years ago another climatic change permitted the reappearance of mountain glaciers in many parts of the Western United States. The present glaciers in and near Glacier National Park were born at that time. They have fluctuated in size but have never approached the dimensions of the glaciers of Wisconsin age. Mainly on the basis of comparison with data assembled by F. E. Matthes for other regions, they are thought to have reached their greatest extent during the middle of the nineteenth century. The glaciers in the park were certainly larger at the beginning of the 20th century (Dyson, 1949b, p. 7-16) than they were 48 years later. However, about 1950 such extensive snow banks began to be recorded during the winter that there is hope that the glaciers may have a period of renewed growth. Matthes (1942) states that retreat of the mountain glaciers at the rates observed during much of the first half of the 20th century might well mean their obliteration within a few decades. He emphasizes, however, that in Europe and elsewhere, where glaciers have been under observation for much of historic time, fluctuations are the rule. Whether or not it has already begun, resurgence of the glaciers in and near Glacier National Park is to be expected before the present ones have disappeared.

RECENT EROSIONAL FEATURES

When the glaciers of Wisconsin time retreated, they left the bottoms of the mountain valleys broadened and mantled by glacial debris. Streams, which were at first fed by melt water from the glaciers, reworked the debris and produced flood plains. Since then in nearly every valley the flood plain has been cut into by a narrow gorge one to several score feet in depth that contains the present stream. These inner gorges, rarely more than 75 feet deep except near and along the major branches of the Flathead River, commonly penetrate short distances into bedrock. They are too narrow to show distinctly in the contours of the topographic maps. In a few places such as upper Nyack Creek, the rock revealed in the gorges has been shown on the geologic maps (with slight exaggeration in width) (pls. 1, 2). Attention was first called to inner gorges of this character by Deiss (1943b, p. 1164), who worked south of the Flathead region. They are not uniformly developed, in part because of the difficulty experienced by some streams in cutting through thick and irregularly distributed glacial debris. Along some stretches of such tributaries as MacDonald, Nyack, and Ole Creeks, the inner gorges are quite large enough to attract the attention of travelers. They are not conspicuous in many parts of the Flathead region, but are, nevertheless, present there. Among the streams in that region that display them may be mentioned Graves and Sullivan Creeks in the southwest part of the Nyack quadrangle and Badger Creek and its tributaries in the eastern part of the Marias Pass quadrangle.

The inner gorges along the main branches of the Flathead River are over 100 feet deep and have in places been obscured as a result of the meandering of these powerful streams. Parts of the gorge walls have been cut back and even obliterated in places. Local flood plains have formed close to the modern rivers and within the inner gorges. Where these are broad enough, the river has meandered. The detailed topographic maps of the South Fork illustrate these features (U. S. Geol. Survey 1937, 1939, 1947, 1950). They also show the inner gorges along the lower reaches of some tributaries, although not as effectively as if the contouring had been carried a little farther above present channels. Similar features are shown by the detailed maps of the main river and of the Middle Fork. For several miles above West Glacier (Belton) the meanders of the Middle Fork are deeply incised in a rock gorge and thereby differ sharply from the wanderings of the river over modern flood plains farther upstream and from most meanders of the other branches of the Flathead River. Presumably the incised meanders originated on a former flood plain at an earlier date, perhaps when the stream flow was checked during one of the readjustments in drainage commented on above.

The present drainage pattern presents features of some interest. As already noted, it is in many aspects inherited from the pattern whose development began as a result of uplift after the deformation near the beginning of the Tertiary period. As no general peneplain has been produced and no regional mantle of deposits, either sedimentary or volcanic, has been laid down since that deformation, the concept of superposition is not applicable to any large part of the drainage pattern. Deiss (1943b, p. 1162-1163) has offered suggestions of that kind to explain features in the mountains south of the eastern part of the Flathead region. He calls attention to streams that flow toward the Great Plains directly across the major structure. His explanation is that these streams originated on an erosion surface that had attained maturity after the major deformation and which was subjected to broad upwarp near the end of the Tertiary period. He assumes that northeastward-flowing streams were established on a plain on the site of the present mountains, maintained themselves during uplift, and were accelerated when the mountain area rose above the level of the present plains to the east. If the Blackfoot surface, or a surface closely allied thereto, was very extensively developed in the area described by Deiss, it is possible that some streams that originated on the more thoroughly beveled parts of that surface might have survived in the manner he postulates. Both in his area and in the eastern part of the Flathead region, some of the major streams find their way to the plains through gorges that cut directly through large ridges in a manner that seems to call for some such explanation as the one he offers. However, that explanation, if valid at all, can be applied only to limited areas. Bear Creek, Ole Creek, and others that head close to the border of the plains but nevertheless flow southwest transverse to the regional structure do not fit Deiss' hypothesis. Ole Creek in part coincides with a transverse fault, and some of the others may have been excavated along similar lines of weakness. Transverse streams that flow either northeast or southwest are so numerous and demonstrated transverse faults are so rare that excavation along lines of weakness is not entirely satisfactory as a general explanation—especially as the faults are old and recent revivals of them are unproved. Probably the present drainage system has multiple origin. Many of the streams have survived, with modifications from those of early Tertiary age, related in one way or another to the structure of the rocks; some may have originated on the Blackfoot surface; many of the more abnormal-appearing ones may result from the linking together of parts of streams of diverse origin through piracy.

In a region of long-continued uplift, piracy is likely to be widespread. Probable examples of it are numerous, and the process is still going on. For example, Badger Creek and its forks, with numerous right-angled turns and an exit to the plains through an impressive gorge, seems clearly to be made up of parts of several streams. How the stream managed to cross the steep mountain front in the northeast corner of T. 29 N., R. 11 W., is not known. Some factor, possibly difficulty in traversing the glacial deposits so liberally dumped on the nearby plains, checked the flow of the creek in fairly recent time. Before it reestablished flow northeastward, the stream deposited a broad expanse of alluvium just inside of the mountain border. This alluvium is still dotted with marshes and small ponds, resulting from imperfect drainage. The interference with the development of Badger Creek thus recorded may explain why its tributaries have failed to extend themselves beyond Muskrat and Badger Passes in competition with the tributaries of the Middle Fork of the Flathead. The long distance the latter has to wander through the mountains would seem to put it at a disadvantage. Even so, its headwaters have penetrated country underlain by hard limestone, and it seems likely that the stream will soon acquire the upper tributaries of the South Fork of Badger Creek. The somewhat greater precipitation on the southwest side of the Continental Divide is a contributing factor. There is also a contest between Badger Creek and the South Fork of Two Medicine Creek. The undisturbed glacial deposits on the divide between them suggest that neither has made much headway against the other since the mountain glaciers retreated. The North Fork of Badger Creek is in an area of highly disturbed rocks. It may have been aided in excavating its canyon by a transverse zone of fracture. Many other oddities in the drainage pattern of the Flathead region suggest readjustments in the streams during the Quaternary period. Among these is Lodgepole Mountain, an island of hard rock surrounded by soft deposits.

The Glacier National Park region contains similar features. The transverse streams flowing southwest have made greater inroads on the mountain block than their counterparts with northeasterly flow—possibly because of the difference in precipitation on the opposite sides of the Continental Divide. Their headwaters are now spreading up longitudinal valleys. If this tendency persists, it seems possible that they may become so linked as to split the part of the Lewis Range in the southern part of the park by means of a series of longitudinal valleys, permitting the Livingstone Range to extend to the valley of Bear Creek. If so, the Livingstone Range would continue to be cut by McDonald Creek but would be a more impressive topographic feature than it now is.