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

STRUCTURE (continued)

INTERPRETATION OF THE STRUCTURE

GENERAL STRUCTURAL FEATURES

In the foregoing, available data in regard to the structural features present are summarized. Much remains unknown or obscure, but enough facts have accumulated to form a basis for discussion and to cast serious doubt on some of the concepts suggested by previous workers, who were under the disadvantage of having only the results of preliminary reconnaissances to work with. The concepts advanced below constitute attempts at interpretation of the observed structural features. Speculations as to the fundamental reasons for the formation of the structural features involve consideration of the origin and constitution of the earth and are beyond the scope of this report.

The accumulation of great thicknesses of sedimentary rocks in the two regions described and the broad undulations of the crust that have interrupted that accumulation from time to time set the stage for the events that gave rise to the structural features observable in the present landscape. More than half a billion years were occupied in preparation for the single period of drastic deformation thus recorded. Readjustments in the rocks deformed at that time are still in progress. In one sense the period of sharp deformation corresponds to what has commonly been called the Laramide revolution. Terms of this kind result from our desire to classify and correlate imperfectly understood parts of gradational processes not strictly subject to such treatment. Thus, the expression Laramide revolution is convenient, provided one keeps in mind that it refers not to a single disturbance but to a series of orogenic episodes separated in time and space (Gilluly, 1949). The major structural features in and near Glacier National Park result from one such episode. Processes more or less directly contributory to the production of the Lewis overthrust and allied structural features probably began long before the part of geologic time to which the Laramide revolution is commonly assigned. Repercussions of that disturbance have continued to the present day.

In the Glacier Park and Flathead regions, the outstanding structural feature is the Lewis overthrust. In front of the present trace of that great fracture is a broad disturbed zone containing many folds and thrusts. Above the main thrust is a massive block that is itself folded. Faults have broken this block and are among the conspicuous and puzzling features of the regional geology. In an attempt to understand the Lewis overthrust and related features, it is necessary to consider (1) the long-held concept that the Lewis is an erosion thrust in which a great block of rock slid over the then-existing ground surface, (2) the date of thrusting, (3) the concept that one or more soles similar to those of the Northwest Highlands of Scotland were produced, (4) the depths involved in the thrust movements, (5) the amount of horizontal displacement, (6) the part played by the folds and, more especially, the faults in the overriding block, (7) the present shape of the Lewis thrust, and (8) adjustments after the thrusting. Each of these will be considered below, but as a preliminary summary, it may be stated that the results of the present investigation indicate that the Lewis thrust is a major fracture that originated deep within the earth and carried a great block of rock a long distance eastward. The thrust is not believed to have reached the surface of the ground anywhere in the vicinity of the region studied and may have played out well below the surface. In any case, the east edge of the thrust is now eroded away. The rocks overridden by the thrust were folded and perhaps broken in the early stages of the deformation and were further crumpled and fractured as the great mass of the overthrust block passed over them.

The different steps in the formation of the Lewis overthrust are shown diagrammatically in plate 4. Steep faults roughly parallel in strike to the overthrust cut the overriding block and may record late stage adjustments after the major thrust had taken place. Plate 4 is necessarily diagrammatic and generalized. It was drawn with conditions in Glacier National Park in mind, and the last of the diagrams (D) is generalized from a block near the center of the park. The dates assigned to the three drawings are relative rather than precise. For example the broad warps shown in the first diagram (A) are indicated as of Late Cretaceous age because a fairly continuous record of Cretaceous deposition without angular unconformities is known in nearby areas such as the Flathead region. However, at least in localities west of the park, it is possible, even probable, that the uplift and warping that constituted early events leading to the formation of the Lewis overthrust may have been started earlier in the Mesozoic era. The second diagram (B) shows the relatively intense folding that followed the warping. Its date is close to the beginning of the Tertiary. Diagram C shows conditions after the thrust had occurred. Here the rocks caught beneath the thrust are more highly contorted than they were at the time represented by diagram B. This resulted in part from pressures during thrusting and may have been accompanied by some downward movement of the deformed mass as the heavy overthrust block passed over it. In some localities, notably north of Marias Pass, relatively intense crumpling of the beds is so localized beneath the thrust plane that genetic relationship seems demonstrated. Compression and contortion in the decidedly incompetent Cretaceous strata beneath a thick slab of highly competent Belt rocks would be greater than might be expected in many thrust zones, where the rocks above and below the thrust plane were more nearly equal in strength.

Diagram C is necessarily noncommittal as to the manner in which the thrust terminated eastward. The most easterly part of the thrust that is shown consists of several, spreading fractures, in contrast to the single fracture shown (very diagrammatically) farther west. Continuation of the process of splitting up into groups of progressively smaller fractures would permit the Lewis overthrust to die out eastward without anywhere reaching the surface. Whether it did this or actually cracked the surface of the ground somewhere far to the east of the present mountains is unknown. No evidence in support of the second alternative has been recorded nor can it be expected as the whole of the eastern part of the thrust has been eroded away, at least in the latitude of Glacier National Park.

EROSION THRUST CONCEPT

The Lewis overthrust was first recognized and named by Bailey Willis (1902, p. 331-343) as a result of a brilliant reconnaissance that contributed greatly to knowledge of the geology of northwestern Montana. He thought that the thrust plane had emerged near the present mountain front and moved over the then-existing surface of the ground east of the break and that the surface at that time was a peneplain. His ideas have been modified by later investigators, notably Billings (1938), who concludes that the Lewis overthrust is not an erosion thrust, that the part of it now visible was probably a subsurface fault, and that the thrusting is older than the erosion surface over which Willis supposed it to have moved. While his own observations were confined mostly to a small area in the vicinity of Chief Mountain, he utilized much information accumulated by others since Willis made his reconnaissance.

Billings notes that, while in places the thrust approximately corresponds in altitude to ridge tops regarded by Willis as remnants of the supposed erosion surface or peneplain, there are numerous places where it is much higher than these remnants (pl. 4, diagram D). Billings rejects the possibility that an older, largely obliterated peneplain may have furnished the surface on which the thrust moved. The erosion thrust concept would be invalidated if no sufficiently flat surface was available for the thrust block to move over. Rough topography would probably halt or break up the moving block. Thus the lack of accord between the probable position of the thrust surface and the ridge crests discussed by Billings casts serious doubt on Willis' interpretation. Further, Billings calls attention to the faults and folds in the disturbed zone in the plains, which are truncated by all remnants of old erosion surfaces there. This fact strongly implies that all erosion surfaces (peneplains or psuedopeneplains) of which traces remain on the high plains east of Glacier National Park are of later origin than the deformation that culminated in the Lewis overthrust.

Another point made by Billings is that if, under any circumstances, a thrust did break through and move along the surface, debris from the thrust block should have been dumped in front of it as it moved over the surface of the ground. Billings searched for such debris without avail, and notes that no previous investigators have reported any; this is one of the most cogent arguments against Willis' original concept of an erosion thrust.

The studies made in 1948-50 did not include detailed observations of the thrust trace in Glacier National Park, where both Willis and Billings worked. However, the mapping in the Flathead region adds evidence opposed to Willis' concept. It would be agreed by all that the steep dips of nearly all of the thrust planes in that region, including the Lewis, imply that they formed at depth; and if any part of the Lewis overthrust emerged from the ground as an erosion thrust, it must have been far to the east of the outcrops of that fault.

Remnants of the supposed peneplain that Willis appears to have regarded as the surface on which the Lewis overthrust emerged near Glacier Park are also present on the plains of the Flathead region. The trace of the Lewis thrust is well within the mountains and generally crops out low on valley slopes. Mountains east of the outcrops of the thrust tower thousands of feet above the highest remnants of old erosion surfaces on the plains as can be seen from the structure sections on plate 2. In order for the thrust to reach these remnants, it would have to bend sharply downward east of the mountains. Figure 20 shows that there are irregularities in the thrust surface and some tendency for the surface to bend downward to the east in the Glacier Park region. Similar irregularities are present in the Flathead region, but the flexure that would be required to extend the Lewis overthrust from its present outcrop over the mountains east of it and down to a position corresponding to the erosion remnants capping ridges rising from the present plains would be fantastic. The inevitable conclusion is that the erosion remnants on the plains bear no genetic relation to the Lewis overthrust. If this is true of the Flathead region, it is hard to imagine how it could fail to be true of the adjacent Glacier Park region. In the Flathead region, just as in the Glacier Park region, all such remnants truncate the folds and faults in the underlying rocks and are younger than these structural features and the Lewis thrust.

The above discussion casts serious doubt on any hypothesis that the Lewis overthrust emerged at the surface anywhere near the localities where it is now exposed. Almost certainly, at the time it was formed, this portion of the thrust surface or zone was deep within the crust. We do not know what happened to the front of the thrust. To the north and south, along the strike, the thrust petered out in minor fractures and folds, but over a hundred miles in each direction was required to accomplish this. In the direction in which the thrust moved, it is not so easy to visualize how a similar process could take place. At the border of Glacier National Park the oldest rocks in the region rest on Upper Cretaceous rocks with a stratigraphic hiatus of more than 30,000 feet. When thrusting occurred, the thickness of deposits of post Late Cretaceous age was small, probably 1,000 feet. Thus, the concept that the unhealed fracture emerged at the surface somewhere to the east is a natural one. However, this concept has no direct evidence to support it. The alternate idea that the fracture zone broke up into successively smaller fissures and terminated without reaching the surface is also plausible. Although the thrust persists far along the strike in both directions, the stratigraphic hiatus is greatly reduced within a short distance. In the central part of the Flathead region (pl. 2), beds of the Missoula group are thrust over Hannan limestone. Similar changes in the shape and attitude of the thrust surface or zone east of Glacier National Park could produce similar decrease in the stratigraphic hiatus. Decrease in the thickness of the Belt series eastward may help to close the gap between the beds above and below the fracture. In the Sweetgrass Hills about 100 miles east of the park and only a short distance beyond the area included in plate 3, it seems probable that Paleozoic rocks rest directly on gneiss of pre-Belt age (Ross, 1950, p. 87), and in Alberta (Burwash, R. A., 1957, p. 101) a similar situation has been found to exist close to the main mountain front. It is not necessary to assume complete juncture of the beds above and below the fracture zone. Perhaps the outer edge of the overthrust block came to rest against a cushion of crumpled Mesozoic rocks.

DATES OF DEFORMATION

The Lewis overthrust must be younger than the Willow Creek formation (Paleocene) (Russell and Landes, 1940, p. 93), which is folded with the other units in the disturbed belt associated with the thrust. On the other hand, the overthrust is older than the beds that have filled structural depressions in the overthrust block. These beds constitute the greater part of the unit mapped on plates 1 and 2 as "old alluvium and associated deposits." The oldest part of this unit has yielded fossils that on the basis of present information are regarded as of Eocene age. On this basis the Lewis overthrust took place in the latter part of the Paleocene or during the Eocene epoch. Of course, preparation for the thrust, including part of the folding, took place earlier, and adjustments of various kinds continued after the thrusting had occurred. The adjustments took place mainly through fracturing of the rocks that were disturbed by the thrusting and by recurrence of movement along the fractures thus produced. The old alluvium and associated deposits have been tilted and locally even crumpled as a consequence of the recurrrent adjustments. The earliest deposits were probably of Eocene date, but some Pleistocene strata are deformed, and minor movements may have occurred more recently. Earthquakes thought to be along old faults have occurred from time to time in northwestern Montana in historic times. None of these very recent disturbances are known to have originated along the faults here discussed (Erdmann, 1947, p. 80-81), but the possibility of slight movements along the faults in the vicinity of Glacier National Park cannot be eliminated entirely.

The concept advanced above that the Lewis overthrust took place late in Paleocene or during Eocene time is in approximate agreement with other estimates for the date of thrusting in the Glacier National Park and neighboring regions. The concept would make the thrust one of the effects of the Laramide revolution, which is the conclusion reached by all previous students of Glacier National Park and its vicinity, although this is expressed in various ways by the different writers on the regions that have been cited above. The reports by Canadian geologists dealing with areas in and near southern Alberta are in accord with this dating, although few of them specifically assign dates to the structural features they describe. Loris Russell (1952, p. 126) stated that the major orogeny in the foothills and plains was post-Paleocene and pre-Oligocene. He regards the postthrust deposits in the valley of the northern part of the Flathead River as of Eocene, probably Middle Eocene, age.

THE CONCEPT OF A SOLE

The discussion above implies that the Lewis overthrust and related fractures were formed under much pressure. The structure has been likened by J. D. MacKenzie (1922), for areas in Canada, and by C. F. Deiss (1943b, p. 1147-1162), for the Saypo quadrangle, to that of the Northwest Highlands of Scotland. The resemblance to the regions here reported on appears to be not nearly as close as the discussions by MacKenzie and by Deiss imply. As the regions are close to known oil fields their structure has practical as well as theoretical interest.

In the Scottish mountains (Caldwell, 1890; Peach and others, 1907, p. 463-476) major, nearly flat thrusts, called soles (Billings, 1942, p. 183) are overlain by blocks cut into segments by numerous, steeper thrusts thought to merge at depth with the soles. The major overthrust farthest back from the front of each disturbed and fractured block is inferred by Peach and his coworkers, in conformity with Cadwell's laboratory experiments, to have been the first to form. They note, however, that at an early stage in the investigation the thrust farthest back from the front was regarded as the last to form, a hypothesis that is still adhered to by E. M. Anderson (1942, p. 102). In subsequent advances this thrust and those in front of it moved forward along the lowest basal thrust or sole which itself was produced by the pressure that resulted in these advances.

In northern Montana, as Deiss has pointed out for the Saypo quadrangle, minor steep thrusts are locally so closely grouped that the structure is imbricate and has enough resemblance to that above the soles in the Scottish Highlands to have attracted attention. However, the imbricate masses are below, not above, the Lewis overthrust. The rock mass above the Lewis thrust has locally crumpled, but on the whole is far from being imbricate. Even in the imbricate masses below the Lewis thrust details of the structure differ from those in the Scottish Highlands. In the Highlands the beds composing each thrust slice in the imbricated masses are approximately parallel to the bounding thrust planes, whereas in Montana the beds are commonly at variance in attitude with the thrusts. Further, in many places in the mountains of the Flathead region, where exposures are especially good, the minor thrusts are fractures on or near anticlinal crests. Some of these can be seen to die out downward rather than to merge at depth with a sole. No such relation to folds has, however, been recorded for the Lewis overthrust. The mylonite zones and other features that testify to extreme compression in the Scottish Highlands are absent in Montana. Also, several major thrusts are recorded in the Scottish Highlands; but in the part of Montana under discussion, the Lewis is the only one known to be of major magnitude.

In the papers cited above neither MacKenzie nor Deiss record direct evidence of the presence of master soles. Deiss pictures such soles in several of his structure selections—but at altitudes a little more than 1,000 feet above sea level, which is far below the limit of observation at the surface. The well near the mouth of Blackleaf Canyon, referred to above, p. 80, extended to a much greater depth than the soles in Deiss' sections without penetrating below the disturbed and faulted zone. In T. A. Link's descriptions of structural features in front of the Rocky Mountains north of Glacier National Park (Link, 1935, 1949) several of the thrusts are referred to as soles. Some of these may be of greater magnitude than any, other than the Lewis, so far recognized south of the international boundary. However, his descriptions and structure sections indicate that the faults he calls soles are not such dominant features of the structure as the soles in the Scottish Highlands are postulated to be. Link's structure sections, based largely on data from wells, corroborate the data from the single deep well in the Flathead region by showing that the disturbed zone in front of the present outcrop of the Lewis overthrust extends to depths far below sea level.

In summary, it now appears that the structure is not as nearly as analogous to that inferred from Cadell's experiments as it was once supposed to be. Possibly some of the thrusts east of the Lewis overthrust deserve to be termed "sole faults" with reference to the minor thrusts immediately above them, but no major underlying regional sole is known. Present data encourage the idea that the Lewis overthrust is the only major, persistent thrust within the regions mapped and the thrusts east of it are subordinate and, on the whole, discontinuous fractures caused by pressures set up during the movements that culminated in the Lewis overthrust. Many are mere fractures along the crests of more or less overturned anticlines. The disturbed zone contains many irregularities in structure and many faults. It includes incompetent rocks and began to be deformed long before movement along the Lewis overthrust began. As the great block of strata above the Lewis thrust was carried forward, the rocks beneath (those of the present disturbed zone) must have been further deformed, and the results of that deformation extend to depths below present sea level. Many of the thrusts within the disturbed zone may have formed during the period of active movement along the Lewis thrust and in doing so may have aided in that movement. This agrees with the suggestion (Peach and others, 1907, p. 472) that "The wedges of piled-up strata showing imbricate structure may be said to have acted like rollers for the transport of advancing masses on higher thrust planes," and that "eventually" friction may have accumulated to such an extent as to produce sharp plication of all the structures overlying the sole. The suggestion just quoted does not seem in close agreement with Cadell's conclusions, but it receives support from the descriptions and structure sections in the chapter that follows it (Peach and others, 1907, p. 477-492).

On this basis, no evidence exists to support the idea that in the Glacier National Park or Flathead regions drilling would penetrate a sole at depth and reach relatively undisturbed beds beneath. In the oil fields in Alberta east of the outcrop of the Lewis and its branches, some of the slices between thrusts in the disturbed zone have proved to be large enough and to possess within themselves the necessary features of structure and of access to source beds so as to retain oil in commercial quantity. In the disturbed zone south of the international border, no such conditions have yet been found although oil fields exist a short distance to the east. Conceivably the broad zone of Cretaceous beds that stretches diagonally across the southern part of the Marias Pass quadrangle would warrant examination by oil men. In general, however, the disturbed zone east of the outcrop of the Lewis overthrust in the two regions discussed in the present report presents many complications and difficulties to petroleum prospectors. Drilling in those parts of the mountains where the Lewis overthrust has not yet been eroded away would seem to be so hazardous as to be futile in the present state of knowledge.

DEPTH OF FRACTURE

The disturbance that resulted in the Lewis overthrust must have originated fairly deep in the earth's crust. Rocks of Late Cretaceous age lie beneath the thrust in Glacier National Park, and these are conformable with beds of probable Paleocene age farther east. The thrust cuts the lowest exposed formation in the Belt series. Thus, it appears that at the time fracture began the whole known thickness of the Belt series, most or all of the Paleozoic and Mesozoic strata and some Cenozoic strata were present in the vicinity of the present eastern border of the park. Deformation and erosion have made it impossible to measure the thickness of all the rocks at this locality, but a conservative estimate would place the aggregate at about 40,000 feet of beds. This estimate includes about 10,000 feet of strata of post-Belt age, corresponding approximately to the thicknesses of corresponding beds in the Flathead region. Similar thicknesses of Paleozoic and Mesozoic strata are known in nearby parts of Canada east of the Lewis overthrust (Webb, 1951).

The rocks caught beneath the thrust include a large thickness of Upper Cretaceous strata. This implies that erosion following the retreat of the last of the Mesozoic seas had not accomplished a great deal in the area near the present eastern border of Glacier National Park before the thrusting took place. If, as is probable, folding preceded the development of the Lewis thrust, areas within the site of the present mountains may have been vigorously eroded. The presence of beds of variegated clay and soft sandstone, with limestone lenses locally (Stebinger, 1916, p. 124-128), of supposed Paleocene age (Russell and Landes 1940, p. 93), east of the park is not in harmony with the concept of intensive denudation in the park area at the end of the Mesozoic. Hence, diagram C in plate 4 shows some of the Mesozoic and older strata still present in the overthrust block, although none are left in that block in diagram D.

Perhaps erosion after the Cretaceous rocks were uplifted and before the thrusting took place was largely confined to valleys in uplifted areas, and any resulting decrease in load above the thrust zone may have been more than compensated for by thickening of the crust by folds. Certainly the beds above the Belt series yielded to folding before the thrusting began.

Immediately north and south of Glacier National Park, the Lewis thrust cuts horizons far higher than the Altyn limestone that is disrupted by the thrust in the park. However, to the south especially, the dip of the thrust is so steep that the distance beneath the surface must increase rapidly. West of the Glacier Park and Flathead regions, the upper part of the Belt series and all Paleozoic and Mesozoic strata have been eroded, but formations broadly like those described in the present report are thought to have once extended over much or all of northwestern Montana, and many of them are thought to have thickened west of the longitude of Glacier National Park. That this is true as far as the Belt series is concerned is indicated by the large thicknesses reported near the Idaho border (Gibson, Jenks, and Campbell, 1941) and in at least one area in British Columbia (Rice, 1941). The complex structure and extensive erosion in western Montana and in British Columbia hinder the obtaining of data on aggregate thicknesses. Perry (1945) indicates that the Jurassic rocks thickened westward and Reeside (1944) records the same thing for the Cretaceous units. In nearby parts of Canada the thickening of the column from the Cambrian upward was great, according to recent isopach maps (Webb, 1951). Eardley's maps (1951, p. 14, pls. 2-16) show various fluctuations but in general support the concept of basins of deposition that extended into western Montana during much of Paleozoic and Mesozoic time.

The concept that seems to fit the known data best is that the fracturing that led to the Lewis overthrust took place at a depth of some miles, perhaps as much as 10 miles, beneath the then-existing surface of the ground. The estimate is rough because of scanty information and because both the surface of the ground and the fracture zone presumably had significant irregularities. Neither uplift nor erosion was uniform, and the attitude of the fracture zone that constituted the incipient overthrust may have had variations similar to those of the present overthrust.

UPLIFT

Once thrusting began, the block above the Lewis overthrust moved upward and eastward. One of the factors that interferes with attempts to estimate the results is that, at least locally, fracturing was distributed through a zone of considerable thickness. The descriptions given above show that in places thrust surfaces are grouped in zones hundreds or even thousands of feet in vertical extent. A more serious cause of uncertainty is the lack of quantitative data as to the competence of the rocks beneath the thrust zone to resist the pressures developed during the fault movement. The block above the fracture zone did not merely slide forward and upward at angles corresponding to the dip of the thrust surface. The amount of uplift that might be estimated on such an assumption would be decreased by the amount that the rocks beneath the thrust failed to resist the pressures exerted on them. Only a faint idea of this can be obtained by direct field observation. Near Marias Pass the soft Cretaceous rocks are crushed and intricately crumpled through a range of 1,000 feet or more (pl. 1 and fig. 21). Most of this deformation is a direct result of pressures during thrusting, but the amount of compression thus recorded may be only a fraction of the downward movement that occurred. Some of the overturning and fracturing recorded in the rocks of the eastern part of the Flathead region and in similar localities probably took place during the major thrusting. In addition to the compression recorded by crenulated zones, fractures, and related structural features, the mass beneath the fracture zone may have been shoved downward as a whole. How much downward movement of this sort occurred depends on factors beyond the limits of observation. It seems safe to assume, however, that downward movements were sufficient for a significant decrease of the net uplift. Even so, uplift sharper and more concentrated than anything previously recorded in the region must have taken place. The result would be accelerated erosion and the carving of mountains. The differences in the topography shown in diagrams C and D is intended to bring this out.

That erosion after the thrusting was vigorous and effective is proved by the Tertiary deposits that remain. M. D. Billings (1938, p. 270) has called attention to the fact that gravel of supposed Oligocene age derived mainly from rocks of the Belt series mantles remnants of a surface in the Cypress Hills in southeastern Alberta near longitude 110° described by W. C. Alden (1932, p. 4-8). This is evidence that Precambrian rocks were exposed to erosion during the geologic epoch succeeding that in which thrusting occurred. So far as recorded, the gravel immediately east of Glacier National Park contains no material derived from post-Belt rocks; so the latter must have been removed from the site of the park before the late Tertiary, when that gravel began to be deposited.

DISPLACEMENT

Evidence as to the amount of horizontal displacement along the Lewis overthrust is scanty. The most frequently cited distance is 12-15 miles, as a minimum (Campbell, 1914, p. 12; Clapp, 1932, p. 25; Dyson, 1949a, p. 14); this is based on Campbell's assumption that the westward swing of some such distance of the overthrust trace near Marias Pass is a measure of the displacement. The bend in the trace in this locality is related to the sharp change in the dip of the thrust rather than to the distance that the block overlying the thrust has moved forward. The stratigraphic hiatus along the exposed trace of the thrust is, as indicated above, markedly different from one locality to another. Everywhere, however, the hiatus is so great as to suggest that the horizontal displacement was larger than Campbell's estimate. Under one of several hypotheses advanced by T. A. Link (1935, p. 1466), the Lewis overthrust has a displacement of at least 44 miles. This is in harmony with C. E. Erdmann's (1947, p. 78) estimate of "* * * not less than 40 miles." Erdmann's statement was in a summary given in a report on dam sites, and his reasons were not stated. Pierre de Bethune (1936) estimated 40 miles for an area in Canada, based, however, on a somewhat radical interpretation of the structure. At a much earlier stage in the study of the region, R. A. Daly (1912, p. 91) suggested a displacement of "* * * at least 40 miles."

The abrupt westward deflection of the thrust trace north and south of Glacier National Park, referred to in the descriptions given above, shows that the overthrust block within the park is at least 15 miles wide, which is what Campbell had in mind. Surely the block has been much eroded, and it originally extended eastward over at least part, perhaps all, of the disturbed zone. Westward, likewise, it did not terminate immediately beyond the present exposure of the thrust trace. The disturbed zone near the park is about 20 miles wide, and an area of similarly disturbed beds has been reported in Alberta (Williams, M. Y., and Dyer, 1930, p. 88-89) over 60 miles from the mountain border. With these facts, plus the great length of the trace of the Lewis overthrust in mind, the estimates of over 40 miles of horizontal displacement quoted above seem conservative. A much larger displacement is possible.

STRUCTURAL FEATURES ABOVE THE THRUST

A large part of the mountain mass here described has been carved from the great body of rocks above the Lewis overthrust; that is, from the block that was shoved forward over the fractures that make up that thrust. Even though it is so deeply carved and in many places so devoid of soil cover that the rocks are locally very well exposed, much remains doubtful about the genesis of the structural features. The rocks now present are thick bedded and so competent that, except in a few localities, they have escaped the crumpling and intricate faulting characteristics of much of the far less competent rock beneath the main thrust zone. Most of the major folds are broad and open. They might well have originated from simple compression without complexities related to faulting. Perhaps the shallow syncline in Glacier National Park and the uplifts that border it originated in that fashion before the fracturing related to the Lewis overthrust had proceeded far enough to have significant effect.

The extensively crumpled beds near the western border of the park (figs. 26, 28) and similar but less impressive features in other localities may be more closely related to the overthrusting. No overthrusts have been demonstrated in the principal zone of sharp deformation in the western part of the park, but the dips are steep and in part overturned toward the northeast. Little more deformation would be required to give rise to thrusts like those in the eastern part of the Flathead region. Some of the scattered minor zones of sharp deformation that have been described differ in that they show that some westward movement has taken place. The significance of these is a matter for debate. Perhaps they correspond to local irregularities in a region where the dominant pressures were toward the east and northeast.

Perhaps, as some geologists have postulated (Chamberlain, 1925; Flint, 1924; Clapp, 1932), they are related to large thrusts of northeasterly dip. These investigators base their concept primarily on the wedge theory of diastrophism, which attributes the uplift of the mountain ranges or of groups of ranges to the formation of huge downward-pointing wedges bounded by thrusts that dipped under the uplifted wedges from both sides. As noted above, little direct evidence in support of the application of this theory to northwestern Montana has been published. In Canada, as noted above, overturned and overthrust zones have been recognized through a broad region that extends far to the west and north of the vicinity of Glacier National Park. Some of the thrust features dip east, but most students agree that the major pressures came from the opposite direction. Thus, recent studies in Canada give little support to the wedge theory.

The significance of the remarks just made to the two regions described here lies in their bearing on the interpretation of the long, steep faults of northwesterly trend that border many of the master stream valleys. The group represented by R. T. Chamberlain, R. F. Flint and C. H. Clapp would regard the long faults as thrusts, mostly or entirely of easterly or northeasterly dip. Recent Canadian work, such as that reviewed by C. S. Evans (1933) and T. A. Link (1935, 1949), does not rule out thrusts of more or less easterly dip but shows that the principal known thrusts are of westerly dip. If the long faults are thrusts at all, the dip must be easterly as can be seen readily from plates 1 and 2. The simpler and, on the whole, more logical explanation is that they are normal faults. According to this view, those on the northeast sides of the master valleys would dip southwest and those on the southwest sides of the valleys would dip northeast. The scanty direct evidence as to the attitudes of the fault surfaces is in accord with this interpretation. Few of the faults on the southwest sides of valleys are included in the mapped regions. Where they exist, the valleys would be grabens. Topographic depressions like that occupied by the northern reaches of the Flathead River are certainly grabenlike in that the bedrock buried in their floors belongs to units stratigraphically much higher than that on their flanks. The net effect has been that of sharp down drop in elongate zones. The depressions thus produced are now partly filled by relatively younger materials. On the whole, the theory of grabens bounded by normal faults seems best to explain the data at hand.

In the above discussion, necessarily inconclusive, emphasis has been laid on the wedge theory of diastrophism. This is mainly because so much of the previously published discussion has involved consideration of that theory. Other theories, such as the ramp hypothesis that has been applied to certain African and Asiatic valleys (Willis, 1927 1936; Rich, 1951, p. 1119-1222), or J. L. Rich's idea of deformation through sliding off a dome above a magma mass could be entertained; but in the absence of more complete information, detailed discussion would not be profitable.

There are a number of faults and zones of weakness in the block above the Lewis overthrust not involved in the discussion given above. Most of these are of northeasterly trend and are inferred from such things as lack of harmony between the rocks exposed on the two sides of a mountain valley. Most, probably all, of the faults of northeasterly trend are of moderate to small displacement. They are minor features in comparison to the great faults of northwesterly trends. Some may be tear faults. All, presumably, are related to the great overthrust beneath them at least to the extent that it set up strains in the block above it.

PRESENT SHAPE OF THE LEWIS THRUST

The Lewis overthrust departs drastically from a geometric plane in two respects. First, it is a fracture zone that in many places is surely hundreds and perhaps locally over a thousand feet thick. Second, this zone, or the principal fracture surface, is irregularly curved. Limitations of scale and of available information have masked these features on plates 1 and 2.

In spite of the inevitable generalization, the two geologic maps record essentially all available data as to the fractures that together make up the Lewis overthrust. They indicate that most of the fractures that make up the fracture zone are subordinate in varying degrees to a main fracture represented in essential correctness by a single line on the map. Features like those east of Two Medicine Lake are exceptional. At that locality the fracture zone has at least two major components and has a maximum mapped width of about three-quarters of a mile.

Departures from a plane surface that result from differences in the inclination of the thrust surface (or zone) are of more genetic significance. Within Glacier National Park the dip near the exposed trace is so low that in most exposures it seems almost flat. Both north and south of the park, the dip steepens very sharply. As these changes in dip involve equally wide differences in the stratigraphic units cut by the thrusts, they must correspond to original variations in the shape of the fracture zone. The steeply dipping parts of the thrust zone are in younger rocks and represent a smaller stratigraphic hiatus than the gently dipping part of the thrust zone, which lies between them. No satisfactory explanation of this anomalous situation is at hand. One possibility is that the low-dipping segment within the park, which involves the lowest part of the Belt series known in the region, has been shoved forward and upward along tear faults on either side. This explanation calls for a large fault approximately along the valley of Bear Creek, extending northeast through Marias Pass, and a similar fault in Canada (Link, 1935, p. 1462-1463; Allan, 1937). No such faults are recorded. The mapping along Bear Creek (pl. 1) is directly opposed to the hypothesis as all contacts cross the valley without measurable offset. None of the postulated faults of northeasterly trend, such as that along Ole Creek, has sufficient displacement to accord with the hypothesis of major movement in tear faults. The most probable alternative hypothesis is that the original fracture zone was so extremely irregular in shape as to account for the broad differences in attitude and stratigraphic relations now present at the outcrop.

Minor variations in attitude such as those represented in figure 20 do not involve such significant differences in the stratigraphic breaks along the fault zone as the larger features just commented on. They have been attributed (Willis, 1902, p. 332; Billings, 1938, p. 263) to folding after the thrusting had taken place. In a locality in Canada just north of Glacier National Park, Hume (1933, p. 9) came to the same conclusion. It is at least equally possible that they, like the larger irregularities, are original. This second hypothesis is favored by the lack of system displayed by the contours in figure 20 and by failure of the rocks in the block above the thrust to show comparable features. Folds corresponding to the contours in figure 20 would be roughly at right angles to the regional folds and faults. As there is no evidence of major deformation in that direction, folds of that trend would be results of adjustment during or following the thrusting.

The variations in the shape of the thrust zone discussed above are related to present exposures of that zone in the eastern parts of the two regions mapped. Other variations must exist in parts of the zone west of these exposures. Suggestions have been made that such variations, presumably the result of folding after the thrust took place, have brought the thrust to the surface along the upper valley of the Flathead River. Folds of this character would parallel the regional structure, and it would be difficult to determine whether they formed during the thrusting, immediately after it, or as a result of distinctly later and unrelated pressures. The question of the date is significant in connection with problems of origin, but whether or not the Lewis overthrust is exposed at the surface somewhere west of its well known trace at and near the mountain front has several connotations. It affects interpretation of the structure in the vicinity of the upper reaches of the Flathead River and the character, thickness and origin of the deposits there, and it also has a bearing on the possible presence of oil.

In Canada the concept of thrusts folded approximately parallel to their strike has attained wide acceptance. G. S. Hume (1933, p. 7-12) believes that the Lewis overthrust in Waterton Lakes Park (immediately north of Glacier National Park) is warped, and he quotes an oral statement by V. R. D. Kirkham that the valley of the northern part of the Flathead River may be a window in that thrust. Pierre de Bethune (1936) had similar ideas about the area farther north. T. H. Link (1935, p. 1464-1466) mentioned the concept of a window in the Lewis thrust as one of several possible explanations of the structure along the Flathead River but expressed doubt as to its validity. J. C. Scott (1951) has summarized data on folded thrusts in the foothills of the Rocky Mountains in Alberta and shows conclusively that in several localities thrust faults have been sharply folded. His paper cites 23 previous papers, many of which present similar ideas having varying degrees of probability. With such widespread evidence that thrusts in Alberta have been drastically folded, the Lewis thrust in the Glacier National Park region might well be expected to have been folded similarly. This does not necessarily follow because, as has already been pointed out, conditions in that region appear to be different from those farther north. In the Glacier National Park region, a single thrust of exceptionally low dip dominates the structure. It is overlain by a block of competent rock. East of that block the far less competent rocks beneath the thrust are folded and faulted. Few details are known, but no thrusts at all comparable to the Lewis have been found east of that thrust either in the Glacier National Park region or in the better known area in the Flathead region.

If, in accord with conditions farther north, the Lewis overthrust in the regions here described has been drastically folded, the most probable places to look for evidence of that folding is in the big valleys of northwesterly trend. The valley of the northern part of the Flathead River (locally termed "the North Fork") would seem to be an especially favorable place to look. The evidence at hand is far from conclusive but, so far as it goes, is opposed to the concept that any of the northeastward-trending valleys are windows in a folded thrust. This concept requires that the faults on either side of each of the valleys would dip toward the mountains, whereas they seem to dip in the opposite direction—toward the valleys. If the valleys are windows, beds of Mesozoic and perhaps also of Paleozoic age, which are comparable to those known to underlie the thrust farther east, should be present in them. None have been found. On the contrary, beds later than the thrusting underlie the mantle of Quaternary deposits in the valleys. To be sure, early Canadian maps (Link, 1932; Daly, 1912, map 74A, sheet 1) show beds of Mesozoic age in the valley of the Flathead extending south to or across the international boundary. These were sought for without success during the present investigation and do not appear on recent geologic maps of Alberta and British Columbia (Canada Geol. Survey, 1928, 1948, 1951).

ADJUSTMENTS AFTER THE THRUSTING

The block of rock that overlies the Lewis overthrust is cut by faults both approximately parallel to and approximately normal to the strike of the thrust. Available data leave much in regard to the faults open to question, but the most probable hypothesis is that most are steep normal faults. Those of northwesterly trend include many with thousands of feet of vertical displacement, whereas those of northeasterly trend have such small displacements that they are difficult to recognize. If these concepts are accepted it seems evident that much or all of the movement on both sets, but especially on those of northwesterly trend, is later than the thrusting. The rock remaining above the Lewis thrust consists almost entirely of beds of the Belt series. The very fact that this rock is so competent that much of it did not yield enough to be closely folded before and after the thrusting implies that the block as a whole did not accommodate itself to irregularities in the thrust zone over which it was shoved. Hence, strains must have been set up in it. Relief of such strains after much or all of the thrust movement had ceased might well result in normal faulting. The faults under discussion are inferred to have resulted from adjustments of this character. None of the apparently normal faults carried any rock of Paleozoic or Mesozoic age down into positions now exposed. Such rocks might be buried under the Cenozoic deposits in the larger valleys, but this is improbable as none of the Cenozoic rocks are known to contain detritus eroded from them. Thus, much or all of the normal faulting may have taken place long enough after the thrusting, so that erosion had had time to lay the strata of the Belt series bare. When the thrusting began, much of the Mesozoic material was still in place. In terms of geologic time the interval between thrusting and relief of strains by normal faulting was not great. Some of the Cenozoic deposits in the valleys are probably at least as old as Oligocene.

Faulting on the large scale believed to have taken place where the major valleys of northwesterly trend now are must have interfered seriously with whatever drainage existed at the time. As a result the principal valleys bordered by faults are floored by extensive deposits. These include some coal and some fine-grained sediments; so parts of the valleys at times contained swamps and lakes. On the other hand, coarse sediments are abundant enough to show that through-flowing streams occupied the valleys much of the time. There must have been repeated shifting along the faults with resulting modifications in the drainage, for the beds of early Tertiary to early Pleistocene age that constitute the valley fill are all tilted, and some are crumpled.

SUMMARY OF STRUCTURAL INTERPRETATION

The greatest accumulation of sediments in the regions described took place during the latter part of the Precambrian era, but the deposition was not followed by marked diastrophism. From the close of the Precambrian through the Mesozoic era, sedimentation continued with interruptions that probably resulted from broad crustal upwarps. At some time late in the Mesozoic, preparations began for the single period of intense deformation recorded in the rocks of northwestern Montana. The first movements may have been broad upwarps similar to the earlier ones, but as diastrophism continued the rocks were folded. During the Paleocene the folding continued until, late in Paleocene or early in Eocene time, a major fracture zone developed, presumably much to the west of the present location of Glacier National Park and at a depth of some miles below the then-existing surface of the ground. The block of rocks above the fracture zone was thrust forward as a unit toward the northeast. Some of the anticlines beneath the overthrust block may well have been broken before the major overthrusting took place, but fracturing and crumpling continued in them as the overthrust block passed over and pressed down on the deformed strata, the upmost of which were much less competent than the lowest and thickest part of the overthrust block. In places intense crumpling and crusting of incompetent, shaly beds took place immediately beneath the moving block. The fracture zone itself varied in original shape and in the number of component fractures. It may have been further folded in the course of the thrusting although evidence in support of this is inconclusive. Direct measurement is impossible, but it is reasonable to estimate that the overthrust block was shoved northeast at least 40 miles. The zone of fracture was well below the surface in the longitude of the eastern border of Glacier National Park at the time that the overthrust movement ceased. Whether or not that zone, or a part of it, reached the surface somewhere farther east has not been determined. Possibly pressure dissipated eastward, and the fracture zone broke up into minor fissures and finally feathered out without emerging at the surface.

When the overthrusting ended, the block of rock above the fracture, or thrust zone, was left in a state of strain and had been raised sufficiently to make much of it subject to active erosion. Normal faults relieved the major strains, but complete equilibrium may not have been attained even yet.