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

STRUCTURE (continued)

MOUNTAIN-BUILDING STRUCTURAL FEATURES

In the Paleocene or somewhat afterwards, conditions changed drastically from those outlined above. Broad warpings of the earth's crust so gentle that the resulting structural features have to be inferred rather than seen were succeeded by violent movements in which the rocks were tightly folded and eventually extensively fractured. Thrusts and normal faults of the first magnitude resulted from a series of earth movements that have not entirely ceased even yet. In the summary that follows, the structural features are not described strictly in the order of their age. Many of the features formed before the Lewis overthrust, but that great fracture is such a dominant feature of the region that it seems desirable to outline its extent and character first. Associated folds and fractures, both older and younger than the Lewis thrust, are described in relation to it.

LEWIS OVERTHRUST

By far the largest and most famous of the thrust faults is the Lewis overthrust, named by Bailey Willis (1902, p. 331). Most of the visible structural features of the two regions described in the present report are related, more or less directly, to this great fault. The total exposed length of the Lewis thrust and fractures essentially coextensive with that thrust is more than 300 miles, with possible minor interruptions. It has been mapped in essential continuity from latitude 50° to latitude 47°30'. Beyond these parallels the same thrust zone continues both north and south. Such apparent interruptions as exist are in part the result of inadequate information, in part the result of changes in attitude and of cross faults. The zone is essentially continuous from latitude 57° south to latitude 47°. For Canada, data on the Lewis thrust have been summarized by Mackenzie (1922, p. 97-132) and are shown on a geologic map of part of Alberta (Canada Dept. Mines, Geol. Survey 1928), a more recent map of the province (Canada Geol. Survey, 1951), and in more generalized fashion on the recent tectonic map of Canada (Derry, D. R., 1950a, b). For this country, the geologic map of Montana (Ross, Andrews, and Witkind, 1955) has been drawn upon.

At the border of Glacier National Park, the Lewis overthrust is at the eastern edge of the mountains; but both to the north and to the south, it swings rather abruptly westward into the mountains. On the north the bend in the thrust trace begins close to the international boundary (Hume, 1932). On the south the trace turns abruptly in the north central part of T. 31 N., R. 13 W, continues southwest for 15 miles, then, equally abruptly, resumes a southeasterly trend fully 10 miles within the mountain mass. Thus, the mountains in the park are carved in an eastward bulge in the block above the Lewis overthrust.

Mackenzie notes that in Canada the dip of the overthrust is moderate to low. Willis (1902, p. 237, fig. 4) shows that south of Chief Mountain east of Glacier National Park the westward dip ranges from 3° to 7°45'. Figure 20 shows by means of contour lines the attitude of the principal thrust surface as mapped on plate 1. The area within which the overthrust can be seen is too narrow, in a direction normal to the strike, for the contours to have much quantitative significance. They do however, indicate that the thrust surface is far from being a geometric plane, the average strike is close to N. 30° W., and the dip is to the southwest at low angles. In a few places it may be as much as 20°, but generally it is less than 10°, and long stretches are nearly flat. Figure 20 suggests that the thrust bends downward to the east, so that the part eroded away may have been even flatter than that now exposed. Farther south, in the Flathead region, the opportunity for measurement is even less, but it is obvious from plate 2 that the dip is relatively steep. The maximum dip in the Flathead region may be as much as 50° southwest, and the strike is about N. 30° W., essentially the same as in Glacier National Park. Still farther south, in the Silvertip and Saypo quadrangles (Deiss, 1943b, p. 1123-1168; 1943a, p. 205-262, also unpublished geologic maps of Saypo and Silvertip quadrangles by Charles Deiss), the dip continues to be steep, but the strike swings to more nearly north. South of the Saypo quadrangle present information is scanty, but there is some evidence that the average dip decreases and the strike varies in different localities.

FIGURE 20.—Sketch contours on the major surface of the Lewis overthrust near the eastern border of Glacier National Park.

The Lewis overthrust within the country described in the present report is, in a broad way, a single, continuous fracture, but in detail there are many departures from this simple picture. On and near Chief Mountain and Sherburne Peak in the northern part of plate 1, at least two thrusts of some magnitude are present. Similarly, in the vicinity of Schafer Meadows in the southern part of the Flathead region, two thrusts are mapped. One, interpreted as the main Lewis overthrust, passes around the east side of Lodgepole Mountain and continues southward along the valley of the Middle Fork of the Flathead. The other, apparently much shorter, is exposed on the lower slopes of Union Peak and continues up Roaring Creek and past Argosy Mountain. Willis' (1902, p. 322-335, figs. 5 and 6) descriptions and drawings and the field notes of Campbell's men show that a variety of minor thrust fractures, too small to show on plate 1, accompany the Lewis thrust in Glacier National Park. G. C. Ruhle (1949, p. 55) notes that at one place on Yellow Mountain repeated thrusts have piled a thickness of strata, each slice not more than 500 feet thick, to a thickness of 2,400 feet. At another place he records that a single assemblage of strata low in the Belt series is repeated nine times by thrusts. Plate 1 shows several short fractures cutting or bordering the main thrust. Most of these are of normal displacement and represent adjustment subsequent to the thrusting, but some may be thrusts. The old field maps of Campbell's men from which these short faults were copied do not in all cases indicate their character or the direction of displacement. All the fractures noted above, and similar but unrecorded features, are best regarded as components of the Lewis thrust zone.

The Lewis thrust in Glacier Park underlies a block of rocks belonging to the Belt series and rests on Cretaceous rocks. Even though slopes are steep, exposures of the rocks immediately beneath the thrust are commonly unsatisfactory. Extensive talus slopes are common, and small landslides in the soft Cretaceous rocks are plentiful, so that many details are hidden. The slides gave trouble in the construction of roads and continue to contribute to the difficulty of maintaining roads in the eastern part of the park. In some exposures the Cretaceous rocks beneath the thrust are nearly parallel to those of the Belt series above, but in others the Cretaceous rocks are intricately crumpled and broken. This latter feature is especially well displayed along the steep mountain front north of U. S. Highway No. 2 from near Blacktail to False Summit (fig. 21). In the Flathead region the beds beneath the overthrust are more competent, and comparatively little crumpling has been observed.

FIGURE 21.—The trace of the Lewis overthrust as seen from Marias Pass. The dark mountains on the skyline consists of Appekunny argillite. The light-colored massive rock near the middle of the slope is Altyn limestone, with the Lewis overthrust immediately beneath it. The lower slopes, in part mantled with talus, are composed of intricately contorted Cretaceous strata. Photograph by Richard Rezak.

In Glacier National Park the Flathead region and areas just south of the latter the rocks immediately above the Lewis overthrust belong to the Belt series. In the park they are close to the bottom of the exposed part of that series, whereas farther south they are at higher horizons and Paleozoic rocks are present short distances west of the thrust trace. In Canada, Paleozoic rocks appear above the main thrust roughly 45 miles northwest of the point where the thrust crosses the international boundary. The block of Paleozoic strata begins at the west end of the sharp westward bend in the thrust trace corresponding to that at the southwestern corner of Glacier National Park. (Canada Dept. Mines, Geol. Survey, 1928, Hume, 1933). In contrast, the rocks beneath the thrust in Glacier Park are largely of late Cretaceous age, whereas much of the rock in corresponding position to the north and south is Paleozoic. Thus, the greatest stratigraphic hiatus in or near the territory covered by the present report is in Glacier National Park, especially its northern part.

THRUSTS IN THE EASTERN DISTURBED ZONE

Throughout the extent of the Lewis overthrust, a belt of disturbed strata lies east or northeast of the trace of the main overthrust. In the latitude of Glacier National Park, this belt is in the plains east of the mountains and, according to Eugene Stebinger (1916, p. 281-305), extends into the Blackfeet Indian Reservation for about 20 miles. Within this disturbed belt the rocks are mostly of Cretaceous age and have been intricately folded and thrust faulted. In many places, according to Stebinger, the rocks have been so much crushed and broken that it is impossible to identify the different formations. This, coupled with the fact that much of the terrain is mantled with unconsolidated material, has made it impractical with present data to show formational contacts on plate 1. The amount of deformation, especially of thrusting, decreases eastward according to Stebinger. Stebinger places the eastern border of the disturbed zone, in the latitude of Glacier National Park, about at the western border of the area underlain by the Willow Creek formation. However, the Willow Creek formation appears to be conformable with and gradational into the Cretaceous rocks beneath it, and Stebinger's structure sections show it to be folded with them although all of the rocks so far to the east of the mountains are relatively slightly deformed. Stebinger concludes that deformation in the disturbed belt is related to the Lewis overthrust, which is along the western margin of the belt.

The disturbed belt east of the Lewis overthrust continues a long distance to the northwest in Canada, where T. A. Link (1949) has summarized many structural details. Link speaks of the larger faults in the region he describes as "major sole faults." It is clear, however, that none of these underlies all the thrusts of the disturbed zone. That is, he does not use the term "sole" in the sense employed by J. D. MacKenzie (1922) and others. This difference in terminology needs to be borne in mind in connection with the following discussion of the origin of the structural features. Link says that the structure is complex and includes tight folds commonly overturned eastward but locally and on a minor scale westward. Warping, upbowing, and folding of overthrust faults and sheets are recognized features of the area, less common near the eastern edge, but more pronounced westward toward the Rocky Mountains. He describes "underthrusting or apparent overthrusting from the east" as a near surface feature. So much geologic work in the search for oil has been done in the area Link describes that he has many more facts at his disposal than are available for the disturbed belt in Montana or than were available to MacKenzie.

South of Glacier National Park part of the disturbed belt is in the mountains, but it does extend some miles out into the plains. Here it is close to 30 miles wide and includes most of the eastern half of the Flathead region. The mountainous area in the western part of the belt is not highly deformed although rocks regarded as belonging mainly to the Kootenai formation predominate and the more competent Paleozoic strata are subordinate. Its northern part contains broad areas in which no thrusts have been recognized. Farther south the rocks in the mountains are more intensely broken, but, even here, thrusts of mappable size are spaced, on the average, fully a mile apart. In the intensely deformed parts of the disturbed belt, the spacing is much closer than this. The part of the Flathead region that is in the plains was not studied during the present investigation, but Stebinger's work shows that faults, mostly thrusts, are closely spaced there. For example, his structure section drawn north of east from the bend in the Blackfeet Indian Reservation south of the railroad station of Glacier Park (Stebinger, 1917, pl. 25, sec. E—F) shows 46 thrusts in a distance of 30 miles. Here, under a cover of almost unconsolidated Cenozoic material, relatively noncompetent beds of late Cretaceous age predominate.

The character of the folds and thrusts east of the Lewis overthrust within the Flathead region can be seen from the map and structure sections of plate 2. (See also figs. 22, 23.) A few thrusts are continuous for more than 10 miles, but most are much shorter. Some of the thrusts converge both along the strike, and along the dip, but many are discontinuous in both directions. Most trend northwest and are rather closely parallel to each other and to the folds. A few, such as those between Family Peak and Mount Fields, in unsurveyed T. 28 N., R. 11 W., trend nearly east. Accurate measurements of the displacements are difficult to obtain, but it is clear that none have displacements comparable in magnitude to the Lewis overthrust. The dip of the thrusts is varied but generally rather high. Dips of 30°-50° are common. In a few places, such as along the lower reaches of Schafer Creek west of the Lewis overthrust, nearly flat thrusts are exposed. Another may be present on the top of Goat Mountain in unsurveyed T. 28 N., R. 12 W., but the available evidence for this is not conclusive. There are, of course, spots where the faults are so intricate and closely spaced that they are necessarily generalized on plate 2.

In the mountains south of the Flathead region, deformation east of the Lewis overthrust is severe and intricate as shown by C. F. Deiss' (1943b, p. 1123-1168) structure sections. In one typical east-trending structure section in the mountains of the Saypo quadrangle, he shows 20 thrusts in a distance of less than 9 miles. Deiss postulates that the thrusts decrease in dip downward "* * * and merge with one or more soles along which the imbricated fault blocks rode eastward." His structure sections show the postulated sole some distance above sea level, but an oil well in Blackleaf Canyon a short distance north of the northern boundary of the Saypo quadrangle was drilled to more than 2,000 feet below sea level without passing out of disturbed and fractured rocks. The theoretical concept of a sole under the disturbed belt east of the Lewis overthrust is not supported by any recorded observations.

FOLDS OF THE DISTURBED ZONE

Many details are lacking in regard to the folds in the plains. Eugene Stebinger (1917) shows that they have a fairly uniform northwesterly strike. Most are tightly compressed and overturned toward the northeast, but some minor folds appear to be overturned toward the southwest. As already noted, faults are closely spaced, some are close to the crests of overturned anticlines.

Within the mountains the folds of the disturbed zone in the Flathead region are better exposed. In the southwestern part of the Heart Butte quadrangle, thrusts are so closely spaced that the folds cannot be shown satisfactorily on plate 2. Here the folds trend approximately parallel to the thrusts, about N. 25°-30° W., and are tightly compressed and overturned towards the northeast (fig. 22). Nearly all the anticlinal crests are broken by thrusts, or by incipient thrusts, many of which are too small to map. Within this part of the region almost every mountain peak contributes evidence that the steep thrusts represent failure at and near anticlinal crests (fig. 23). Whereas all the major folds have the features just indicated, there are irregularities of many kinds. Among these may be cited the peculiar folds illustrated in figure 24, which are overturned toward the west. Some of the fault slivers, notably those in incompetent beds of the Ellis group (fig. 18) are crumpled and fractured without apparent system. These are not directly associated with anticlinal crests.

FIGURE 22.—View south from below Bennie Hill Lookout, Heart Butte quadrangle, Flathead region. Shows closely folded and overturned Paleozoic strata, mostly Hannah limestone.

FIGURE 23.—Crooked Mountain, Heart Butte quadrangle, Flathead region. Shows an anticline in Hannan limestone overturned toward the northeast and broken by small thrusts. The principal thrust is concealed in the talus at the foot of the cliffs below the peak. Retouched to emphasize structure.

FIGURE 24.—Irregular folds in Cambrian limestone south of Swift Reservoir, Heart Butte quadrangle, Flathead region. Most of the folds are pinched at the crests and open in the troughs. Also, they are overturned to the west, opposite to the regional structure.

In the relatively broad area of Lower Cretaceous rocks in the southwest corner of the Marias Pass quadrangle and in areas farther north, underlain mainly by Cretaceous and Mississippian rocks, the folds are more gentle as can be see from structure section B—B', and parts of the other sections on plate 2. In these areas some anticlinal and synclinal axes are plotted on plate 2, but other, less persistent folds are present. Even in these areas most of the folds tend to be assymetric, with the steeper flanks on the northeast. Strikes are commonly between N. 45° W. and N. 55° W.

STRUCTURAL FEATURES IN THE OVERTHRUST BLOCK

The part of the block of rock above the Lewis overthrust that extends from the exposed trace of the thrust westward as far as the eastern side of the northern part of the valley of the Flathead River is composed mainly of strata of the Belt series. The small masses of Paleozoic rocks included in it in the Flathead region do not materially alter its character as a mass of competent strata that would yield reluctantly to deformative stresses. In harmony with this, the first impression one gets of the block is that it consists of flat to gently inclined beds that have successfully resisted the successive attacks of diastrophic forces. Closer inspection shows that this impression is only partly correct—the whole block has been flexed, and parts of it have been buckled and fractured.

In Glacier National Park the most readily seen and persistent fold is the shallow syncline whose axis extends from near Boulder Glacier through Flattop Mountain between Clements Mountain and Mount Cannon to near Tinkham Mountain, a distance of almost 40 miles. Plate 1 shows the axis, and figure 27 illustrates the character of the gentle fold. The average trend is N. 40° W., but in the southern part the syncline is both more curved and less clearly defined than it is in the general vicinity of Flattop Mountain, where it is 10 miles wide, a fact that is illustrated in structure section C—C', plate 1. The syncline is bordered on both sides by less uniform groups of folds. Each group constitutes a poorly defined compound anticline or anticlinorium bounding the shallow syncline, as can be seen from the structure sections on plate 1 and from the traces of axial planes of the component folds plotted on that map.

East of the major syncline the folds in the compound anticline are open. No extensive zones of crumpled or overturned beds are recorded outside of the locally thick zones of fracture constituting the Lewis overthrust itself. Local crinkles and minor thrusts are present; for example, the field notes of Corbett and Stebinger refer to two probable thrust faults in the upper part of the drainage basin of Belly River. One of these is on the ridge south of Helen Lake, but has not been mapped. The other is on the point of the ridge west of Mokowanis Lake and is shown on plate 1. Both are in Siyeh limestone and about 5 miles southwest of the trace of the Lewis overthrust. In addition, two small thrusts, one of which is shown on plate 1, are visible in cliffs near Ptarmigan Creek. Undoubtedly other minor thrusts than those recorded are present here and there.

The most severely deformed zones in the rocks of the park are west of the major syncline. Crumpled and locally overturned beds, with minor overthrusts in a few spots are visible near Brown Pass. The high parts of the Livingstone Range north and south of Brown Pass have not received enough study to determine the extent of the deformation. Farther south, complex folds of greater magnitude are present in numerous places, notably on Nahsukin Mountain, Rogers Peak, McPartland Mountain and mountains immediately southwest of it, and on and near Mount Thompson. Closely crumpled but not overturned beds are conspicuous near the head of Coal Creek south of Marthas Basin. The rocks along the lower part of Muir Creek are intricately contorted on a small scale. The two adjacent ridges that include Rogers Peak and McPartland Mountain contain folds of especially great magnitude and complexity. The extent of deformation here is fully as great as in most localities in the crumpled rocks beneath the Lewis thrust. Figure 28, taken from the northwest and at fairly close range, shows details of the folds on McPartland Mountain. Figure 26, taken from Sperry Glacier, shows a cross section of most of the crumpled zone from the edge of the major synclines exposed on Heavens Peak westward. Stanton Mountain, just west of the area shown in figure 26, is included in the sharply folded zone. In Mount Thompson, 13 miles southeast of McPartland Mountain, the rocks are similarly intensely deformed, as plate 1 indicates. Figure 25 shows some of the tightly folded rocks in this locality. The Appekunny argillite on the slopes of Mount Thompson has responded to the intense pressures by development of more slatey cleavage than it displays in most localities. Two Medicine Pass is another locality where crumpled beds are conspicuous.

FIGURE 25.—Contorted Grinnell argillite on the north wall of the cirque on the south slope of Mount Thompson. Glacier National Park. Photograph by Eugene Stebinger.

FIGURE 26.—View northwest from Sperry Glacier showing part of the intensely folded rocks west of the major syncline. The view extends from the right edge of Heavens Peak almost as far as Stanton Mountain. (Retouched to emphasize structure.)

FIGURE 27.—View southeast from near Fifty Mountain Camp, Glacier National Park, showing the gentle syncline that dominates the structure in the central part of the park. This major fold is visible from few viewpoints. The Missoula group floors the top of Flattop Mountain in the foreground and forms most of the pinnacles on the skyline. Most of the rest of the rock in the view is Siyeh limestone. The flat structure of Flattop Mountain shows in the foreground, contrasting sharply with the glaciated pinnacles and cliffs south of it. Flattop Mountain is the most conspicuous remnant of an old, high erosional surface in the region. Photograph by Richard Rezak. (click on image for a PDF version)

FIGURE 28.—Overturned Grinnell argillite on McPartland Mountain, Glacier National Park, viewed from the northwest. Photograph by R. H. Chapman. (click on image for a PDF version)

The observations summarized above show that from Nahsukin Mountain southeast to Two Medicine Pass irregular and somewhat discontinuous zones of intense deformation mark a disturbed and uplifted belt on the western flank of the central syncline. This belt is about 3 miles wide and fully 35 miles long. It may continue northward more than 10 miles farther into Canada. In most places the folds are overturned toward the northeast, but here and there the opposite is true. In some places, such as Nahsukin Mountain, the contorted beds are so unsystematic as to give no clue regarding the directing stress.

While individual folds and contorted zones are discontinuous, the compound anticlines or anticlinoria on both flanks of the central syncline extend from the Canadian boundary to the southern part of Glacier Park. The compound anticline on the west crosses from the Livingston Range to the Lewis Range near the upper end of Lake MacDonald and merges into a group of folds in the Flathead region. In the southern part of the park, these folds are crossed by faults, but there and farther south they do not have the fractured crests so plentiful in the disturbed zone east of the trace of the Lewis overthrust. In Glacier Park this compound anticline is roughly 15 miles back of the outcrop of the Lewis overthrust, but the corresponding folds in the Flathead region constitute the first major structural feature to the west of that great fault. This comes about as a result of the sharp deflection in the trace of the overthrust near Marias Pass. The major syncline and the compound anticline east of it do not continue as far south as that pass.

Within the Flathead region the rocks west of the outcrop of the Lewis thrust are flexed into open folds that locally are overturned toward the northeast. An example is illustrated in structure section B—B (pl. 2) west of Battery Mountain in the southern part of the Nyack quadrangle. Here many of the beds within a narrow zone are so tightly crumpled as to suggest an incipient high-angle thrust of south. westerly dip. No fault displacement has been detected, however.

In the northern part of the Flathead Range, especially along the range crest from Mount Liebig northward to Great Bear Mountain, the Siyeh limestone is more intricately deformed than in any other part of the Flathead region. This part of the range has not been mapped, but the sharply contorted beds in it are plainly visible from distant vantage points.

In the general vicinity of Mount Bradley, Hematite Peak, and Twin Mountain, east of Long Creek, the rocks are sharply and irregularly folded. One short anticline is shown on plate 2, but in most places the folded rocks do not seem to fall into any systematic structural pattern. The disordered folds in this zone may account for the apparently abrupt ending of the fault mapped along Bergsicker Creek.

STEEP MASTER FAULTS

Long faults of large displacement are major features of the structure along the valleys of the principal branches of the Flathead River both in Glacier National Park and in the Flathead region. The character of these faults is obscured by the fact that long segments of them are buried under alluvium and other kinds of unconsolidated material. Some students regard the faults as thrusts; others, as normal faults. Conclusive evidence is not available, but the concept of normal displacement seems the most logical one. Similar faults are outstanding features of the geology farther to the west; so the problem is of regional significance. Numerous opinions have been expressed; but, in many localities, the question as to character and direction of displacement remains open. Existing data on the subject are reviewed below. As a reflection of doubt, the conventional symbols indicating relative displacement have been omitted, in most places, from plates 1 and 2, and many of the faults are represented noncommittally on the structure sections as vertical. On plate 4, which is diagrammatic, the fault along the upper reaches of the Flathead River is represented as normal. This conforms with the ideas gained during the present investigation, but is not to be regarded as proved.

Both in the two regions here reported on and in neighboring areas in western Montana and in Canada, direct observations of the faults are possible in few places. Even where the faults cut slopes in bedrock, few actual fault surfaces, susceptible of measurement, are known. In consequence, the mapping had to be guided in part by float and by the interpretation of topographic forms. In most places the result has been that the fault trace as mapped accords with the concept of a steep fault dipping valleyward. As the pre-Tertiary rocks under the valleys are, where known, stratigraphically higher than those in the mountains beyond the fault zones, the valleys would be grabenlike and bounded by normal faults. Unfortunately, the data on which the fault lines are based are too scanty to permit the shapes of the mapped fault lines to be regarded as conclusive evidence regarding the actual shapes of the faults. Similar difficulties in interpretation have confronted all those who have studied the steep faults in and west of the two regions here described. Consequently, in spite of much discussion in the literature, positive evidence remains scanty. In a later section of the present report, which concerns interpretation of the structure, pertinent papers are cited, including a number that relate to parts of western Montana and Canada. That section may be anticipated here by the statement that most of the geologists who have worked in Glacier National Park and contiguous areas regard the steep faults along major valleys as of normal displacement. However, some of those who have worked farther west think the comparable faults there are thrusts, and most of the more recent workers in Canada favor that view. Probably a single and simple explanation will not hold for all of the faults. Available information on the faults under discussion in and close to the Glacier National Park and Flathead regions is summarized below.

The longest and apparently the most dominant of the steep faults mapped during the present investigation trend N. 20° W. to N. 45° W., with local variations beyond these extremes. One of the largest extends from beyond the international boundary north of Kintla Lake southeastward past the southern border of plate 1. This fault borders the valley of the Flathead River 6 or 8 miles northeast of that stream, continues in the hills northeast of the Middle Fork of the Flathead, and crosses Bear Creek near the railroad station of Blacktail. Throughout this long stretch the pre-Cenozoic rocks southwest of the fault belong to units several thousand feet higher stratigraphically than those northeast of the fault. In most places the fault plane is concealed beneath deposits of Cenozoic age. In the few places where not so buried the shape of the fault trace as mapped indicates a steep southwesterly dip. Even in these places exposures are not good. If true, both the dip and the relative displacement fit the concept of a major normal fault, but so much of the fault trace is concealed that the possibility of a very steep reverse fault of northeasterly dip cannot be eliminated. Some support to the hypothesis of a reverse fault is afforded by intricately crumpled beds exposed between the fault and the mouth of Muir Creek and by small steep faults, in part clearly reverse, in railroad cuts about halfway between the stations of Red Eagle and Hidden Lake along the Middle Fork below the mouth of Coal Creek. One such exposure is shown in figure 29. All the faults in the railroad cut are steep, curved, and appear to have displacements of only a few feet. Some, as figure 29 indicates, are thrusts; others are indeterminate.

FIGURE 29.—A railroad rut along the Middle Fork of the Flathead below the mouth of Coal Creek showing exposure of minor fractures.

In the valleys of the northern part of the Flathead River and of the Middle Fork of the Flathead, the Tertiary rocks are disturbed. Figure 19 shows some of them tilted about 30° NW. In other localities, such as along the lower reaches of Lincoln and Coal Creeks, they are locally contorted. C. E. Erdmann (1944, p. 62-67, 112-114; 1947, p. 134) noted inclined Tertiary strata in several places along the Flathead and its Middle and South Forks. The beds in the bank of the Flathead near the mouth of Kintla Creek that, as noted in the stratigraphic descriptions earlier in this report, have yielded Oligocene fossils are also tilted. Thus the Tertiary and later beds that underlie the valleys bordered by the steep master faults are extensively deformed. The movements that affected the valley fill of Tertiary and later age (the old alluvium of pls. 1 and 2) must have been related to the bordering faults. Hence, disturbances along these faults have occurred far more recently than the date of the Lewis overthrust. Erdmann (1947, p. 139-141, 188-190), who is among those who think of the long faults as thrusts, calls attention to two normal faults bounding a horst cut into by the Flathead River near the mouth of the Canyon Creek in sec. 23, T. 32 N., R. 20 W.

Prospecting for oil near the international boundary (Link, 1932, p. 789; Erdmann, 1947, p. 210-212) has yielded little information of value in the present connection. Small quantities of oil have been reported from wells and seepage in localities short distances east of the probable position of the major fault on the northeast side of the valley of the upper Flathead River. The nearest outcrops consist of Appekunny argillite. Presumably the oil rose along fault fractures from a source in concealed rocks of post-Belt age, but it is impossible to determine from available data whether these fractures belong to an east-dipping thrust that passes beneath the exposed rocks of Precambrian age or to branches of a normal fault. One test well explored the ground near a seep in the Cenozoic deposits close to the river below the mouth of Kintla Creek. This well was west of the fault but yielded no data as to the pre-Cenozoic rocks there.

With respect to the long fault east of the northern part of the Flathead River and near the western border of Glacier National Park, the field notes of the geologists in M. R. Campbell's parties show that they regarded the fault as normal. They give no specific details in support of the assumption, which was doubtless based on such general features as topographic forms and the fact that rocks in the hills west of the river are stratigraphically higher than those in the mountains east of the river. Most of the mountains west of the river are beyond the limits of plate 1, but they are underlain largely by beds of the Missoula group, while the Ravalli group crops out east of the river valley (Ross, Andrews, and Witkind, 1955). J. T. Pardee (1950, p. 397-398.) has the same opinion and says that at the international boundary the valley has a graben structure. He thinks the fault on the west side of the graben, although less well defined than that on the east, extends south almost as far as Lake McDonald. This fault was not studied during the present investigation. Most of it is west of the area represented on plate 1. Perhaps the fault that is doubtfully indicated east of the Apgar Mountains is a segment or branch of that which Pardee speaks of as on the west side of the graben. Plate 1 shows that the fault inferred to exist east of the Apgar Mountains is sketched mainly on the basis of topographic forms, and its character is indeterminate. As indicated below, it seems to be a northward continuation of the big fault southwest of the Flathead Range.

J. L. Dyson's (1949a, p. 17 and 19) structure section shows a normal fault on the east side of the valley of the Flathead. His text refers to this structure noncommittally as of "the high-angle variety," but he adds that the faults bounding the Apgar Mountains and Belton Hills are normal. The fault east of the Belton Hills is, as can be seen from plate 1 of the present report, a segment of the fault along the northeast side of the valleys of the Flathead River and the Middle Fork of the Flathead.

In the general vicinity of Blacktail, the long steep fault described above is in the same topographic depression as the Lewis overthrust, and the two fractures must lie very close together. To the south they diverge. The steep fault appears to die out, but the Lewis overthrust continues southeast past the southern border of the Flathead region. South of Blacktail, in the Flathead region, the steep fault has been traced on plate 2, with some uncertainty, as far as the confluence of Lynx Creek with Twentyfivemile Creek.

The above summary shows that most geologists who have worked near the western border of Glacier National Park favor the concept that the valley of the northern part of the Flathead River is a graben bounded by normal faults. The Canadian geologists who have worked in southwestern Alberta are more inclined to regard the faults as thrusts. In an early summary, Link (1932) said that "* * * the generally accepted structural interpretation * * *" of the valley of the Flathead was that of an "* * * inverted wedge or modified ramp * * *" bounded on either side by thrusts dipping valleyward. In later papers he (1935, p. 1464-1465) gives different explanations of the structure along the Flathead but continues to favor thrusts of some kind rather than normal faults. This paper and that of G. S. Hume (1933), which expresses similar ideas, are further discussed in the interpretive part of the present report. Pierre de Bethune (1936), whose fieldwork was about 25 miles north of the international boundary, also favors the concept of thrusting for structural features bordering the valley of the Flathead.

Another great fault of northwesterly trend extends from McGees Meadow past the lower end of Lake McDonald, down the upper part of the valley of Emery Creek, between Firefighter Mountain and the Flathead Range, and thence along the northeast side of the valley of the South Fork of the Flathead River past the southern border of plate 2. The general relationships of this fault, as shown on the maps, are similar to those of the fault just described. That is, the dip appears to be southwest, and the rocks now west of the fault are higher stratigraphically than those east of it. If the dip was definitely known, these relationships would prove that the fault was normal and of great vertical displacement with the down-dropped block on the South Fork side. However, the dip has not been thus established because the fault surface itself is hidden under Cenozoic materials and vegetation. No contorted rocks or minor thrusts like those close to the fault described above have been recognized near this one, with the exception of the thrust near the southern border of the Flathead region southeast of the lower Twin Creek. Pardee (1950, p. 398-399, pl. 1, sec. B—B') agrees that the fault northwest of the South Fork is normal, but he does not draw it the full length of the valley. Farther north a fault trending west of north is shown. The limestone west of the upper reaches of Lower Twin Creek is shattered in such a way as to support strongly the concept of faulting, but that east of the stream appears to be conformable with the argillite above. Both masses are represented as belonging to the Siyeh limestone, but this is open to some question. They might be limestone bodies low in the Missoula group.

An inferred fault is mapped from north of Abbott Ridge along the southwest side of the valley of the South Fork of the Flathead River past the southern border of plate 2. This fault is inferred in part from the topography, in part from the presence of Tertiary deposits east of Abbott Ridge and in the depression north of that ridge through which the road passes, and in part from the presence of green argillite at the place where the road crosses the South Fork about a mile and a half northwest of Coalbank. If the green argillite is correctly interpreted as part of the basal unit, greenish calcareous argillite of the Missoula group, some such fault as that sketched is required. C. E. Erdmann in his study of the Hungry Horse dam site (Erdmann, 1944, p. 79-80) found two shear zones so close to the position of the inferred long fault as to add to the evidence of fault movement in the locality. These zones are on either side of the position of the inferred fault as plotted on plates 1 and 2. He reports that the northwestern of the two is well exposed in a roadcut "* * * near the north center of sec. 36, T. 30 N., R. 19 W. It begins just below the top of the Siyeh limestone downward through a stratigraphic range of at least 1,400 feet." His description indicates that much disturbance has taken place in this zone. The limestone is much jointed and sheared and contains gouge seams up to 6 feet thick, some of which includes a jumble of limestone blocks as much as 5 by 10 feet in exposed dimensions. He interprets the zone as a result of thrust from the southwest but evidently regards the aggregate displacement as small because he shows no fault or displacement along the line of the shear zone (Erdmann, 1944, pls. 10, 11). The second shear zone crosses Hungry Horse Creek 1,300 feet west and 2,150 feet north of the southeast corner of sec. 30, T. 30 N., R. 18 W. Erdmann says it apparently separates an area of low dips from a more thoroughly compressed zone east of it. Some movement and brecciation have taken place and small drag folds nearby "* * * indicate strong compression from the west."

A concealed fault of northwesterly trend is shown in the southwest corner of plate 2. This is a segment of a much longer fault mapped by Clapp (1932, pl. 1) which bounds the Swan Range on the southwest. Just as in the case of the faults on the northeast sides of the valleys of the Middle and South Forks of the Flathead, topographic features are compatible with the idea that it is normal, with a southwesterly dip. The sharply truncated ends of many of the spurs from the range crest might readily be supposed to be triangular fault facets dipping valleyward. This feature, which might otherwise be regarded as conclusive proof of a normal fault of southwest dip, is negated by Davis (1921, p. 87-89). He pointed out that a large glacier once occupied the Flathead Valley and ground off the spur ends of the Swan Range. Davis does not oppose the idea of a fault that outlined the Swan Range but implies that such features as fault facets would have been obscured or obliterated by the action of the glacial ice. Coarse glacial debris is mingled with the hillwash on the southwest flank of the range. Pardee (1950, p. 395-396, pl. 1) agrees in part with Davis but thinks fault facets rise above the level of the glacially truncated spurs in places. Pardee maps and describes the fault bordering the Swan Range as a normal fault of westward dip. He cites Clapp's concept (1932, p. 24-27, pl. 1) of a thrust fault of eastward dip in this position and calls attention to Clapp's suggestion that later normal faulting may have occurred. However, Clapp thought of the normal faults as minor features and of the steep thrust (or reverse fault) as dominant, whereas Pardee regarded normal faulting as dominant. It is unlikely that a normal fault of westerly dip would coincide in position and strike with an earlier reverse fault of opposite inclination. Along the part of the southwest face of the Swan Range seen during the present investigation, glacial deposits mantle some slopes to heights above the truncated spur ends. Much of the hillwash mapped on plate 2 includes glacial debris, and the hillwash extends to between 500 and 1500 feet above Flathead Valley. It would be difficult to map the glacial material separately or to determine how much of it came from a glacier in Flathead Valley. Some, of course, came from glaciers originating high in the Swan Range.

One bit of evidence opposed to the view that the fault on the southwest flank of the Swan Range is normal has been found. The north end of this range is at Bad Rock Canyon, a rock-cut gorge west of the region mapped during the present investigation but which was visited in the course of the work. Roadcuts in Bad Rock Canyon show fractured folds in Grinnell argillite that are overturned toward the west or northwest. Such features might have been formed in connection with a steep reverse fault of northeast dip along the southwest flank of the Swan Range. Taken alone, they are insufficient to prove such a fault or to disprove the assumption of a normal fault of opposite dip in this position.

Because the evidence in regard to the long, steep fault summarized above is inconclusive, data in regard to other faults bordering major valleys in the same general region were sought. Much has been written, but the difficulties encountered in assembling positive evidence during the present investigation are shared by all of the investigators. The principal studies have been along the Rocky Mountain Trench, a conspicuous topographic feature in Canada and western Montana some distance west of Flathead Valley (pl. 3). Early workers (Daly, 1912, p. 25-26, 137-139, 600; Schofield, 1920, p. 73-81) regarded the faults as normal and the trench as grabenlike. Shepard (1922, p. 16-139) originally regarded most or all of the faults in the vicinity of the Rocky Mountain Trench as steep thrusts dipping away from the topographic depression on both sides. However, nearly all of his drawings show the faults dipping west, or, more strictly, southwest. In his first paper, Shepard (1922) concludes with the remark that the Rocky Mountain Trench is less of a unit than Daly and Schofield thought. It was produced "* * * partly by normal erosion, partly by erosion along lines of weakness resulting from intersection of numerous faults, and partly by the escarpment of a fault, inferred to be a thrust." In a later paper, (Shepard, 1926) concerned largely with an area 120 miles north of the international boundary he regards the trench as an eroded horst, bounded on the east by a west-dipping thrust, rather than a normal fault. In this paper he discounts Daly's idea of a graben.

Several other geologists who discuss the structure of the region from the trench eastward as far as Glacier National Park think that some of the faults are steep thrusts with easterly dips. Among these Wilson (cited by Chamberlain, 1925, p. 759-760 and Flint, 1924, p. 411-415) reports small east-dipping faults on the west flank of the Mission Range (pl. 3) and between there and the Continental Divide in the southern extension of the Lewis Range. Clapp's small-scale generalized map (1932, p. 24, pl. 1) shows steep, east-dipping thrusts at the borders of several ranges from longitude 114°30' to the western border of Glacier National Park, although his abbreviated text cites little evidence in support of his interpretation. Pardee (1950, p. 393-395, pl. 1) maps the trench in Montana as bordered by normal faults. Most of those cited above who advocate east-dipping thrusts do so primarily on theoretical grounds.

C. S. Evans 1933, p. 145-170 A II) has summarized published information on the Rocky Mountain Trench in connection with his study of a part of the trench about 120 miles north of the international boundary. At one locality in his area where opportunities for observation are especially good, Evans says that thrust faults and overturned folds in the mountains dip away from the trench on both sides. His interpretation is that the structures result predominantly from thrust toward the northeast. The east-dipping structural features east of the trench would thus be underthrusts. Evans thinks there may have been two periods of deformation that resulted in a much-sheared zone in which the rocks yielded readily to erosion. At an early stage the trench may have been in part a structural depression, but the present form results largely from erosion in rock weakened by deformation, including fracture. The difference between Evans' conclusions and the earlier ones of F. P. Shepard (1922) is in interpretation rather than observation. Both speak of thrusting and overturning toward the trench from both sides. Evans, however, rejects Shepard's suggestion, made after a second study (Shepard, 1926, p. 640), that the trench was eroded along a horst. D. R. Derry (1950a, b, p. 42-43) in a summary issued in connection with the tectonic map of Canada favors the concept of a steep west-dipping reverse fault zones without any east-dipping faults. The observations cited above are of interest because they give details not available for much of the Rocky Mountain Trench or for areas between there and Glacier National Park, but it must be remembered that the area involved is far west and north of the park.

T. A. Link (1935, p. 1464-1466; 1949, p. 1475-1501) and A. J. Goodman (1951) discuss structure in the valley of the Flathead River and in the foothills in nearby parts of Canada. They make use of the extensive information on the region that has accumulated in the course of exploration for oil and agree that east-dipping faults and folds overturned to the west do occur in the Canadian foothills. This fact lends some support to the idea that similar structural features may be present farther west, where less information is available. As will be shown below, Link and Goodman differ in their interpretations of the east-dipping structural features, leaving doubt as to the significance to be attached to them in the present connection.

In summary, review of published reports, some of which include summaries of papers not here specifically cited, shows that most of the more detailed studies so far accomplished that are pertinent to the present discussion have been done in Canada. Thrusts in and bordering the individual ranges (mostly west-dipping) are favored by most of the writers, but enough of them advocate normal faults, so that the possibility cannot be ignored, even for the specific areas described. None of the detailed work is in and most of it is distant from the regions described in the present report.

STEEP MINOR FAULTS

Within the two regions discussed in the present report there are numerous short, steep faults not directly connected with the long ones described above. Such data as are at hand in regard to these are summarized below.

Plate 1 shows two short faults of northwesterly trend. One of these is southwest of Upper Kintla Lake close to the international boundary, and the other is south of Marthas Basin in the southern part of the park. Both appear to be normal faults and have apparent downthrows on the northeast side, a relation opposite to that on the long faults of similar trend. The fault near Marthas Basin is within a zone nearly half a mile wide in which the Grinnell argillite is crinkled into numerous, closely spaced steep folds. There seems, however, to be no direct relation between the fault and the folds.

Steep faults of northeasterly trend are shown on plates 1 and 2, and others may be concealed in valleys of similar trend. Two inferred faults of this trend are shown in the southwest part of plate 1. While these are buried under Cenozoic deposits and no direct evidence of either is known, they appear to be required by the offsets in the long faults of northwesterly trend, especially the one that follows the northeast side of the valley of the South Fork of the Flathead and may die out in McGees Meadow. However, the position of this latter fault is sketched mainly on the basis of the topography. Detailed work may show that it curves more than is at present appreciated or modify the position of the fault trace in other respects.

In the northern tier of sections in unsurveyed T. 28 N., R. 18 W., the rocks are disturbed and show such abrupt variations in attitude that a sharp structural break must be present. This is represented on plates 1 and 2 by a fault striking a little north of east, north of Clayton Creek. Available data are insufficient to determine either the displacement or the exact strike of this fracture, which may be merely a branch of the longer one in the valley of Clayton Creek. This longer fault also is an inferred fault about which little is known. The distribution of the mapped masses of the basal part of the Missoula group near Clayton Creek requires some such a fracture. Further, rocks in outcrops that project through the glacial debris in the headwater basin of Noisy Creek are fractured and have variable attitudes. This zone of disturbance is in line with the inferred fault along Clayton Creek.

Several short faults of easterly and northeasterly trends in the plains in the Heart Butte quadrangle are plotted on plate 2 on the basis of observations by Eugene Stebinger (1916). These appear to result from minor adjustments that broke some of the longer thrust faults of northwesterly trend.

The principal fault of northeasterly trend recorded in the mountains of the Flathead region east or northeast of the Middle Fork of the Flathead is that along Ole Creek in Glacier Park. This fault is mostly concealed under alluvium but has a branch that cuts the Appekunny argillite near Sheep Mountain. The fault along Ole Creek seems required by the lack of accord in the geology on the two sides of the valley. This is more evident when irregularities in the attitude of the bedding are observed in the field than it can be by inspection of the map alone. The displacement must be small, as is proved by the fact that, as the map shows, none exists along the lower reaches of Ole Creek. A zone of fracture is exposed north of the narrow outcrop of Altyn limestone nearly opposite the place where Debris Creek joins Ole Creek. The same thing is true of the fault sketched in dashed line south of Flotilla Lake in the southern part of the Marias Pass quadrangle (pl. 3). Inspection from vantage points shows that the beds along the line sketched are disturbed and fractured, but large scale studies would be required to determine the exact position and character of the fault. Farther west a fault is mapped along Bergsicker Creek. This one crosses the Flathead Range and is lost to view under the Cenozoic deposits in the valley of the South Fork of the Flathead. The lack of accord in the geology on the two sides of this fault is obvious from a glance at plate 2. Possibly the fault crosses the valley of the South Fork and extends up Sullivan Creek, but evidence to support this idea has not been found. Disturbance in the beds on the shoulder of Mount Baptiste is evident from a distance, but it has not been traced otherwise. Two short faults that strike more nearly north than those just mentioned are mapped at the head of Aeneas Creek.

The faults described in the preceding paragraph are the only ones of their trend and character for which evidence has been found in bedrock exposures. Long, straight stream valleys trend approximately parallel to these faults. It is possible that some or all such valleys follow lines of faulting or of shearing that facilitated erosion but are not so exposed as to have been detected.

The above summary calls attention to various faults in both regions that have been detected and mapped with different degrees of precision. It is probable that many fractures have escaped detection. Many of the larger tributary streams that flow northeast or southwest are approximately parallel to each other and normal to some of the master streams and to faults that parallel those streams. A few faults along the valleys of such tributaries have been mapped, but studies to date have not yielded enough evidence to warrant sketching faults along most of them. The drainage pattern is so regular as to suggest structural control. The suggestion made by Billings (1938, p. 263) that major valleys are developed along the crests of folds in the plane of the Lewis overthrust may have some validity, but it is supported by little specific evidence so far. Some of the valleys may have been alined along faults not yet detected, and others may be along zones of weakness in the rocks that correspond to fractures or shear zones without significant displacement. Large stretches of the valley sides are mantled by soil, talus, and glacial and other debris that permit few outcrops to be found. Also, particularly in the Flathead region, the forest cover is so dense on some slopes and valley bottoms as to interfere seriously with observation. As noted below, the concept that some of the tributary valleys are carved along zones of weakness and fracture in the underlying rocks would help to explain topographic features whose origin is otherwise puzzling.