GLACIER Origin of the Scenic Features of the Glacier National Park

EFFECT OF GEOLOGIC STRUCTURE AND HARDNESS OF THE ROCKS.

As the mountains in the Glacier National Park have been carved by streams and glaciers from an uplifted land mass, it follows that the forms which they have assumed as well as their height above drainage level depend largely on the kind of rock and its ability to withstand the action of the elements.

In a general way the mountain range is composed of harder, more resistant rocks than either the plains on the east or the valley of Flathead River on the west. In fact, the very existence of a range of mountains here is due to that difference in hardness, for if the rocks of the mountains had been as soft and as easily eroded as the rocks of the lowlands, the mountains long ago would have been cut down by the streams until they merged with the more nearly level surfaces on either side, and if such had been the case there would have been no mountain scenery and no reason for the creation here of a national park.

In Glacier Park the mountains are composed largely of sandstone, shale, and limestone, which originally were much like the rocks in the surrounding regions, but the mountain rocks are very old; in fact, they are some of the oldest known sedimentary rocks on the globe, and since the time when they were laid down as mud and sand in the ancient ocean that covered this part of the globe they have been hardened by pressure and by the formation of minerals in the minute spaces between the grains of sand and mud of the original rocks until to-day the sandstones are turned to quartzites, the limestones hardened to a very resistant rock, and the shale into argillite. Of the rocks forming the mountains the limestones are decidedly the most massive and resistant to erosion and they form generally the highest peaks.

MAP OF GLACIER NATIONAL PARK. (click on image for a PDF version)

The distribution of the highest summits, as shown on the map, is due primarily to the geologic structure, and secondarily to the hard rocks. During the time of uplift and overthrust, as explained on another page, the originally horizontal beds of rock composing the mountains were pushed forward toward the northeast with such resistless power that they buckled, forming a broad, shallow, downward fold or syncline having a breadth nearly equal to that of the mountain mass. Owing to this synclinal structure the hardest (limestone) rocks outcrop on the two edges of the fold, and by their resistance to erosion have formed the crests of the Lewis and Livingston Ranges. This structure also is a contributing condition in the formation of the topographic basin of Flattop Mountain. The situation here may be explained by figure 12 which represents the rocks as they would appear if a deep trench were cut across Flattop Mountain southwestward from Ahern Pass and the observer were looking toward the northwest.

The bed of heavy limestone, which in general makes the crest of the mountain in Ahern Pass, and also in Swiftcurrent Pass to the south, dips to the southwest and passes under Flattop Mountain, reappearing in the summit of the range to the west. The red argillite lying above this heavy bed is somewhat softer, and in the bottom of this trough of rock or syncline has been partly removed, forming the topographic basin of Flattop Mountain. Another factor that has been instrumental in the formation and preservation of this basin and of the mountain wall which surrounds it is a bed of lava. This is a very hard rock and the traveler is painfully conscious of its presence wherever it is crossed by one of the trails. It forms a great black band near the base of the red argillite noted above and is well shown in Swiftcurrent Pass and in Granite Park. Although it is known to geologists as a diabase rock, it was called by the early travelers and prospectors "granite," and from the great ledges and bowlders strewn about the surface came the name "Granite Park." It is crossed by the trail from Swiftcurrent Pass to Ahern Pass, by the Mineral Creek trail near the head of that stream, and on Flattop Mountain by the trail to Waterton Lake.

As this rock was originally poured out as molten lava it might seem to be in conflict with the statement, made on a preceding page, that none of the mountains of the park are of volcanic origin. Both statements are strictly correct, as the outpouring of this mass of lava had absolutely no relation to the mountain uplift. The sheet of lava covered an area originally at least as large as the park, but in many parts it has been cut away by the streams, leaving no trace of its former existence. The reason why geologists are so certain that the lava had no connection with the formation of the mountains is that it was poured out on the surface of a flat land or on the bed of a lake or the ocean in which the argillite was being deposited and at the same time as the deposition of that material. This occurred countless ages before there were any mountains in this region, and since that time the sheet of lava has been bent and folded with the limestone and argillite. The evidence that it was originally a surface flow is to be found in the vesicular or porous texture, the ropy structure, and flow lines so well shown in Granite Park and on Flattop.

FIG. 12.—SKETCH CROSS SECTION THROUGH FLATTOP AND CHIEF MOUNTAINS SHOWING BROAD SYNCLINE UNDER FLATTOP AND THE FAULT UNDER CHIEF MOUNTAIN. Vertical scale greatly exaggerated. (click on image for a PDF version)

In Gould Mountain, Wilbur Mountain, and at various other places in the park the traveler will also notice another black band, which may be confused with the lava bed mentioned above, but which has a very different history. Although originally it came from the earth's interior it was not poured out on the surface like the sheet of lava, but while molten was forced into the rocks from below wherever it could find a place. Generally it formed a layer between the other beds of rock, but in places where it is now exposed, as on the steep part of the Swiftcurrent trail, it cuts vertically through the beds for several hundred feet, forming a dike. This rock is called a diorite, and it is hard, like the lava, and where present at the surface renders trail making difficult. The intrusion of this mass occurred long before the existence of the mountains, and hence, like the lava, had nothing to do with their formation.

East of the Lewis Range the spurs composed of green argillite are somewhat lower than the main crest, for that rock is softer and more easily eroded than the limestone which gives to the crest its altitude and its ruggedness. As the green argillite forms the outer points of all these spurs, it gives rise to a sort of high terrace along the mountain front, but below the terrace and separating it from the plains there is in general a very steep slope, in places a precipitous wall of massive limestone which is one of the most pronounced topographic features of the park. It is well shown on Swiftcurrent Creek and in Flattop Mountain¹ lying between Boulder Creek and Upper St. Mary Lake (figure 13). It is also conspicuous in Yellow and Chief Mountains (title-page) farther north. The limestone forms such a precipitous front because it is underlain by very soft shale or sandstone that weathers much more rapidly than the limestone, which consequently is undermined and tends to break off in large blocks, leaving vertical or nearly vertical faces. The soft shale and sandstone underlying the hard limestone do not belong to the group of rocks forming the mountains, but are the same strata that form the plains and are much younger geologically than the mountain series. The abnormal relationship of the older rocks lying upon the younger rocks is due to a great thrust fault, or break in the strata, which has completely changed the order of succession of the rocks and has caused much of the beauty of the mountain front. The mode of formation of this great fault or break in the strata is illustrated by figure 14, which represents the rock strata as they would have appeared in a deep east-west trench provided the spectator could have watched long enough, possibly thousands of years, to have seen the movement take place. The spectator is supposed to be looking north, and in section A he sees the edges of the rock strata lying flat as they were originally deposited in bodies of water, some in the ocean and some in shallow lagoons and lakes.

---

¹This must not be confused with Flattop Mountain west of the Continental Divide.

---

FIG. 13.—FLATTOP MOUNTAIN ON EAST SIDE OF THE RANGE, AS SEEN FROM THE RIDGE NORTH OF SWIFTCURRENT VALLEY. The fault is located at the base of the cliffs. Photograph by Bailey Willis.

After the rocks were laid down in a horizontal position great pressure was exerted from the west and what is now the mountain mass was crowded forward or eastward against the immovable rocks of the plains. As the force was resistless, there was no escape except by bending and wrinkling, and it is supposed that one large fold and several minor ones were produced, as shown in section B.

The pressure, though relieved slightly by the corrugation, still persisted, and the folds were greatly enlarged, as shown in section C. At this stage the folds had nearly reached their breaking limit, and had the pressure continued the tendency would have been for the rock strata to break along the lines of least resistance, which is indicated in the various folds by broken lines. As time went on, the pressure continued and the strata broke in a number of places, as indicated in the diagram, and the rocks on the west side of the folds were pushed upward and over the rocks on the east, as shown in section D. What are now the mountain rocks are represented by patterns of cross lines; these rocks have been shoved over the plains rocks (represented in white), producing an overthrust fault.

As the rocks at the west were thrust eastward and upward they made, in all probability, a mountain range, but they did not at any time project into the air, as indicated in diagram D, because as soon as the rocky mass was uplifted above drainage level, streams began to wear it away and to cut deep canyons in its upland surface, and they also reduced the soft rocks of the plains to a nearly level surface. The rocks of the mountains, owing to their more resistant character, still tower above the plains, and where they overlie the soft rocks the mountains are terminated by a vertical wall of limestone (as shown in section E). This explains the absence of foothills, which is such a conspicuous feature of this mountain front and one in which it differs from most mountain ranges.

Naturally on such an abrupt and exposed front the streams have cut deep gorges though the hard mountain rocks and down into the soft plains rocks that underlie them, so that the actual trace of the fault on the surface is an irregular line zigzagging out and in from spur to valley, forming a saw-tooth outline.

In places along the fault line the streams have cut through the overthrust mass, leaving isolated outliers of the hard mountain rocks far from the main line of the range. The most noted example of this kind is Chief Mountain, on the northeastern boundary of the park, shown on the title-page of this paper, which is formed of a single block of the mountain limestone completely isolated from other rocks of its kind and resting directly on the soft sandstone and shale of the plains. The plane of the fault is clearly indicated in the photograph by the upper line of the forest. The mountain stands as a single monolith 1,500 feet in height, facing the plains as though it were a sentinel standing guard over the hunting ground of the Red Man. The Indians called this mountain the "Old Chief," from its commanding attitude, and it is still known by this name.

Near the south line of the park Debris Creek, flowing to the west, has been so active that it has cut entirely though the overthust mountain mass and is now flowing for a distance of 3 miles on the underlying soft plains rocks. In other words, the stream has eaten a hole though the mountain rocks to the younger strata beneath.

As shown on the map the fault separates the mountain from the plains rocks along the eastern front throughout the entire extent of the park. It crosses the Great Northern Railway at Fielding on the west side of the summit and it passes into Canada where the North Fork of Belly River crosses the international boundary line. Although it is present everywhere along this line it is much more conspicuous in the northern than it is in the southern part of the park, because in the north the limestone which immediately overlies it is much thicker than in the south; in fact, in many places at the southern end of the park the limestone is wanting and at other points it is so inconspicuous that it may not be noticed by the traveler if his attention is not specifically directed to it.

At Summit station the traveler sitting on the north side of the car can see the fault as a nearly horizontal line about two-thirds of the distance from the base to the top of the mountain. The rocks above and below the fault are dark argillite and shale, but here and there the fault between them is marked by thin irregular masses of yellow limestone which in certain views are very conspicuous.

FIG. 14.—DIAGRAM REPRESENTING PROBABLE MODE OF ORIGIN OF THE OVERTHRUST FAULT ON THE EAST SIDE OF THE MOUNTAIN RANGE.

From the railroad northward the limestone becomes more prominent and beyond Two Medicine Valley it is everywhere present, forming precipitous slopes or perpendicular walls. It is this limestone which crosses the valley below Upper Two Medicine Lake, the fault occurring at its base just below Trick Falls. In Cut Bank Valley it is very prominent, making the rugged cliffs which surround the Great Northern camp and a picturesque fall in the creek 1-3/4 miles above the camp. This limestone, with the fault at its base, is very conspicuous in Divide Mountain; it makes the rocky rim that holds Red Eagle Lake in place and forms the barrier over which the stream cascades below the lake.

One of the most striking displays of this bed of heavy limestone is seen at The Narrows on Upper St. Mary Lake. The softer rocks above and below have been cut away by the streams and glaciers, leaving the two points of this resistant bed projecting from the shores. The great fault is just at the base of the limestone and passes under the site of the Great Northern camp.

Much of the beautiful scenery of Swiftcurrent Valley is due to this massive bed, which forms nearly vertical cliffs on each spur and ridge, ranging in height from 500 to 1,000 feet. The perpendicular front of Flattop Mountain is particularly impressive from the trail connecting Upper St. Mary and Sherburne Lakes. The fault swings in just at the base of the cliff on Point Mountain and crosses Swiftcurrent Valley just below the falls at the outlet of Lake McDermott. It then runs diagonally up the ridge on the north, everywhere marked by the base of the cliffs and separating the yellow limestone from the black shale of the plains rocks underlying it.

From Swiftcurrent Valley north the limestone forms a barrier to every stream it crosses. It rises as a wall several hundred feet high where it crosses both forks of Kennedy Creek and the three forks of Belly River, and it makes almost unscalable mountain fronts on the spurs and ridges between these streams. The limestone composing Chief Mountain has already been mentioned, and it is almost as prominent and striking a feature in Yellow Mountain and in the mountain masses bordering the valley of Belly River.

Although the evidence regarding the existence and character of this fault is incontrovertible, there is still one question unanswered, and this is, "How far has the overthrust mountain mass moved?" To those who think of the "everlasting hills" as one of the immutable features of this earth such a question may seem startling indeed, but to the geologist it is the normal question, for to him all surface features are in a state of change and the only reason we do not see them change is because the action is so slow that to ordinary senses it is imperceptible.

That the mountain mass of the park has been thrust far to the northeast from its original position is clearly shown by the diagram on page 26, but the full extent of that movement may never be known, for it is difficult, if not impossible, to locate the original place from which the mass was overthrust. Nevertheless some idea of the extent of the movement may be obtained by measuring the distance between the point where the fault line is cut by one of the streams (B, figure 15) and a line connecting the easternmost termination of the fault on the spurs of the mountains, In figure 15, B is the point where the fault crosses Swiftcurrent Creek, C represents the position of the fault on Flattop Mountain, and A represents the position of the fault on Yellow Mountain. The distance from B to the line AC as measured by Mr. Bailey Willis in 1901 is 7 miles. The limestone bed originally lay deep in the earth and much farther to the southwest than the point B, but how much farther is a question that can never be answered. Geologists feel certain, however, that it has been pushed from its original position to point B and then 7 miles farther northeastward to its present position on Yellow and Flattop mountains.

FIG. 15.—DIAGRAM ILLUSTRATING THE METHOD BY WHICH THE AMOUNT OF VISIBLE OVERTHRUST ON SWIFTCURRENT CREEK WAS DETERMINED.

Recently the writer has calculated the extent of the movement in Marias Pass, along the line of the Great Northern Railway, and found it to be at least 15 miles. From these measurements it is certain that the whole mass involving rocky strata thousands of feet in thickness and weighing countless millions of tons has been shoved toward the northeast at least 15 miles, and were the original position of the mountain mass known it might prove to be a much greater distance.

In other parts of the world as great or greater overthrusts are known to geologists, but in no case is there one so extensive in which the observer can actually see the trace of the fault throughout the greater part of its length. On account of the great movement and the excellency of the exposures this great fault, known to geologists as the Lewis overthrust, is remarkable and destined to become a classic in geologic literature.

FIG. 16.—SHEPARD GLACIER, AS SEEN FROM UPPER GLENNS LAKE ON BELLY RIVER. Steep slopes form the bounding walls of great cirques cut at the head of the Valley. Photograph by Bailey Willis.

CAUSES THAT HAVE PRODUCED THE PRESENT TOPOGRAPHY.

The effect of hardness of rocks and the geologic structure on shaping the topography of the park has been that of a quiescent condition, simply modifying or controlling the active forces that have done the real work of carving the uplifted mass into deep valleys and cirques and leaving sharp peaks, serrate ridges, and flats at high altitude. The processes that have been thus active are weathering, erosion by streams, and erosion by glaciers.

Weathering.—By weathering is meant that slow change that is constantly taking place in the outer shell of the earth's crust due to the solvent action of the moisture in the atmosphere, the solution by percolating waters, and the mechanical breaking down of the rocks by changes in temperature. The moisture in the atmosphere is continually affecting the rocks by dissolving their more soluble constituents and there are practically no rocks that are free from its action. The process is a slow one as measured by years, but everyone can doubtless recall many cases of polished marble or even granite monuments losing their luster when exposed to the weather and in course of time the deeply cut inscriptions upon them. This process is constantly in operation and, although it is most active at or near the surface, it causes the decay of the rocks in many regions to a depth of a hundred feet or more below the surface. In high rugged mountains the actual decay of the rocks is seldom noticeable, for each mineral grain as it is loosened is swept away by the wind and nothing except the roughened surface remains to tell the tale. The mechanical breaking up of rocks on the high barren summits by changes in temperature from the warmth of the midday sun to the freezing cold of night is much more noticeable, and usually the traveller is greatly impressed by what seems to him an unseen and unknown power that is capable of breaking and splintering the most massive rocks so that in time they can be removed by the wind.

The constant effect of weathering is to smooth off the rougher portions of the mountains and gradually reduce their surfaces toward a plain, but the process is so slow that it is scarcely appreciable and in high mountains is more than offset by other processes that tend for a time to increase the ruggedness of the region and to accentuate the mountain forms. These processes will now be described.

Erosion by running water.—Running water is the most powerful agent known in carving mountains and other features of high altitude. Where the streams have a sharp descent the cutting is rapid, but it decreases slowly and becomes less effective as the grade is reduced until the stream becomes sluggish and then it ceases and the stream builds up instead of cutting down its channel. This operation may be witnessed by anyone in small rills or brooks after a hard rain. Each stream is swollen with water which, if the ground or rocks are soft, is heavily charged with mud and sand that act as scouring agents, effectively deepening the channel of the stream. Where the stream reaches a lowland or pond the coarser sediment carried by the water is deposited, being spread out over the immediate flat hand or built out as a delta in a pond or lake. Usually the deposition of such material by a stream has little effect other than to fill up ponds and to build up and smooth over the bottoms of the valley, but under certain conditions the amount of material carried by the streams is so great that important changes of the topography are produced. Thus it seems probable that at one time the St. Mary Lakes did not exist, or, if they were present, that they were very much smaller than they are to-day. At that time the valley below the lakes was very much deeper than it is now, but the sand and gravel brought down by Swiftcurrent Creek was dumped into the valley, filling it up to its present height. This delta of Swiftcurrent made an effective dam, and behind it the water accumulated. When the barrier was first built there was a single body of water filling the valley in what is now the Upper and Lower lakes, but later these were separated in the same manner by deltas built out by Divide Creek from the south and by Wild Creek from the north. In times of flood these creeks carry a large amount of sand and gravel which in former times was dumped into the lake, building out deltas or points of land from both sides, and as these points were nearly opposite, they soon coalesced, separating the lake into two parts. The relationship of these deltas can be seen from high up on the ridge on the north side of the valley where it is crossed by the trail from Lower St. Mary to Sherburne Lakes.

FIG. 17.—BOWMAN LAKE, AS SEEN FROM NEAR THE OUTLET. RAINBOW PEAK ON THE RIGHT. Photograph by M. R. Campbell.

The process of stream erosion, to many readers, may seem to be a slow one, but in reality it is rapid, for the streams, especially in a mountainous country, are constantly at work. There is no cessation, no relief from the constant rasping of the sand grains on the beds of the streams, and as a result the hardest and most massive rocks are rapidly worn away, the streams cutting deep V-shaped gorges. If conditions are just right the gorges at first may have nearly vertical walls and be true canyons, but in a region of considerable precipitation they will sooner or later take on in cross section the shape of a V.

If the valleys and canyons of Glacier Park are closely studied, either on the ground or by means of the contour map published by the United States Geological Survey,¹ it will be seen that only a few of the minor ones show the cross section noted above as indicative of stream erosion. Instead of a sharp V-shaped profile in cross section their bottoms are distinctly flattened and broadened, taking on the form characteristic of glacial erosion. The rounded contour of valleys that have been glaciated is shown in figure 18, which is a view down Swiftcurrent Valley from a point a little below the pass. Despite the fact that practically all the valleys in the park are of this form, geologists do not hesitate to say that all of the valleys of this region were cut to approximately their present depths by streams and that the rounding has been simply a modification of their forms due to the scouring action of ice as explained under the heading "Glacial erosion." In a few places minor valleys and gulches show the V-shaped form due entirely to stream erosion, but these are recent features and have been cut since the ice retreated from the main valleys, leaving the streams free to carve their valleys in their own way.

---

¹This map may be purchased from the Director of the United States Geological Survey, Washington, D. C. for 30 cents.

---

FIG. 18.—SWIFTCURRENT VALLEY, LOOKING DOWN FROM NEAR THE PASS. Grinnell Mountain on right and Appekunney Mountain on left. Valley was originally V-shaped, but has been scoured out by the old glacier until it is a perfect U in cross section. Photograph by M. R. Campbell.

As noted on a previous page the result of stream erosion on a recently uplifted mountain mass is to increase the ruggedness, but after a long time, if there is no more uplift, the streams will have cut as deeply as their outlets permit and then their work is the broadening rather than the deepening of the valleys. The work goes on more slowly than before; but if conditions remain undisturbed the valleys will become broader and broader, the intervening highland more and more reduced, until finally a fairly even surface results. This is the apparent explanation of the topographic basin of Flattop Mountain described on page 14. In the comparatively soft rocks in the center of the syncline the streams long ago widened their valleys until they coalesced, but the hard limestones forming the bounding rim (shown in fig. 19) resisted the action of the elements and remained unreduced. When this basin was formed it is probable that the altitude of the range was not so great as it is at the present time, or else that the surface of the surrounding region was much higher than it is to-day.

FIG. 19.—PART OF THE NORTH RIM OF THE BASIN OF FLATTOP MOUNTAIN, AS SEEN FROM KOOTENAI RIDGE. Photograph by Bailey Willis.

Erosion by ice (glaciers).—Opinions differ somewhat regarding the ability of glaciers to erode. Some maintain that a glacier has little erosive power, whereas others ascribe much of the work done in a mountainous region to ice. It seems probable that neither view is entirely correct and that ice has considerable erosive power, but only in modifying the form of gorges previously cut by running water.

Nearly all of the valleys in Glacier Park show by their forms that they have been occupied by ice, although in many cases no glaciers exist in them at the present time. The form of a valley after it has been modified by moving ice is well illustrated by figure 18. Such valley forms could have been produced only by moving ice which filled the valleys to considerable depths and covered most of the lower lands of the park. There is abundant evidence in the forms of the valleys and in the scratches left by the moving ice to prove that it was at least 2,500 feet deep in the larger valleys and that the crests alone remained above its level. This, of course, means that immense glaciers must have originated in these mountains and flowed out in all directions, extending 20 or 30 miles onto the Great Plains on the east and down the valley of Flathead River on the west. The present glaciers are only the diminutive remnants of the earlier ones, and if the mean annual temperature were raised slightly or the amount of precipitation decreased it is probable that they would entirely disappear. As it is they cling to the north and east sides of the high ridges and peaks where the winter snows find a lodging place and where the summer sun has little effect upon them. There are no longer any rivers of ice, but the work they accomplished is still visible and lends beauty to the mountain scenery, for the rugged mountain tops are accentuated by the rounded graceful forms of the valley walls.

The glaciers have had an even more marked effect upon the topography than that of smoothing the valleys in which they lay. They have the power to cut into the mountain slopes at their heads, forming cirques that add greatly to the ruggedness and variety of the topographic forms to be found here. The process of cutting out these cirques is not well understood, but at or near the point where the névé or snow field changes into the moving ice of the glacier the cirque is formed. Blocks of rock are evidently plucked out by the moving ice even back to the névé and this tends to give to the cirque a nearly level floor. The plucking extends constantly backward, undermining the bounding walls, and these break down in more or less vertical cliffs, giving to the excavation a circular or semicircular form with nearly vertical walls from a few hundred to more than 3,000 feet in height, as shown in figure 20.

FIG. 20.—GREAT VERTICAL CIRQUE WALL ON THE NORTH FACE OP GOATHAUNT MOUNTAIN. Part shown in picture fully 1,200 feet high, and the wall extends below at least an equal distance. To appreciate this great height, compare it with the white outline representing the Washington Monument (555 feet high) on the same scale. Photograph by Bailey Willis.

All of the present glaciers lie in such cirques, but some are much better developed than others. One of the best examples is Grinnell Glacier, which lies on the east side of the Continental Divide on Cataract Creek, one of the tributaries of Swiftcurrent. This glacier has eaten back into the mountains so far that the crest is nearly cut through and the cirque wall is jagged in the extreme. It is known as the Garden Wall, and, as shown in figure 9, it stands in rugged grandeur, a veritable wall a thousand feet in height.

One of the most striking cirque walls is that surrounding Iceberg Lake, a little sapphire gem at the head of North Fork of Swiftcurrent Creek, shown in figure 21. The lake and glacier lie in a beautiful amphitheater about one-half mile in diameter surrounded on three-fourths of its circumference by nearly vertical walls from 2,500 to 3,000 feet high. Four Woolworth buildings, New York's greatest skyscraper, 750 feet high, could be placed one above another in this cirque, and the spire of the uppermost one would just reach the highest point of the bounding wall.

FIG. 21.—ICEBERG LAKE. View shows lake and glacier about one-half mile in diameter, surrounded on three sides by rock wall 3,000 feet high at the highest point. Black band of lava shows on left. A great dike cuts the wall in notch on right. Photograph copyrighted by Kiser Photo Co. Published by permission of the Great Northern Railroad.

Probably the deepest cirques are on Cleveland Mountain, the highest summit in the park. Cirques on either side have walls approximately 4,000 feet high. The one on the west side shown in figure 22 looks as though some immense prehistoric animal had taken a huge bite out of the side of the mountain which previously had sloped regularly from the margin of the valley to the summit of the peak. The "bite" has a fairly level floor and is bounded on the back by a cliff so steep that on it vegetation can not find a foothold and across it only the most hardy of the mountain goats can make their way.

FIG. 22.—GREAT CIRQUE ON NORTHWEST SIDE OF MOUNT CLEVELAND, AS SEEN FROM WATERTON LAKE. Summit of mountain 6,2000 feet above the lake. Cirque wall at least 4,000 feet high and nearly vertical. Photograph by Bailey Willis.

The glaciers have not only produced the cirques which add so much to the picturesqueness of the topography and to the ruggedness of the range, but to them are due directly most of the beautiful lakes, large and small, which are without doubt one of the most attractive features of the region. Some of the lakes, such as McDermott, Red Eagle, Ellen Wilson, Two Medicine, and many smaller ones, occupy rock basins which the old glaciers scoured out in places where the rocks were slightly softer than they were lower down the valley. These may be distinguished by the rocky barrier that crosses at the outlet. The limestone ledge which holds Lake McDermott in place is partly shown in figure 23. Similar ledges cross the outlets of Red Eagle and Two Medicine Lakes. Some may suppose that these ledges were thrust up like a dike, but such is not the case. They have remained in their original positions, but the softer rocks in the valley above them have been carried away, leaving a basin in which the waters of the lake accumulated. In Two Medicine Lake the water does not now, as formerly, flow over the top of this ledge but it has found a crack or subterranean channel and through this it flows and bursts from a hole in the bottom of the ledge. From this circumstance it is known as Trick Falls.

FIG. 23.—MCDERMOTT LAKE, WITH GRINNELL AND GOULD MOUNTAINS IN THE BACKGROUND. The lake lies in a rock basin, and the outlet forms a pretty fall ever the massive ledges of limestone that hold the lake in place. Photograph by T. W. Stanton.

Other lakes, such as Bowman and Upper Quartz, are held in place by great glacial moraines which have been deposited across their valleys like huge dams. These may be known by the hummocky ridges that are plainly visible at the lower ends of both of these lakes. Many other small lakes and ponds are rimmed about by moraines, but they are too numerous to mention.

A third class of lake basins has been formed by the glaciers, but in these the dam is not formed by a moraine, but by the outwash of sand and gravel from the end of the ice. The basins of Logging, Lower Kintla, Lower Two Medicine, and McDonald Lakes are supposed to have been formed in this manner. At the extremities of these lakes there is no visible evidence of a barrier, but when the valley below the lake is examined it will be found to be deeply filled with coarse but well-rounded gravel which the stream flowing away from the ice doubtless carried and deposited, forming a dam just as effectively as though a moraine had been built around the ice front.