GLACIAL HISTORY OF THE YOSEMITE VALLEY
TRANSFORMATION OF THE YOSEMITE VALLEY BY THE ICE
KERNEL OF THE YOSEMITE PROBLEM
What, it may now be asked, was the effect of these repeated ice invasions upon the configuration of the Yosemite Valley? To what extent has the valley been remodeled by the glaciers, and in what measure is its present form due to their action? These are the questions that contain the kernel of the Yosemite problem.
The opinions that have been advanced by previous students of the problem have been briefly reviewed in the introductory section of this paper. Their wide divergence was due principally to three circumstancesto the lack of definite data as to the extent and magnitude of the ancient Yosemite Glacier; to the vagueness of ideas that prevailed as to the nature and the duration of the ice age; and to the absence of reliable information as to the eroding power of glaciersas to whether or not glaciers are capable of excavating deep canyons in hard rock. In other words, it was not known whether the Yosemite Glacier extended but a few miles beyond the valley or all the way down to the foothills of the Sierra Nevada; whether it was in existence a few thousand years or a million years; whether it possessed the power to cut deeply into the solid granite or could do little more than sweep out loose débris.
To-day the situation is quite different, for the maximum extent and maximum thickness of the Yosemite Glacier have been ascertained within narrow limits, and it is clear that the valley was invaded by the ice several times for prolonged periods, the history of its alternating stages of glaciation and deglaciation corresponding in general to that of the major advances and retreats of the great ice sheets over North America and Europe. What is more, glacial research has advanced to a point where it is possible to form rough estimates of the duration of the successive stages of glaciation and deglaciation and of the ice age as a whole. Though the precise number of centuries involved is not known, there is little doubt as to the order of magnitude of the time intervals involved as compared with that of the postglacial interval, whose length in years is no longer a matter of guess.
As regards the capacity of glaciers to excavate deep canyons in hard rocks, little doubt now remains on that score. Though there are still skeptics, it is safe to say that most geologists are convinced that glaciers are indeed powerful eroding agents. It is clear also that much depends upon the character of the rocks involved: in some kinds of rock glaciers erode much more effectively than in others. Probably few glaciated valleys that have come under investigation recently have shed more light on this question than the Yosemite itself, for not only does this valley afford striking illustrations of the degree to which the efficiency of the glacial processes is influenced by the character and the structure of the rocks, but its preglacial depth and form have been determined with a fair degree of accuracy, and hence it is possible to calculate with corresponding accuracy the amount of rock material that was actually excavated from the valley during glacial time.
DEEPENING EFFECTED BY THE YOSEMITE GLACIER
The downward excavating effected by the Yosemite Glacier is clearly indicated by the longitudinal profiles in Plate 27, A. All the space between the profile CC' of the preglacial canyon stage and the profile DD' of the glaciated rock floor of the valley (the bottom of ancient Lake Yosemite) represents excavating done since the beginning of the ice age. True, this is not a measure of glacial cutting alone: it represents glacial cutting and stream cutting combined, for the glaciers were active only at intervals during the ice age, and during the prolonged interglacial stages the Merced River performed its characteristic work. However, it is plain from the very character of the profile DD' and from its extension by the giant stairway to the Little Yosemite and the upper Merced Canyon (pl. 27, A) that glacial cutting was vastly preponderant, for a stairlike canyon profile of this type, characterized by alternating treads and risersthe treads approximately level or bearing shallow lake basins, and the risers cut in the form of precipitous cliffsis a characteristic product of glacial erosion, not of stream erosion. The particular manner in which glaciers produce steps in canyon floors is elucidated on pages 95-96. The tendency of streams is to produce fairly smooth, unbroken valley profilesto eliminate steps as well as basins. The Merced has done nothing since the ice age to accentuate the stairlike profile of its pathway; on the contrary, it has done what it could do to demolish the steps and fill the basins.
Plate 27, A, shows strikingly the inequalities in the depth of glacial excavation. Evidently the ice accomplished much more work in some places than in others. At the lower end of the valley the glacial deepening measures only about 500 feet, but up the valley it increases gradually, reaching a maximum of about 1,500 feet near the head of the valley. Thence it diminishes rapidly to 850 feet at the top of the Vernal Fall and to a minimum of about 250 feet near the top of the Nevada Fall, to increase again gradually to 700 feet in the Little Yosemite. One can not but marvel at these marked inequalities in the deepening effected by the iceindeed, it is hard to say which seems more astonishing, the maximum or the minimum. However, the true significance of these inequalities becomes apparent only when they are considered together with the variations in the lateral cutting.
WIDENING EFFECTED BY THE YOSEMITE GLACIER
The lateral cutting effected by the Yosemite Glacier is strikingly revealed in the cross profiles in Figures 24 to 30. Nothing, indeed, serves better to give a true conception of the thoroughgoing transformation which the Yosemite Valley has suffered by glaciation than this series of diagrams. Each shows exactly to scalethat is, without vertical exaggerationthe cross profile of a certain part of the valley as it is to-day, plotted from the contours of the topographic map (pl. 7), and the corresponding cross profile of preglacial time. The reconstruction of the preglacial cross profile has not been simply a matter of imagination, for the depth of the preglacial Yosemite Canyon has been determined (see profile CC', pl. 27, A), and the character of its sides was governed by a number of elements such as the heights and gradients of hanging valleys and hanging gulches and the forms of truncated spurs and projecting rock monuments. General guidance, further, is afforded by the well-known laws of canyon cutting by streams and of the progressive conversion of sheer canyon walls into slopes of moderate declivity by the processes of weathering and erosion. In some places intimate knowledge of the local structure of the granite has permitted the introduction of sculptural details such as benches, facets, pinnacles, and knobs.
From all these cross profiles it is manifest that lateral cutting has been a more important element in the transformation of the Yosemite chasm than downward cutting. At every point the widening accomplished exceeds the deepening. It is, in fact, mainly through lateral cutting that the narrow V canyon of preglacial time has been transformed into the broad U trough of to-day. It is evident, further, from the very breadth of the U profile that the glacial processes far outstripped the fluvial processes, for whatever trenching the river did in interglacial epochs must have tended to produce an inner gorge, yet of such a gorge there is no vestige whatever. Surely it is not to be supposed that the river at any time possessed volume enough to spread over the whole width of the valley and that it performed a significant share of the lateral cutting. If it had had such volume and power it would have produced many broad-bottomed yosemites in the relatively unresistant sedimentary rocks below El Portal long before it evolved a single valley of that type in the resistant granitic rocks of the Yosemite region.
Comparison of the different cross profiles with one another reveals also the fact that the widening accomplished was no more uniform than the deepening and that in general the glacial processes worked very unequally. They accomplished large results in each of the two chambers of the Yosemite Valley, but considerably less in the portal between those chambers and astonishingly little in the Merced Gorge below the valley and in the gorge below the mouth of the Little Yosemite.
A fair measure of the lateral cutting effected in the main Yosemite chamber is indicated in the cross profile from Eagle Peak to Sentinel Rock shown in Figure 24. Measured at right angles to the sides of the preglacial canyon it amounts to fully 1,800 feet, where as the deepening amounts to approximately 1,200 feet. The space between the preglacial and postglacial profiles as shown in the diagram measures about 1,400,000 square yards; hence the ice excavated from a section of the valley 1 yard long in the direction of its axis about 1,400,000 cubic yards of rock. This measurement, it is true, is made at a point where a jagged spur projected from the face of Sentinel Rock, yet the amount stated may be considered fairly representative of most of the main Yosemite chamber, for there were several other spurs, all of which were planed away by the glacier. The amount is of course merely an approximation, but it will serve for comparison with similar amounts indicated by other cross profiles.
The maximum of glacial excavation is shown in the cross profile from North Dome to Glacier Point and Sentinel Dome (fig. 25). The ice there accomplished truly prodigious results: at no other place did it effect a more complete transformation. The conditions were, however, peculiar to the head of the valley, for the two preglacial gorges of Tenaya Creek and the Merced River there came together, and the two main branches of the Yosemite Glacier were confluent. The hump in the middle of the profile represents in section the tapering spur that sloped formerly from the west base of Half Dome to the junction of the gorges. A more massive spur doubtless projected northward from Glacier Point, and a smaller spur projected from North Dome in the direction of the Washington Column. The quantity of rock excavated by the ice from a section of the valley 1 yard long in the direction of its axis, computed from this diagram, approximates 1,700,000 cubic yards.
Representative of the lower Yosemite chamber is the cross profile shown in Figure 26. Here too, unquestionably, entire spurs were removed by the ice. The lateral cutting, measured as before at right angles to the preglacial canyon sides, amounts to nearly 1,400 feet, and the total quantity of rock excavated over the distance of 1 yard is about 800,000 cubic yards. The cross profile of the portal between El Capitan and the Cathedral Rocks (fig. 27), on the other hand, shows only 700,000 cubic yards of excavation. But the most marked contrast is offered by the profile in Figure 28, which is taken across the Merced Gorge, just below the Yosemite Valley. The work done there by the ice can scarcely be called a transformation. There was only moderate enlargement of the valley section and no real change in general form, the inner gorge, though widened somewhat, remaining strongly in evidence. The quantity of rock excavated per yard here is less than 500,000 cubic yardsthat is, little more than one-half the quantity indicated for the lower chamber and little more than one-third the quantity indicated for the upper chamber.
Another cross profile of special interest is that in Figure 29, drawn across the mouth of the Little Yosemite and over Mount Broderick and Liberty Cap. It shows an abrupt diminution in the glacial excavation, the quantity removed from a section 1 yard long being only about 300,000 cubic yards. Of course this quantity is not comparable on even terms with the quantities shown in the cross profile previously considered, for it represents work done only by the Merced Glacier, not by the entire Yosemite Glacier. But even for the Merced Glacier the quantity stated is very small. The profile in Figure 30, which is taken across the middle of the Little Yosemite, shows an amount of material excavated more than half as large again.
At the head of the Little Yosemite, finally, glacial excavation dwindles to a minimum. There, opposite Bunnell Point, the walls contract and the upper gorge of the Merced beginsa gorge so narrow that, were it not for the smoothness of its walls and the presence of polish, striae, and grooves, it might readily seem a product of stream erosion purely. (See fig. 37.) To obtain a true conception of that gorge in its relations to the upper Merced Canyon as a whole, one should view it from a lofty summit such as Half Dome or Clouds Rest. It is then seen to be a mere inner trench winding across a large billowy mass of bare granite that obstructs the canyon for a distance of 2 miles.
How are these extreme and in places abrupt variations in the depth and breadth of glacial excavation to be explained? Why have the glaciers been able to accomplish so much in some parts of the Yosemite region and so little in others?
It was pointed out by Muir60 that the Yosemite Valley begins just below the junction of three converging canyons from each of which formerly issued a mighty glacier. In reality there were only two glaciers, the Merced and the Tenaya, for no ice stream ever issued from Illilouette Valley, as already explained (pp. 62-63); but this does not invalidate Muir's argument. Similarly, he showed that the Hetch Hetchy Valley and the Kings River Canyon, two of the most typical yosemites in the Sierra Nevada, each begins below the junction of two or more great glaciated canyons. The confluence of two or more great ice streams, which results in an abrupt increase in excavating power, Muir was therefore inclined to regard as the primary cause of the formation of a capacious Yosemite-like valley. No doubt this explanation is correct as far as it goes, for the consolidation of two or more ice streams into one does result as a rule in greater compactness of form and therefore in increased economy of movement and increased excavating capacity per unit mass. However, some yosemites in the Sierra Nevada quite as typical as any of those mentioned. are not associated with the junction of large glaciated canyons; and there are plenty of examples of two or more powerful ice streams uniting without giving rise to a typical yosemite.
The Little Yosemite, notably, has no converging canyons at its head. It has been developed at a place where the Merced Glacier received no significant increments but where, on the contrary, that ice stream dissipated part of its volume by overspreading the flanking uplands. The Grand Canyon of the Tuolumne, on the other hand, affords several good examples of glacier junctions that have not resulted in the production of any yosemites. From the canyon of Return Creek (pl. 2) there issued a tributary glacier 3,000 feet thick, yet its confluence with the Tuolumne Glacier, which was 4,000 feet thick, produced no yosemite. Nor was a yosemite produced below the mouth of Cataract Canyon, which was another large glacier channel. Even the junction of the Piute Creek Glacier, which was 3,500 feet thick, with the Tuolumne Glacier gave rise to only a minor enlargement of the canyonthe incipient yosemite known as Pate Valley.
A comparative study of the principal yosemites of the Sierra Nevada, further, shows that they are by no means proportionate in size to the glaciers that occupied them. Some of the greatest yosemites were the pathways of only moderately large glaciers, and some of the lesser yosemites were the pathways of very large glaciers. Thus the Yosemite itself, the most capacious of all the valleys of its type, was the pathway of a glacier only 37 miles long and slightly over 3,000 feet thick at the time of maximum glaciation; whereas the Hetch Hetchy Valley, though only half as long and half as wide as the Yosemite and about 1,000 feet shallower, was traversed by a glacier 60 miles long and 4,000 feet thick. Moreover, lateral moraines left by that glacier (the Tuolumne Glacier) show clearly that not only was the Hetch Hetchy completely overwhelmed by the ice of the earlier stages, which was 1,000 feet thicker there than in the Yosemite, but during the last glacial stage it was again filled to overflowing by an ice stream comparable in magnitude and eroding power to the earlier Yosemite Glacier, whereas during the same stage the Yosemite itself contained only a shallow and feeble ice tongue that did not reach quite to its lower end.
The Kings River Canyon, though a few miles longer than the Yosemite Valley, is only half as wide and two-thirds as deep. So narrow is its floor that "canyon" seems at least as appropriate a term for it as "valley"; yet the ice streams that occupied it in the earlier and later stages of glaciation respectively had but little less volume and power than the glaciers that occupied the Yosemite Valley in each of those stages.
These facts afford no little food for thought. They show that typical yosemites have been produced not only below the points of confluence of large branch glaciers but also at points where there was no such confluence, and, further, that the dimensions of a yosemite bear no definite relations to the size and power of the glacier that occupied it. Evidently another factor is involveda factor that in large measure controlled the glacial processes and determined how much excavational work they might accomplish at any place. That factor is the structure of the rock.
In order that the influence of rock structure on the excavating efficiency of glaciers may be understood it is desirable first to gain a clear idea of the manner in which glaciers do their work.
HOW DO GLACIERS EXCAVATE?
It is commonly supposed that glaciers erode their beds mainly by grinding and scouring. It is true that with the rock fragments which they hold in their bottom layers glaciers perform considerable abrasive workwitness the polished, striated, and even deeply grooved floors and walls of glaciated canyons. Nevertheless, the efficacy of this abrasive action is not inherently great. Only in soft, friable rocks does it accomplish really large results. In hard, tough rocks, such as granite, it achieves but littlenot enough, in any event, to account for the profound remodeling of entire valleys and canyons. On such rocks, as a rule, the presence of glacier polish is indicative of moderate changes slowly produced.
The process whereby glaciers excavate to best effect in hard rocks is by plucking, or "quarrying" entire blocks and slabs. Because of their very weightsome 30 tons to the square foot for every thousand feet of thicknessand the fact that they are shod with coarse rock waste frozen in their basal layers, glaciers have a strong frictional hold on their beds; and so, as they move forward, though at a rate of only an inch or two a day, they dislodge and drag forth entire blocks and slabs. The peculiar property that ice has of freezing tightly to objects with which it is in contact is probably a potent factor in the process, as has been suggested by Profs. T. C. and R. T. Chamberlin,61 of the University of Chicago.
The blocks and slabs thus dislodged are, however, rarely broken off from sound, unfractured rock. The glaciers take advantage, rather, of the fractures already existing in the rockthe joints by which it is divided into natural blocks and slabs. (See pl. 41, B.) This is true especially of hard, tough rocks such as granite, for even a glacier 3,000 feet thick does not exert pressure enough to disrupt a floor of sound, massive rock of that type. It is clear, then, that the joint structure plays a very important part in glacial quarrying. Without it, in fact, the process is scarcely operative in hard rocks.
Several agencies, furthermore, tend to facilitate the quarrying process, by loosening up the blocks and slabs. Acid carried by water that percolates through the joint fractures dissolves the weaker minerals and lessens cohesion; and water freezing in the fractures pushes the blocks and slabs apart with its momentary but strong expansive force. (See pl. 42, B.) It has even been contended, by those who would attribute only slight erosive power to glaciers, that the quarrying is limited practically to the removal of blocks and slabs previously loosened in preglacial or interglacial time, but there is ample evidence to the contrary. In many glacier channels the quarrying can be seen to have progressed far below the zone of weathering, its depth varying primarily with the thickness and power of the glacier concerned.
The manner in which the glacial quarrying process operates is illustrated by Figure 31. Any joint block in the bed of a glacier, such as that marked A, which is for any reason unsupported or weakly supported on its downstream side, is particularly susceptible of being dislodged, for the force of the glacier is exerted upon it at a small angle forward from the vertical, as indicated by the arrow. The block A and its side companions having been removed, the block B and those flanking it will next be unsupported and ready for removal, and so the process will continue farther and farther up the valley. Its rate of progress will depend upon the power of the glacier, the size and weight of the blocks, and the looseness of the joints.
Muir62 in one of his earlier papers described long trains of glacially quarried blocks which he had observed in the vicinity of Tenaya Lake and which he had succeeded in tracing back to their points of origin on the mountain side. The mountain was not smoothed or rounded but cut in square-edged steps bounded by joint planes. Stepped and hackled rock surfaces are, indeed, the rule in glaciated mountains.
Impressive evidence of the quarrying action of glaciers is to be found also in the presence of angular, sharp-edged blocks in the moraines. Such blocks abound especially along the upper courses of lateral moraines, where they were dropped before they had been carried far enough to lose their angularity by wear. A particularly striking example is that shown in Plate 42, A. The small moraine represented consists almost exclusively of clean-cut, angular blocks, mostly from 4 to 12 feet in length, that were torn from the jointed cliffs of a glacial cirque near by.
Because glacial quarrying in hard rocks is so largely dependent upon the presence of joints, its action naturally is strongly influenced by the spacing of the joints. Where these fractures are close together, so that the rock is divided into small, light blocks or slabs, quarrying will proceed with relative ease and rapidity; there even a small, feeble glacier will be able to accomplish much. Where the joints are far apart50 feet or morethe blocks between them are too large and too heavy even for a mighty trunk glacier to dislodge. Rock so sparsely jointed is virtually massive, so far as the glacial processes are concerned, and can be reduced only by slow abrasion.
Between these two extremes, of course, there are many intergradations, depending upon the distribution and the arrangement of the fractures. In these intermediate rocks in which the jointing is moderately coarse the excavating force of the glacier is as a rule the decisive factor. In dealing with rocks so jointed a small shallow glacier exerting relatively little pressure on its bed may be almost impotent, whereas a glacier of great depth, exerting correspondingly great pressure, may accomplish signally large results. It is, in fact, precisely in those areas where the joint blocks are fairly largefrom 5 to 25 feet to the sidethat a powerful glacier may attain its greatest excavating efficiency, for every block it removes has great cubical content. And this explains why in regions of coarsely jointed rocks there is usually a vast difference between the amount of excavational work done by the great trunk glaciers and that done by the small tributary glaciers.
This brief analysis shows that the quarrying action of glaciers is inherently selective, especially in regions where the rocks are hard and tough. There its effectiveness is dependent in large measure upon the character of the rock structuremore specifically upon the spacing of the joints.
Bearing this in mind let us now briefly examine the jointing of the rocks of the Yosemite region and see in what manner it has affected the action of the glaciers.
THE KEY TO THE SECRET OF THE YOSEMITE'S ORIGIN
No one who gives the rock walls of the Yosemite, the Little Yosemite, and Tenaya Canyon more than a superficial glance can fail to note the marked variations that occur in their joint structure. Not only does the arrangement of the joints differ from place to place, but the spacing varies widely; in certain zones the joints are only a few inches apart, and the rock is shattered into small slabs or mere slivers; elsewhere and more commonly the intervals between joints range from a foot to a score of feet, so that the rock is divided into great angular blocks or thick sheets; and in still other places joints are altogether lacking for distances of hundreds and even thousands of feet, and the rock is wholly undivided. What is more, these variations are sometimes remarkably abrupt, so that structural extremes are brought into immediate juxtaposition. In few other regions where granitic rocks occur is there so great structural diversity or are sharp contrasts in structure so prevalent as in the Yosemite region.
Now the course of the Merced in its larger aspects is by no means related to or controlled by these structural vagaries in the granitic rocks. It must not be supposed that the river follows a belt of jointed rock and that the divides on both sides are composed of massive rock. Though it may be guided here and there for a short distance by a set of joints or by a master fracture, the Merced runs on the whole in disregard of the rock structure. The reason is that when the Merced first established its southwesterly course down the slope of the Sierra Nevada there still was a veneer of folded sedimentary rocks over the granite. It was not until later in its history that the river wore its channel down through the sedimentary rocks and into the granite underneath. Then, being well intrenched and unable to deviate widely from its original course, it perforce had to cut the jointed and the massive rocks alike. The Merced is, indeed, what geomorphologists term a "superimposed" stream; and so are most of the other rivers on the western slope of the range. Therein lies the key to the secret of the origin of the Yosemite Valley and of all the other yosemites in the Sierra Nevada. Each of these capacious U-shaped valleys has been developed in an area of prevailingly fractured rocks in which the agents of erosion worked with comparative facility and in which the glaciers, when they came upon the scene, quarried with extraordinarily great effect. The narrow portals and gorges above and below the yosemites, on the other hand, are cut in bodies of prevailingly massive rock which the glaciers could not quarry and could reduce only by slow grinding.
These facts became evident to the author in 1905, while he was engaged in the topographic surveys for the map of the Yosemite Valley shown on Plate 7. The very task of delineating the cliffs in detail drew his attention to the relationship that exists between their sculpture and their inner structure.63 But the reason for the prevailingly close jointing in the area of the Yosemite Valley was not wholly clear until his colleague Frank C. Calkins in 1913 showed that the chasm is situated in an exceptional locality where many small bodies of relatively basic rocksgranodiorite, diorite, and gabbrohave been intruded into the otherwise vast, unbroken bodies of siliceous granite and monzonite that make up the central parts of the great Sierra batholith (see pl. 51), also that these basic rocks are in general more closely jointed than the siliceous rocks.
Of course it is not possible now to determine exactly the original extent within the valley area of each of these masses of well-jointed basic rocks, as they are in large part removed by erosion. The rock floor of the valley, moreover, is hidden from view by a thick deposit of sand and gravel. However, numerous remnants of the basic rocks are visible in the walls and on the adjoining uplands, and from a detailed survey of these remnants it is possible to form some estimate of the original extent of the intrusive bodies and to appraise the effect they must have had on the quarrying action of the glaciers. Indeed, enough is known of the distribution and extent of the different bodies of jointed and of massive rock, respectively, to account satisfactorily for the development of all the major features of the Yosemite regionfor the division of the Yosemite Valley into two broad chambers; for its peculiar spoon-shaped lower end and its equally peculiar square upper end; for the stepwise ascent of the, giant stairway; for the breadth of the Little Yosemite and the depth of Tenaya Canyon.
CONFIGURATION OF THE YOSEMITE VALLEY EXPLAINED
The central part of the main Yosemite chamber has been excavated from a body of coarsely jointed granodioritethe Sentinel granodiorite. (See appendix.) This rock extends directly across the valley in a belt about 2 miles wide. The Three Brothers and Taft Point mark its western margin; the Royal Arches and Glacier Point its eastern margin. (See pl. 51.) It is divided mainly by vertical and horizontal joints and hence has in many places a distinctly columnar or prismatic structure. This is evident especially in the columnar crags and pinnacles on the wall east of Union Point. In other places, especially in Sentinel Rock (pl. 19, A), it has a smoothly sheeted structure, the partings being nearly vertical. Almost throughout, therefore, this rock material is divided into blocks and sheets of large sizeof a size, in fact, which the Yosemite Glacier during its higher stages could quarry with great efficiency.
From the vicinity of Taft Point west as far as the Cathedral Rocks diorite and gabbro predominate. These of all the rocks of the Yosemite region are the most thoroughly fractured; hence they must have been readily quarried by the glacier, even at times when it had only moderate volume. It is not surprising, therefore, to find that in the area of these rocks the south wall is embayed. Just how far into the valley these well-jointed rocks extended originally can only be surmised, but there is reason to believe that they occupied considerable space, for the valley here attains its greatest width, in spite of the fact that its north wall is composed of prevailingly massive granite.
The narrow portal between El Capitan and the Cathedral Rocks is, as might be expected, framed by promontories of exceptionally massive rock that could not be quarried by the glacier. The great prow of El Capitan consists wholly of this highly siliceous, massive graniteEl Capitan granite it has been appropriately named. The Cathedral Rocks consist only in part of granite of this type and are traversed by numerous sheets and dikes of other igneous rocks, but they are nevertheless for the most part unfractured, the different rock materials in them being intimately welded together. But for this fact the whole promontory surmounted by the Cathedral Rocks and bearing the hanging gulch of Bridalveil Creek would probably have been quarried away by the glacier flush with the south wall of the valley.
The lower Yosemite chamber doubtless owes its great width to the ease with which the glacier quarried in the large bodies of well-jointed gabbro and diorite which extend throughout most of its length. Considerable masses of these dark-hued basic rocks still cling to the north side, west of the Ribbon Fall. Their unstable masonry, crisscrossed by numerous joints, has not remained standing in the form of a sheer wall but has broken down completely, producing the immense talus known as the Rock Slides, over which the Big Oak Flat Road is built.
On the south side glacial quarrying has been less effective, the bulk of the rock being El Capitan granite. As a consequence sheer walls and massive buttresses remain, but there are several recesses which show that the quarrying has been facilitated and guided locally by zones of intense jointing and shattering. The recess dominated by the Leaning Tower is of this kind. It is entirely probable, further, that the abrupt increase in the width of the valley below the portal and the persistent southward trend of the wall extending from the Bridalveil Fall to the Leaning Tower are due in large part to the influence exerted by the same zone of shattering on the glacial quarrying.
Significantly the lower Yosemite chamber contracts abruptly west of the body of fractured gabbro and diorite, and the great barrier which incloses the spoon-shaped lower end of the valley consists of massive El Capitan granite. The benches that flank the Merced Gorge as far west as the Gateway are composed of the same obdurate material, a fact which accounts for the narrowness of the gorge. Close examination shows, however, that this granite is not wholly massive but traversed at long intervals by vertical and horizontal master joints. It is therefore really divided into blocks, but these were much larger than the glacier could dislodge. The ice merely rode over them, grinding their surfaces, as is strikingly revealed by the smoothly curving shieldlike hump of Turtleback Dome, shown in Plate 43, A.
The lower end of the Yosemite Valley (pl. 16, A), though scenically unattractive, is of peculiar scientific interest. Few other localities in the Yosemite region afford more striking evidence of the dependence of glacial quarrying upon the presence of favorable structures in the rock and of the comparative inefficiency of glacial abrasion in massive rock. Though the abrupt contraction of the valley at its lower end might at first sight seem to indicate the place where the glacier usually terminated and beyond which it only rarely advanced, it marks in fact but the western limit of the quarriable rocks in the Yosemite Valley and the beginning of the unquarriable rocks along the Merced Gorge.
The lower end of the valley is of interest, further, because there a considerable share of the glacier's mass had to move upward in order to get out. The central portion of the glacier, of course, passed through the Merced Gorge without moving upward, but the flanking masses (fig. 22) rode up the rock slope at the end of the valley and surmounted the uneven benches that flank the gorge on both sides. Indeed, the deeper the glacier excavated the Yosemite Valley, the higher these ice masses had to climb in order to make their exit. Toward the end of the earlier stages of glaciation they had to climb 1,000 feet.
This almost incredible ascent of the ice is attested beyond possible doubt by the striae and associated glacial markings on the rock slope, especially in the vicinity of Artist Point. It is by a fortuitous circumstance, really, that these markings which date back to the El Portal stage of glaciation remain preserved. They have been protected from the weather by a thin mantle of glacial débris and are now being brought to light again by the gradual washing away of the débris. They are, however, by no means plain and might readily be overlooked, for the rock surface, though smooth on the whole, is closely pitted owing to the etching out of the biotite crystals in the granite. Only on the feldspar and the quartz do bits of polish and striae remain, and it is only under a favorable light that they can be clearly recognized.
Of greatest value as indicators of the direction of ice movement are the so-called "chatter marks," of which a few can be distinguished here and there. These are fine curving tension cracks in the rock produced by heavy boulders that were dragged by the glacier, and they are invariably bowed upstream. Unfortunately, the finest series of these chatter marks are now destroyed, for the ledge on which they were situated was blasted out when the Wawona Road, near Artist Point, was widened for automobile traffic.
The upward movement of the ice thus indicated was induced by the contraction of the spoon-shaped lower end of the valley. It must have required a strong propelling force, but such a force was generated by the piling up of the ice at the head of the valley, where there was a great ice cataract, as will be explained more fully presently. From the head of the valley, in fact, the surface of the Yosemite Glacier sloped forward uninterruptedly, as is attested by the gradual descent westward of the lateral moraines. Evidently the ice mass did not buckle over the obstructing rock barrier at the lower end of the valley. (See fig. 22.)
The square-cut head of the Yosemite Valley is more puzzling than the spoon-shaped lower end. That it was quarried out in large part from jointed rock is not directly demonstrable but can only be inferred from circumstantial evidence. The Half Dome quartz monzonite, in which it is carved, is notably massive over large areas and gives rise to numerous domes. The inclosing walls are prevailingly massive. The Royal Arches (pl. 21, B), on the north side, have a sheeted structure, but the sheeting is of an unusually massive type. The wall at the head of the chasm, which seems to be crossed by a bewildering maze of fractures, upon closer study is found to consist largely of massive rock that is merely exfoliatingcasting off scales from its exposed surface. The rounded conoidal mass which forms the basal part of the wall under Glacier Point (pl. 43, B) is a gigantic monolith comparable for unbroken continuity with the prow of El Capitan.
It seems probable, nevertheless, that the rock excavated from the head of the valley was divided by many fractures and on the whole readily quarriable, for the Half Dome quartz monzonite, though locally massive, is extremely varied in structure, being in some places rhythmically jointed and elsewhere closely sheeted or even intensely shattered. This is readily observed in the Little Yosemite, in Tenaya Canyonindeed, throughout the upper Yosemite region where the monzonite is the country rock. Several pronounced zones of shearing that must have penetrated deeply into the head of the valley are visible in its walls. One of these zones gives rise to the sharp recess that separates the Washington Column from the Royal Arches (pl. 21, B), and another is to be seen in the recess to the northwest of Grizzly Peak (pl. 16, B). A relatively broad belt of fractures that extends northeastward into Tenaya Canyon and facilitated the excavation of that canyon was prolonged in all likelihood some distance into the Yosemite Valley and must have facilitated its excavation also. But the strongest indication of the influence of fractures on the shaping of the valley head is found in the straightness and the orientation of the head wall. Only the presence of a north-south belt of fractures at the extreme head of the valley could have caused the quarrying glacier to produce so straight a wall trending in a direction unrelated to that of the ice movement. Had the rock been prevailingly massive, some parts of it would still remain projecting in the form of irregular salients. There is, further, indirect evidence of control by a north-south belt of fractures in the fact that at the south the wall terminates abruptly at the westward-trending zone of shearing which delimits the massive body of Grizzly Peak. This rock mass projects beyond the great façade, forming an obstruction which to the very last deflected the Merced Glacier and compelled it to twist itself through a narrow, tortuous gorge. All these facts and others of a similar nature would seem to warrant the conclusion that the head of the Yosemite Valley was excavated from prevailingly fractured rocks and that it owes its peculiar square-cut form to the controlling influence exerted by local fracture systems on the glacial quarrying.
Three other facts of prime importance remain to be explainedthe great depth to which the head of the valley has been excavated below the level of the preglacial gorge, a depth not less than 1,500 feet, it would appear from the longitudinal profiles in Plate 27, A; the steady decrease in the depth of glacial cutting from the head of the valley down to the lower end, where it measures only about 500 feet; and the scooping out of the basin of ancient Lake Yosemite in the rock floor of the valley. These matters, however, are all bound up with the questions, How are the stair-like steps with basined treads characteristic of profoundly glaciated canyons produced? It seems best, therefore, to defer their explanation until that question has been answered.
ORIGIN OF GLACIAL STAIRWAYS
The giant stairway from whose main steps the Vernal and Nevada Falls descend, impressive though it may be, taken by itself, is after all only the beginning of a much longer stairway that extends throughout the upper Merced Canyon, from the Yosemite Valley to the base of Mount Lyella stairway 21 miles in length and making a total ascent of 7,600 feet. This fact is clear from the longitudinal profile shown in Figure 22. Some of the steps in that greater stairway are ill-formed, none are as clean-cut as, those at the Vernal, and Nevada Falls, and nearly every one of them has a shallow basin hollowed out in its tread, yet the stair-like character of the canyon profile as a whole is unmistakable. Moreover, the floors of the main Yosemite and Little Yosemite are seen to constitute treads in that greater stairway. They do not differ materially from the other treads save in their greater length and in the fact that the basins in them are completely filled with stream-borne sediment, whereas on the upper treads the basins are filled only in part. The rock floor of the Yosemite Valley really comprises two treads differing but slightly in altitudea short one in the lower chamber and a long one in the upper chamber. The rock sill on which the moraine dam at the El Capitan Bridge rests forms the edge of the upper tread.
Such stairwise ascent by successive steps has long been recognized as a characteristic feature of strongly glaciated canyons, but the precise nature of the process whereby such canyon steps are produced is still a moot question. Several different hypotheses have been offered in explanation, but only two need be here outlinedthose of Willard D. Johnson and E. C. Andrews, both of whom, it is of interest to note were led to their conclusions largely through observations made in the Sierra Nevada.
Johnson64 believed that the glacial remodeling of a canyon is performed in large part by oft-recurring frosts, which split the rock and render its fragments available for removal, and that the glacier itself acts mainly as a transporting agent. Such frost action he believed to be sharply localized on the cross cliffs in the rock bed, because the glacier there breaks in its descent, and the crevasses permit the air, and with it the oscillations in temperature back and forth across the freezing point, to penetrate to the rock, whereas those parts of the rock bed which are covered by the unbroken body of the glacier are protected against such temperature changes. Each cross cliff would therefore be subjected to intense frost sapping, as he termed it, and in the course of time would be cut back by that process, receding gradually up the valley in the manner illustrated in Figure 32. Furthermore, he supposed that the frost attacks the floor at the foot of the cliff as well as the cliff itself, and that its tendency therefore is to produce a nearly level or even slightly backward-sloping stretch in the place of the original steeply graded canyon floor. With continuance of the process each of these nearly level or ponded stretches would be extended headward as the cross cliff above it receded and at the same time would be cut off at its lower end by the recession of the cross cliff below, and thus the canyon floor would eventually acquire a stairlike character. It is only fair to Johnson to add that he himself later abandoned this hypothesis, realizing that crevasses do not as a rule reach down to the bed of a glacier. A short time before his death he announced to the present writer that he was "about to make a violent attack on the Johnson frost-sapping hypothesis," but unfortunately he did not carry out his plan.
Andrews,65 on the other hand, supposed the cross cliffs to recede primarily as a result of the quarrying action of the glacier, which, he argued, must be particularly effective at the edge of each step, as the blocks there are unsupported on the downstream side. The tendency to excavate a basin in each tread he explained as being due chiefly to the great vigor with which the ice cascading from the step above abrades the rock bed at its foot.
According to both of these hypotheses, it is to be noted, the cross cliffs, or steps, would recede headward with some rapidity while glaciation was in progress. They would be essentially unstable, migrant features. The long basined rock floor of the Yosemite Valley would have resulted from the headward recession of a single cross cliff over a distance of 7 or 8 miles. Originating, presumably, as a low inconspicuous break in the floor near the lower end of the valley (a short distance below the constriction between El Capitan and the Cathedral Cliffs, Andrews thought), this cross cliff would have gained progressively in height as it receded almost horizontally into the steeply rising bottom of the preglacial canyon. According to Andrews's conception the cross cliff "must have been at least 2,000 feet in height" when it reached a point opposite the hanging valley of Yosemite Creek and, under the powerful erosive action of the great ice cascade that poured over it, it was cut back so rapidly that "the glacial energy had no time in which to excavate deep basins * * * but merely formed a huge tread with relatively shallow basins upon it." On passing Glacier Point the great cliff split into two parts, one part receding into Tenaya Canyon, the other along the course of the Merced, where it resolved itself into the flight of the giant stairway just before the ice age came to an end.66
It does not seem probable, however, that the glacial remodeling of the Yosemite Valley took place in just that way. The steps of the giant stairway can scarcely be migrant features that have reached their present positions by headward recession through distances of several miles and that would, by implication, promptly resume their rapid headward march if glaciation were renewed. They are composed not of fractured rock that would be readily quarried by a glacier or broken up by frost action but of massive rock that is scarcely susceptible either of being quarried or of being split by frost. The cliff over which the Vernal Fall leaps (pl. 24) consists of a monolith unmarred by a single fracture throughout its height of 300 feet or its breadth of 600 feet. The cliff of the Nevada Fall (pl. 25), which is twice as high and twice as broad, consists of massive rock for a height of about 400 feet from its base, and its upper portion is divided by only a few nearly horizontal fractures spaced from 50 to more than 100 feet apart.
It is significant, further, that all the other steps in the upper Merced Canyon have risers and sills composed of very sparingly jointed or wholly massive rock. Indeed, observations carried over a considerable part of the Sierra Nevada show that the same holds true for all canyon steps carved from its granitic rocks. Even in the areas of sedimentary and volcanic rocks the canyon steps have as a rule sills of more than ordinarily resistant materials. Particularly instructive in this regard is the flight of steps in Bloody Canyon, on the east flank of the Sierra Nevada, which is composed of a variety of rocks, sedimentary, volcanic, and granitic. In nearly every step the slightly raised edge or sill, from whatever material; it may be hewn, is associated with a constriction in the canyon section due to the resistance offered to glaciation by the same obdurate rock in the sides of the canyon. Such a constriction is to be seen at the Vernal Fall; and another is to be seen at the top of the giant stairway, where the mouth of the Little Yosemite is partly blocked by the obdurate masses of Liberty Cap and Mount Broderick. The treads, on the other hand, are almost invariably broad. They are the broadest parts of the canyons and, with their embayed sides and concave floors, constitute roughly spoon-shaped basins situated at successive levels one above another.
It must be clear to anyone who considers these facts that rock structure, or, more broadly, rock resistance, plays an important part in the development of canyon steps by glaciation; that, indeed, it determines in large measure at what points in a given canyon, the individual sills and treads shall develop. That being so, it follows that no hypothesis that aims to explain the production of glacial stairways can be considered satisfactory that fails to take into account this influence of rock resistance. Accordingly, a new explanation suggests itselfan explanation that is in harmony with the principle of selective quarrying already laid down.
Briefly stated, this new explanation is as follows: In a canyon or valley cut in granitic rocks of widely varying structure such as prevail in the Sierra Nevada, a glacier is bound to excavate with locally varying efficiency; where the rock is massive or only sparsely divided by fractures, the glacier, being unable to disrupt the rock, can reduce it only by abrasiona slow and relatively feeble process; on the other hand, where the rock is plentifully divided by natural partings, the glacier will quarry out entire blocks and excavate at a fairly rapid rate.67 From the first, therefore, it will tend to work irregularly, producing hollows in the areas of jointed rock and leaving obstructing humps in the areas of massive, unquarriable rock. The humps, however, will tend to assume strongly asymmetric forms, gently sloping and smooth on the upstream side, abrupt and more or less hackly on the downstream side, for, as will be seen in the diagram in Figure 33, the force of the ice is directed at an angle against the upstream side and so subjects it to intense abrasion; and the force is directed at a small angle away from the downstream side and so exerts there a pull favorable for quarrying.
The smaller knobs of this asymmetric type have come to be known by the quaint name "roches moutonnées," which was given to them by the Swiss mountaineers because, when viewed from up the valley, their rounded forms suggest the backs of grazing sheep. That they owe their peculiar modeling to abrasion on the upstream side and quarrying on the downstream side is quite generally recognized, but that the larger obstructions which occupy the entire breadth of the canyon floor and form the sills and edges of the steps are shaped in essentially the same way appears not to be generally understood. In the upper Yosemite region and the adjoining parts of the High Sierra, however, moutonnée forms of all sizes abound, ranging from mere hillocks and rock waves a few feet high to canyon steps a thousand feet high, and there the inherent kinship of all these features is readily manifest to one who takes the trouble to compare them with one another.
A typical example that is intermediate between a mere roche moutonnée and an entire canyon step and is readily accessible for inspection is the abrupt rise in the canyon floor halfway between the Vernal Fall and the Nevada Fall, which was the site of the historic hostelry known as La Casa Nevada. It has been referred to as the second step of the giant stairway, but it extends really only part of the way across the canyon. The trail leads steeply up to the top of the sill through a notch in the downstream face, which is determined by strong vertical master fractures, and then it leads down again along the gentle back slope of sparsely fractured granite.
The manner in which the basined treads of a glacial stairway are evolved will be most readily understood if the treads are viewed in reference to the slope of the preglacial canyon floor. If the diagram in Figure 34 is so tilted as to make the preglacial canyon floor appear horizontal, the treads will assume the aspect of basins that are strongly asymmetric, being deepest at their upper ends. They are so shaped manifestly because the ice erodes with greatest vigor at their upper ends; it is thicker there than at their lower ends, descends into them with plunging motion, and at the foot of its cascade is compelled to make an abrupt turn, as is indicated by the arrowcircumstances all of which cause the ice to exert particularly great pressure on its bed. Downstream, of course, the pressure diminishes progressively, reaching a minimum at the edge of the step.
Treads are shaped both by quarrying and by abrasion, but it is a fair presumption that the quarrying process is dominant wherever the rock is jointed and the glacier has sufficient power to dislodge the blocks. In Figure 34 several stages in the evolution of a tread are shown in order to bring out the fact that, as the quarrying proceeds, always in the headward direction, numerous minor cross cliffs controlled by joints are likely to be developed. But these are only temporary hackles in the canyon floor that migrate headward and are eliminated in the course of time. The main canyon steps, on the other hand, are seen to be fairly stable features that are cut back only very slowly, their sills and risers being composed of the more massive and therefore more obdurate rock. Likewise they are worn down very slowlymainly by abrasionand the sill of each step therefore controls the general level of the tread back of it. No other hypothesis advanced hitherto has explained what determines the level at which a tread shall be developed.
If is evident, further, that the process is a self-intensifying one, which constantly tends further to accentuate the stairlike profile of the canyon. For the greater the depth of excavation at the foot of a given step becomes the more powerful will be the glacial action there. The limits to which the process can go in any locality are determined, of course, by the force of the glacier, the length of time it is active, and the resistance of the rocks. Finally, there is no inherent tendency in the process to produce level or nearly level treads; it works regardless of gradients. It may produce treads that are approximately level, or that have a gentle slope, either forward or backward, or that have one or several basins scooped out in them, all depending upon the structure of the rocks involved and the degree to which that structure controls the glacial action.
APPLICATION TO THE YOSEMITE VALLEY
When this conception of the way in which glaciers produce stairlike steps in canyons is applied to the Yosemite Valley, it will be readily seen that the basined rock floor of that valley is a typical canyon tread excavated in a long stretch of prevailingly well-jointed rocks dominated at its head by a large body of massive granite from which the ice plunged down abruptly; also that the great depth of glacial excavation at the head of the valleyabout 1,500 feetis due primarily to that plunging movement. But it was not the descent from the giant stairway that endowed the ice with its great excavating power; the energy which the ice derived from that descent it expended largely in excavating the narrow gorge immediately below the stairway. It was the much greater plunge from the platform at the west base of Half Domethe platform above the head wall of the valleythat generated most of the power. (See fig. 35.)
During each of the earlier ice floods the Merced Glacier and the Tenaya Glacier swelled to such great volume that they coalesced into one vast ice field above which only the crown of Half Dome was visible. At those times the ice covered the platform overlooking the head of the valley to a depth of fully 1,000 feet and thence plunged in the form of a mighty cataract, a glacial Niagara. During the last ice flood, which had relatively moderate volume, the two glaciers did not coalesce over the platform but made separate entry into the Yosemite Valley, each through its own portal. The Merced Glacier then descended wholly by way of the giant stairway and the gorge below. Figure 35 affords an impressive comparison of the mighty ice cataract that plunged over the head wall and the relatively modest ice cascades that tumbled from the steps of the giant stairway.
From the longitudinal profile in Figure 35, it is clear further, that even at the beginning of its career the great ice cataract made a descent of not less than 1,500 feet, the platform from which it plunged having an altitude of about 7,000 feet and the preglacial gorge of the Merced River below an altitude of about 5,500 feet. Of course, the descent at first was not nearly so abrupt as it grew to be later, after the valley head had been enlarged by glacial quarrying and the head wall had acquired a steep profile; still the descent was abrupt enough to cause the ice to break and to develop considerable power. (See fig. 15.) As the valley head was excavated to greater and greater depththe platform above, meanwhile, being worn down but littlethe height of the cataract was increased to 2,000 feet and finally to more than 2,500 feet.
The great depth of glacial excavation at the head of the Yosemite Valley is thus amply accounted for; at no other place was there an ice cataract so high and so powerful. Nevertheless, the work is not all to be credited to the ice cataract. A considerable share must have been performed by glacial action of a less spectacular type. During the long periods of moderate ice supply that preceded and followed the maxima there was not ice enough to produce a cataract, and the Merced and Tenaya Glaciers, entering the valley by separate portals, eroded each its own pathway as far down as their junction. Later, after the spur at the head of the valley had been cut away, these glaciers debouched directly into a great ice pool that occupied the valley head, their currents meeting to form one central current of great eroding power. The depth of ice at the point of junction was always considerable. It amounted to about 1,500 feet even during the last glacial stage, when the Yosemite Glacier reached no farther than the Bridalveil Meadow. The transformation of the valley head is therefore to be regarded as having been brought about by glacial action of different kinds, that of the plunging ice cataract being the most powerful though not the most lasting.
It remains to account for the remarkably clean-cut forms of the steps of the giant stairway. These steps have remained standing because they are composed of massive, unquarriable rock. Yet all three have conspicuously sheer, smooth fronts, or risers, and well-defined, straight edges clearly indicative of quarrying controlled by joint fractures. Close examination of the steps reveals, indeed, that the risers of all three are controlled by vertical master joints. The step at the Vernal Fall affords the simplest case; its riser coincides with a single joint plane, belonging to the northwestward-trending system which prevails in many parts of the Yosemite region. This fact explains not only the verticality and straightness of the riser but also its orientation diagonally across the canyon, in disregard of the direction of ice movement. The riser of the second stepthe one above the Diamond Cascadeowes its orientation, also at an angle to the axis of the canyon, to a master joint of the northeastward-trending system. As is obvious from the contouring on Plate 7, this step is directly alined with the narrow gash between Liberty Cap and Mount Broderick, which has been cut along a zone of northeasterly joints. The step from which the Nevada Fall leaps is not so obviously controlled by a master fracture as those just described, for the lower portion of its riser slants at a considerable angle to the vertical and is composed of massive, exfoliating granite. But the upper portion has a verticality and straightness that could not possibly have been produced by the adventitious breaking of massive rock under the stresses imposed by a cascading glacier. Besides, it trends northeastward, roughly parallel to the riser of the second step, and thus it also appears to have been controlled by a master joint of the northeasterly system.
The explanation of the clean-cut character of these three steps is, then, that although the steps themselves are composed of massive rock, there was formerly in front of each of them a body of vertically jointed or sheeted rock that was readily quarried by the glacier. In other words, at each step the glacier quarried directly up to the massive rock. To the cliff of the Vernal Fall, notably, there still clings a large fragment of a vertical rock sheet that tells the story clearly.
DEVELOPMENT OF THE LITTLE YOSEMITE VALLEY
The position of the Little Yosemite at a level 2,000 feet above the main Yosemite, which gives it the appearance of a hanging valley, although it is in fact the path of the master stream, seems no longer anomalous in the light of the foregoing explanation of the development of glacial stairways. Its nearly level floor is one of the treads in the long glacial stairway that extends throughout the upper Merced Canyon. The general level of that floor was determined mainly by the sill of massive granite at the mouth of the valleythat is, by the top of the body of granite from which the precipice at the Nevada Fall is hewn. So exceedingly resistant to glaciation was that body of massive rock that it has been worn down only about 250 feet below its preglacial level. (See pl. 27, A.)
However, not only the mouth of the Little Yosemite but a large share of its floor appears to be composed of massive granite which the glacier found difficult to excavate. In the few places where the floor is not concealed by sand and gravel, as in the vicinity of the Sunrise Trail and opposite Sugar Loaf, it is seen to be made of massive granite; and throughout most of the valley the cover of sediment is so thin that the crests of the small frontal moraines left by the Merced Glacier (see pl. 29) project above it.
The great breadth of the Little Yosemite, on the other hand, was produced, unquestionably, by effective lateral quarrying in masses of well-jointed rock. The lower half of the valley, which is almost as broad as the Yosemite itself, probably was broad even in preglacial time, for its sides have only moderate declivity; but this statement does not invalidate the preceding oneit rather implies that the glacial as well as the preglacial processes of erosion were favored in lateral cutting by the prevailingly fractured condition of the rock.
The two bosses that obstruct the mouth of the Little YosemiteMount Broderick and Liberty Caphave survived the onslaughts of the glacier by virtue of the exceedingly resistant nature of their massive rocks. Rounded and smoothed on the up-valley side (pl. 44, A), sheer and angular on the down-valley side (pl. 44, B), they are typical roches moutonnées, but they are so enormous compared with most knobs of that kind that it would seem more appropriate to liken them to elephants than to sheep. Liberty Cap stands nearly 1,000 feet above the floor of the Little Yosemite; Mount Broderick about 600 feet.
That Mount Broderick consists essentially of one great monolith is evident at a glance, but the massive nature of Liberty Cap is not perhaps so readily apparent. Its rounded back is cut by several horizontal master joints, and its sheer front is hackled by a series of nearly vertical fractures that begin at a prominent shear plane near the base (pl. 44, B); still, a study of the boss in its entirety leaves little doubt that none of the joints and fractures mentioned penetrate far into its interior. They are indicative, rather, of structural features that traversed the rock which formerly enveloped Liberty Cap and which has been quarried away by the glacier.
The control exerted on the glacial quarrying by fractures in the surrounding granite is clearly revealed also in the sheer fronts and equally sheer sides of the two bosses. As a glance at Plate 7 will show, their fronts are controlled in the main by northwestward-trending fractures and their sides by northeastward-trending fractures.
It is manifest, further, that the narrow cleft which separates Mount Broderick from Liberty Cap has been gouged out along a zone of northeastward-trending fractures. Its sharp V shape might seem suggestive of stream erosion rather than of glacial action, yet there can be little doubt that this cleft is a product mainly of selective quarrying. Prior to the ice age, probably, Mount Broderick and Liberty Cap formed part of a continuous ridge, or spur, that projected southeastward from the base of Half Dome. Instead of a cleft there was then only a shallow saddle from which ravines descended in opposite directions. The Merced Glacier, when it overtopped the ridge, naturally singled out the unresistant, slivered rock in the narrow zone and quickly deepened the saddle to an acute notch. Though downward its excavating action was, facilitated by the fractures, sideward it was restrained by the massive, unquarriable bosses, and so perforce the glacier gave rise to a strikingly unglacial-looking cleft. Its scorings remain displayed on the smooth walls of the cleft, most vividly on the side of Mount Broderick.
Running water doubtless has played some part in the cutting of the cleft but only a minor part. Whenever the Little Yosemite was filled with ice as far down as its mouth, some water must have escaped through the unfinished notch, supplementing the glacier's action in a feeble way. And when finally it was cut down to the floor of the valley, the cleft probably became the temporary channel of a larger stream derived from the melting glacier. To-day it is traversed by only an insignificant streamletthat which issues from the spring-fed pool known as Lost Lake.
At the south base of Liberty Cap is another, smaller cleft, or gorgenamely, that through which the zigzag trail leading to the Little Yosemite is laid. This gorge also dates back to glacial time, but as it belongs to a subordinate class of sculptural features which are appropriately treated together, its explanation will be deferred to another place (p. 113).
The upper half of the Little Yosemite is narrower and steeper sided than the lower half, because in it lateral quarrying by the glacier was impeded by flanking bodies of massive rock. The walls of this part of the valley are exceptionally massive; few cliffs elsewhere in the Yosemite region exhibit a more complete absence of fractures. In the Cascade Cliffs, for instance (pl. 45, A), the rock is divided by only one approximately horizontal master joint at a height of about 600 feet from the base, and under Bunnell Point (pl. 45, B) the wall is unmarred by a single fracture throughout its height of 2,000 feet. Innumerable scales cling to the faces of all these cliffs, it is true, but they are purely surficial features due to exfoliation, and the partings behind them penetrate the rock to only slight depth. Indeed, the very presence of these scales affords proof of the wholly massive nature of the rock bodies to which they adhere, for only massive rock exfoliates in this manner.
The almost complete absence of fractures in the walls of the upper Little Yosemite accounts also for the equally complete absence of vegetation on them. They afford no roothold for either trees or bushes and in consequence are strikingly bare. Moreover, as they are also devoid of prominent sculptural details or angular features, they present a singularly blank, expressionless aspect. The prevalence of smooth, gently curving surfaces is due in part to the grinding action of the glacier, in part to exfoliation of the rock.
The two promontories on the south wall known as the Cascade Cliffs (pl. 11, B) obviously are stubs of spurs that were neatly cut off by the glacier. They consist largely of massive rock (the western cliff exhibits imperfectly developed diagonal joints), but the parts that have been removed from them probably consisted of jointed rock. Indeed, the straight, sheer faces of the promontories clearly betray the influence of vertical joints of the northeastward-trending system, and there is thus reason to suppose that the points of the spurs were traversed by other joints of that system which facilitated and guided the glacial quarrying.
The north wall of the Little Yosemite has been planed off less severely than the south wall. From it projects, near the head of the valley, that remarkable dome-crowned spur of bare granite known as Sugar Loaf. (See pls. 31, A, and 45, B.) Although repeatedly overridden by the glacier, this spur, which is massive throughout save for a single horizontal master joint, still stands 1,300 feet high and partly blocks the valley.
Less remarkable scenically but even more instructive is the broad spur that projects southward from Moraine Dome, opposite the Cascade Cliffs. This spur, too, has been repeatedly overriden by the Merced Glacier, as is strikingly attested by the lateral moraines that curve across its back. (See pl. 29.) The topmost of the series, the great embankment that marks the highest level attained by the Merced Glacier during the last glaciation, is fully a quarter of a mile back from the end of the spur, and accordingly it is clear that the spur, even in its present blunted form, projected at least a quarter of a mile into the body of the glacier.
To obtain a true appreciation of the stout resistance which this spur has offered to reduction by the glacial processes one must view it also in its relations to the margins of the very broad Merced Glacier of the earlier ice stage. This may be best done in a cross section such as that in Figure 36, which shows that this earlier glacier spread over the upland to the north of the spur for a distance of fully 1-1/2 miles. Southward the glacier spread over the upland back of the Cascade Cliffs to a distance of three-quarters of a mile, and thus it had a T-shaped cross section resembling somewhat the figure of a mushroom with massive stem and broad, flat cap.
The great disparity in breadth between the Little Yosemite Valley and the earlier Merced Glacier thus revealed is most enlightening to the student of glacial phenomena. For it shows that even so capacious and so perfect a U trough as the Little Yosemite does not afford an accurate gage of the maximum volume attained by the glacier whose pathway it was. The Little Yosemite appears to have been elaborated mainly by the prolonged and doubtless oft-repeated action of glaciers of moderate volume, and whenever during times of particularly great snow accumulation the ice exceeded such volume, it found the trough too small and deployed over the adjoining uplands, as a river at high stages deploys over its flood plain. The exceeding stubbornness of the massive granitic rocks evidently prevented these relatively rare and short-lived ice floods from enlarging the cross section of the valley to suit their volume. The same has happened in other yosemites in the Sierra Nevada. The Hetch Hetchy, notably, was completely overwhelmeddeeply buried, evenby the later as well as by the earlier Tuolumne Glacier. Its cross section, therefore, affords no measure whatever of the maximum volume attained by that glacier. The Yosemite Valley, on the other hand, is conspicuous as one of the few U troughs that had sufficient capacity to accommodate the entire volume of its glacier, even at times of maximum ice accumulationbecause in its generally fractured rocks glacial quarrying proceeded with greater facility than in the rocks of the other valleys.
Rock resistance to glaciation reached a climax at the head of the Little Yosemite. The narrow upper gorge of the Merced (fig. 37), which begins there, is carved. into a vast body of massive Half Dome quartz monzonite that obstructed the path of the Merced Glacier in precisely the same way in which the body of massive El Capitan granite into which the lower Merced Gorge is cut obstructed the path of the Yosemite Glacier. Over this unquarriable body of monzonite the Merced Glacier was compelled to lift a considerable part of its mass, just as the Yosemite Glacier had to lift part of its mass over the unquarriable body of granite at the lower end of the Yosemite Valley. The two gorges, it will be seen, are analogous features; both are essentially stream-worn gorges but slightly enlarged by glaciation. They differ mainly in that the upper gorge, having been glaciated in the last ice stage as well as in the earlier, still has smoothed and in part polished walls, whereas the lower gorge, having remained untouched by the later ice, has lost all evidence of glacial abrasion from its walls.
The question may here be raised, Why is not the head of the Little Yosemite, like that of the main Yosemite, marked by a great cross cliff over which the ice plunged abruptly in glacial time? Is it not to be inferred from the explanation given on page 97 of the mode of development of the head of the main valley that such a cross cliff is a necessary feature for the development of any spacious Yosemite-like valley chamber with approximately level floor? The answer is that during glacial time there were indeed great ice cascades at the head of the Little Yosemite. They plunged from the broad, uneven benches of massive rock that flank the narrow gorge above and that come abruptly to an end at the head of the valley; but the cliffs bordering those benches do not extend at right angles to the axis of the valley but flare out irregularly, doubtless in obedience to the local structure.
DEVELOPMENT OF TENAYA CANYON
Tenaya Canyon owes its great depth unquestionably to the fact that throughout its length both stream and glacier have eroded with comparative facility owing to the presence of a narrow belt of fracturing in the Half Dome quartz monzonite. In preglacial time Tenaya Creek, accelerated by the successive uplifts of the Sierra Nevada, deepened the canyon almost as rapidly as the Merced River deepened the Yosemite Valley. The valley of Snow Creek in consequence remained hanging a thousand feet above the canyon bottom. Later the Tenaya Glacier effectively quarried the fractured rocks, adding greatly to the depth of the canyon, though but little to its width. To-day Tenaya Creek is again rapidly intrenching itself. Since the ice age that little stream has cut into the canyon step at the base of Mount Watkins a gorge 100 to 300 feet in depth, in striking contrast to the almost negligible depth to which the relatively powerful Merced River has cut the massive rock of the giant stairway.
The irregularity of the trenched steps in Tenaya Canyon also is due to the prevailingly fractured state of the rock, which has permitted the glacier to quarry more or less effectively throughout the length of the canyon. Nor are these ill-formed, trenched steps at all comparable in point of permanence to the clean-cut, monolithic steps of the giant stairway. Were glacial conditions to return some day, doubtless they would be cut back appreciably and also reduced in height by vigorous quarrying; whereas the steps of the giant stairway, being unquarriable, would be but slowly worn by abrasion. Indeed, in case of renewed glaciation, probably all of Tenaya Canyon would again be excavated to an appreciably greater depth, whereas the Little Yosemite would be excavated very slightly.
Entirely different from the low, irregular steps just mentioned is the abrupt 600-foot step that marks the head of Tenaya Canyon and the beginning of the upper Tenaya basin. As is evident from its smooth, unbroken front, down which the Tenaya Cascade glides and evidently has been gliding for thousands of years, without carving a notch, this great step is composed of the most massive, durable kind of rock. Like the steps of the giant stairway it was only moderately reduced in height by glacial abrasion, whereas the fractured rock masses in front of it were quarried away in wholesale fashion. It differs from the steps of the giant stairway mainly in that its edge is rounded instead of square, its top convexly curved, and its front concavely curved, in profile as well as in plan. Nor is there any clear evidence in its smooth-flowing outlines of direct control by master fractures.
Tenaya Canyon has been cited as a fine example of a glacial U trough, but this characterization is justly applied only to its broad-floored lower portion in which Mirror Lake is situated. (See pl. 8, A.) There, evidently, the Tenaya Glacier was able to quarry both broadly and deeply, owing to the presence of considerable masses of jointed rock. There also it gouged out in the rock floor a shallow basin analogous to the glacial basins in the main Yosemite and the Little Yosemiteat least such a basin appears to be indicated by the long stretch of level, sandy floor above Mirror Lake. Of that basin Mirror Lake might seem at first to occupy the unfilled remnantindeed, it has been long regarded as a lake of glacial origin, but, as is explained in greater detail on page 105, it is now known to have been formed after the ice age by a dam composed of rock débris that fell in avalanches from the walls of the canyon.
Throughout its middle course Tenaya Canyon is narrower than near its mouth and its cross section resembles that of a sharp-keeled boat, as may be seen from the summits of the Quarter Domes. (See pl. 46, A.) This peculiar shape is due in part to the incision of the narrow inner gorge in the bottom of the canyon, in part to the angle at which great bodies of massive rock slope out from the bases of the walls toward the middle. It is the smoothness, the continuity, and the prevailing steepness of these flanking rock slopes, no less than the ruggedness of the central gorge, that render Tenaya Canyon so extremely toilsome and hazardous to traverse. They call for ingenuity and daring on the part of those who would attempt to fight their way through the entire length of the chasm.68
Toward its head, again, Tenaya Canyon widens out somewhat and assumes an approximate U shape. There its floor ascends by two low yet well-defined steps, each with a rounded edge of massive granite and a filled lake basin on its tread. This part of the canyon is directly overlooked by the great cliffs that inclose its head. It was the bowl into which the converging ice cascades plunged from above. That it was not gouged out to greater breadth in spite of the tremendous eroding power developed by the plunging ice is explained, manifestly, by the exceptionally resistant nature of its massive walls.
Tenaya Canyon as a whole presents the most stupendous exhibit of massive granite in the entire Yosemite region, perhaps in the entire Sierra Nevada. (See pls. 46, B, and 40, B.) Not even the Grand Canyon of the Tuolumne nor the Kings River Canyon is hemmed in by walls so uniformly and continuously massive. It is no wonder, therefore, that Tenaya Canyon has remained on the whole narrow in proportion to its depth. It affords in that respect a direct antithesis to the Little Yosemite: in the lower part of that valley the ice was permitted to quarry broadly but not to great depth; in Tenaya Canyon the ice was permitted to quarry deeply but not to great breadth.
SUMMARY OF CHANGES PRODUCED BY THE GLACIERS
The changes that were brought about in the configuration of the Yosemite region by the repeated ice invasions of the glacial epoch may be summed up as follows:
In the Yosemite Valley itself, where the rocks were prevailingly well jointed, glacial quarrying was particularly effective and accomplished conspicuously large results. Both downward and sideward the Yosemite Glacier quarried, trimming off projecting spurs, cutting back the craggy slopes of the preglacial river canyon to sheer, smooth cliffs, and transforming the brawling cascades descending from the hanging side valleys into leaping falls of astounding height. Throughout the length of the valley the inner gorge of the Merced was wiped out of existence and in its stead there was produced a broadly concave, basin-shaped rock floor. Even the features of the mountain valley of the Pliocene epoch were largely destroyed, and thus from a tortuous V canyon the Yosemite was enlarged to a spacious, moderately sinuous U trough with approximately parallel, spurless sides and with a long lake basin in its bottom. Only between El Capitan and the Cathedral Rocks, composed of massive rock which the glacier was unable to quarry away, did it leave a marked constriction. In the areas of sparsely jointed granite immediately above and immediately below the Yosemite Valley the ice was able to effect but moderate changes in the form of the canyon. There in consequence the inner gorge of the Merced remains preserved and still presents the characteristics of a trench worn by the river in the bottom of an old mountain valley.
In the Yosemite Valley the depth of glacial excavation decreases from a maximum of about 1,500 feet at the head to a minimum of about 500 feet at the lower end. The reason for this decrease in glacial deepening down the valley is found in the fact that during all phases of glaciation the ice was thicker and therefore had greater excavating power at the head of the valley than at the lower end, and during the maximum phases it plunged into the head of the valley in the form of a mighty cataract.
Above the Yosemite Valley the Merced Glacier sculptured the steps of the giant stairway from local bodies of extremely massive rock. By quarrying headward directly up to the vertical master joints delimiting those bodies and by grinding them down from above, it gave them their marvelously clean cut, steplike forms.
In the lower half of the Little Yosemite lateral quarrying was favored but downward quarrying was impeded by the structure of the rocks, and as a consequence that part of the valley was given great breadth but relatively little depth. Liberty Cap and Mount Broderick, being composed almost wholly of massive, unquarriable rock, were left standing at its mouth as two roches moutonnées of gigantic size. The upper half of the Little Yosemite has remained narrower than the lower half because it is hemmed in by extensive bodies of massive rock. The head of the Little Yosemite was, like the head of the main Yosemite, the site of a powerful ice cataract and to that circumstance largely owes its great depth and peculiar configuration. In the Little Yosemite as in the main Yosemite the depth of glacial excavation decreases steadily down the valley and for the same reasons.
In Tenaya Canyon downward quarrying was favored by the presence of a longitudinal belt of fractures, but lateral quarrying was restricted by flanking masses of undivided rock. As a consequence that canyon was cut down almost to the level of the Yosemite Valley and was given even greater depth than that valley, but it was left comparatively narrow and in part shaped like a sharp-keeled boat rather than like a U trough. Only an imperfect stairway was developed in its floor, but at its head a gigantic step was fashioned from a body of massive granite under the cascading action of the Tenaya Glacier.
The Yosemite, the Little Yosemite, and Tenaya Canyon, then, owe their present, configuration very largely to glacial action. All three had been cut in preglacial time to considerable depth by the streams flowing through them and had acquired the aspect of rugged V canyons, but so thoroughgoing was the remodeling action of the glaciers that now only a few vestiges of their preglacial forms remain. Each chasm had a glacial cataract at its head, yet each differs from the other two in general shape and proportions because in each the structure of the rocks was different and influenced the glacial processes in a different way.
Last Updated: 28-Nov-2006