POSTGLACIAL HISTORY OF THE YOSEMITE VALLEY
EVOLUTION OF CLIFF SCULPTURE
The detail sculpture of the walls of the Yosemite is by no means all of postglacial origin. Much of it was produced during the last stage of glaciation, when the valley was filled with ice only to one-third of its depth and when its cliffs were exposed to particularly intense frost action and to frequent wear by snow avalanches. Much of the sculpturing dates also from the relatively long interglacial stage that preceded, and the deeper recesses doubtless were initiated long before the ice age, while the Yosemite was passing through the mountain-valley and canyon stages of its earlier development. Nevertheless all its sculptured features may appropriately be treated together here, the purpose being rather to explain how they have been formed than to trace their origins.
RELATION OF FORM TO STRUCTURE
That the verticality of the walls of the Yosemite is due in large measure to the tendency of its granitic rocks to split along vertical partings, or joints, was first noted by Prof. Joseph Le Conte.71 The influence of the joints in determining sheer cliff profiles and angular rock forms was recognized also by Becker72 and Turner,73 who studied the jointing of the granite in the Yosemite region and the adjoining parts of the Sierra Nevada somewhat more closely. But until the topographic map of the valley (pl. 7) was made, there was available no detailed cartographic representation of the Yosemite cliffs on which the different joint systems and their effect upon the sculpture might be traced in detail. It is proposed here with the aid of this map and the available photographs to make a rapid survey of the cliffs, noting the more conspicuous instances of sculptural forms controlled by joint structure.
Sentinel Rock (pl. 19, A) affords an outstanding example. Its sheer, smooth cliff face is determined largely by a single nearly vertical joint plane. The granodiorite of which it is formed has a sheeted structure, the nearly vertical joints, which trend in a northeasterly direction (see pl. 7), being developed to the exclusion of almost all joints in other directions. The influence of these northeasterly joints is evident also in the splintered crest of the spur of which Sentinel Rock forms part, in the sculpture of the rock mass at its base, in the steps over which Sentinel Creek cascades, and in the truncated spurs for some distance along the south side of the valley.
In the neighborhood of Taft Point (pl. 8, B) oblique joints dipping 40°-50° W. begin to influence the cliff sculpture, and all the spurs are consequently more or less conspicuously asymmetric. The sloping west side of each spur coincides with one or more oblique joint planes, the precipitous east side is defined by a northwestward-trending vertical master joint, and the point is truncated by a sheer facet of the northeasterly joint system. All the minor sculptural features that diversify the spurs are similarly bounded by fractures of the three sets mentioned and are of asymmetric, rhombic pattern. Even stream erosion is here guided by structural planes. The streamlet that descends through the asymmetric gorge east of Profile Cliff cuts not vertically but obliquely downward, sliding side-wise, so to speak, along the plane of an inclined master joint. The overhang of Profile Cliff and of the entire west wall of the gorge is due to the undercutting action of the streamlet.
The astounding fissures that gash the overhanging wall and edge of the gorge near Profile Cliff have developed along vertical joints of the northeasterly system. Probably the rock immediately adjacent to these partings was peculiarly susceptible to weathering, having been minutely sheared and slivered by faulting movements that took place under great pressure shortly after the granite had solidified. As a consequence the joints, which were at first mere cracks too narrow for a knife blade to enter, have become enlarged to gaping abysses too wide for a man to stride across. Their enlargement is the more surprising in view of the fact that no streams have ever made their way through them, the surface of the upland here sloping to the west, away from the edge of the gorge.
Oblique master joints, dipping to the west, account also for the asymmetric forms of the Cathedral Rocks. (See pl. 3.) Eastward-dipping joints, on the other hand, determine the angle of the rock slope on the west side of Bridalveil Creek. The V-shaped gulch of that stream owes its wonderful symmetry largely to the fact that the oblique joints controlling the inclination of each of its two sides dip toward one another at approximately the same angle. The finest example of asymmetric sculpture called forth by oblique master joints is presented by the massif of the Three Brothers. (See pl. 19, B.) Its three successive roofs slant with architectural regularity, all at a uniform angle, because they are determined by joint planes having the same westerly dip. The gables, on the other hand, are carved along vertical joint planes trending in a northeasterly direction. (See pl. 7.)
In the embayment east of the Cathedral Rocks (pl. 8, B), where diorite and gabbro are the prevailing rocks, the jointing is intense and chaotic, and the sculpture in consequence is intricate and irregular. The wall of the valley here is so thoroughly dissected by deep, ramifying gulches that it is reduced to a bewildering mass of craggy, slivered spurs. Some of the outstanding pinnacles consist of Bridalveil granite, sheets of which cut through the other rocks. Of this resistant and relatively massive granite are made also the twin shafts of the Cathedral Spires (pl. 18), which alone have survived the wholesale dismantling that has taken place roundabout them.
In the lower Yosemite chamber vertical joints predominate. Exclusive control by eastward and northward trending vertical joints is seen in the sheer square-cut cliffs that frame the recess of the Ribbon Fall. Easterly and northerly joints, combined with northeasterly and northwesterly joints, have influenced the shaping of the cliffs on the south side. So strong has been their directive control that in several places, notably below Crocker Point, the cliffs have been quarried back by the glaciers and have since been chiseled in detail irrespective of the positions of the drainage lines on the slopes above. In the Merced Gorge northwesterly and northeasterly joints of approximately equal strength have given rise to the clean-cut, prismatic forms along the edge of Turtleback Dome. (See pls. 43, A, and 7.)
In the upper Yosemite chamber the most impressive example of a rock face determined by a single master joint is found in the great cliff at Glacier Point, shown in Plate 20, B. Though not without minor irregularities, it coincides throughout most of its height and length with one vertical joint plane trending almost due east. (See pl. 7.) The projecting slabs at its top (pls. 9, B, and 20, B) are explained by the local development of horizontal fractures that interfere with the continuity of the vertical master joint. They were left overhanging when the last vertical rock sheet fell from the cliff face.
Below the main cliff at Glacier Point is a second, even higher cliff (about 1,200 feet) also defined by an eastward-trending master joint; and a short distance to the west is the singularly straight, narrow cleft through which the Ledge Trail leads steeply to Glacier Pointa cleft caused by weathering and erosion along a nearly vertical master fracture, probably a fault, that traverses the rock in a southeasterly direction.
In striking contrast to the colossal rock forms just mentioned are the finely chiseled, castellated, and columnar features in the vicinity of Moran Point and Union Point, which have been called forth by numerous intersecting joints, vertical, oblique, and horizontal. The best-known representative of this type of sculpture is the Agassiz Column (pl. 47, B) which stands, precariously balanced on a crumbling base, near the Glacier Point Short Trail, just below and east of Union Point. Every facet of this rock is determined by a joint plane.
Further scrutiny deserves to be given to the view in Plate 47, B, for the clear relations between sculpture and structure in the cliffs about the Yosemite Falls that are revealed in its background. The alinement of the bushes in rows gives prominence to the few fractures that traverse the otherwise massive, bare rock and shows plainly that the two terraces to the right and the left of the Lower Yosemite Fall are defined by horizontal master joints, whereas the incline west of the Upper Yosemite Fall coincides with a zone of oblique master joints. These features are of peculiar human interest because, were it not for them, the building of a trail to the top of the Yosemite Falls (see pl. 7) would have been an extremely difficult feat to accomplish. As it is, the trail builder has taken advantage of the upper rock terrace to carry the trail from Columbia Rock to the embayment of the upper fall, and the upper flight of zigzags he has constructed on the incline produced by the oblique joints. (See also pl. 9, B.)
The 1,500-foot precipice over which the Upper Yosemite Fall (pl. 22) leaps is determined, like the cliff at Glacier Point, by an eastward-trending master joint. It is not strictly vertical but inclined at an angle of about 80°. Only the west half, which is composed of Sentinel granodiorite, approximates a plane, smooth wall, the east half, which is composed of El Capitan granite, being diversified by overlapping rock sheets. A narrow ledge defined by a horizontal master jointMuir's famous Fern Ledgeextends westward across the face to and slightly beyond the path of the fall, about 450 feet above the base. This ledge is the most advanced part of the cliff, yet it is cleared by the main body of the fall, owing to the parabolic descent of the water. Below the ledge the cliff overhangs or leans out at angles ranging from 10° to 30° and at its base there is a deep cavern developed along another horizontal master joint.
A fuller description and interpretation of the features of this remarkable cliff can not be given here; it must suffice to point out that the fall has failed to carve a recess because the cliff as a whole has receded by the scaling off of immense rock sheets. The stubs of a considerable number of these are to be seen in the projecting buttress under Yosemite Point. The most remarkable remnant of such a sheet, however, is the gigantic tapering rock monument known as the Lost Arrow (pl. 22), which clings to the cliff east of the fall. It has a total height of about 1,500 feet, and its upper third stands detached, like a pinnacle, the parting behind it having been enlarged to an open cleft, doubtless as a result largely of the destructive action of the spray from the fall which freezes in it in the winter.
It goes almost without saying that every one of the major waterfalls of the leaping type in the Yosemite region is, like the Upper Yosemite Fall, associated with a cliff determined by a vertical or nearly vertical master joint. The Lower Yosemite Fall (pl. 23, C) leaps over a northwestward-trending cliff; the Bridalveil (pl. 23, A) and Illilouette Falls, each over a northeastward-trending cliff; the Vernal Fall (pl. 10), again, over a northwestward-trending cliff. The precipice over which the Nevada Fall (pls. 10 and 25) descends is not so clearly determined by a master joint as the others just named, but its straight front and northeasterly trend unquestionably betray the influence of a northeasterly joint plane. At nearly every fall, moreover, the cliff cuts at an angle diagonally across the path of the stream and without reference to the direction in which the ancient glaciers moved: the directive control exercised by the rock structure is supreme, and its effect is unobscured by either stream or glacial erosion. On the other hand, the Cascades (pl. 23, B), the only major falls that are not of the leaping type, descend over irregularly fractured rock that does not stand in a smooth, clean-cut wall.
INFLUENCE OF ZONES OF FRACTURING
Among the most puzzling features of the walls of the Yosemite are the deeply cut recesses and gulches that occur at places where no streams or only insignificant streamlets come down from the upland. A striking example is the capacious recess west of El Capitan (see pl. 7), whose dimensions are out of all proportion to the little upland vale that drains into it. Not only is it many times as large as the recess cut by the Ribbon Fall, but it is larger than the embayment at the Upper Yosemite Fall. Nor can it be attributed to the eroding action of a powerful torrent that came from a melting glacier on the upland during the ice age, for the crown of El Capitan and the hills to the north and northwest of it have never borne any glacial ice.
Whitney,74 denying the possibility of its having been formed by the ordinary processes of erosion, sought to account for the recess and the other reentrant angles in the valley walls as by-products of the catastrophic downfaulting of the bottom of the Yosemite which he had postulated. Examination of the recess however, leaves no doubt that it has been developed in a zone of intensely fractured rock by the destructive action of ground water, frost, torrential rains, and snowslides. Several master joints converge toward the head of the recess, and between them the rock is cut by a multitude of cross joints, so that even its more resistant portions stand up only in the form of a slivered, comblike crest. Naturally rock so thoroughly broken would yield much more readily to the dismantling agents than the exceptionally durable masses of undivided rock immediately adjacent. The lesson is made the more telling by the fact that the recess is carved in granite of the same kind as that of which the towering mass of El Capitan is composed. The contrast in sculpture is called forth by the contrast in structure.
Other sharply incised recesses that are similarly etched out, so to speak, along narrow zones of intense fracturing separate the three summits of the Cathedral Rocks (pls. 3 and 7) from one another. None of these recesses receives any drainage at its head. They cut across the backbone of the ridge and head against a slope that drains in the opposite direction. Selective quarrying by the Yosemite Glacier, which overwhelmed the Cathedral Rocks during the earlier glacial stages, may account in part for the cutting of these recesses, but it seems more probable, to judge from the large pikes of débris at their mouths, that they have been sculptured mainly by the processes enumerated above during the long interval that has elapsed since the departure of the earlier ice.
The largest and deepest recess of this type is the gulch of Eagle Creek, west of the Three Brothers. It is larger than Indian Canyon, yet, unlike that canyon, it receives no drainage from any upland valley. Like the recess west of El Capitan, however, it has been developed along a zone of shattering, in which the sculpturing agentsin particular storm waters and snowslides, which converge in its funnel-like formhave worked with especial ease.
To this class of recesses belong also the mysterious notches, gulches, and alcoves cut in the lips of the hanging valleys near to but not at the waterfallsfeatures whose significance has long been a subject of speculation. Such are the broadly open notch west of the Upper Yosemite Fall (pl. 22), the alcove with overhanging roof at the head of the recess into which the Lower Yosemite Fall plunges, and the alcove at the head of the Illilouette Gorge, which is carved in the mountain side irrespective of the position of the Illilouette Fall. (See pl. 7.) Similar features occur on the steps of the giant stairway over which the Merced River descends. The cliff at the Vernal Fall is cut obliquely past the fall and heads southeast of it in an alcove with overhanging roof; and north of the Nevada Fall, at the base of Liberty Cap (pls. 25 and 10), is a gorge several hundred feet deep that is carved about 800 feet back into the lip of the Little Yosemite Valley.
The late Professor Branner75 offered the explanation that each of these notches, gorges, and alcoves marks the place where during the later part of the ice age a torrent flowing along the margin of the glacier in the valley above had poured over the brink. And he regarded these features as prima facie evidence of the superior cutting power of streams over glaciers. It is now realized, however, that the matter is not so simple as Professor Branner supposed, that the location and character of each notch, gorge, or alcove are determined by a zone of fracturing in the rock, and that several of these features have been produced without the agency of any glacial torrent. Each case differs somewhat from the others and must be considered by itself.
That the broadly open notch west of the Upper Yosemite Fall (pl. 22) was once the path of a stream flowing along the western margin of the Hoffmann Glacier can hardly be doubted, for what appears to be an old stream channel leading to the notch is traceable along the west side of the hanging valley for a quarter of a mile, and 1,600 feet below the brink, at the foot of the incline on which the zigzag trail is built, there appears from beneath the débris a deeply cut stream channel, now dry, that joins the gorge of Yosemite Creek a short distance above the lower fall. However, as already explained (p. 111), the incline and notch at the top are developed in a zone of strong oblique jointing. (See pls. 47, B, and 9, B.) This zone must have been peculiarly susceptible to glacial erosion as well as to stream erosion, and, as the general modeling of the notch and the incline is suggestive of glacial quarrying rather than of stream cutting, it seems a proper inference that these features were carved largely by the ice of the Hoffmann Glacier, which was confluent with the Yosemite Glacier during the earlier stages of the ice age. Erosion by a torrent followed when the Hoffmann Glacier had shrunk to a narrow ice tongue, but it lasted probably for a short period only and accomplished relatively small results.
Another recess, more sharply cut than that below the notch west of the Yosemite Fall, exists east of the fall, between Yosemite Point and the Castle Cliffs. (See pl. 9, B.) According to Branner's idea, this recess might be supposed to have been cut by a torrent that flowed along the eastern margin of the Hoffmann Glacier, but there is no trace of a channel that leads to its head. On the other hand, the rock in the recess is divided by numerous vertical joints and shows evidence of having been and of still being plucked away along this zone of weakness, presumably by torrential rains and snowslides chiefly.
The recess into which the Lower Yosemite Fall leaps (pls. 23, C, and 7) appears to have been eroded past this fall, yet there is no warrant for assuming that its sharp, undercut head marks a former site of the fall, for on the sloping ledges above the alcove there is no vestige of a stream channel. Here again an examination made at a time when the fall was dry disclosed the fact that the recess has grown and is still growing headward along a narrow zone in which the rock is finely sheared and slivered by vertical fractures. Evidently the recess was initiated by Yosemite Creek and has been extended in the direction of the zone of shearing, irrespective of the position of the fall, by the wasting away of the slivered rock. In this process the flying spray from the fall doubtless has played and still plays a prominent part; by permeating the slivered material and by freezing in the interstices it materially hastens disintegration. The production of an alcove is explained by the fact that the vertical zone of shearing stops at a horizontal master joint above which the rock is massive.
The headward growth of the Illilouette Gorge (pl. 7) for a distance of 500 feet past the Illilouette Fall has a similar explanation. There is no evidence that any stream ever flowed along the margin of an ice tongue in the hanging valley above and poured down at the head of the gorge. Nor can the former existence of such a glacial stream properly be assumed, for, as has already been shown, the Illilouette Glacier at no time extended in the form of an ice tongue down to the lip of the valley. The facts are that the Illilouette Gorge has been excavated along a southwestward-trending zone of fracturing, partly by the ice of the Merced Glacier, partly by the ordinary processes of weathering and stream erosion, and that it has continued to grow headward, past the fall, by the caving away of the rock in the zone of fracturing. Here, as in the recess at the lower Yosemite Fall, the spray from the falling water has been a potent factor in hastening disintegration.
The alcove southeast of the Vernal Fall (pls. 10 and 7) has been produced in a closely analogous manner. No glacial torrent was ever concerned with its making, but it has been "stopped out," to use a miner's term, at a place where the rock is traversed by many fractures. The overhanging roof of the alcove, on the other hand, is composed of relatively massive rock.
The gorge north of the Nevada Fall (pls. 25 and 7), through which the tourist trail to the Little Yosemite Valley is laid, is more nearly the kind of feature that Branner had in mind than any of those just considered. It was cut, in all probability, by a torrent that flowed along the northern margin of the Merced Glacier at a time when that glacier still occupied the Little Yosemite but had already melted back from the lower steps of the giant stairway. However, the fact must not be disregarded that the gorge coincides with a zone of fracturingthe only zone of that kind which traverses the otherwise massive granite at the mouth of the Little Yosemite. Were it not for the presence of this favoring structure, it may well be doubted whether the glacial torrent in the short period of its existence would have succeeded in trenching so deeply. No corresponding gorge exists on the south side of the Nevada Fall, though doubtless there was a glacial torrent also along the southern margin of the Merced Glacier. But the granite on the south side of the valley is extremely massive, and this fact suffices to explain the absence of a gorge.
It is to be borne in mind, further, that the zone of fracturing north of the Nevada Fall must have facilitated glacial erosion as well as stream erosion. The selective manner in which glaciers excavate in rocks of variable structure has already been explained (pp. 90-91). The most impressive product of such selective action by a glacier is the deep gorge between Liberty Cap and Mount Broderick, which was excavated along a northeastward-trending zone of vertical fractures. It is entirely probable, therefore, that the gorge north of the Nevada Fall was excavated in large part by the Merced Glacier and then was deepened somewhat further by the temporary glacial torrent.
In conclusion attention is invited to the gorges and gulches that are carved at the mouth of the hanging valley of Snow Creek. (See pl. 7.) The deep central gorge, through which Snow Creek descends at the present time, doubtless has been the path of the stream for a long timeever since the El Portal stage of glaciation. The gorge has two heads, however, and it is a significant fact that the east head, which is dry, is just as deeply and sharply incised as the west head, through which Snow Creek now flows. There is, moreover, in the floor of the hanging valley a faint remnant of an old channel that leads to the east head of the gorge. It is to be inferred, therefore, either that Snow Creek formerly entered the gorge by the east head, or else that a torrent flowing along the east margin of the Snow Creek Glacier (in the El Portal stage) followed that path.
Half a mile west of the gorge of Snow Creek is a relatively shallow gulch through which the Mirror Lake and Soda Springs Trail ascends. This gulch in all likelihood was carved by a glacial torrent that flowed along the west side of the valley, but it may also be a temporary channel of Snow Creek itself. Indeed, it seems probable, in view of the great breadth of the valley, that during those phases of glaciation when the Snow Creek Glacier extended to or nearly to the brink of Tenaya Canyon, the waters coming from the ice coursed to the lip of the valley by diverse and varying paths, not only because they were crowded aside by the ice body, but also because they were obstructed by masses of morainal débris. The gulches were cut, however, only in those places where the otherwise massive granite is traversed by numerous joints.
At the extreme east side of the hanging valley, near Mount Watkins, is still another sharply incised gulch. This gulch may possibly have been the path of a glacial torrent at one time or another, but it seems more probable, in view of the absence of any channel leading to its head, that it has been carved, like many other features of its type, solely by the processes of disintegration and erosion previously mentioned along a narrow zone of intense fracturing.
FORMS CARVED IN MASSIVE ROCK
In striking contrast to the angular, faceted types of sculpture that are called forth in jointed rocks are the smoothly rounded forms produced in unjointed, massive rocks. The domes of the Yosemite region are the outstanding representatives of this class. Each of them consists of a wholly undivided body of granite, a gigantic monolith. The same is true also of the whale-back spurs, cylindrical ridges, and conoidal buttresses, which are closely allied to the domes. Indeed, the Yosemite region is peculiarly rich in such distinctive forms carved from massive granite; it contains a greater and more varied assemblage of them than any other area of similar extent in the Sierra Nevada or, perhaps, anywhere on the earth.
There have been various misconceptions, among scientists as well as laymen, regarding the origin of these smoothly rounded forms. Whitney76 conceived the domes "to have been formed by the process of upheaval itself," for he could "discover nothing about them which looks like the result of ordinary denudation." Galen Clark,77 the old pioneer of the Sierra, who became the first guardian of the Yosemite State Park, evidently influenced by Whitney's conception, supposed the domes to have been forced up by the pressure of gas from underneath at a time when the granite was still in a semiplastic condition, and he imagined the Yosemite to have been created by the bursting open of several domes in a row. But Clark's and Whitney's views need scarcely more than passing mention here, for they do not take into account the now well-established fact that granite is inherently a material of deep-seated origin that has crystallized under a confining roof or crust of older rocks and that has not flowed out upon the surface of the earth (p. 25). The domes in any event can not be original surface forms but must have resulted from the action of erosive processes of some sort, after the granite had become uncovered.
A belief that has long been widely prevalent is that the domes are essentially gigantic "roches moutonnées," worn round by mighty overriding glaciers. The late Prof. Joseph Le Conte78 was among those who gave voice to this belief. He supposed that "the whole surface of the region with its greater and smaller domes had been molded beneath a universal ice sheet or confluent glacier which moved onward with a steady current careless of domes." Le Conte had no means of testing the soundness of this idea, but there is now abundant proof that the Sierra Nevada was never overwhelmed by a "universal ice sheet" but bore only glaciers of local origin, and the depth to which these glaciers covered the Yosemite region is known within a narrow margin of error from the survey of the moraines. (See pp. 73, 82). Several of the domes, it can be stated positively, were never overtopped by the ice of the glacial epoch: Sentinel Dome stood wholly above the highest level of glaciation; all of the bare, domelike part of Mount Starr King has remained untouched by the ice; and the crown of Half Dome rose like a rocky isle fully 500 feet above the surface of the Merced and Tenaya Glaciers, which coalesced about it. Yet these domes are among the most conspicuous and most typical domes of the Yosemite region. That glacial abrasion is not a necessary factor in the development of domes is demonstrated, further, by the fact that the celebrated Stone Mountain, near Atlanta, Ga., which is a granite dome of precisely the same type as those of the Yosemite region, stands 300 miles south of the southernmost limit reached by the continental ice sheet in the eastern part of the United States.
It is now generally recognized by geologists that the domes owe their rounded forms to the exfoliation of massive granitethat is, the casting off of successive curving shells or scales from their exposed surfaces. Every dome bears a number of such curving shells, arranged concentrically about one another like the layers of an onion, and the outer ones, it is evident, break up in the course of time and drop off. On Sentinel Dome (pl. 21, A) which is readily accessible, this concentric shell structure may be studied at close hand. The shells are seen to vary from about half a foot to several feet in thickness. The outer shells are as a rule the thinnest. On Half Dome some of the shells are 6 to 10 feet thick, and in the Royal Arches shells measuring 10 to more than 100 feet in thickness are displayed, but these massive shells are truly exceptional. The aggregate thickness of all the shells on a dome varies widely but seldom exceeds 100 feet.
The cause of exfoliation is at the time of this writing still somewhat of a mystery. That the shells burst loose from the core of a dome because of expansive stresses in the granite is clear from the facts of observation (see pl. 48, A) as well as from the principles of mechanics, but how the expansive stresses originate is a matter of doubt. External heating of the rock by the rays of the sun is unquestionably a factor, for the quarrymen on Stone Mountain, who produce shells artificiallyby blasting and pneumatic pressurefind that their process is effective only in the summer, when the rock is warmed by the sun. But neither solar nor seasonal warming nor the two combined, are found, upon mathematical analysis, to be capable of producing stresses powerful enough to disrupt granite to depths of as much as 100 feet. Neither can the swelling of the granite as a result of hydrationthat is, the chemical union of water with the constituent mineralsbe a competent cause. It can be at most only a feeble subsidiary cause, for microscopic examinations of rock taken from some of the outermost and oldest shells of exfoliating bodies of granite reveal only signs of very moderate hydration. Not to go into an exhaustive discussion of all the processes that may possibly be concerned in the production of exfoliation shells, it may be stated that for the present the most probable cause is held to be simply the relief from pressure experienced by the granite as the superincumbent masses of rock are removed by erosion. As the load diminishes naturally expansive stresses are liberated within the granite. In jointed rock such stresses are relieved by slight readjustments along the numerous fracture planes, but in massive rock the stresses accumulate until at last they cause rupture along partings approximately parallel to the exposed surface.
Whatever may be the ultimate cause of exfoliation, the manner in which the process operates to produce smoothly rounded forms is sufficiently clear from the numerous examples at hand. Its tendency is first to eliminate projecting corners and angles and to replace them by fairly sharp curves, as shown in Figure 38. With the dropping off of the succeeding shells these sharp curves are replaced by more and more gentle curves, and thus in the course of time a smoothly and continuously rounded surface is evolved. A striking illustration of the successive steps in the process is afforded at the brink of the great precipice of the Upper Yosemite Fall, near the top of the Lost Arrow. (See pl. 48, B.) Doubtless there was originally a square edge, produced by the dropping away of the huge sheet of rock of which the Lost Arrow is the principal remnant; but in the course of the 200,000 years or more that has elapsed since the El Portal glaciation (the later ice did not touch the cliff) the square edge has been transformed by progressive exfoliation to a gently curving one. At one place, however, there still remains part of what was presumably the second shell to be formed, characterized by a sharply curving outer surface. The next shell under it is much more gently curved, and the next more gently still.
Similar but on a larger scale are the exfoliation features displayed at the edges of the Quarter Domes. (See pl. 48, C.) As is clear from Plate 7, each of these two domes owes its peculiar configuration to the controlling influence on the side facing Tenaya Canyon of a northeasterly and a northwesterly master joint that intersect each other at approximately right angles. Each dome doubtless had, at the end of the El Portal glaciation, a fairly sharp point and angular edges, but through exfoliation these points and angles have been blunted and rounded off. On the lower Quarter Dome (pl. 48, C) a sufficient number of the older shells remain in place to give some indication of the original point and edges.
In both of the examples above noted the exfoliation features were produced since the El Portal glaciation. The length of time involved in their development is therefore known approximately, and thus some idea may be had of the rate at which exfoliation shells are formed. In each instance 2,000 centuries was required for the moderate rounding off by exfoliation of an originally angular edge. The postglacial interval, which comprised only about 200 centuries, was too short, in many places, for the production of a single shell. The crown and back of Mount Broderick (pl. 44, A), which were stripped of all their shells by the overriding Merced Glacier of the Wisconsin stage, still show no signs of renewed exfoliation over the greater part of their surfaces. The severely glaciated sides of the Little Yosemite likewise are devoid of shells over large areas. Only here and there a new, thin shell has recently been detached from them.
From these facts it is evident that the domes of the Yosemite region have been a long time in the making; they are among the oldest features of its landscape. Their sweeping curves attest their great antiquity. Domes such as Sentinel Dome and Mount Starr King, which stand above the general level of the Yosemite upland, doubtless were in process of exfoliation as far back as the Miocene epoch, at least 8,000,000 years ago, when the upland was still an undulating lowland and when the Merced River flowed in a broadly open, shallow valley. It follows that these domes, as they now appear, are the much reduced, rounded-off remnants of originally much larger rock masses.
It is probable that all the domes of the Yosemite region were at first more or less angular and irregularly shaped, for they were originally surrounded by jointed rock. Some suggestion of their original forms still remains in their present outlines. None of the domes are wholly symmetrical. Every one of them is elongated, in some one direction and more or less pronouncedly one-sided. (See pl. 7.) They really grade over into whaleback ridges such as Mount Watkins, Boundary Hill, and the summit of El Capitan. Straight sides are a common feature of them, and there is little doubt that the trend of these sides is inherited from master fractures that bounded the original rock masses. North Dome, for instance, appears to be derived from a rock mass bounded by north-south master joints, and the cylindrical ridge immediately north of it evidently was delimited originally by northeasterly master joints. Mount Watkins, Half Dome, Mount Broderick, Liberty Cap, and the whaleback spurs of the Starr King group also betray in their outlines the controlling influence of northeasterly joints.
In the features of Half Dome the student of dome development will read the most interesting story. The curving back of the dome (pls. 49 and 50, A) is evidently a product of long-continued exfoliation. Shells have been formed on it and have dropped from it in succession for millions of years. That the process must have been extremely slow may be judged from the fact that the huge shell which at present envelops almost the entire back of the dome has been in place so long that furrows several feet in depth have been worn in it by the rock grains washed from the crown. The central parts of the shell have so many of these furrows that they present a ribbed or fluted appearance. (See pl. 50, A.)
On the broad crown of Half Dome, where gravity and snowslides are not effective removing agents, the shells remain in place still longer. They remain, in fact, until, mainly as a result of daily heating by the sun, in less measure because of recurrent frost action, they disintegrate into mere slivers and grains of rock. As a consequence the dome bears on its summit a great accumulation of old shells, the aggregate thickness being estimated at 90 to 100 feet.
The straight, sheer front of Half Dome is by comparison a rather new feature, yet it too has suffered from exfoliation. Its general trend and its angle of declivity (about 82°) were determined by a zone of nearly vertical joints extending in a northeasterly direction. This sheeted structure terminated in the shoulder at the northeast end of the cliff face, as may be seen in Plate 50, B, and it has given rise to a sharply incised notch at the southwest end. (See pl. 7.) Doubtless the thin sheets were readily plucked away by the Tenaya Glacier, which during the earlier stages of glaciation reached within 500 feet of the top of the dome. (See fig. 23.) Then, the body of the monolith being exposed, it began to exfoliate in plane sheets parallel to the zone of joints but curving in under the old shells at the top. Perhaps the Tenaya Glacier plucked away some of these newly formed sheets also, but it seems more probable that the exfoliation took place largely during the interval following the El Portal glaciation. The ice of the Wisconsin stage did not reach the base of the cliff. (See fig. 23.) In any event the sheets scaled off in relatively rapid succession, owing to their verticality and the intense frost action that prevails on the shaded northwest side of the dome, and as a consequence the old shells on the summit were left overhanging, and there was formed the cornice which projects from the top of the cliff.
This modern interpretation of the evolution of Half Dome, it will be seen, finds no room for any assumed demolition of one-half of the dome. No such assumption appears justified by the facts known about the structure of the rock on the northwest side of Half Dome, nor would it accord with the present conception of the erosional origin of the Yosemite Valley and Tenaya Canyon. Had there been another half of the dome consisting of a gigantic monolith, it would be still in existence to-day, for neither the Tenaya Glacier nor the agents of normal erosion that shaped the preglacial valley of Tenaya Creek could have demolished it.
A rare type of sculpture produced characteristically in exfoliating granite and associated with the domes consists of successive arches recessed one within another. The back of Half Dome (pl. 50, A) presents an imperfect example that is nevertheless instructive because it shows clearly how such arches originate through the caving off of the lower portions of shells. The remaining portion of each shell naturally tends to assume the shape of an arch, because the arch is, as architects well know, the form of structure best adapted to bearing a heavy distributed load. The finest example is afforded by the Royal Arches. (See pl. 21, B.) They are fashioned on a colossal scale, the main arch rising to a height of 1,000 feet (measured to its underside) and having a span of 1,800 feet. Many of the shells range from 10 to 80 feet in thickness, and several of them unite near the top of the main arch to form one shell nearly 200 feet thick. The Yosemite Glacier was the principal sculptor; during the last stage of glaciation it plucked away the lower portions of the shells, which had previously been loosened by exfoliation from a partial, low-set dome that bulged out into the valley. The great strength of the massive shells has permitted the forming of exceptionally high and broad arches, and the homogeneity of the granite has given rise to unusually perfect, smooth curvature. A short distance west of the Royal Arches is another set of arches, sculptured like wise by the Yosemite Glacier but from much thinner shells. (See pl. 16, B.) Because of their proximity to the Royal Arches they receive little attention, yet they are a good average example of the type as it occurs in different parts of the Sierra Nevada.
In conclusion it is to be pointed out that the shells produced by exfoliation are not invariably convexly curved. As has been shown, on the cliff face of Half Dome they are plane, and the same is true in other places, notably on the south side of the Little Yosemite. There are, however, also examples of concave exfoliation, the shells produced having hollow outer surfaces. Wherever a powerful glacier, after plucking away all the exfoliation shells, gouges into a structureless mass of granite in the bottom or side of its channel, it tends to grind out a smoothly concave surface. When later exfoliation begins anew, it produces concave shells. Imperfect examples of concave shells are to be seen in the salient of El Capitan (pls. 3 and 17), which is itself an imperfect dome, not wholly massive throughout, that has been vigorously gouged into by the Yosemite Glacier. Much finer examples of concave exfoliation are afforded by Mount Watkins, whose southeast side was gouged into by the Tenaya Glacier, and by the shallow glacial cirques that scallop the great cliff front of Clouds Rest. (See pl. 40, B.) Concave exfoliation on a large scale is exhibited in the canyon of the Merced River above the Vernal Fall. The sides and floor of that canyon exfoliate in more or less concave shells parallel to its U-shaped cross profile. Emerald Pool occupies the central basin gouged by the Merced Glacier in massive granite. Many similar examples of glacial U canyons and cirques whose walks and floors exfoliate in concave curves exist in the High Sierra above the Yosemite region. It is not to be supposed that these canyons and cirques were excavated in rock that was traversed originally by concentric U partings; on the contrary, the partings have been developed since the canyons and cirques were gouged out.
A host of other remarkable rock forms due to exfoliation might here be mentioned, not only in the Yosemite Valley but in the Little Yosemite, in Tenaya Canyon, on the Yosemite upland, and in the adjoining parts of the High Sierra. Indeed, it may be conservatively estimated that nearly half of the landscape features of these regions owe their modeling to exfoliation. But the foregoing brief explanations must suffice; they will at least furnish the key to the interpretation of all those features.
Last Updated: 28-Nov-2006