USGS Logo Geological Survey Professional Paper 329
Reconnaissance of the Geomorphology and Glacial Geology of the San Joaquin Basin, Sierra Nevada California


The west slope of the Sierra Nevada, of which the San Joaquin Basin forms a part, is a region of exceptional interest to the geomorphologist, both in its broader aspects and in its details. Although furrowed by many deep canyons, it nevertheless retains on its undissected intercanyon uplands considerable remnants of ancient erosion surfaces, as well as lofty monadnock ridges with occasional tabular summits. These features, together with the evidences of Pleistocene glaciation in the higher portions, present a decipherable, if partly obscured, record of the geomorphologic history of the range. Moreover, as the rise of the Sierra Nevada was intimately related to the earth movements that affected the Great Basin, to the east, and as, furthermore, the growing height of the range brought upon that province the stark aridity that has so profoundly influenced its erosional development, it will be apparent that the story to be read in the features of the west slope of the Sierra Nevada has bearings that extend far beyond the confines of the range.

The precipitous east front of the Sierra Nevada, though extremely impressive, is to the geomorphologist less instructive than the gently declining west slope. The product of essentially one episode in the life history of the range, and that a relatively recent one, the east front tells a rather short and simple story. The west slope, on the other hand, recounts a long history reaching back into the beginning of the Tertiary period; and it even retains sculptural and drainage elements which were inherited from an ancient mountain system of Appalachian type, that occupied the site of the present Sierra Nevada during the Cretaceous period.

The west slope of the Sierra Nevada is the only large area of elevated land in the United States west of the Rocky Mountains that presents such a comprehensive geomorphic record in fairly readable form. Primarily because of its great areal extent it has escaped thorough dissection in spite of strong uplift and consequent vigorous erosional attack. By comparison with the Sierra Nevada, the back slopes of the Basin Ranges are of very limited extent, measuring, as a rule, only 4 to 6 miles from crest to base. The Sierra back slope, however, measures 40 to 60 miles from crest to base, and—in the distance of 430 miles through which it maintains this breadth—is traversed by rivers as long as 100 miles. Even if the northernmost and southernmost portions, which are complicated by lines of dislocation, are subtracted, the west slope remains by far the largest sharply delimited unit area of uplifted and tilted earth surface in the western United States. The Cascade Range, it is true, covers a greater area than the Sierra Nevada, but most of its elevation is due to irregular upwarping. Its surface configuration in addition reflects the vagaries of volcanic action and scarcely offers an advantageous field for geomorphologic studies of a broad regional nature.

Three factors besides areal extent have retarded the dissection of the west slope of the Sierra Nevada: first, the resistance to erosion of the harder metamorphic rocks and the sparsely jointed, massive granitic rocks that prevail over large areas; second, the general absence, except in the northernmost and southernmost parts of the range, of lines of dislocation that facilitate localized trenching; and third, the protection against gullying erosion which its soil and its vegetation afford, and which probably always have afforded, except in the crestal parts which were severely glaciated during Pleistocene time and have since remained barren because of their altitude. In the latter respect the Sierra Nevada stands in marked contrast to the naked, craggy flanks of the Basin Ranges, which are fully exposed to the processes of denudation.


Before narrowing consideration to the San Joaquin Basin, it will be profitable to consider the various sections of the west slope of the Sierra Nevada as a whole, both to the north and to the south of the basin, in regard to the problems they present with reference to geomorphologic analysis.

The northern two-fifths of the Sierra Nevada, extending approximately as far south as the northwestern boundary of Yosemite National Park, is in large part mantled with volcanic agglomerate and lava, with the result that its earlier erosion surfaces are almost completely obliterated. The buried valleys and ridges can be glimpsed here and there in the sides of the newer canyons that trench through the volcanic cover, but the glimpses thus afforded hardly suffice to permit reconstruction of the ancient landscapes with any degree of confidence. Moreover, several northwestward-diverging faults of considerable throw disrupt the unity of the west slope at the northern end of the range, thereby complicating the geomorphologic record.

The southern one-fifth of the range, from the Kern Canyon south to the Tehachapi Pass, is traversed by several lines of faulting which, though not seriously affecting its unity, nevertheless control some of the principal drainage lines and have influenced erosion. Furthermore, this part of the range does not slope gently toward the western foothill belt, but has the aspect of a profoundly dissected horst that breaks off abruptly at its western as well as its eastern margin.

The upper Kern Basin is probably unsurpassed for simplicity of sculpture; probably no other part of the range retains erosion surfaces which so clearly exhibit the effects of incomplete cycles of erosion as does this basin, in which Lawson (1904, 1906) made his classic morphologic analysis. But the precise correlation of these erosion surfaces with those in the central part of the range presents difficulties, for in several respects the Kern River is anomalous among the master streams of the Sierra Nevada. Its course appears to have been determined largely by structural factors, rather than by tilting of the Sierra block; for it flows southward along a zone of faulting, in a course that trends parallel to the crestline of the range instead of normal to it, as in the case of the other master streams. Furthermore, the Kern River traverses that part of the range wherein a transition takes place from tilted block structure at the north to horst structure at the south; and in the lowermost part of the Kern Basin there are other structural irregularities. The tributaries of the Kern River trend in various directions, without apparent system, rather than in patterns decipherable in terms of the early geomorphologic history of the basin. It appears, therefore, that the Kern River Basin is hardly typical for the range as a whole, but presents a rather special case, complicated by several local factors.

It is in the central and south-central parts of the range that the geomorphologist finds the conditions most favorable for his purpose. More specifically this area extends from the northwestern boundary of the Yosemite National Park southeastward to that spur ridge in Sequoia National Park known as the Great Western Divide, and it comprises the drainage basins of the Tuolumne, Merced, San Joaquin, Kings, and Kaweah Rivers. This part of the Sierra Nevada, 150 miles long in the direction of its axis and averaging 60 miles in breadth, appears to have been essentially a single massive rigid block, and has remained relatively free from deformation or fracturing except locally, as at its western margin. The drainage system of the west slope of this portion of the range, so far as known, has not been dislocated or rearranged in consequence of either faulting or warping, except in a minor way as in certain parts of the foothill belt. Neither have any of its streams been eliminated or significantly diverted by extensive volcanic outpourings such as prevailed in the northern parts of the range. Local extrusions have not been wanting in various places and at different times, but the drainage changes caused by them have been inconsequential.

One drainage basin, that of the Merced River, has remained free from volcanic outpourings, containing only one diminutive crater. As a consequence, within its confines, the processes of stream erosion have worked without interference, except by the glaciers of the Pleistocene epoch, ever since the earliest recognized cycle. Nevertheless, the Merced Basin is not the simplest field for geomorphologic study on the west slope, for its central feature, the Yosemite Valley, is of a peculiarly enigmatic sort (Matthes, 1930a). It is a chasm of aberrant type in whose interpretation the geomorphologist can hardly hope to succeed without having first gained an insight into the history of the more normally shaped canyons of the west slope. Furthermore, the lower half of the Merced Basin, being composed of upturned strata of thin-bedded metamorphic rocks that differ sharply in resistance to erosion, is dissected by a maze of valleys and gulches, and retains on its skeleton ridges few recognizable remnants of the more ancient erosion surfaces. The profiles of the Merced River for each of two earlier erosion cycles can be reconstructed with some confidence throughout the upper half of the basin, which is composed almost wholly of granitic rocks. However, these profiles cannot be extended for any great distance down through the lower half, for lack of reliable diagnostic features in its greatly dissected topography. It was because of these baffling circumstances that the author deferred publication of his conclusions regarding the history and mode of development of the Yosemite Valley until he had had opportunity to test their validity by comparative studies of other major canyons in the range, not only of those that possess Yosemite-like widenings but also those that are normally shaped throughout.

Reconnaissance for this purpose in the Tuolumne Basin (1916, 1917, and 1919), which adjoins the Yosemite region on the north and northwest, afforded an opportunity to examine the Hetch Hetchy Valley, which is most closely analogous to the Yosemite Valley. The outstanding lesson of Hetch Hetchy was that the geomorphic history is by no means recorded with equal clearness in all parts of the Sierra Nevada, for in some parts it may be extremely obscure in spite of the grand scale on which the features of the landscape are modeled. The Tuolumne Basin, though rich in glacial features, did not prove especially helpful in defining successive landscapes produced by successive cycles of erosion, for it contains few well-preserved remnants of ancient surfaces of erosion, and is poor in clean-cut hanging valleys whose discordance is demonstrably due to preglacial canyon cutting. Over considerable parts of the Tuolumne Basin the granitic rocks have the same unusual structure as in the Yosemite region, and have given rise to similar aberrant sculptural forms. The Hetch Hetchy Valley is another yosemite, although of smaller dimensions than its prototype.

The Table Mountain region (fig. 1), visited in 1921, on the divide between the lower Tuolumne and lower Stanislaus canyons, proved to be an unexpectedly fruitful field because the gravels in some of its lava-entombed stream channels yielded well-preserved fossils that afford a means for determining the age of one of the older erosion surfaces. (See p. 23.) Farther north, however, the older topographic features are largely buried under the mantle of volcanic rocks.

Just south of the San Joaquin Basin, the writer's investigations have indicated that in the basins of the Kings and Kaweah Rivers, likewise, there are conditions which interfere with geomorphologic analyses of a broadly regional character, notably that these basins exhibit relatively few of the well-preserved hanging valleys which are of critical importance in deciphering the successive events of the history of the range.


The broad San Joaquin Basin appears in some ways to be the most revealing drainage basin of the west slope in regard to features that aid in the interpretation of regional geomorphologic problems. Briefly stated, its importance in this respect stems from the following combination of circumstances:

The erosional features of the basin are unobscured by volcanic flows except in relatively small areas.

The granitic rocks, exceptionally extensive in this basin, are prevailingly massive and exert profound influence on the topography. The San Joaquin Canyon is, therefore, consistently narrow, and exhibits neither anomalous Yosemite-like widenings nor, for the most part, the unusual sculpturing characteristic of the yosemites. Furthermore, the resistant nature of the massive granitic rocks accounts for the preservation of an exceptionally fine record of the successive cycles of erosion, and an extraordinary wealth of clean-cut hanging side valleys.

The granitic rocks are practically continuous to the foot of the range. Consequently, the ancient erosion surfaces and the array of hanging valleys likewise extend not merely throughout the manifestly glaciated upper course of the San Joaquin Canyon but also through the unglaciated lower course—in fact to within a few miles of the foothills.

The San Joaquin River has an essentially normal southwestward course. Thus flowing down the slope of the Sierra block, the river must have received the entire effect of the rejuvenation caused by each tilting movement. In contrast, the tributaries of the San Joaquin River for the most part have northwesterly or southeasterly courses, substantially at right angles to the master stream as well as to the direction of tilting of the block. The flow of the tributaries, therefore, must have been only mildly accelerated by the tilting of the Sierra block. In this contrasting relationship of master stream to tributaries, the drainage pattern of the San Joaquin Basin is typical for most of the basins in the central and southern part of the Sierra Nevada, but probably in no other basin is the relationship so well shown.

Finally, it may be noted that in the San Joaquin Basin the erosion surfaces form broad uplands which bear lofty monadnock ranges. The erosion surfaces of the uplands have remained essentially undisturbed because of the resistant nature of the bodies of massive granitic rock which underlie them. The monadnock ranges, by their position and prevailingly northwest ward trends, are representative for the Sierra Nevada as a whole, and significant in the interpretation of its early geomorphic history. Moreover, by their great height above the erosion surfaces which surround them, they testify, more clearly than does the evidence which the author has observed in other parts of the range, to the relatively great antiquity of that most ancient erosion surface which is still identifiable from small but sharply defined remnants preserved on a few of the summits.


A primary purpose in undertaking studies in the San Joaquin Basin was to test the soundness of conclusions reached in the Yosemite region regarding the geomorphologic development and the glaciation of that region, and to determine the bearing of these conclusions on the history of the Sierra Nevada in general. It is appropriate, therefore, to summarize the results of the author's study of the Yosemite Valley (Matthes, 1930a). (See fig. 11.)

FIGURE 11. Idealised profile showing features produced by successive cycles of erosion in the Yosemite region. E, tabular remnant of Eocene surface on high residual peak; R, residual peak no longer bearing remnants of Eocene surface; M, Miocene surface preserved on intercanyon uplands; M', longitudinal profiles of hanging valleys of Miocene cycle (glaciated in the High sierra); M", reconstructed cross profiles of main valleys of Miocene cycle; Pl, main valley of Pliocene cycle (glaciated); Pl', longitudinal profile of hanging valley of Pliocene cycle; Pl", reconstructed cross profile of main valley of Pliocene cycle; Pleistocene gorge of Pleistocene cycle (unglaciated). Vertical scale same as horizontal scale. From 16th International Geological Congress, Guidebook 16, p. 35, 1933. (click on image for a PDF version)

The most significant conclusions relate to the origin and significance of the hanging side valleys of the Yosemite chasm. Detailed mapping of the morainal system of the ancient Yosemite Glacier indicated that that glacier never extended more than a mile beyond the site of El Portal. The hanging side valleys of the Merced Canyon below El Portal do not hang, therefore, because of any glacial deepening which that canyon has undergone. The explanation was offered that they hang because their streamlets have been unable to trench as rapidly as the Merced River since the rejuvenation of the Merced by the last uptilting of the Sierra Nevada. The streamlets were handicapped not only by their comparatively small volume but also by the fact that their courses trend northwestward and southeastward, substantially at right angles to the direction of the tilting, and therefore have remained essentially unsteepened, whereas the Merced's course trends southwestward, directly down the slope of the Sierra block, and therefore has been appreciably steepened.

Projection of the longitudinal profiles of these hanging valleys forward to the axis of the Merced Canyon shows that they are closely accordant in height. Their profiles indicate a series of points on a former profile of the Merced with respect to which the side streams had graded their courses prior to the last uplift. This old profile can be extended upward into the glaciated part of the Merced Canyon above El Portal and even into the profoundly glaciated Yosemite Valley, accordant points being furnished by a number of hanging side valleys (due allowance being made for the effects of glacial erosion on those valleys).

However, not all hanging valleys of the Yosemite region are accordant with this set. Several constitute a separate set indicating another old profile of the Merced at a level 600 to 1,000 feet higher than the first. Others point to an old profile of the Merced about 1,200 feet lower than the first. There are thus three distinct sets of hanging valleys produced in three cycles of erosion. Those of the upper set, like those of the middle set, were left hanging as a result of rapid trenching by the Merced induced by an uplift of the range, there having been two such uplifts. Only the valleys of the lower set hang because of glacial deepening and widening of the Yosemite Valley, the cycle in which they were cut having been interrupted by the advent of the Pleistocene glaciers.

During the remote cycle of which the hanging valleys of the upper set and the undulating Yosemite upland are representative, the Yosemite Valley itself was broad and shallow, postmature in form. That early stage in its development, accordingly, is called the broad-valley stage. The deeper hanging valleys of the middle set were graded with respect to a deeper Yosemite Valley of submature form which must have had the aspect of a mountain valley. That stage in its development is therefore called the mountain-valley stage. The short, steep hanging valleys of the lower set and certain topographic features associated with them show that during the third cycle of erosion the Yosemite Valley was a roughly V-shaped canyon with a narrow inner gorge. This stage, which immediately preceded the glacial epoch, is therefore called the canyon stage.

In the Yosemite region, paleontological evidence for the age of the upland erosion surface is lacking, but farther north, in the Table Mountain district between the Tuolumne and the Stanislaus Rivers (figs. 1, 12), lava-entombed stream channels correlative with the upland have yielded well-preserved impressions of leaves and a few mammalian remains. These fossils, according to determinations by Ralph W. Chaney and Chester Stock, date back to the later part of the Miocene epoch (Matthes, 1930a, p. 1, 28; 1933a, p. 35, 70). Accordingly, the upland erosion surface in the Yosemite region is thought to be of late Miocene age. Uplift of the Sierra Nevada at the end of the Miocene epoch initiated the next cycle, during which the mountain-valley stage was evolved. That cycle lasted presumably through most, if not all of, the Pliocene epoch. the canyon stage was produced in all probability during the Quaternary period.

FIGURE 12.—Section through Table Mountain region, Sierra Nevada. Pl, Pliocene surface of erosion developed on unresistant slates and volcanic rocks of the Mariposa formation: M, residual mountains bearing of Miocene surface; M', reconstructed hills on Miocene flanking the ancient valley through which the latite of Table Mountain flowed; T, reconstructed portions of Table Mountain flow. Beneath the columnar latite of Table Mountain is the gravel of the buried river channel, which has been mined for gold. The Stanislaus and Tuolumne Canyons were cut during the Pleistocene cycle. Vertical scale is twice the horizontal scale. From 16th International Geological Congress, Guidebook 16, p. 36, 1933. (click on image for a PDF version)

The excellent preservation of the hanging valleys of the upper set, in spite of their great age, is explained by the exceedingly resistant nature of the massive granite that underlies them. The valleys of the middle set were carved in prevailing jointed rocks that were less resistant to stream erosion, and the gulches of the lower set were carved in closely fractured rocks in which the streams eroded with relative ease.

The gradients of the two higher profiles of the Merced furnish data from which the amplitude of each of the two great uplifts of the Sierra Nevada can be roughly calculated. The uplift at the end of the Miocene epoch added about 3,000 feet to the height of the range; the uplift at the end of the Pliocene epoch added 6,000 feet more. These determinations rest on deductions made after the reconnaissance of the San Joaquin Basin was completed, and which took into consideration the Tertiary profiles of the San Joaquin River and the rivers of the northern Sierra Nevada as well as those of the Merced River (Matthes, 1930a, p. 44).

Of the earlier geomorphologic history of the Yosemite region a glimpse is afforded in the explanation of the origin of the southwesterly course of the Merced River and the arrangement of the lesser tributaries at right angles to it. The Merced River established its course conformably to the southwesterly slant of the Sierra region, presumably early in the Tertiary period, when there still existed remnants of a system of northwestward-trending mountain ridges of Appalachian type, which had been formed at the end of the Jurassic period by the folding of sedimentary and volcanic strata of Paleozoic and Mesozoic age. As it grew headward, the Merced River probably captured the drainage from the longitudinal valley troughs between these ridges. Below El Portal, on the lower slope of the Sierra Nevada, where the folded strata still remain in a broad belt, the lesser tributaries of the Merced are for the most part adjusted to the northwesterly strike of the beds. In the Yosemite region and the adjoining parts of the High Sierra, from which the folded strata are now stripped away, broadly exposing the granitic rocks, the northwesterly and southeasterly trends of many of the streams are largely an inheritance by superposition from the drainage system of the now vanished older mountain system. The northwesterly trend of the Clark Range, the Cathedral Range, and certain stretches of the main crest of the Sierra Nevada are probably likewise inherited from that ancient mountain system. Though the majority of these monadnock ranges have sharp or splintered crests, a few retain tabular summits, the isolated remnants of an erosion surface which is clearly far more ancient than the Miocene uplands and therefore probably Eocene in age.




In the foothill belt, the mountainous surface of the Sierra Nevada declines to the level of the Great Valley of California and becomes covered by overlapping sediments. The relief diminishes in the first place through the progressive decrease in the eroded depth of the main canyon and secondly through the burial of the canyon floor and the tributary valleys to greater and greater depths. The mountains accordingly stand out with less and less height, until only their tops emerge from the engulfing sediments at varying distances from the range foot, like rocky islands fringing a mountainous coast indented as a result of recent submergence. A typical example of such an outlying mountain top is Round Mountain (altitude 870 feet) which rises about 450 feet above the plain. (See Round Mountain quadrangle.) Smaller knobs of this type occur to the south and southeast of it.

A third cause for diminishing relief in the foothill belt seems to be in the diminishing height of the mountains above the hard-rock valleys. Black Mountain (altitude 3,621 feet), at the inner margin of the foothill belt, stands 2,500 feet above the valley of nearby Dry Creek; but the hills to the southwest rise only about 1,200 feet above the same valley, and still farther to the southwest most of the hills are only 300 to 700 feet high. Owens Mountain, southeast of Friant (see Friant quadrangle), dominates the foothills roundabout with an altitude of 1,611 feet and stands 1,180 feet above the neighboring valleys, but its height is exceptional. Immediately south of it, the country drops off to levels of 600 and 500 feet, and only small rocky knobs, a few tens of feet high, rise above the level of the plain, which here is about 450 feet above sea level.

A conspicuous feature of the foothill belt is the series of flat-topped hills, or mesas, capped by volcanic rock, that extends in a southerly direction athwart the southwesterly course of the San Joaquin River. Like other volcanic mesas of this type farther north in the foothill belt of the Sierra Nevada, these are collectively known as Table Mountain. (See Millerton Lake and Academy quadrangles.) The partly disconnected mesas have an aggregate length of 16 miles, of which somewhat more than 6 miles lies on the northwest side of the river and the balance on the southeast side. Where the San Joaquin River transects the series of mesas, the river bed is fully 1,700 feet below the surface of the ancient lava flow. West of the river near the outer fringe of the foothills, is another lower and less well preserved series of flat-topped hills known as Little Table Mountain. (See Lanes Bridge quadrangle.) This series is only 5 miles long and extends in a south-southeasterly direction, ending about 4 miles below Friant in a bluff overlooking the river.

The plain immediately in front of the debouchure of the San Joaquin River consists of a great alluvial fan built by the river. This fan extends westward and southwestward for about 30 miles out from the foothills. At Friant, near the apex of the fan, the surface is 460 feet above sea level. Thence it slopes down to an altitude of about 200 feet at its outer margin. So gentle is this slope that to the traveler's eye the surface appears utterly flat. The city of Fresno stands near the middle of the fan, about 17 miles below the apex and 6 miles south of the river.

A number of radiating and usually double-crested ridges on the surface of the fan indicate natural levees along old channels which the river, or branches of it, have occupied at different times. Although the eye hardly detects these features, they stand clearly revealed on the detailed topographic maps, on which the contour lines are drawn at intervals of 5 feet. (See Friant, Lanes Bridge, Clovis, Bullard, Herndon, Biola, Gravelly Ford, Kerman, Kerney Park, Fresno, and Malaga quadrangles.)

These ridges are of the type which Kirk Bryan (1923, p. 29) noted in the Sacramento Valley and has called "channel ridges." Most of the old stream channels are now dry, but some have been improved and are being used as irrigation ditches.

In geologically Recent time the San Joaquin River has cut a trench in its great fan. This trench is 155 feet deep at Friant; thence downstream its depth diminishes progressively to only a few feet at the margin of the fan. The trench varies in width from 1/2 to 2 miles and contains several terraces marking successive stages in its cutting.


The canyon of the San Joaquin River (figs. 13-15) is neither among the deepest nor shallowest of the canyons in the west slope of the Sierra Nevada, but is in an intermediate class. It is not to be compared with the canyon of the Kings River, which attains the prodigious depth of 8,000 feet, nor with the canyons of the Tuolumne and Kern Rivers, which are between 4,000 and 5,000 feet deep. On the other hand, the San Joaquin Canyon is not to be classed with the trenches of the Feather, Yuba, American, Mokelumne, and Stanislaus Rivers, in the northern half of the range, which scarcely exceed 3,000 feet in depth. It falls, rather, in that middle class to which the Merced Canyon also belongs. The latter, though less than 3,000 feet deep in its lower course, reaches depths of 4,000 feet and locally, of 4,800 feet, in the Yosemite region.

The San Joaquin Canyon ranges for the most part between 3,000 and 4,000 feet in depth below the flanking uplands, but it exceeds 5,000 feet in depth, on the east side at least, for several miles in its middle course above and below the mouth of Big Creek. Its depth increases to 5,800 feet opposite the platform west of Huntington Lake, and a few miles farther up it reaches a maximum of 6,800 feet opposite the westernmost summit of Kaiser Ridge. On the west side, the nearest summit of the Chiquito Ridge rises 5,100 feet above the river, but the canyon side there is broken at a height of 2,200 feet above the river by a broad upland bench. Immediately above this deep portal between the Chiquito Ridge and the Kaiser Ridge the depth of the canyon again diminishes to an average of 3,000 feet, and that depth is maintained as far as the junction of the Middle and South Forks.

Of special geomorphologic significance is the fact that the San Joaquin Canyon has a distinct 2-story form throughout the greater part of its length (fig. 13). It has the appearance of a deep mountain valley of mature form with moderately steep, forested sides, in whose broad floor a narrow, sheer-walled inner gorge of approximately equal depth is cut. This 2-story aspect begins a short distance above the foothills, about a mile above Auberry. There, a well-defined gently sloping bench, representing a strip of floor of the mountain valley, appears at a height of 1,200 feet above the river. This bench extends 15 miles along the east side of the canyon to the tributary canyon of Big Creek. It varies in width from a hundred yards to a mile, the average being well over a half mile. Uninterrupted by any deep-cut side gorges, it afforded an excellent location for the San Joaquin and Eastern narrow gage railroad which formerly led to Cascada Station (Big Creek Post Office).

FIGURE 13.—View down San Joaquin canyon from B.M. 4112 on road to Hogue Ranch. The railroad, barely discernable on the far side of the canyon (see arrow), follows a sloping bench (Pliocene surface) above the gorge (Pleistocene canyon), and a corresponding bench on the near side of the gorge appears in the lower part of the photograph.

The height of this bench above the canyon bottom increases progressively upvalley to 1,800 feet in the Jose Basin and to 2,200 feet just below the side canyon of Big Creek. The upland in turn rises fully 2,000 feet above the bench. Some of the mountains on the upland surface rise higher yet. Music Peak stands 2,800 feet above the bench; Mount Stevenson 3,000 feet.

On the west side of the canyon a similar bench begins near Oat Mountain, and thence extends up 16 miles to the hanging valley of Shakeflat Creek. It is more irregular than the bench on the east side, its surface being diversified by alternating spurs and shallow transecting valleys, yet its average height above the canyon bottom corresponds to that of the eastern bench and, like the latter, increases gradually upvalley. To the west, wooded slopes of moderate declivity rise 1,500 to 2,500 feet to the uneven surface of the upland. Neither of the two benches persists clearly defined all the way up to the junction of the Middle and South Forks. The eastern bench is but feebly developed above Big Creek, and the western bench is scarcely traceable above Shakeflat Creek. Nevertheless, most of the inner part of the canyon beyond these points is steeper sided and more gorgelike than the outer part.

The inner gorge6 is narrowest and most conspicuously sheer walled in that part of the San Joaquin Canyon where the flanking benches are most prominent. Indeed, in the 5-mile stretch from Mill Creek to a point below the mouth of Big Creek (fig. 14) it is so constricted that in its natural state, before the road was built, the gorge was utterly impassable. Even the engineers who surveyed the course and profile of the San Joaquin River for the U. S. Geological Survey, although experienced in making their way up the rough, bouldery beds of mountain torrents, were balked by this exceedingly narrow portion of the San Joaquin's gorge, and were obliged to carry their line over the flanking benches. In several places, the walls of bare massive granite descend to the rock channel of the brawling stream with marvelously smooth convexly curved profiles that are steepest at the bottom.

6In his field notes Matthes describes features that suggest that the inner gorge of the lower san Joaquin canyon is a two-storied affair, but he does not incorporate this suggestion in the rough draft which formed the basis for this report. To put his observations on record, they are herewith quoted:

"Below the mouth of Big Creek, [San Joaquin Canyon is] narrow, V-shaped. River runs in rock cut channel. Granite prevailingly massive, exfoliating on a large scale. well-defined shoulders about 500-600 feet up. Apparently indicative of an old valley floor. These shoulders, however, are themselves about 1,000 feet below mouths of late-Pliocene hanging valleys. Lowermost part of shoulders prominent on west side of canyon, north of Italian Bar (2,100 feet). Also on south side, east of mouth of Mill Creek. Road built on this shoulder. Shoulders of corresponding height (about 400 feet above river) north of mouth of Big Creek. Would seem to indicate renewed tilting of Sierra late in Pleistocene. Farther down river these shoulders are not marked or [are] entirely absent for long stretches, for instance in stretch below Italian Bar, as far as mouth of North Fork, still even in this stretch the slopes immediately adjacent to river bed are much steeper than higher up."

FIGURE 14.—View in San Joaquin Canyon below mouth of Stevenson Creek, looking downstream. A few exfoliation shells cling to the massive granite walls. The white streak on the left is a streamlet that glides down the smoothly curving surface of the rock.

Equally impassable is the inner gorge for a stretch of 1-1/2 miles above Ross Creek, and again for a distance of 4 miles below the junction of the Middle and South Forks. In the last named stretch (fig. 15) the walls, though relatively low, are exceptionally sheer and close together. They grow rapidly higher toward the junction, where they measure 1,400 feet and are continued in the still higher walls of the gorges of the Middle and South Forks, which gorges are of essentially the same, sheer-walled type. Viewed from any of the high surrounding mountains, these gorges all seem like great mysterious moats dug by cyclopean hands deep in the granite floors of the broad mountain valleys. The extraordinary aspect of the junction is further enhanced by Balloon Dome, the solitary monumental dome of smooth, bare granite 600 feet in height, which rises from the upland spur between the two tributary gorges.

FIGURE 15—View up gorge of the San Joaquin River toward Balloon Dome (altitude 6,900 feet), the great monolith which stands between the canyons of the Middle Fork and the South Fork at their juncture. The massive granitic rocks which here form the walls of the gorge are exfoliating in many places.

Although impressive as a scenic feature, the inner gorge of the San Joaquin River is far surpassed in depth and grandeur by the gorge of the Middle Fork. That gorge, with its smooth walls of exfoliating granite 3,000 to 3,500 feet in height, streaked with silvery ribbon cascades that glide down from lofty hanging valleys, affords an unusually fine example of the massive type of cliff sculpture that finds its noblest expression in the Yosemite region. The gorge maintains its great depth and rugged grandeur for 15 miles, a distance of more than twice that of Yosemite Valley. As far as the vicinity of Fish Creek, it makes many turns that correspond to the crooked course of the river. Two to three miles broad from rim to rim, the gorge occupies practically the entire width of the canyon, there being little left of the outer mountain valley beyond its rims. Into it empty two great tributary gorges 2,000 to 2,500 feet in depth, the gorges of the North Fork and Fish Creek, with the result that a climax of imposing canyon topography is seen that is surpassed in but few other parts of the Sierra Nevada.

Above the mouth of Fish Creek, the canyon of the Middle Fork makes a right angle turn, and immediately beyond, it undergoes a radical change in form and aspect. Its walls flare apart, and a longitudinally ridged rock floor, a mile wide, spreads out between them. There is still an inner gorge cut in this rock floor, but it is a relatively small, subordinate feature. Northward this gorge becomes shallow rapidly, and in about 4 miles it comes to a head, one half of a mile below Rainbow Falls. Thence upward the canyon assumes the form, essentially, of a simple glacier trough with broad level floor and parallel spurless sides. It maintains depths, below the flanking uplands, which decrease from 1,500 feet to 1,000 feet a short distance from its head. The latter divides, the east branch leading steeply up to the Agnew Pass, the west branch to the outlet of Thousand Island Lake.

In the midst of this broadly open part of the canyon, which with its charming meadows and stately pine groves is the very antithesis of the constricted, forbidding gorge below, is situated the Devils Postpile. This mass of columnar basalt forms a solitary hump in the canyon, about 300 feet high and elongated along the axis of the valley of Middle Fork. This hump is the first of several such obstructions. For several miles above Pumice Flat, the broad trough shape of the canyon is hardly evident to one traveling through its depths because of the succession of rock ridges 300 to 500 feet high, that occupy its floor. These ridges are composed not of basalt, but of metamorphic and granitic rocks. Between them are timbered fiats and lush meadows, among which the Agnew Meadows are by far the largest.

In its upper course the canyon contracts to a V-shape, but otherwise it retains the aspect of a glacial canyon. Its sides are notably smooth, spurless, and parallel, and on the west side a series of hanging valleys containing rock-rimmed lakes and lakelets overlook the canyon at heights varying from 500 to 1,400 feet. This upper stretch of the canyon, which trends southeastward, parallel to the main divide of the Sierra Nevada, and the broadly trough-shaped stretch below, which has a southward trend, are both remarkable for their nearly straight courses. The river follows a sinuous path through the bottom, but the sides of the canyon in both of these stretches have but faint curvature, in contrast to the walls of the gorge below the elbow-bend at Lion Point.

Two tributaries of the Middle Fork of the San Joaquin River—the North Fork and Fish Creek—likewise have notable canyons. North Fork Canyon heads on the southeast face of Mount Lyell (13,090), at the extreme north end of the San Joaquin Basin, and between Electra Peak (12,462 feet) on the west and Mount Davis (12,308 feet) on the east. It extends southward along the west base of the great Ritter Range though gradually diverging westward from it. North Fork Canyon is troughlike throughout, its west walls rising 1,500 to 2,500 feet to irregular, glaciated uplands, and its east walls, which are continuous with the slopes of the Ritter Range, rising to impressive heights of 5,000 feet or more in the Minarets and neighboring summits. The southerly course of the North Fork Canyon is zigzag, rather than direct, reflecting similar irregularities in its river, whose course has short southwestward sections alternating with longer southeastward ones. Glacial modification of North Fork Canyon is evident not only in its cross section, which is typically U-shaped, but also in the rounding of its many bends. The canyon is, therefore, better described as sinuous than as sharply angular. On the east, many short, straight tributaries descend steeply from the west slope of the Ritter Range; and on the west longer tributaries, with less steep gradients, enter the North Fork through prevailing southeastward courses.

Fish Creek Canyon, which trends northwestward for the most part, parallel to the adjacent main crest of the range differs from the North Fork Canyon in being remarkably straight and spurless. Its floor lies 2,500 to 3,000 feet below the glacially scoured uplands to the west, and 4,000 feet below high points on Mammoth Crest, to the east. The 3-1/2-mile section above its junction with Middle Fork is a steep-walled, narrow gorge; farther upstream, the canyon opens and becomes broad-floored, particularly beginning at the 7,800 foot level. In its upper part the valley has the aspect of a trough-like glaciated canyon, and its many small tributaries have scores of alpine lakes in their upper reaches and cirques.

The South Fork of the San Joaquin approaches its junction with the Middle Fork through a sheer-walled gorge that is even narrower than that of the Middle Fork and impassable so far as is known to the author. It attains a maximum depth of 2,200 feet opposite Balloon Dome and thence it shallows headward very rapidly, so that in a distance of about 10 miles it is reduced to an insignificant trench a few hundred feet in depth (fig. 16). As a consequence, the gorge of the South Fork is not comparable in scenic grandeur to the gorge of the Middle Fork. It has instead the aspect of a mysterious cleft or abyss in whose depths the river is all but hidden from view.

FIGURE 16.—View southeastward from end of ridge north of Hoffman Meadow, looking up the shallowing, upper section of the Pleistocene canyon of the South Fork of the San Joaquin River. On either side of the canyon are remnants of the broad Pliocene valley.

The South Fork makes an extremely rapid descent in this gorge. It falls 2,300 feet in 12 miles—that is, nearly 200 feet to the mile on an average. Actually the descent is largely concentrated in the lower third of the gorge, the fall there being 1,250 feet in 4 miles. The steepest stretch is but a short distance above the junction; the stream there descends about 400 feet in 1 mile.

The head of the gorge is ill defined. About 14 miles above the junction, near the mouth of Rattlesnake Creek, the gorge is still 500 feet deep, and for several miles farther upstream it continues as a shallow furrow in the floor of the broad mountain valley of the South Fork. That mountain valley is, in comparison even with the deeper and wider parts of the gorge, of immense size. Throughout the first 12 miles above the junction it is 4 to 5 miles wide, measured from the base of the mountains on one side to the base of the mountains on the other. It is not level floored, however. The bases of the mountains just referred to are 1,000 feet above the edges of the central gorge. The flanking peaks on the south side rise 4,000 feet above the valley, those on the north side from 4,500 to 5,000 feet.

The conditions along the South Fork are exactly the reverse from those along the Middle Fork. On the latter the inner gorge widens at the expense of the outer valley and attains imposing dimensions; on the South Fork the inner gorge dwindles to a mere furrow and the outer valley is the dominant feature.

Headward the great mountain valley of the South Fork contracts by degrees, at the same time assuming more and more the form of a simple U-shaped glacial trough with parallel, spurless sides. However, its floor is level only in places, being obstructed at intervals, like the floor of the upper Middle Fork Canyon, by knobs and ridges of resistant rock. Such is the case notably above the junction of Mono Creek, where a ridge 600 feet high and a multitude of lesser ridges and knobs, all of granite, make a curiously broken topography, extremely rugged on a small scale and correspondingly difficult to traverse. The most conspicuous obstruction is a ridge of granite 1,000 feet high that separates Poison Meadow from Jackass Meadow. It is known as the Jackass Dike although it is not a dike at all in the geologic sense. This ridge stands directly in the axis of the valley, forcing the river to make an eastward detour. Of unusual interest is the ridge of smooth, almost flawless massive granite between 100 and 200 feet high that projects squarely across the valley floor just below Florence Lake, leaving a small gateway through which the river can pass.

In the vicinity of Florence Lake the U-shape and the glacial aspect of the valley become pronounced and unmistakable. Upland shoulders develop on both sides, thereby delimiting the U-shape more sharply; and associated with these shoulders are many small hanging valleys nearly all of which contain typical glacial tarns. At the Blaney Meadows the U-shaped trough is still further accentuated by the presence of lateral moraines that extend like continuous embankments along the upland shoulders. These shoulders stand about 2,300 feet above the valley floor, and the adjoining peaks rise 3,500 to 4,600 feet about it.

Above the Blaney Meadows the trough narrows, so that there is room for only a few strips of bottom land along the stream, making the term canyon more appropriate than the term valley. Each of the three head branches of the South Fork Canyon—Goddard Canyon, Evolution Valley, and Piute Canyon—is of the same general type. All are narrow troughs with smooth, spurless, parallel sides; and all have, like the upper course of the Middle Fork Canyon, nearly straight or at best gently curving courses. Sharp bends and strong windings such as characterize the main canyon of the San Joaquin River are conspicuously absent from them.

Goddard Canyon, which is the pathway of the main headwater branch of the South Fork, is the most nearly straight of the three. It maintains a depth of about 2,000 feet below the upland shoulders to within a mile of its head, where the floor rises abruptly to the level of the upland and terminates in a beautiful amphitheater containing Martha Lake, a nearly circular lake three quarters of a mile in diameter. Throughout, its extent, Goddard Canyon is flanked by peaks that rise from 3,000 to 3,500 feet above its floor.

Evolution Valley, which hangs 600 feet above the main valley (fig. 17), is correspondingly less deep below its upland shoulders. These shoulders, moreover, are much broader than those of Goddard Canyon—so broad that they constitute platforms from which the flanking peaks rise as separate entities, far removed from the central trough (fig. 20). As a result the valley as a whole has a widely open aspect which, together with the breadth of its floor and the presence of several natural meadows, justifies the appellation of "valley" rather than of "canyon," although the total depth below the flanking peaks is fully as great as that of Goddard Canyon.

FIGURE 17—View from valley or the South Fork of the San Joaquin River looking eastward at the mouth of Evolution Valley, which hangs 600 feet above the floor of the main valley. Photograph by G. K. Gilbert.

Evolution Valley comes to a head abruptly in a small cirque to the east of the majestic peak named The Hermit; but this valley head does not mark the extreme limit of the hydrographic basin of Evolution Creek. The basin extends 5 miles farther to the southeast at a higher level, the true head being at Muir Pass, beyond which are the sources of the Middle Fork of the Kings River. Evolution Creek drops into the head of Evolution Valley by a spectacular cascade 1,000 feet in height. It spills from the lip of an upland valley of rare scenic beauty that contains a series of vividly colored, gemlike lakes, and is guarded by a group of imposing alpine peaks—the Evolution group, which is dominated by Mount Darwin, 13,841 feet in altitude. This remarkable upland valley has come to be known as Evolution Basin (fig. 23). It is a noteworthy fact that, although it forums an upper story, so to speak, to Evolution Valley, and has its lip at an altitude of 10,990 feet, it still lies 2,850 feet below the flat summit of Mount Darwin.

Piute Canyon differs from the two canyons just described mainly in that it splits into two forks extending almost at right angles to each other. Both forks, however, consist of typical smooth-sided and nearly straight troughs, flanked by broad upland shoulders. The northeast fork, known as French Canyon, heads in a shallow amphitheater below Pine Creek Pass; the southeast fork similarly heads in a shallow amphitheater below Piute Pass. These passes have altitudes of 11,000 and 11,400 feet respectively.

The uplands flanking these troughs attain remarkable breadth in some places and are covered by many lakes and tarns. No less than 9 good-sized lakes and 12 small tarns lie on the upland to the east of French Canyon; 5 lakes and several tarns lie on the upland to the west of it, and 1 lake and several tarns lie at its head. The upland to the north of the southeast fork, which is known as Humphreys Basin, comprises 10 square miles of area and contains a dozen lakes and many tarns. Its central feature is a lake over 1 mile in length, well named Desolation Lake.

Two other canyons are tributary to the South Fork, namely, the canyons of Bear Creek amid Mono Creek. Both are essentially of the trough type and have distinctly U-shaped cross sections; both have upland shoulders in their lower courses, but are almost devoid of such features in their upper courses. Both are 3,000 to 4,000 feet deep below the flanking peaks.


In describing the San Joaquin Canyon and its main branches, cursory mention has been made of hanging valleys that debouch into the canyons. These hanging valleys are a characteristic feature of the basin, and, compared to the other basins of the west slope of the Sierra Nevada, they are unusually numerous. Indeed, so few of the lesser tributary valleys of the San Joaquin River and its main branches are continuously graded down to the level of the master stream throughout their length, that it may be said that hanging valleys are the rule rather than the exception. Though the basin lacks the many spectacular waterfalls that distinguish the Yosemite region, it does possess an array of cascades that pour from the mouths of the hanging valleys (figs. 18, 19).

FIGURE 18.—Falls of Stevenson Creek on east side of San Joaquin Canyon. Exfoliation shells are evident in the massive granitic rock which holds up this hanging tributary valley.

FIGURE 19—Big Creek Cascade, viewed from vicinity of Cascada Station below Huntington Lake. Kerckhoff Dome (not labeled on topographic map) in background. Photograph by R. A. Parker.

Of special note is the fact that the hanging valleys are found not only in the glaciated upper course of the canyon but also throughout the unglaciated lower course—in fact, into the foothills zone, within a few miles of the mouth of the canyon. indeed, the hanging valleys are most numerous in this lower section, and here also are found some of the best preserved examples.

The hanging valleys are as a rule, of moderate depth, broad, and of gentle gradient—that is, they are mature or even postmature in form. Several terminate abruptly at the brinks of the canyons, with lips as yet scarcely notched, so that their waters tumble from them in spectacular cascades; others are trenched by incipient gulches for short distances from their mouths. Still others are so trenched for distances of several miles, but even these have untrenched upper courses long enough to show unmistakably that they belong to a family of mature valleys graded to a former higher level of the master stream.

As became clear in the field, and is evident also from the topographic map, most of the hanging valleys appear to fall into two distinct sets, or tiers. Those of the lower set are associated with the benches and shoulders which flank the San Joaquin Canyon. Those of the upper set form part of the billowy topography of the upland surface.

The hanging valleys of the lower set are more numerous, and, generally speaking, are better preserved. Some are remarkably clean-cut, being as yet almost unnotched, and the waters that issue from them cascade and glide down the walls of the inner gorge still unrecessed. This is notably the case in the 6-mile section of the San Joaquin Canyon above the mouth of Italian Creek, wherein the shoulders and benches are most strongly developed and the inner gorge is narrow and steep sided.

Though the hanging valleys of the lower set terminate above the canyon floors at different heights, they nevertheless vary within certain limits and according to a significant pattern. Their heights are least in the foothills and increase upstream, as the following examples show. The valley of Fine Gold Creek has a profile showing a discordance, with reference to the floor of San Joaquin Canyon, of about 400 feet; Big Sandy Creek, 7 miles farther upstream, has a somewhat greater discordance. About 15 miles above the mouth of Big Sandy Creek, Backbone Creek and Bald Mill Creek hang about 1,100 and 1,200 feet respectively. Still farther upstream are several valleys, mostly on the west side, clearly related to the flanking benches and shoulders of the canyon, which hang above the canyon floor at closely accordant heights ranging, for the most part, from 1,400 to 1,700 feet, generally close to 1,600 feet. Representative of this group are the hanging valleys of Saginaw Creek, Italian Creek, Hookers Creek, Clearwater Creek, Ross Creek, Fish Creek, and Shakeflat Creek. Shakeflat Creek is just within the glaciated area.

Big Creek, in leaving the upland at Huntington Lake at close to 7,000 feet, descends steeply about 4,850 feet through narrow Big Creek Canyon to reach the San Joaquin River (fig. 19). Its branch, Pitman Creek-Tamarack Creek, similarly leaves a broad open valley on the uplands south of Chinese Peak at about 7,000 feet, and descends about 4,900 feet to the San Joaquin River. These hanging valleys are representative of the upper tier.

In the lower, unglaciated section of the San Joaquin Canyon a few streams are noteworthy in that they first cascade from lofty upland valleys into the outer canyon of the San Joaquin River, and then, some 2,000 feet lower, they plunge abruptly from a well-defined bench into the narrow inner gorge. These streams are hanging with reference to both the upland and the flanking benches. The most conspicuous example is Stevenson Creek, which descends in a brawling cascade from a broad, shallow upland valley lying at about 5,100 feet, now partly drowned by Shaver Lake; and then, from a bench at about 3,400 feet, it makes a second descent, with a steep drop of about 1,600 feet, into the inner gorge of the San Joaquin River (fig. 18). The situation in the case of Jose Creek is somewhat similar. Four branches of this stream (the longest ones heading about a mile above Ockenden) cascade from shallow vales on the gently undulating plateau known locally as Pine Ridge, leaving its edge at an altitude of about 5,400 feet; and then, after uniting into a single stream and flowing several miles on a relatively gentle gradient, at an altitude of 2,900 feet the waters again cascade wildly from the lip of the Jose Basin, making a descent of about 1,500 feet to the bottom of the inner gorge.

Farther up the San Joaquin Canyon, within the outermost glaciated section, where flanking benches and shoulders are feebly represented or lacking, most of the hanging valleys are part of the upland surface and therefore are referable to the upper set. Such is the case with the valleys of both Rock Creek and Jackass Creek. The latter, for example, leaves a mature valley on the uplands and descends steeply about 2,900 feet to join the San Joaquin River.

Along the Middle Fork are numerous hanging valleys, most of which also belong to the upper set. Granite Creek tumbles into the Middle Fork from the fairly clean-cut lip of a hanging valley about 2,400 feet high. Stairway Creek, like Stevenson and Jose Creeks, has headwaters that descend steeply from upland valleys, and in its lower course it leaps from the unnotched lip of a hanging valley about 1,900 feet high. Directly opposite Stairway Creek, a streamlet cascades down about 2,300 feet from the brow of Junction Bluffs; and east of Lion Point another streamlet makes a precipitous descent of 2,600 feet from the edge of the upland. Crater Creek leaves its shallow upland valley at 8,600 feet and makes a steep descent of 2,600 feet to join the Middle Fork; the last 700 feet of this descent is made as it drops into the head of the inner gorge. Farther up the Middle Fork, the height of the hanging valleys decreases by degrees, as the upper valley becomes shallower. Shadow Creek makes a descent of only 500 feet, and the streamlet that issues from Garnet Lake drops only 600 feet.

In the lower South Fork one finds only a few hanging valleys of notable height. The streamlet which drains Cow Meadow hangs over 2,000 feet; Hoffman Creek falls about 1,800 feet from its upland vale; and Rube Creek makes a plunge of 1,500 feet. Farther up, the discordances diminish rapidly as the South Fork gorge shallows headward. Four Forks Creek and Rock Creek first cascade from upland valleys, and then in their lower courses they descend steeply about 1,000 feet to join the South Fork. A few miles farther upstream, Rattlesnake Creek cascades from a valley only 500 feet high. Beyond this point new forms of topography appear; the broad upper valley of the South Fork, as it gradually contracts to a U-shaped glacial trough, possesses a set of lofty hanging side valleys.


The great canyons of the San Joaquin River and its main branches are deeply trenched in the plateau-like uplands distinctive of the basin. These uplands are extensive in the central part of the basin, which in this respect is typical of the middle slope of the Sierra farther north. The uplands that flank the Yosemite Valley on both sides are of this general type, and so are the uplands on both sides of the Hetch Hetchy Valley and adjoining stretches of the Tuolumne Canyon. Lindgren (1911, p. 37-38) has noted the prevailingly plateaulike character of the middle slope still farther north in the range. In the southern Sierra Nevada, also, in the drainage basins of the Kings, Kaweah, and Kern Rivers, there are similar areas on the middle slope.

In the San Joaquin Basin, the uplands form extensive plateaulike areas, commonly several miles wide, mostly at altitudes of 7,000 to 8,000 feet and having an aggregate area of about 100 square miles. They are readily traced westward and decline in altitude to 5,000 or even 4,000 feet at their western margins within a few miles of the foothills, where they break off in declivities of 2,000 feet or more. Traced eastward, the uplands become higher and, for the most part, smaller and less continuous. Nevertheless they are recognizable as distinctive benches, shoulders, and other areas of undulatory surface which extend to the base of the High Sierra crests, at altitudes of 9,000 feet or higher.

Where most typical, as on the middle and lower slopes of the Sierra, the uplands are not strictly flat, but undulating, with a relief generally less than 500 feet and only exceptionally 1,000 feet. The valleys on the upland, where transected by the canyons of the San Joaquin River and its branches, are abruptly cut off and left hanging, or descend with greatly steepened gradient, as described in the foregoing section. The uplands bear many broad, grassy meadows (figs. 47, 48), which from their topographic situation and considerable extent obviously belong in a different category from the generally more restricted meadows which in places occupy the canyon floors (fig. 44).

The somewhat monotonous aspect of the uplands, resulting from the approximate concordance of their billowy timbered ridges, is relieved by scattered eminences that stand above their general level. These include isolated knobs and mountain groups, a few hundred feet high, common in the central part of the basin, and lofty ridges—actually mountain ranges several thousand feet high—especially characteristic of the upper part of the basin.

Some representative upland areas will be described its main branches are deeply trenched in the plateau of the San Joaquin Basin. Uplands slope southward and southeastward from the Chiquito Ridge and Kaiser Ridge and have hills and scattered knobs which range from a few hundred feet to as much as 1,000 feet in height. These eminences have no definite trend, but many of the shallow upland valleys parallel the crests of the High Sierra. The uplands in this section, as well as elsewhere in the basin, have many bright green meadows in some of the broadly open valleys, interrupting the somber green mantle of coniferous forests that otherwise extends uniformly over the entire area. It is noteworthy that most of the uplands in this area were never overridden and modified by ice, even in the more extensive earlier glacial stage. (See glacial map, pl. 1.)

The lower San Joaquin Canyon, which bisects this area from northeast to southwest, separates the uplands adjoining Chiquito Ridge from those adjoining Kaiser Ridge. The uplands west of the canyon begin, at the base of Chiquito Ridge, at about 7,000 feet, and extend southwestward 5 to 10 miles; then, at 5,500 feet, they terminate in an abrupt escarpment 1,500 to over 2,500 feet high. This escarpment, named South Fork Bluffs at the north, runs south-southeastward in a remarkably straight line as far as the San Joaquin River.

The uplands east of the canyon slope southwestward from altitudes of more than 9,000 feet at the head of Big Creek, to 4,500 feet on Bald Mountain, northeast of Big Sandy Valley. They thus decline 4,500 feet in a distance of 25 miles, or at a mean rate of 180 feet to the mile. Southward and southeastward the plateau country extends without significant break into the drainage basin of the Kings River.

Particularly noteworthy is the fact that the upland on Bald Mountain reaches within a few miles of the foothills, where it breaks off first in slopes of moderate declivity, then in a precipitous escarpment known as the Big Sandy Bluffs, the total drop being between 2,000 and 2,500 feet. It is up through a recess in this escarpment that the celebrated "tollhouse grade" of the road from Tollhouse to Shaver Lake reaches the upland. To the southeast of the tollhouse grade the escarpment is continued by the steep front of the flat-topped ridge known as Burro (Burrough) Mountain. The general trend of the escarpment is southeastward, and associated with it is a linear ridge, named Backbone Mountain, that trends in approximately the same direction.

North and northeast of Chiquito Ridge are other upland areas of essentially similar configuration, lying, for the most part, at altitudes of 6,000 to more than 7,000 feet. Such are the Beasore and Mugler Meadows, in the headwaters of Chiquito Creek; and the various meadows in the area drained by the middle courses of Jackass Creek and Granite Creek. In the basin of Jackass Creek, where the many natural meadows afford summer pasturage for large herds of cattle, resistant rocks form residual knobs such as Jackass Rock, Squaw Dome, and Cattle Mountain, which stand 600 to 900 feet above the undulatory upland (fig. 6).

Another upland tract occupies the triangle isolated by Kaiser Ridge on the south, the San Joaquin Canyon on the northwest, and the South Fork Canyon on the northeast. It slopes from a maximum altitude of 7,500 feet along the northeast edge to about 6,000 feet at the southwest corner. This tract includes Cow Meadow and Hoffman Meadow, and from it rises Mount Tom, almost 2,000 feet high.

Farther to the north and northeast, upland tracts border the North Fork Canyon and, especially, the Middle Fork Canyon. For example, on the southeast side of Middle Fork Canyon a billowy tract at 7,500 to 9,000 feet extends southward from Junction Bluffs. Across the canyon, a similar tract slopes southward from 9,000 feet at the Granite Stairway to 8,500 feet at Lion Point. Also, there is a significant upland stretching, with undulatory surface, from 8,700 feet at the north edge of Fish Valley to more than 9,000 feet at the base of Mammoth Crest and Mammoth Mountain. Both Mammoth Pass, altitude 9,300 feet, and Minaret Pass, altitude 9,200 feet, are at the general level of this upland surface and form part of it.

Other upland tracts are found in the areas drained by the South Fork and its tributaries. Those within the arc formed by the South Fork and Mono Creek include Onion Spring Meadow, Warm Creek Meadow, and other meadows lying, for the most part, at 7,000 to 8,000 feet. Along the principal valleys tributary to the South Fork, remnants of the upland surface form marginal benches and shoulders. In some places, as along Mono Creek Canyon and Evolution Valley, these features are remarkably regular and continuous (figs. 20, 21). In most places they undoubtedly have been more or less modified by glacial erosion. At the headwaters of Big Creek, in the Black Mountains-Hot Springs Pass area, upland tracts give rise to a number of extensive meadows, such as Long Meadow (figs. 47, 48) and Rock Meadow.

FIGURE 20.—View down Evolution Valley from high rock sill west of Evolution Lake. Upland beaches, remnants of the Miocene erosion surface, extend almost continuously along both sides of the U-shaped trough, which is a Pliocene valley modified by glaciation.

FIGURE 21.—Well-preserved bench, a remnant of the Miocene erosion surface, on the south side of Evolution Valley. The cascade descends from the mouth of a hanging valley.

Finally it should be noted that many of the valley heads at 9,000 feet or higher that indent the mountain ridges of the High Sierra are probably the erosional equivalents of the upland surfaces so extensively and well preserved at lower altitudes. This is indicated by their appropriate situation, altitude, and configuration, and is evident despite the fact that they have been resculptured in varying degrees into capacious, flat floored cirques through intense and repeated glacial erosion. In this category may be placed the cirque basin of Thousand Island Lake (altitude 9,850 feet) at the head of the South Fork (fig. 38), the upper most reaches of Fish Valley (the 5-mile section lying above 9,000 feet), and Evolution Basin (the 6-mile stretch above 11,000 feet, leading to Muir Pass), and probably many other cirques such as the Pioneer Basin at the head of Mono Creek and the expansive Humphreys Basin between the forks of Piute Creek.


It might be anticipated that on the back slope of a tilted block range like the Sierra Nevada, furrowed as it is by deep transverse canyons, one would find a preponderance of transverse crests. Yet in the higher parts of the San Joaquin Basin, as also in those of the Tuolumne and Merced basins and many other sections of the Sierra Nevada, subsidiary crests trending northwest-southeast, parallel to the axis of the range, greatly predominate, indeed, such longitudinal crests are characteristic features of the upper part of the range, adding greatly, by their bold sculpture, to the scenic beauty of that alpine region, which has come to be known as the High Sierra. But nowhere in the range are longitudinal crests more numerous, more continuous, and more closely spaced than in the upper part of the San Joaquin Basin. Rising 2,000 to 3,000 feet above the plateaulike uplands, and 4,000 to 5,000 feet above the floors of adjacent canyons, they stand high enough to be considered mountain ranges. A number of the crests, as will be noted below, do not lie entirely within the San Joaquin Basin, but extend far beyond its limits. The presence of the longitudinal crests accounts also for the preponderance of longitudinal stream courses among the tributaries of the San Joaquin River.

One of the most noteworthy of the longitudinal crests is the Ritter Range in the northern part of the San Joaquin Basin. Its summits range from 11,000 to over 13,000 feet in altitude, including Mount Lyell (13,090 feet), Banner Peak (12,957 feet), Mount Ritter (13,156 feet), and the Minarets, that imposing group of summits which Californian mountaineers have aptly styled "the king and queen of the Sierras7 and their retinue." The Ritter Range, exceptional in that it greatly overtops the drainage divide of the Sierra Nevada, about 6 miles to the east (see fig. 4), extends from Mount Lyell on the northwest to Iron Mountain on the southeast, a distance of 12 miles; but as will be clear from the topographic map, it is continuous northwestward with the Cathedral Range, which divides the upper Merced and Tuolumne Basins and which is also 12 miles in length. The two ranges, therefore, form a continuous crest whose aggregate length is 24 miles.

7The popular use of the plural form Sierras, though it may seem inconsistent with the singular form of the name Sierra Nevada and High Sierra, is nevertheless fully justified by the multiplicity of serrate crests in the higher parts of the range. The term Sierra Nevada Mountains, on the other hand, is not tolerated in California, it being manifestly tautological.

The Glacier Divide (fig. 22), in the southeast part of the San Joaquin Basin, is appropriately named for the many small ice bodies that cling to its shady and precipitous northeast flank. It is only 6 miles long, but as will be evident from the topographic map, it forms a northwestward extension of the sinuous crest that constitutes the main divide of the Sierra Nevada at the head of Evolution Creek (fig. 23). This crest continues southeastward far beyond the limits of the San Joaquin Basin and, as may be seen on the Bishop and Mount Whitney quadrangles, extends without significant break around the heads of the Kings and Kern Rivers. It has a total length, including the Glacier Divide, of 76 miles, and bears no less than 90 peaks ranging from 12,000 to over 14,000 feet in altitude, including Mount Whitney (14,496 feet), the culminating summit of the Sierra Nevada.

FIGURE 22.—View southeastward along Glacier Divide, showing marked asymmetry of crestline resulting from more intense glaciation on the northerly slopes of this ridge. The northerly slopes bear a number of small glaciers, two of which show in this view.

FIGURE 23.—Panorama looking southward from granite hill near outlet of Evolution Lake (altitude 10,990 feet). At the left is the flank of the sharp-crested ridge bearing Mount Darwin and Mount Wallace. The conical peak in the center is Mount Spenser. At the right is a glaciated ridge that separates Evolution Basin from the head of Evolution Valley below.

Noteworthy also is a series of alined crests farther to the west, whose principal units within the San Joaquin Basin are Chiquito Ridge, Kaiser Ridge, and the LeConte Divide.

Chiquito Ridge (fig. 8) is the crest lying west of the San Joaquin Canyon. It controls the southeastward courses of Chiquito Creek and its West Fork. This ridge, more than 10 miles in length, trends northwestward to the border of the San Joaquin Basin and continues within the Merced Basin as far as Mount Raymond, on whose west spur stands the Mariposa Grove of Big Trees. Its total length, thus measured, is 14 miles. Its summits rise a little above 8,000 feet in altitude, yet they stand 3,000 to 4,000 feet above the valleys of Chiquito Creek and its West Fork.

South of the San Joaquin Canyon is Kaiser Ridge, the great crest that controls the course of the South Fork of the San Joaquin River and walls off the High Sierra portion of the basin. A western segment of Kaiser Ridge (termed Kaiser Crest on some maps) rises to the north of Huntington Lake (fig. 24). Another segment continues the crest to the east of Kaiser Pass. Together the two segments form an arcuate ridge 19 miles long, bearing first toward the east, then east-southeast, and finally south toward the Hot Spring Pass. The principal summits on this divide exceed 10,000 feet in altitude and stand 4,000 feet above the valley of the South Fork.

FIGURE 24.—View southeast across Twin Lakes. The cirque wall in the background is cut into Kaiser Ridge, only the north side of which was effectively eroded by the ice in the later glaciation. The near shore and the low cliffs on the right are composed of white crystalline limestone; the island and the far shore are composed of granite.

From Hot Spring Pass, a sinuous mountain rampart runs eastward with increasing elevations to Mount Henry (altitude 12,197 feet), where it joins the LeConte Divide. The latter, a spectacular crest bearing a row of 12,000-foot peaks, extends in a straight line 9 miles southeastward to Mount Reinstein (altitude 12,595 feet), at the extreme southeastern corner of the basin. The crest continues southeastward 8 miles farther, into the Kings River Basin, where it is known as the White Divide.

This succession of crests, from Mount Raymond at the north to the south end of the White Divide, has an overall length of more than 60 miles, and two-thirds of this length lies athwart the San Joaquin Basin. Though the line of crests shows many irregularities of trend, as might be expected in view of its great length, the northwestern direction nevertheless predominates.

There are other northwestward-trending crests in the San Joaquin Basin, but none are comparable in length and continuity to those described. A nameless crest between the South Fork of Bear Creek and the South Fork of the San Joaquin River, bearing Mount Senger and Mount Hooper, is 9 miles long. Another nameless crest between Goddard Canyon and Evolution Creek, bearing Mount McGee (fig. 25) and Emerald Peak, is only 6 miles long, but if its continuation in the Kings River Basin, here called the Black Divide, be added to this length the entire length is 15 miles. About 10 miles southwest of the Ritter Range is a crest bearing Madera, Gale, and Sing Peaks; and northwestward from Triple Divide Peak, where it enters the Merced Basin, this crest is continued in the much more impressive Clark Range. The over-all length of this crest is also about 15 miles.

FIGURE 25.—Section of a High Sierra crest: the Mount McGee group of summits and the cirques at the head of McGee Creek, viewed from a point south of Evolution Valley.

As compared with the prominent and persistent longitudinal ridges which have been described, the transverse crests are not only fewer but of subordinate importance in the configuration of the San Joaquin Basin. One of these, the Goddard Divide (fig. 26), follows part of the southeast edge of the basin, where it separates the headwaters of the South Fork of the San Joaquin River from those of the Middle Fork of the Kings River. From Mount Wallace on the main Sierra crest, this divide trends southwestward to Mount Reinstein on the LeConte Divide, linking the two crests, which here are only 7 miles apart. Noteworthy features of the Goddard Divide are Mount Goddard (altitude 13,555 feet) and Muir Pass, by which the John Muir Trail enters the basin of the Middle Fork of the Kings River. At the north edge of the San Joaquin Basin, a nameless transverse crest straggles along the 11-mile interval between Mount Lyell and Triple Divide Peak, separating the headwaters of the North Fork of the San Joaquin Basin from those of the Merced Basin.

FIGURE 26—View from base of Red Mountain, looking up Goddard Canyon to Goddard Divide and, at left, Mount Goddard (altitude 13,555 feet). Photograph by G. K. Gilbert.

A few minor crests form curves or hooks. The most prominent examples are the Silver Divide, about 10 miles long, and the Mono Divide, 7 miles long, both in the High Sierra region between the South and Middle Forks. Both branch off from the main Sierra crest in southwestern directions but gradually curve around to the northwest.


Most ridges in the High Sierra part of the San Joaquin Basin have splintered crests or sharp summits due to the widening of adjacent valleys and the enlargement of cirques by glacial action. However, there are at least two summits in this basin which, like a number of other peaks in parts of the Sierra Nevada (Matthes, 1933a, p. 34-35; 1933c; 1937a, p. 8-9) have tabular summits. These peaks, Mount Darwin and Mount Wallace (fig. 27), are both on the main crest of the range, near the southeast corner of the basin. Mount Darwin has two detached summit platforms, 13,841 feet and 13,701 feet in altitude, respectively, which stand nearly 3,000 feet above the level of the adjacent Evolution Basin. Mount Wallace has a summit platform at an altitude of 13,328 feet.

FIGURE 27.—View from John Muir Trail in Evolution Basin: Mount Darwin at left, Mount Wallace at right. These two mountains have nearly level summits that are remnants of an ancient erosion surface. The level to which the ice rose in Evolution Basin is indicated by the upper limit of corrasion. Above this level the cliffs are deeply riven by frost and gullied by snow avalanches.

These tabular summits were preserved because the preglacial Mount Darwin and Mount Wallace were full-bodied mountains, broad enough so that—though trimmed back by the erosional effects of glaciers heading on their slopes—they nevertheless were not reduced to attenuated points, like the neighboring peaks. The flat summits that remain, appear to have escaped glaciation altogether and exhibit only the smoothing effect of long-continued nivation and concomitant solifluction.8

8These summits apparently were not visited by Matthes. They are probably very similar to the summit of Mount Whitney, which Matthes has described in detail (1937a, p. 16-18; 1950b, p. 83-89).


The main drainage divide of the Sierra Nevada at the head of the San Joaquin Basin is of unusual interest. It does not follow the same northwestern crest throughout its length, but is in places offset abruptly from one crest to another. Thus at the Piute Pass the divide is offset sharply northeastward to the crest which bears Mount Humphreys. Six miles farther, at the northwestern end of that crest, the divide again is sharply offset to the southwest, across the Pine Creek Pass. Thence the divide pursues a winding course along alternate northwestern and northeastern crests, as far as Mount Stanford, at the head of Mono Creek. There it turns abruptly southwestward along the crest which, farther to the southwest, bears the name Silver Divide. At Red and White Mountain, it finally resumes a northwestward course, and this it maintains for a distance of 27 miles to Agnew Pass at the northern corner of the San Joaquin Basin. The divide makes many minor turns and windings, especially in the stretch southeast of the Mammoth Pass, but these are due largely to deeply inset glacial cirques.

Throughout the greater part of its extent, the divide at the head of the San Joaquin Basin has a strongly serrate skyline marked by a long succession of jagged peaks. Even the Piute and Pine Creek Passes are scarcely deep enough to mar the general sawtooth effect. But at the Mammoth Pass, the divide abruptly loses its spectacular aspect and, dipping from altitudes of over 11,000 feet to a level of 9,300 feet, assumes a subdued and relatively monotonous appearance. Mammoth Mountain, it is true, again lifts the divide for a space above the 11,000-foot level, but beyond that mountain the divide again sags to about 9,200 feet in the low Minaret Pass. From this long sag it ascends by gentle slopes to the crest of the relatively smooth ridge that culminates in San Joaquin Mountain and terminates at the Agnew Pass.

Just outside the San Joaquin Basin is another summit which may belong to the same category as Mount Wallace and Mount Darwin. This summit, not labeled on the topographic map, has been designated "Mono Mesa" by J. T. Howell, and is a granitic tableland with an altitude of 12,240 feet, located at the head of Mono creek, 1 mile east of the intersection of Fresno, Mono, and Inyo counties. Howell suggests that "it represents what is probably California's oldest land surface that is unaltered by erosion."

Opposite the low sag at the Minaret Pass the valley lands to the east of the Sierra Nevada attain an unusually high level—about 8,000 feet above the sea and as a consequence the eastern escarpment, which at Mount Humphreys stands 9,500 feet high, here dwindles to a mere 1,200 feet. The escarpment, in this its lowest stretch, is masked by great accumulations of pumice derived from nearby volcanoes, and so presents only an unimpressive slope of moderate declivity. As if to dwarf it completely into insignificance, the sky piercing summits of the Bitter Range (fig. 4) rise a few miles to the west.


It is commonly stated, in brief generalized sketches of the Sierra Nevada, that the range consists essentially of one great crustal block tilted toward the southwest, with its northeastern edge elevated so as to form the crestline. That statement, which implies that the crestline, or main drainage divide, coincides with the top of the greateast escarpment, is literally true for some parts of the range, but it does not hold good for many other parts, notably the part here under consideration.

To any one who traces in detail on the topographic map the course of the main drainage divide, or crestline, of the range around the headwaters of the San Joaquin River, it must be evident that as a geomorphic feature it is distinct and essentially independent from the escarpment. In some places the main divide lies many miles to the southwest of the escarpment. Indeed, throughout its total length of 70 miles at the head of the San Joaquin Basin the divide coincides with the top of the escarpment only in two short stretches: in the 3-mile section between Mount Humphreys and Mount Emerson, and on the top of Mammoth Mountain.

From the crest between Mount Humphreys and Mount Emerson the escarpment falls off steeply 5,000 feet to what is known as the Buttermilk Country. The Buttermilk Country is still between 3,000 and 4,500 feet above Owens Valley and separated from the latter by several miles of the jumbled Tungsten Hills; but these hills are only details of the much-faulted and fractured front of the range, and the fact remains that the main escarpment of undisturbed rock rises directly from the Buttermilk Country. Near Mount Emerson southward, the escarpment is abruptly offset 12 miles northeastward, and as a consequence to the southeast of Bishop Creek, a large triangular area of upland country intervenes between escarpment and main divide. It is thus evident that in this section the main drainage divide of the range is in no wise determined by the top of the escarpment, but is an ancient erosional divide that antedates the creation of the escarpment by tectonic movements.

Northward from Mount Humphreys, the main divide and the escarpment diverge. Basin Mountain and Mount Tom, which presumably indicate approximately the original brink of the escarpment, stand 1-1/2 and 3 miles respectively to the northeast of the divide, attached to it by spurs that bear remnants of an unbroken surface corresponding in general altitude with the upland to the west of the divide. North of the canyon of Pine Creek, moreover, the distance between divide and escarpment increases abruptly, and another large upland area intervenes between them. This upland area attains a maximum breadth of 6 miles at the point where the escarpment makes a right-angle turn, from north-northeast to west-northwest. The configuration of the upland area is, like that of the upland to the southeast of Bishop Creek, essentially analogous to that of the High Sierra portion of the San Joaquin Basin, except that it drains northeastward instead of southwestward. The upland area just mentioned narrows westward and tapers to a point on Mammoth Mountain. Thence to the Agnew Pass the divide remains close to the top of the escarpment but it nevertheless does not coincide with the latter, being separated from it throughout by a narrow strip of gently sloping upland.

At Agnew Pass, the main divide finally deviates widely from the generally northwestward course of the main escarpment. It swings westward and southwestward to the Ritter Range follows that range from Mount Davis to Mount Lyell, and returns to the escarpment by a northeasterly course, thus making a detour around the head of the Rush Creek Basin, which drains into Mono Lake. The maximum departure of the main divide from the escarpment, on Mount Lyell, is 8 miles. The Rush Creek Basin is not merely a canyon with branching headwater gulches cut into the escarpment, but an upland area of High Sierra aspect, and its central valley, though cut to considerable depth, hangs 1,800 feet above the foot of the escarpment at Reversed Creek.

From these facts it is clear that at the head of the San Joaquin Basin the main drainage divide of the Sierra Nevada bears no constant relation to the east escarpment of the range, but is, except at two points, an independent feature and one of wholly different origin.


The geomorphologic features of the San Joaquin Basin closely resemble in their broad relationships those of the neighboring Yosemite region. However, as previously noted, they are distinguished by relative simplicity, being singularly free of those conditions which make Yosemite a special case and which complicate the records of other basins.

Briefly, the salient features may be summarized as follows: The San Joaquin Canyon is essentially a narrow, youthful gorge cut into the floor of a relatively broad, mature valley. This in turn is flanked by extensive undulating uplands surmounted by knobs and mountainous crests. At least one crest has peaks that bear distinctively tabular summits.

It is evident that these features are the products of four partial cycles of stream erosion, each initiated by a major uplift. At higher levels, erosional records have been obscured, as a result of repeated glaciation. The erosional records will be analyzed first, and in the next chapter the superimposed effects of glacial action will be considered. In general, it will be found advantageous in the present analysis, to work from the youngest to the oldest features of the landscape, and so far as possible to begin with the unglaciated lower part of the range, thence discussing the glacially modified higher parts to the east.


Expressive of the latest partial cycle of stream erosion are the narrow, V-shaped gorge or "inner canyon" of the San Joaquin River and its branches. These features are regarded as being entirely of Pleistocene origin. The youthfulness of the gorge, as shown by its narrow-shape and the steepness of its cliffy sides, is evident, and the fact that it is still being deepened and extended headward clearly shows that enough time has not elapsed since the last uplift took place to enable the stream to intrench itself deeply all the way to its sources. At the same time there are ample indications that most of the gorge cutting was done prior to the great extension of the El Portal glaciers. On the basis of these relationships, the Pleistocene gorge of the San Joaquin Canyon may be correlated with the "canyon stage" of the Yosemite region.

The benches and shoulders which flank the Pleistocene gorge in the lower and middle reaches of the basin, giving the San Joaquin Canyon a "two storied" aspect by reason of the "outer valley" which they produce, are remnants of a mature valley developed in a partial cycle of erosion that preceded the canyon stage. These remnants are best preserved in the 6-mile section between the mouth of Italian Creek and the mouth of Big Creek; farther upstream they are less continuous and distinct but nevertheless traceable; and in the main branches of the San Joaquin, above the heads of the gorges, they are continued by the actual upper valleys themselves,—here untrenched by streams but glacially deepened and remodeled to a greater or less extent into U-shaped troughs. The maturity of these valleys is indicated by their broadly flaring cross sections and their low, smooth gradients (deduced, in the case of the San Joaquin Canyon itself, from the extended gradients of the lateral valleys that were left hanging in the latest cycle of erosion). Evidently in this partial cycle of erosion the San Joaquin River extended its main branch valleys much farther headward than it has in the present (canyon) cycle—in fact almost to the crest of the range. The head of Middle Fork valley, for example, probably was then about 3 miles above Agnew Meadows; that of the South Fork at about the position of the present head of Evolution Valley. The maximum depth of the San Joaquin valley at this stage was in the vicinity of the mouth of Stevenson Creek, where it was cut about 1,800 feet below the uplands.

To have attained its mature character the valley must have required, considering the durable character of the rocks in which it was produced, a period several times—and probably not less than five times—as long as that which was required for the carving of the Pleistocene gorge. It may be inferred that the partial cycle of erosion in which the valley was developed embraced all, or nearly all, of the Pliocene epoch. Accordingly, the Pliocene valley of the San Joaquin River may be assigned to the "mountain valley stage" of the Yosemite region.

The undulating uplands into which the Pliocene valleys have been cut are erosion surfaces referable to a still earlier cycle. Remnants of this surface are extensive and well preserved, in spite of their antiquity, because they are developed on the prevailingly massive, and therefore obdurate, granitic rocks which underlie a great part of the basin. The remnants clearly indicate that the landscape produced in this cycle was one of late mature aspect, within a relief over wide areas of less than 1,000 feet. Locally, however, monadnocks—not only scattered hills and mountain groups but even ranges a dozen to a score of miles long, most of them trending in a northwestern direction—give it much greater relief, for example 1,500 feet in the vicinity of Shaver Lake and over 2,500 feet at the present crest of the Sierra Nevada.

In the High Sierra region, the uplands reached to the base of the monadnock ranges and occupied the intervals between them. The head of Middle Fork valley lay essentially where it now does, in the position of Thousand Island Lake; that of the South Fork valley, in Evolution Basin; and other principal valleys headed in comparable situations.

The forms of the uplands are so much more mature than those of the Pliocene valleys that there is good reason to assign for their development a correspondingly greater span of time. Furthermore, they are clearly correlative with the uplands of the Yosemite region—indeed, they are traceable across the divide and continuous within the uplands of that region. From evidence already set forth (see p. 23) it appears that the Yosemite uplands, representing the "broad valley stage," are of late Miocene age. Accordingly, the upands of the San Joaquin Basin are also thought to have been produced in a cycle of erosion which began some time back in the Miocene and continued until the latter part of that epoch.

Finally, the tabular summits of Mount Darwin and Mount Wallace are, from the evidence of the great height to which these summits rise, much older than the Miocene uplands. These small but significant platforms are undoubtedly the remnants of a once extensive erosion surface, and they are clearly referable to that most ancient erosion surface recognizable in the region, of which traces survive on certain summits in the Yosemite region and elsewhere in the Sierra Nevada. This erosion surface is thought to be Eocene in age.


The hanging lateral valleys of the San Joaquin Canyon and its main branches present one of the striking features of the basin. In the glaciated area, especially in Middle Fork Canyon, the question arises as to how much of the discordance of the hanging side valleys is due to glacial erosion and how much of it is due to stream erosion; but in the lower canyon where, below the limits of glaciation, the landforms are essentially all stream-sculptured, obviously glacial erosion may be ruled out as a complicating factor. It is here, in the 40-mile stretch of canyon beginning a few miles below the mouth of Chiquito Creek and extending to within a few miles of the foot of the range, that the hanging valleys are especially well developed and numerous.

The explanation is offered that the lateral valleys of the lower San Joaquin Canyon are hanging because their streams have been unable to keep pace with the accelerated trenching of the San Joaquin River following its rejuvenation by uptiltings of the Sierra Nevada. The disadvantage of the side streams may be attributed not only to their smaller volume but also, and more particularly, to the fact that their courses—trending southeastward or northwestward, at right angles to the direction of tilting—were essentially unsteepened, whereas the course of the San Joaquin River—trending southwestward, parallel to the direction of tilting—was appreciably steepened.9

>9Erwin (1934, p. 59) observes, for the Minaret District, that the relative cutting power of the streams, as a factor in the production of hanging valleys, has depended upon the volume of the streams, the resistiveness of the bedrock, and the relation of the stream courses to the tilt of the range. He analyzes the hanging valleys of the district with these factors in mind. For most hanging valleys at high altitudes, glacial erosion must also be taken into consideration.

Most of the hanging lateral valleys are in two tiers or sets, the one much higher than the other. Valleys of the upper set are features of the plateaulike uplands, that is they hang within reference to the mature Miocene erosion surface. Valleys of the lower set are associated within the flanking benches and shoulders of the San Joaquin Canyon, that is, they hang within reference to the Pliocene erosion surface. These two sets of hanging valleys are quite distinct; those of each set are remarkably accordant in level, but no amount of adjusting enables one to bring the two sets into accord.

The hanging valleys, no less than the associated erosional surfaces, are clearly indicative of two separate and distinct stages in the erosional history of the region. They attest to two rejuvenations of the master stream induced by two uptiltings of the Sierra Nevada. Valleys of the upper set were left hanging as a result of the uplift which initiated the Pliocene or "mountain valley" stage of erosion; those of the lower set, as a result of the uplift which initiated the Pleistocene or "canyon" stage of erosion. Valleys of the upper set, being developed on massive granitic rock of an exceedingly resistant nature, have been preserved in spite of their great age. Valleys of the lower set, being developed on rock which is less massive and therefore less resistant to stream erosion, during the Pliocene stage became graded to the new base level of the master stream, but in the latest or Pleistocene stage have become hanging again through the cutting of the canyon. The two sets of hanging valleys are analogous to the two upper tiers of hanging valleys in the Yosemite region.


The hanging valleys of the San Joaquin Basin provide data from which it is possible to reconstruct the longitudinal profiles of the San Joaquin River and its two main branches for each of the two stages of erosion indicated. Similar stream profiles were reconstructed for the Merced River and were useful in analyzing the erosional features of the Yosemite region (Matthes, 1930a, p. 33-45; pl. 27). In the San Joaquin Basin the reconstruction of profiles is particularly successful because hanging valleys are exceptionally numerous, and they occur throughout a longer section and farther down toward the foothills than in the case of the Merced River.

The method used is essentially as follows (Matthes, 1930a, p. 35-36): The longitudinal profile of a hanging tributary valley may be projected with a fair degree of accuracy to the axis of the master valley, thereby restoring the destroyed lower part of the profile of the hanging valley and determining the approximate point at which the tributary stream formerly joined the master stream. The method assumes, of course, that there was originally no break in the profile of the tributary valley, but this assumption is justified in view of the mature form of the hanging valley and the smoothly concave character of its profile as far down as it is preserved. These two characteristics, which all the hanging valleys have in common, show that they were developed in a protracted cycle of erosion during which side streams evolved courses smoothly graded down to the level of the master stream.

If this method is applied to those hanging valleys of the San Joaquin River which careful study indicates will yield reliable data, two series of points are obtained, indicative of the former level of the San Joaquin River at the end of the "broad valley" or Miocene stage of development and the "mountain valley" or Pliocene stage of development. These two series of points may be plotted on the longitudinal profile of the present San Joaquin River, and when connected they give the restored profiles of the river, as shown on pl. 2.

The Miocene and Pliocene profiles cannot be reconstructed empirically from the topographic map alone, by one unacquainted with the ground. On the contrary, judgment based on close field observation is required, as well as sufficient understanding of topograhic methods to enable one to appraise the accuracy of features shown on the map. The writer has had sufficient experience in the topographic mapping of difficult mountain country to realize that much has to be sketched in from planetable stations with only such control as is afforded by intersecting points. Traversing along stream courses is as a rule precluded because it is too time consuming, and the topographer, if at all experienced in this work, will make the most of such views as his planetable stations afford to sketch the areas of relatively simple configuration such as uplands, especially those that are only sparingly timbered. Any data relating to hanging valleys based on the topographic map must, therefore, be critically evaluated before being used. This is illustrated, in following paragraphs, by analysis of some of the hanging valleys considered in reconstructing the Miocene profile of the Middle Fork of the San Joaquin River.

Granite Creek, the first tributary of the Middle Fork above the junction with the South Fork, affords a particularly reliable altitude point for the restoration of the Miocene profile of the Middle Fork. A careful plot of its profile shows that in spite of local modifications due to glacial action, including the development of a glacial stairway in its upper course, the general curve of the preglacial profile is readily reconstructed. Differences in the personal judgment of several individuals attempting to make such a reconstruction probably would not prevent their obtaining closely accordant results. The extension of the curve forward to the axis of the main canyon involves but little extrapolation. The total length of the upland course of Granite Creek (following the West Fork and Post Creek, which is really the main headwater branch) is 12.5 miles; the extension of the profile from the edge of the upland to the axis of the main canyon measures 1.85 miles.

The hanging valley of Granite Creek has been well preserved, despite its great age, because of the great resistance the stream has had in trenching the massive granite that underlies its lower course. The granite that extends from Soldier Meadow down to the edge of the upland and over a considerable area to the southwest is so sparsely fractured that vegetation still is very scant in places. Luxuriant meadows, it is true, occur on alluvial ground in the shallow valleys, and groves of forest trees stand on the gentle slopes where granite sand offers at least a thin soil for their roots but the hills and ridges from which the granite sand is being stripped are still largely bare. The massive structure of the granite, further, is conspicuous in the sides of the canyon of the Middle Fork, where it gives rise to smooth, unscalable cliffs.

The East Fork of Granite Creek does not afford data that can be used for restoration of the Miocene profile of the Middle Fork. As will be clear from the map, the East Fork of Granite Creek was originally a tributary to the North Fork of the San Joaquin, its course being southeastward either through the depression now occupied by the Cora Lakes or, farther south, through the deeply incised channel north of Green Mountain. Its old valley has been so thoroughly remodeled by recurrent glacial action that very probably the stream has followed different channels at different times during the Pleistocene. Its present course southward through the notch west of Green Mountain is undoubtedly due to a diversion brought about by the glacier that filled its old valley. The notch itself was in all probability produced by an ice lobe that spilled southward over a low divide west of Green Mountain.

The point on the Miocene profile of the Middle Fork indicated by Granite Creek has an altitude of 6,600 feet or more; and a very closely accordant point is indicated by the profile, of the small unnamed tributary of Granite Creek that flows past the north end of Squaw Dome.

Although the gently sloping upland back of the Junction Bluffs, on the southeast side of the Middle Fork, and the similar upland back of Lion Point, on the northeast side, are almost certainly good-sized remnants of the Miocene surface, they afford no data that can legitimately be used in the restoration of the Miocene profile of Middle Fork. Only short streamlets traverse these wooded uplands, and as the contouring must necessarily have been sketched by eye from planetable stations on distant summits, it can hardly be relied upon to give trustworthy results.

The upland course of Crater Creek presumably has a fairly reliable profile curve which may be extended to the axis of the Middle Fork. The detail with which its crooked wandering course is mapped affords some guarantee of accuracy in altitudes. The fact that the streamlet reaches the edge of the upland between the two Red Cones, which must have been located by planetable intersections and doubtless have well-determined elevations is, further, some guarantee of the accuracy of the altitude of the stream to the edge of the plateau. Due allowance must be made for glacial erosion at the head of the valley, but this cannot have been extensive. As the contour lines on the topographic map plainly show, the lower part of the upland course of Crater Creek is incised below the original profile of the Miocene cycle, and is therefore disregarded.

Fish Valley may be considered, as a final example. This valley has been so profoundly altered by glaciation that its earlier profiles cannot be determined with confidence. But several of the lower tributaries of Fish Creek, notably Silver Creek, the west fork of Silver Creek, the unnamed streamlet to the west of the west fork, and the streamlet that descends from the upland to the north, into the head of Fish Valley, furnish mutually accordant points that define probably with a fair degree of accuracy a stretch of 3 miles of the Miocene profile of Fish Creek. The curve of this profile, extended downstreamward, accords well with the profile of the Middle Fork of the San Joaquin as determined by the two nearest tributaries, Stairway Creek below, and Crater Creek above.

The Pliocene profile of the San Joaquin River, as is shown on plate 2, has been reconstructed, from altitude points provided by the lower tier of hanging valleys, for a distance of about 60 miles below the junction of the Middle and South Forks. This profile extends, therefore, to within a few miles of the mouth of the San Joaquin Canyon. From its lowest point, at Fine Gold Creek, upstream to the junction of the Middle and South Forks, the profile steepens gradually and quite regularly.

The Pliocene profile has been reconstructed for a distance of about 12 miles up the Middle Fork to a point provided by the hanging valley of Stairway Creek. If the profile be extended beyond this point, through the glacially remodeled and deepened upper valley of the Middle Fork, it reaches to a point about 3 miles above Agnew Meadows, which is believed to be the approximate position of the head of the Pliocene valley.

The Pliocene profile has been reconstructed up the South Fork for a distance of about 14 miles, from points provided by the hanging valleys of Four Forks Creek and Rattlesnake Creek. If the profile be extended on up the South Fork, through the glacially remodeled and deepened upper valley, it meets the floor of Evolution Valley at an accordant level. This valley, which now hangs about 600 feet above the floor of the main valley (fig. 17), appears to be the glacially modified head of the Pliocene valley (Matthes, 1925, p. 41).

Similarly, the Miocene profile of the San Joaquin River has been reconstructed, from altitude points provided by the upper tier of hanging valleys, for a distance of about 27 miles below the junction of the Middle and South Forks. The lowest point on this profile, at Jose Creek, is about 40 miles from the mouth of the canyon. Throughout this section, the slope of the profile is quite uniform.

Above the junction, the Miocene profiles of both the Middle Fork and the South Fork are better documented than the corresponding Pliocene profiles, as there are more altitude points, some even within the upper valleys.

The Miocene profile has been reconstructed for a distance of 30-1/2 miles, up the Middle Fork to a point provided by the hanging valley of Garnet Lake. There is a marked steepening of the gradient as far as Crater Creek, but above this point the profile flattens, as far as the uppermost point, at Garnet Lake. If the profile is extended a short distance beyond this point, it reaches the basin of Thousand Island Lake at an accordant level. This basin, which lies about 1,650 feet higher than Agnew Meadows, is apparently the glacially modified head of the Miocene valley.

The Miocene profile has been reconstructed 34 miles up the South Fork to a point provided by the hanging valley in which Heart Lake is situated. Throughout this distance, the profile steepens progressively. If the profile is extended on up through Evolution Valley, with a gradient that continues to steepen, it joins the floor of Evolution Basin at an accordant level. This basin, which hangs more than a thousand feet above the head of Evolution Valley, appears to be the glacially modified head of the Miocene valley (Matthes, 1925, p. 41).

It is to be noted that if the reconstructed Middle Fork and South Fork profiles of either the Miocene or the Pliocene set are compared for the sections above the junction wherein both sets are represented, those of the Middle Fork are steeper than those of the South Fork. In these sections, the two branches of the San Joaquin River are oriented at right angles to each other, the Middle Fork flowing southwestward, directly down the west slope of the range, and the South Fork flowing northwestward, along the slope and parallel to the crest of the range. As a result of this difference in orientation, the ancient profiles of the Middle Fork have been steepened by the uptiltings of the Sierra Nevada, whereas the corresponding profiles of the South Fork have remained unsteepened or have been steepened to a negligible amount. The decrease in slope of the Miocene profile of the Middle Fork, between Crater Creek and Deadman's Pass, probably reflects the change in trend of the river, which in this section has a southerly course; and the further flattening of this profile above Deadman's Pass probably reflects the southeastward course of the river in this section.

These relationships in the San Joaquin Basin are consistent with those which have been found farther north in the Sierra Nevada, where the reconstructed Tertiary profiles of other rivers similarly show steepening in those sections of their courses that trend parallel to the direction of tilting, and lack of steepening in those sections where their courses are at right angles to that direction (Lingdren, 1911, p. 46-48; Matthes, 1930a, p. 43-44).


The hills, mountain groups, and long mountain ranges standing above remnants of the Miocene erosion surface are monadnocks and, in all probability, are features inherited from the system of northwestward trending ridges that occupied the place of the present Sierra Nevada in Late Jurassic and Cretaceous time. The folded sedimentary and volcanic beds have been largely worn away from most of these ranges, as at Kaiser Ridge and only the cores of granitic rock remain; but in a few instances, notably the Ritter Range and the LeConte Divide, the metamorphic beds still remain in large volume.

Recognition of the mountain crests as features inherited from the older mountain systems carries with it, by implication, the recognition of the northwestward-trending valley troughs between them as similarly inherited features, it would appear, therefore, that the South Fork of the San Joaquin River follows a valley that dates far back into Cretaceous time, and that the same is true of the other northwestward-trending valleys, such as those of Chiquito Creek, Granite Creek, the North Fork, and the upper course of the Middle Fork.

In the case of the South Fork, the headward part of the river still flows over the metamorphic rocks, to whose structure it long ago became adjusted and in which Goddard Canyon has been excavated. Throughout the rest of its course, however, the South Fork flows over granite. The South Fork in this section, and also the other longitudinal streams in the basin which similarly are now cutting in granite, are believed to have acquired their northwesterly or southeasterly courses largely by superposition from the metamorphic rocks. Their trends over the areas of granitic rock still reflect the pattern of the ancient drainage which had become extensively adjusted to the structure of the Appalachian-type mountains (Matthes, 1930a, p. 2).

The canyons of the North Fork and the upper part of the Middle Fork also are cut in metamorphic rocks, and their courses conform in large part to the structures of these rocks (Erwin, 1934, p. 55). The southeasterly trend of the canyon of the Middle Fork, from the vicinity of Agnew Pass to a point about a mile below Agnew Meadows, accords with the strike of the upturned beds and the parallel system of strike faults. The course of the river coincides approximately with the trace of one of these faults for a distance of several miles. Several small tributary streams on the east side of the canyon likewise follow the strike throughout the major part of their courses. The canyon of the North Fork, taken as a whole, has a more southerly trend than the strike of the beds, but when examined in detail its course is found to consist of four long southeastward-trending segments that have southwesterly trends, across the strike. Most of the tributary streams on the west side of the North Fork flow southeastward.

Similar structural control is evident in the Ritter Range, whose major landforms, and also certain of its minor ones, trend parallel to the strike of the metamorphic rocks (Erwin, 1934, p. 9). The great height of the range doubtless is due to the superior resistance to erosion of the rocks which make up its bulk. The LeConte Divide probably illustrates the same relationships.10

10The investigations of Erwin (1934) and Mayo (1937, 1941) made subsequent to this reconnaissance, indicate that joint patterns also have been highly influential in determining drainage and topographic patterns in the central and southern parts of the Sierra Nevada. Mayo (1941, p. 1053) has found that the steep, primary joints are features inherited from the oldest structural patterns of the range.

By mapping the joints that dip more than 60°, Mayo (1941, p. 1050-1053, pl. 2) has shown that throughout a considerable part of the Sierra Nevada, including a portion of the San Joaquin Basin, there are four principal joint sets. These trend, respectively, west-northwest, northwest, north-south, and northeast.

The northwest joints are especially well developed and important as a topographic influence. For example, according to Mayo (1941, p. 1052) "in the Mt. Goddard quadrangle, many canyons, such as First and Second Recesses, and Goddard Canyon, trend northwestward . . . Throughout the area the sculpturing of many a bold peak and sharp ridge has been partly controlled by the orientations and spacings of northwest joints."

Adjustment to north-south joints may explain the course of the Middle Fork in the section south of Pumice Flat (Mayo, 1937, p. 181).

In the metamorphic rocks of the Ritter Range, Erwin (1934, p. 56) noted that the northeast set of joints is best developed, and it is reflected in topographic features of the sharp and rugged Minarets.

Mayo (1941, p. 1052) states that the northeast joints "are followed by many of the master streams that drain the western slope of the Sierra Nevada and by most of the canyons that furrow its eastern scarp. Along the crest of the range . . . swarms of northeast joints are responsible for jagged irregularities in the skyline of the northwest-trending ridges."

Over wide areas, however, correspondence of this kind between landforms and the constituent rocks, or their structures, is not evident. Presumably in these places the lithologic dissimilarities either were not such as to cause noteworthy topographic contrasts, due to differential weathering and erosion, or else the contrasts were obliterated during the long Tertiary cycles of erosion. Thus in the northern part of the basin, none of the projecting tongues of metamorphic rock, whether in the western, southern, or southeastern margins of the metamorphic area, finds expression in the primary or even in the secondary features of the topography. One cannot infer the windings of the contact between metamorphic and granitic rocks from the positions and forms of ridges or valleys. The latter features appear to have been carved indifferently across both kinds of rock, and the trace of the contact wanders among and over them without attendant changes in style or topography, except perhaps in the minutely sculptured details.

In connection with the course of the San Joaquin River, the broad sag in the crest of the range at Mammoth Mountain appears to be of critical significance. Mayo (1941, p. 1068, 1081, pl. 3) points out that this sag lies at the southwestern corner of the "Mammoth reentrant" in the eastern front of the Sierra Nevada. The reentrant coincides with a complex of intersecting faults. The latter probably played an important part in determining not only the reentrant but also the development of the sag and the localization of the intrusive rocks in it. This sag is thought to be the mouth of a Tertiary valley which extends far to the eastward and whose stream joined the Middle Fork when the waters from the country to the east still drained westward across the Sierra region. The Middle Fork flowed in a relatively shallow valley slightly below the level of the sag and below the shoulders that now flank the canyon. It seems probable, in view of the depth and breadth of the sag, that the valley indicated by it was then occupied by a large river—the San Joaquin—and that that part of the present South Fork, which pursues a general southeasterly course from Thousand Island Lake to the Devils Postpile, was merely a tributary of the master stream. The San Joaquin River, therefore, appears to be a "beheaded" stream, its original upper course having been cut off as a result of the faulting which produced the Sierra escarpment. Several other Sierra streams besides the San Joaquin, that now begin in gaps in the crest of the range, doubtless were similarly decapitated.11

11The interpretation of the San Joaquin River set forth above is the one which Matthes published (1930b). In unpublished and undated notes he has also given the following interpretation, which assumes that the San Joaquin River has had much the same history as that of the Merced River (Matthes, 1930a, p. 30-31). According to this interpretation, the San Joaquin River established its course conformably to the southwestern slant of the region, presumably early in the Tertiary period, when the first upwarpings or tiltings occurred that determined the southwestward direction of nearly all the master streams of the Sierra Nevada. As it grew headward it did so at the expense of the streams of the older drainage system, which were successively captured. The dividing ridges probably did not stand in the way of this capturing as much as might be supposed, for they were discontinuous, with many gaps. Among the captures made by the San Joaquin were those of the South Fork and the upper part of the Middle Fork. The sag in the crest of the range may simply mark a low divide; it may indicate an ancient valley through which the upper course of Middle Fork flowed northeastward, before it was captured; or it may record the valley of a former tributary to the Middle Fork.

Recent stream capture, of the type postulated by this interpretation, is illustrated on a small scale by Stevenson Creek. The former head of this creek flowed northwestward into the Shaver Lake Basin, but it was captured by the small stream of Blue Canyon, which flows south-southeast into the Kings River.


At the head of the San Joaquin Basin the main drainage divide, or crestline, does not coincide with the top of the great east escarpment of the Sierra Nevada except at two very short stretches. Throughout the rest of its 70-mile course, the main drainage divide bears no constant relation to the escarpment, but is separated from it, in places by intervals of many miles, and is a wholly independent feature in regard to altitude and trend.

It is evident that the main drainage divide is an erosional feature, and is of much greater antiquity than the escarpment: an ancient divide which parted northeastward-flowing waters from southwestward-flowing waters long before tectonic movements gave rise to the escarpment. The forms and gradients of the upland valleys to the northeast of the divide are closely similar to the corresponding features of the upland valleys on the southwest side, showing that prior to the formation of the escarpment, the Sierra region fell off rather symmetrically in both directions. Ancient erosional features like those referred to above, of Miocene and Pliocene age, are found only at high levels of the eastern slope.

The east escarpment is probably largely of early Pleistocene origin, and is a feature produced as a result of the downfaulting of the Owens Valley graben. There is reason to believe that it did not assume its imposing height until after the Sierra Nevada had undergone its first period of glaciation. For, whereas the moraines of the Sherwin, Tahoe, and Tioga stages of Blackwelder (1931) lie at the months of the deep canyons that gash the east mountain front, certain moraines of Blackwelder's earliest, the McGee stage, are situated on a high shoulder overlooking the escarpment, where they could have been deposited only before the canyons were cut to their present depth. (Matthes, 1933a, p. 35-39; 1950b, p. 42-50.)

<<< Previous <<< Contents >>> Next >>>

Last Updated: 27-Jul-2009