Evidence is plentiful that the glaciers are rapidly retreating. Lateral morainic material perched 150 feet above the sides of Chickaloon Glacier indicates a recent dwindling in the size of that glacier. Some of the small glaciers no longer join the trunk ice stream, but have become independent glaciers, whose troughs are elevated above the surface of the main glacier. (See Pl. IV, D, p. 42.) After the recession of the ice a growth of moss encroaches on the canyon walls which confine the glaciers. Where recent retreat has taken place a considerable interval of bare wall intervenes between the present position of the glacier and its former position, as indicated by the mossgrown profiles on the rock walls.

In addition to the glaciers at the head of Chickaloon Creek, a small glacier occurs at the head of Kings Creek and another at the head of Little Susitna River. Of the glaciers on the south side of Matanuska Valley, Mendenhall says: "

Along the middle course of the river several small ice streams are to be seen at the heads of the canyons which furrow the southern valley wall, but the largest glacier of the valley is the Matanuska Glacier with a frontage of 3 miles along the river. The course of this stream can be traced from the top of Glacier Point for 25 or 30 miles to the southeast. Its source here is probably in the névé which follows the crest of the Chugach Range to the east, and gives rise to the various glaciers which flow down to Prince William Sound on the the south and to the Tazlina and Klutina rivers on the north. While its east front is buried under moraine which supports a considerable spruce forest, the west front is clear of débris to the edge. A well-marked medial moraine divides the main stream into two parts, representing branches which unite near the source. Local conditions of precipitation controlling the rate of flow of these tributaries seem to be slightly different, so that the western branch is somewhat more active than the eastern one.

Within 10 miles of its source the Matanuska River receives the waters from another glacier, smaller than that just described and more actively retreating. Its front has now receded several miles from the river and the draining stream flows through a gravel-filled valley with a very high gradient.

At the head of Knik River lies the Knik Glacier, whose width is 23 miles. It is not much smaller than the Matanuska Glacier. Knik Glacier forks a few miles back from its terminal end, one fork extending eastward about 20 or 25 miles, probably to the same ice field that is drained by the Matanuska Glacier, the other fork extending southwestward. The latter fork again divides. In its western prong is a large lake 5 miles long and 13 miles wide, whose waters are apparently impounded by the extension of the lateral moraine of the eastern prong across the mouth of the trough of the western prong. Bergs were seen floating in the lake. According to local report the breaking of this morainal dam was the cause of the disastrous flood which swept down Knik River some years ago.

"Mendenhall, W. C., A reconnaissance from Resurrection Bay to Tanana River, Alaska, in 1898: Twentieth Ann. Rept. U. S. Geol. Survey, 1898-99, pt. 7, p. 327.

The front of Knik Glacier rests upon a gravel plain covered with well-rounded stream wash. Irregular heaps and ridges of this gravel 12 feet high lie in front of the glacier for 400 feet. Pits and kettle holes 6 to 8 feet deep are common. On the western side of the glacier an accumulation of large angular blocks of graywacke and slate, constituting the dump of the superglacial material, forms a narrow peripheral moraine nearly a mile long. In profile as exposed on its extreme western edge the glacier can be seen resting upon bed rock, which here begins to rise rather steeply from under the gravel covering of Knik Valley. The recent history of the Knik Glacier, as interpreted from the above features, shows a recession from a temporary advance over the gravel deposit of the valley.

PAST GLACIATION.

The existing glaciers are the shrunken remnants of a system of ice streams which had a vastly greater extent in the recent past. At the time of its maximum expansion this mighty system included the enormous piedmont glacier of the Copper River basin, and the trunk glaciers occupying Matanuska, Knik, and Susitna valleys, which were probably confluent at the head of Cook Inlet.

a

During the occupancy of the lowlands by these great ice streams, the Talkeetna Mountains were subjected to severe alpine glaciation, to which the ruggedness of the present topography is largely due. (See Pl. IV, C', p. 42.) In higher portions of the range the glacial erosion has often proceeded to its most destructive lengths. Knifeedged divides are numerous throughout the region, and instances of the obliteration of divides and the coalescence of opposing cirques are frequent. All the existing glaciers occupy the heads of U-shaped troughs. The finest example of such troughs is that along the upper course of Talkeetna River, which is perfectly straight for 18 miles.

The former presence of a great glacier once filling Matanuska Valley is indicated by abundant evidence. At the head of the valley the morainal ridges of the Copper River basin curve round into Matanuska Valley. An isolated hill about 500 feet high, projecting above the lake-covered flat at the head of the valley, shows an abrupt stoss side toward the Copper River basin, and a long, smooth lee slope toward Matanuska Valley. Along the upper course of the valley on the flank of Sheep Mountain glacial gravels of various exotic types are found at an elevation 1,500 feet above the valley floor. These facts indicate that the ice of the southwestern portion of the Copper River basin drained into Matanuska Valley in a stream not less than 1,500 feet thick. Passing farther westward we find that

• Mendenhall, W. C., Geology of the central Copper River region, Alaska; Prof. Paper U. S. Geol. Survey No. 41, 1905, p. 72.

the summit of the irruptive boss on the west side of Kings Creek is smoothed and glacially scored; and that between Eska and Granite creeks on the flanks of the valley a heavy lateral moraine, consisting of large angular bowlders of various granitic rocks, rests 2,400 feet above the river. Near tide water the buttes in the middle of the gravel plain of Matanuska Valley are striated and strongly fluted on their summit, 700 feet high.

This evidence is supplemented by that furnished by the physiography. A marked feature of Matanuska Valley, as seen from Knik Arm, is the appearance of triangular faceting cut on the left wall of the valley. This feature is due to the truncation of the projecting ridges between the streams tributary to the main valley, whereby the spurs have been replaced by steeply sloping planes with triangular outlines. This feature, it may be mentioned, has been developed in almost geometric perfection in the region of Turnagain Arm on Kenai Peninsula, but nowhere on the same large scale as in lower Matanuska Valley.

The evidence is equally strong that Knik Valley was formerly occupied by a large ice stream. Along its middle course glaciated surfaces are found 1,500 feet above the gravel floor of the valley. A minimum estimate is thus fixed for the thickness of the glacier.

On the Susitna side of the Talkeetna Mountains evidence of glacial action even more pronounced than that on the eastern side is visible on every hand. The mountains are bordered by a more or less irregular bench several miles wide, approximating 2,500 feet in elevation. Its surface is noticeably glaciated, rounded hillocks, smoothed outcrops, glacial débris, and numerous undrained lakes characterize its topography. Streams entering the Susitna pass at the mountain borders through beautiful glacial U-shaped valleys, their smaller tributaries entering these valleys through sharply incised, vertical-walled canyons, an evidence of recent cutting. The rise from this broad bench to the mountains eastward, when compared with the gentle slope of the bench to the Susitna, is steep.

It is probable that an ice sheet that occupied the entire Susitna Valley and was fed from the surrounding mountains once moved southward to Cook Inlet. The tributaries from the Talkeetna Mountains to Susitna River all present clear evidence of having been occupied by ice streams. Glacial U-shaped valleys and erratics occupying positions no longer within the reach of ice movements are traces of the mantle that covered much at least of the present drainage system.

It is probable that after the recession of the huge ice mass occupying Susitna Valley glacial action continued in the smaller streams for a considerable period. This is indicated by the fact that the gla

cial U-shapes on the western slope of the Talkeetna Mountains are considerably lower than the broad ice-sculptured bench 2,500 feet high.

PHYSIOGRAPHIC DEVELOPMENT.

The region under discussion falls naturally into three physiographic provinces: Matanuska Valley, Talkeetna Mountains, and Copper River basin.

MATANUSKA VALLEY.

According to Mendenhall, Matanuska Valley is probably due to simple erosion somewhat modified by ice action. In a subsequent report he has suggested that the southern edge of the Copper River basin is probably a fault or a sharp flexure. The westward prolongation of this fault would coincide with the southern wall of Matanuska Valley. Martin remarks that the Matanuska Valley suggests a "graben " or sunken area within parallel faults, but notes that this appearance may be due to the more resistant character of the rocks bounding the valley. The exact mode of origin of Matanuska Valley can not yet be regarded as established. It is beyond question true, however, that if of tectonic origin the valley has subsequently been largely modified by stream erosion. In proof of this is the fact that it possesses the form of a normal stream-wrought valley, narrow in its upper course and broad in its lower. From a width of 1 mile at the Copper River basin Matanuska Valley widens to 20 miles near Cook Inlet. Knik Arm appears to represent a slight drowning of lower Matanuska Valley, from which the region is at present recovering.

There is some evidence available that recent movement has taken place along the northern side of the valley. Along its upper course Sheep Mountain presents a steep, rugged front facing the river. At the base of the ridge the strata are shattered and an extensive disruption of the injected dikes has taken place. Extensive alluvial breccia cones spread out into the valley for a mile or more from the mouths of the ravines. The steep slope of these cones (6° to 10°) and their symmetrical development make them noteworthy features of the topography. The formation of such detrital cones is held by workers in the Western States to be due to a sudden break in the stream profile at the point of deposition, usually by recent faulting, rather than to arid conditions, which was first invoked in their explanation.

The present Matanuska River occupies a wide trench sunk into an older, broader valley averaging 5 miles or more in width. The

Op. cit., p. 239.

Mendenhall, W. C., Geology of the central Copper River region, Alaska: Prof. Paper U. S. Geol. Survey No. 41, 1905, p. 87.

Op. cit., p. 17.

trenching of the old valley appears to have begun in preglacial time, but was interrupted by the glacial advance. The ice occupancy effected some erosion, notably on the south side of Knik Arm, but left pronounced irregularities of relief in the floor of the valley, such as Glacier Point and the bed-rock bosses at the mouth of the Matanuska. On the recession of the ice a heavy mantle of gravel was laid down. A succession of slight uplifts followed, as expressed in the finely terraced character of the affluents of the Matanuska, and the streams began to remove the gravels from their channels and to deepen their trenches. The excavation of the gravels has been incomplete along the Matanuska, where long stretches of stream bluffs consist of gravel banks. These features suggest that either the preglacial drainage had begun to swing in its course or that glacial erosion has modified the valley floor, and that the irregular bed-rock relief thus produced has been obliterated by the gravel deposits subsequently laid down. The former supposition is regarded as the more probable, and the great bend of the Matanuska appears to be an incised meander inherited from the preglacial drainage. The present overloaded condition of the Matanuska is due to the vigorous trenching of the deep glacial gravels at its headwaters.

TALKEETNA MOUNTAINS.

The internal structure of the Talkeetna Mountains suggests that they belong to a province geologically distinct from that of the Chugach Mountains. The latter, as shown by Schrader, Spencer, and others, represent a belt of intense crustal shortening. In the Talkeetna Mountains, however, the strata are comparatively undisturbed. The younger rocks especially have been but gently affected by orogenic movements. The faulting is normal, implying a stretching of the earth's crust.

The early Tertiary history of the Talkeetna area has already been sketched. It was shown that, coeval with the deposition of the Kenai, a peneplain of considerable perfection was developed over the region of the Talkeetna Mountains. The older age limit of this planation is fixed by the fact that the surface is in part cut across a Lower Cretaceous limestone. The later limit is determined by the fact that the fluviatile conglomerates, resting upon the peneplain, appear to represent the similar conglomerates occurring at the top of Kenai or in the Matanuska basin. According to this view the conglomerates at the top of the Kenai represent the period at the close of lacustrine conditions, and the conglomerates in the Talkeetna area are the fluviatile accumulations of the rivers which transported the materials of the Kenai sediments. During the later stages of peneplanation and the formation of the fluviatile conglomerates, the land