University of Washington Publications in Geology The Geomorphology and Volcanic Sequence of Steens Mountain in Southeastern Oregon

STEENS MOUNTAIN ANDESITIC SERIES
(continued)

THE GREAT FLOW

GENERAL FEATURES

In the valley of Cottonwood Creek and to the south on the northern wall of Alvord Creek valley, a tremendous flow of andesite caps the stratified tuffs which overlie the upper flow of basic andesite. In the Alvord Creek locality, the flow shows a thickness of about 900 feet (fig. 40), but within a mile to the north, in the valley of Cottonwood Creek, it was found to have thinned to about 500 feet (fig. 41). Northward from that point it is not exposed, although stratigraphically at least the upper part should outcrop in the valley of Willow Creek. In the valley of Mann Creek, however, about seven miles farther north, a similar lava shows a thickness of about 400 feet. Although here the base is not visible, an adjacent cinder cone, at the base of the scarp at the mouth of Little Mann Creek, suggests that the bottom of the flow is not far distant (fig. 58).

Fig. 40. Northern side of the Alvord Creek valley adjacent to the scarp. At the base on the right the upper flow of basic andesite is exposed. Above it in the center of the view the great andesitic flow rises about 900 feet.

Fig. 41. View of the northern fork of Cottonwood Creek. showing at its base the upper flow of basic andesite. Above it, the tuffaceous bed is visible beneath the northern extension of the great flow. The poorly defined exposures in the upper center are formed by the subsequent thin andesitic flows. On the left, the basaltic series forms the cliff at the top.

With increasing thickness, the flow shows coarser jointing and on the northern wall of the valley of Alvord Creek it exhibits coarse rectangular columns that are approximately ten feet in diameter (fig. 42). These great columns are cut locally by irregular platy jointing, which towards the base is roughly horizontal, while higher in the exposure it is usually somewhat inclined. In both the valley of Cottonwood Creek and that of Mann Creek, the jointing is fairly uniform, but in the upper part of the exposure north of Alvord Creek it becomes lost to a great measure in irregular zones of alteration.

Fig. 42. View of the northern wall of the valley of Alvord Creek showing the maximum exposure of the great andesitic flow, which here has a thickness of about 900 feet. The columns average approximately 10 feet in diameter. The local absence of columnar jointing in the upper part may be explained by deuteric alteration.

The upper surface of the flow is invariably characterized by a well-defined shoulder, which usually supports a scattered growth of junipers. This topographic break is due to the erosion of a soft grayish clay-like material, which appears to have been derived from the devitrification of a glass, probably owing in part to the effect of volcanic emanations. Although the minor vesicularity of unaltered vitreous remnants forms the only indication of the top of the great flow, its extrusive origin is more conclusively indicated by the relation of the overlying thin andesitic flows, which lap against its inclined surface. These flows lack any indication of marked induration, and also fail to contain the hornblende which characterizes the upper surface of the great flow. To the south of Alvord Creek, the relation of the great extruded mass is not so clearly defined, partly owing to the fact that it is, at least locally, overlying one of the vents from which it welled.

On the southern wall of the valley of Alvord Creek, towards the scarp, the jointing becomes more symmetrical and the lava exhibits magnificent hexagonal columns that tower more than 300 feet above the talus slope (fig. 43). Although the upper exposures are not so distinct, the columns appear to extend upwards for at least twice that height. At the eastern margin of the exposure, they curve eastward and become progressively reduced in size from over five feet to approximately six inches in diameter. In like manner, immediately to the north, at an elevation several hundred feet lower and considerably below the base of the great flow, a waterfall in Alvord Creek has been formed by similar curving columns which dip northward to an unexposed cooling surface (fig. 44). These two localities, showing curving columns, mark points adjacent to the rounded margin of a vent, which is directly in line with one exposed to the south in the valley of Little Alvord Creek. The wall rock presumably was formed by the Alvord Creek Beds, which are exposed to the east.

Fig. 43. Columnar andesite forming the great neck on the southern wall of Alvord Creek valley. The columns are at least 300 feet in height. At the basal margin they curve until normal to a roughly vertical contact.

Fig. 44. The curving columns at the northern margin of the andesitic vent in Alvord Creek.

Within about 200 yards to the east of the main exposure, a small elliptical intrusion of dark aphanitic andesite cuts the tuffs. The outline is roughly about 100 by 150 feet across (fig. 45). Two similar, far smaller intrusions cut the stratified tuffs a few hundred yards to the south. These three volcanic necks lie almost precisely on a common north-south line, which parallels their major axes. In like manner, both the great flow and the upper andesitic series, which will be subsequently described, may have been extruded from a series of vents distributed at intervals along one main line of tensional weakness.

Fig. 45. View of the southern wall of Alvord Creek valley, showing the great andesite vent on the right and the largest of the series of smaller intrusions on the left.

Farther to the south on the northern side of the valley of the south fork of Alvord Creek, the inclination of the columns at the base of the great exposure suggests that it again approaches the eastern margin of the vent, but erosion is not sufficient to disclose a crosscutting relation. Here and immediately to the north, the andesite appears to have welled quietly from a broad elongate vent. To the south, the surface of the flow locally decreases in elevation, forming a depression which is filled with later andesitic flows. Poor exposures prevent the great flow from being traced continuously southward, but scattered outcrops indicate that it gradually develops explosive characteristics. Perhaps merely by chance, these are best defined where the underlying vent cuts the previous eminence formed by the uparched Alvord Creek Beds and the Pike Creek Volcanic Series, which abuts against the southern dip slope of the tuffaceous sediments.

On the northern side of the valley of Little Alvord Creek, directly above the vent for the rhyolite of that name, the southern continuation of the great flow of andesite forms a large exposure about 250 yards across with a semicircular lower margin which reaches to within 50 feet of the uppermost rhyolite tuff (fig. 46). About 100 yards to the east, a small outcrop shows the andesite in contact with the capping perlitic phase of the lower biotite-dacite flow of the Pike Creek series. This contact appears to be crosscutting, although it is not very conclusive.

Fig. 46. The rounded exposure in the center is formed by the southern margin of the great andesitic vent on the northern side of the valley of Little Alvord Creek. The upper part of the rhyolitic vent outcrops on the lower left at about 800 feet below the basaltic flows which form the cliff at the top.

The great exposure of andesite with its rounded lower margin extends upwards for over 700 feet, In its lower part, it shows coarse columnar jointing. Upwards it grades into a vertical platy rock, which is associated with great masses of reddish vesicular breccias (fig. 47). The uppermost facies, beneath the basalt, is largely decomposed to a soft grayish rock which is identical to that capping the great flow farther to the north and which is also presumably derived from devitrified glass. In this zone there were also observed a few scattered blocks of unaltered hornblende-free andesite, which may be of pyroclastic origin.

Fig. 47. Reddish breccias in the center of the andesitic vent north of Alvord Creek. On the right is a platy injection with vertical jointing. A figure in the foreground on the left furnishes the scale.

In the valley of Little Alvord Creek, some of the vesicular breccias, towards the western portion of the rounded exposure, exhibit locally a horizontal distribution. This fact, together with the horizontal platy jointing, which here is especially distinct near the base of the exposure, suggests that the andesite welled out as a flow far below the one which caps the series to the south. In the center of the exposure, however, the broad zones of breccia, which parallel the highly inclined platy structure, testify to an underlying locus of extrusion and suggest a relatively late stage of explosive activity. It is possible that the discordant relation with the biotite-dacite may also be explained by a volcanic explosion, but no supporting evidence was observed. On the other hand, to the west in the valley of Little Alvord Creek, both the lateral extent of the andesite and its horizontal structures suggest a flow, the relation of which would be more easily accounted for, at least in part, by the abrupt ending of one of the viscous dacitic flows.

Where Little Alvord Creek cuts this horizon at about a quarter of a mile to the west of the large semicircular exposure, it discloses such irregular platy jointing as to suggest another locus of extrusion. This locality is about a mile and a half due north of some small intrusions of dark glassy andesite in the western part of the valley of Pike Creek. It is possible that together they mark the position of another series of vents on a north-south line of weakness. If so they are contemporaneous in their activity with the main ones to the east. It is more probable, however, that the irregularity in the jointing has been caused by the same rapid increase in gradient that appears to have been responsible for the steep flow structure in the dacite immediately to the east.

PETROLOGY OF THE GREAT FLOW

In spite of its thickness, the great andesitic flow is predominantly aphanitic. Locally, however, small phenocrysts of both feldspar and amphibole are visible megascopically in a dark gray groundmass, which exhibits a fairly smooth fractured surface. The phenocrystic facies was observed invariably in the upper part of this massive lava. In the valley of Alvord Creek it also forms the horizontal columns that indicate the margin of the crosscutting andesite. To the south of Little Alvord Creek, a similar lava forms the flow capping the Pike Creek series. Towards the center of the main exposures, the porphyritic texture gradually disappears. The rock assumes a greenish or reddish shade and simultaneously develops a far rougher surface when fractured. The massive type invariably grades upwards into a light gray, highly altered facies containing small irregular cavities. Unfortunately, this light gray porous rock is so easily eroded that the surface of the flow is invariably masked by soil. The local survival of unaltered remnants indicates, however, that it has been derived from the devitrification of a glass.

In thin sections of the porphyritic type, the andesine commonly shows a marked seriate development with individual crystals ranging up to 2 or 3 mm. in length, but with the average more nearly .1 or .15 mm. These laths usually show irregular alignment. Locally the larger ones are chiefly fragmental. Although the presence of glassy inclusions in the coarser crystals frequently renders their zonal growth apparent, they are relatively homogeneous in composition. The plagioclase as a rule forms about 60 per cent of the rock. The texture is characteristically hyalopilitic. It can seldom be classed as pilotaxitic, for there is usually a cryptocrystalline semi-opaque ground, which is probably derived from a decomposed glass. It consists largely of indistinct feldspathic material filled with opaque dust which is thought to be formed of magnetite and kaolin.

With greater depth, the texture in some specimens becomes slightly coarser and the smaller laths average approximately .2 mm. in length. As a rule in this phase, however, the feldspar shows no marked increase in size, but instead of exhibiting a clean cut outline it is irregular and ill-defined. The fine groundmass thus formed may appear blotchy in thin section on account of localized alteration, which does not otherwise affect the texture.

The needles of hornblende range up to 5 mm. in length and locally form about five per cent of the rock. This mineral is variable in its optical properties. Some appears definitely to be of a basaltic variety, with a dark brown pleochroism, a very low extinction and a high index of refraction. In other specimens, the pleochroism of crystals of similar magnitude varies to a decided green, while the extinction angle ranges as high as 14°. Since similar variations have been artificially induced in hornblende¹¹ it is possible that physical conditions may have locally affected the properties of the mineral. Commonly the dark brown crystals are partially replaced by a marginal rim of magnetite or of brownish opaque iron oxides. In many sections only pseudomorphs survive. The elongate concentrations of magnetite, thus formed, may be observed in various stages of disintegration. Even the remnants are lacking in the central altered zone where the individual crystals of plagioclase are most poorly defined.

---

¹¹Virgil E. Barnes, "Changes in Hornblende at about 800° C.", Am. Mineralogist, vol. 15, pp. 393-417, 1930.

---

Pale greenish augite appears invariably to be present as a minor constituent. It was observed as irregular grains less than .5 mm. in diameter. A few thin sections show scattered, well-shaped crystals of hypersthene with a maximum length of about .8 mm. In contrast to the amphibole, these pyroxenes are unaltered. Aside from the hornblende, no regularity was detected in the distribution of these mafics, which form the only accessory minerals observed.

In spite of the small variations, an analysis of a remarkably fresh specimen from the uppermost hornblendic zone coincides almost exactly with that of a hornblende-free rock from the base of the columns about 900 feet below (table VII). A comparison of these analyses with those of the basic andesite, however, indicates that the lava became progressively more salic (table VI).

TABLE VII