The Geology and Petrography of Crater Lake National Park, 1902
DESTRUCTION OF MOUNT MAZAMA AND FORMATION OF THE CALDERA.
After the eruption of the pumice came the revolution which removed the large core and the upper 6,000 feet of Mount Mazama and gave rise to the caldera. There are only two ways in which this change could have been effected—either by an explosion which blew away the top of the mountain or by a subsidence which engulfed it.
The occurrence of vast quantities of pumice for a distance of 20 miles in all directions about the base of Mount Mazama is evidence of a most tremendous explosive eruption at that point, an eruption the equal of which has not yet been found anywhere else in the Cascade Range. Fine material was blown out in vast quantities at the same time and by drainage gathered into the surrounding valleys which it fills to an extent unknown, as far as I have observed, on the slopes of any of the other great volcanoes of the range.a This impressive evidence shows conclusively that a late, if not the final, eruption of Mount Mazama was explosive, and of such magnitude as to suggest that the removal of the mountain and the origin of the caldera may be counted among its effects. This suggestion, however, is not supported by the evidence resulting from a study of the ejected material and its relation to the lava flows of the rim. The fine material filling the valleys, and the pumice throughout its great area, are hornblendic in character and were erupted in connection with the dacites of the rim. Their eruption, therefore, was of the usual type, and not of the kind which removes mountains. As far as may be judged from the size of the pumice deposits now exposed in the rim, the greatest eruption of that sort of material from Mount Mazama occurred before the extrusion of the dacite of Llao Rock, and is evidence that the greatest explosion occurred long before the destruction of Mount Mazama.
aAs far as my own observation goes the above remarks apply to Lassen Peak, Mount Shasta, Mount Pitt, Mount Thielsen, Diamond Peak, and Mount Hood.
There is another matter of importance bearing directly on the explosive theory of the caldera which renders that theory wholly untenable and fully corroborates the conclusion derived from a study of the character and distribution of the pumice. The lava exposed on the inner slope of the rim is chiefly andesite, and its relation is such as to indicate that solid sheets of andesitic lava formed by far the larger part of Mount Mazama. If the caldera resulted from an explosion this mass of andesitic flows would have been broken to fragments and blown out, so as to fall around the caldera and form a rim of fragmental material. From the size of the lake and the remaining portion of Mount Mazama it is possible to compute approximately what the size of the rim formed in this way would be. But before we can do this it is necessary to consider the size and shape of the caldera, especially of that part which lies beneath the lake.
To determine the configuration of the bottom of Crater Lake a large number (168) of soundings were made under the direction of Major Dutton, who summarizes the results as follows:a
“The inferred configuration of the bottom may be conceived of as a nearly plane surface for the most part, upon which stand three abruptly rising prominences. The largest of these rises above the surface of the water and discloses itself as a large cinder cone. This one [Wizard Island] stands near the western margin of the lake. * * * The other two prominences are disclosed only by the plummet, for their tops are submerged, one at a depth of about 450 feet, the other at a depth of about 825 feet. The depth of the floor upon which these prominences stand varies from 1,600 to 2,000 feet. At the deepest cast the wire gave a reading of 1,996. To this should be added a small but unknown correction for the stretching of the wire, which will make the true depth of this cast fully 2,000 feet. So far as known to me this is the deepest fresh water in the United States.”
aEighth Ann. Rept. U. S. Geol. Survey, Part I, 1869, p. 157.
On the map (Pl. VI) the principal soundings are noted. From this data, together with information from Mr. W. G. Steel, who was present when the soundings were made, the presence of two sublacustrine cones is inferred, and it is clear that a large mass of lava spread from the Wizard Island vent over the lake floor. The great depression toward the eastern margin of the lake may not have been filled up any after the caldera was formed, but it is evident that the depth of the western portion has been greatly reduced by the material erupted from the three small vents upon its floor. It appears well within the bounds of reason to assume that 1,500 feet is not greater than the average depth of the original caldera below the present level of the lake.
The area of the caldera, as marked out by the crest of the rim, is over 27 square miles, and its original volume, making some allowance for the subsequent refilling from the craters on its floor, is about 12 cubic miles. If to this we add 5 cubic miles for the part of the mountain above the caldera—and this is a conservative estimate—we get 17 cubic miles of material, for whose disappearance we have to account. If this material were blown out by a great explosion and fell equally distributed upon the outer slope of the rim within 3 miles of the crest it would make a layer over 1,000 feet in thickness. This mass would be so conspicuous and composed of such fragmental material that its presence could not be a matter of doubt. There can be no question concerning its complete absence, for the surface of the outer slope of the rim exposes everywhere either glaciated rock, glacial moraine, or pumice, all of which are features which belonged to Mount Mazama before its destruction, and no trace of a fragmental rim, such as is referred to above, was found anywhere.
The phenomena on the outer slope of the rim lend no support to the view that Mount Mazama was blown away and the caldera produced by a great volcanic explosion. In fact, they completely negative such a view, and we are practically driven to the opinion that Mount Mazama has been engulfed. Major Dutton, who studied the rim of Crater Lake with a training gained among the active volcanoes of the Hawaiian Islands, recognized the wide distribution of the pumice, but the absence of a well-defined fragmental rim kept him from attributing the origin of the caldera to an explosion. On the other hand, he fully appreciated the difficulty of proving that it originated in a subsidence.a He says:
“In the Hawaiian calderas the evidences of sinkage are conspicuous. They are not confined to the deeper floors of the pits, but are also seen in the partial subsidence of great blocks or slices of the walls immediately enclosing them, and in irregular sunken spots in their vicinity, also in the marks of powerful shearing or faulting action in the walls themselves. They appear to be correlated to the remarkably quiet habits of the Hawaiian volcanoes, to their habitual modes of eruption, and to the special structure of the volcanic piles, which do not rise in steep conical peaks, but are very broad and flat. At Crater Lake, neither in the walls themselves, nor in the immediate neighborhood back of the crest line, have any traces of sinkage been observed as yet. Nothing can at present be pointed out which suggests the Hawaiian mode of origin, beyond the fact that a vast crater is before us. The general structure and habits of the Cascade volcanoes are indicative of a more vigorous style of volcanic action than the Hawaiian.”
aScience, Vol. VII, 1886, p. 181.
Let us examine briefly some of the possible conditions of engulfment. The width of the caldera across the crest of the rim is about half a mile greater than at the level of the lake, and Mount Mazama could not have subsided on the present inner slope of the rim as one solid mass without profoundly splitting the rim at one or more points. No such cross fractures of the rim have been observed.a The mountain, therefore, appears to have collapsed, one part after another in succession, probably from the center outward as the support was removed. The present inner slope of the rim may not in all cases or even generally be the one formed at the time of the collapse. In some cases, however, the inner slope was formed at that time. Of this we have evidence in the behavior of the flow at Rugged Crest. It was one of the final flows from the slope of Mount Mazama. Before the central portion of the flow, where thickest, had congealed within the solid crust, Mount Mazama sank away and the yet viscous lava of the middle portion of the stream flowed down over the inner slope of the andesitic rim into the caldera. The liquid interior of the flow having withdrawn, the crust caved in and formed Rugged Crest, with its peculiar chaotic valley of tumbled fragments, columns, and bluffs. Other explanations of the peculiar reversed flow of Rugged Crest have been sought but without avail. The facts are so simple and so direct that they appear to preclude any other hypothesis.
aThe dikes are in fractures across the rim, but they were aim formed before the great catastrophe occurred.
It would be apparent from the facts also that the collapse of the mountain was at least moderately sudden, for it is not at all probable that the Rugged Crest flow was long exposed before reaching the present level of the lake and the caldera beyond.
We may be aided in understanding the origin of the caldera by picturing the conditions that must have existed during the eruption of the Rugged Crest dacite from the upper slope of Mount Mazama. At that time a column of molten material rose in the interior of the mountain until it overflowed at the summit, or burst open the sides and escaped through the fissure. The rent of the mountain side was caused in such cases by the pressure of the column of molten material it inclosed. The molten lavas being heavy the pressure of the column within the mountain was very great, and increased rapidly with the height of the volcano. During the final activity of Mount Mazama there must have been within it a column of lava rising to a height of over 8,000 feet above the base of the Cascade Range. It is possible that this great pressure, aided perhaps by some other forces, made an opening formed low down upon the mountain slope, allowing the lava to escape. The subsidence of the lava within the mountain left it unsupported and caused its collapse. Phenomena of this sort are well known in connection with the Hawaiian volcanoes. In 1840, according to Prof. J. D. Dana, there was an eruption from the slopes of Kilauea, 27 miles from and over 3,000 feet below the level of its summit. At Kilauea the summit of the lava column is well exposed in a lava lake. In connection with the eruption of 1840 the lava of the lake subsided to a depth of 385 feet, and the irregular walls surrounding it were left without support and broke off and fell into the molten material below. During the intervals between the eruptions of Kilauea the molten column rises toward the surface only to be lowered by subsequent eruptions. The subsidences, however, are not always accompanied by an outflow of lava upon the surface. At other times it may gush forth as a great fountain hundreds of feet or more in height as if due directly to hydrostatic pressure.
That Mount Mazama disappeared and the caldera originated through subsidence seems evident, but the corresponding effusion upon the surface, if such ever occurred, has not yet been found. It is hardly conceivable that 17 cubic miles of material, much of it solid lava, could collapse, be again melted, and sink away into the earth without a correlative effusion at some other point.
The bottom of the caldera at its deepest portion is at an elevation of 4,200 feet above sea level, and it is not to be expected that the point of escape would occur at any higher level. Klamath Marsh, which lies at the eastern base of the Cascade Range, is over 200 feet higher than the bottom of the caldera. This would indicate that the effused mass should be sought on the western slope of the range, where the 4,200-foot contour occurs, along Rogue River, at a distance of less than 12 miles from the rim of the lake. The correlative lavas might perhaps be expected to be dacites closely related to the final flow of Mount Mazama, but on Rogue River the lavas are generally basalt, and there is no suggestion of the escape of such an enormous mass of lava as recently as the time of the great collapse. Whether or not we are able to discover the corresponding effusion, there seems no reasonable doubt that Mount Mazama was once a reality and that it was wrecked by engulfment.