Toward the close of the period of andesitic eruptions, flows of dacite escaped from fissures far down the south and east flanks of the volcano, and explosions of dacite pumice alternated with flows of andesite from the summit vents.
Perhaps from the beginning, Mount Mazama supported many glaciers. Even the oldest visible lavas are underlain by glacial moraines. In many places, the caldera cliffs reveal layers of bouldery till and fluvioglacial sand interbedded with volcanic rocks. The constructional forces building the volcano struggled incessantly with the erosive force of ice. Clearly, the glaciers advanced and retreated many times. Their greatest advance came after the dacitic eruptions mentioned above. At that time, many tongues of ice were more than 10 miles long, and one extended 17 miles from the summit. In some of the canyons, the thickness of the glaciers exceeded 1000 feet. Save for a few projecting arêtes, the whole of Mount Mazama was mantled by an uninterrupted sheet of ice.
After the period of maximum glaciation, when the ice had retreated from the divides and was confined to the canyon bottoms on the upper slopes of the volcano, a semicircular arc of vents opened on the north flank, approximately 5000 feet below the summit. From this Northern Arc of Vents, which determines the position of the north wall of Crater Lake, viscous flows of andesite and dacite were erupted. About the same time, a cluster of acid andesite and dacite domes rose near the east base of Mount Mazama, and many basaltic cinder cones were active on the lower slopes. Explosions of dacite pumice, partly in the form of glowing avalanches (nuées ardentes), also took place from the summit region.
A long period of quiescence ensued. The glaciers retreated until only three small tongues extended beyond what is now the rim of Crater Lake. Even the longest stretched less than a mile and a half beyond the caldera rim. The slopes of the volcano were almost barren of vegetation.
The climactic eruptions then began. At first they were mild, but soon they increased in violence. During the initial stages, fine dacite pumice was blown high above the summit vents, to be drifted eastward by the wind. As the explosions became more violent and the pumice lumps increased in size, the wind veered toward the northeast. No less than 5000 square miles were buried beneath the ejecta to a depth of more than 6 inches. The finer dust spread over a vastly larger area. So much pressure was thus released that the gases in the feeding magma escaped from solution with unusual rapidity. The pumice was no longer projected high above the vents, but escaped in prodigious amount, boiling over the crater rims and rushing down the sides of the volcano at a tremendous rate, flowing after the manner of glowing avalanches. Most of this pumice was confined to the canyons, down which it raced for distances up to 35 miles. Where no canyons existed, the pumice flows spread as incandescent sheets. Those that swept down the east and northeast sides of the volcano deployed onto the flats bordering the Klamath Marsh. Such was their power that they traveled 25 miles from their source, though half their journey lay across a plateau. Twenty miles from the vents, the flows include bombs of pumice 14 feet across. Just before the glowing avalanches ceased, the ejecta changed from dacite pumice to crystal-rich basic scoria. When the eruptions had come to an end, the glacial canyons were transformed into broad plains dotted with countless fumaroles. Each canyon had become a “Valley of Ten Thousand Smokes.” The abundance of charred logs within the deposits offers vivid testimony to the destruction of the forests.
When the culminating eruptions were over, the summit of Mount Mazama had disappeared. In its place, there was a caldera between 5 and 6 miles wide and 4000 feet deep. How was this formed? Certainly not by the explosive decapitation of the volcano. Ofthe 17 cubic miles of solid rock that vanished, only about a tenth can be found among the ejecta. The remainder of the ejecta came from the magma chamber. The volume of the pumice fall which preceded the pumice flows amounts to approximately 3.5 cubic miles. Only 4 per cent of this consists of old rock fragments; 10 to 15 per cent consists of crystals, and the rest is made up of pumiceous glass. The volume of the pumice flows is approximately 8 cubic miles.
Of this amount, only 15 to 20 per cent consists of old lava fragments. The remainder represents new magma in the form of crystals and glass. Weak, dying explosions deposited approximately a quarter of a cubic mile of fine ejecta, chiefly crystals and minute chips of rock. Accordingly, 11.75 cubic miles of ejecta were laid down during these short-lived eruptions. In part, it was the rapid evacuation of this material that withdrew support from beneath the summit of the volcano and thus led to profound engulfment. The collapse was probably as cataclysmic as that which produced the caldera of Krakatau in 1883.
