The Main Pumice Fall

     Distribution and Thickness

The culminating activity of Mount Mazama began with the expulsion of finely divided dacite pumice, projected high above the summit of the volcano and then drifted afar and winnowed by the winds. From the distribution of this pumice (map, figure 16), it follows that during the eruptions the prevailing winds were from the same quarters as at present. At first the winds blew toward the east, but as the explosions increased in violence the winds veered toward the northeast. Two features are thus explained: first, the pronounced eastward bulge of the isopachytes toward Summer Lake and the absence in that direction of much coarse pumice, particularly in the lower part of the deposit; second, the prevalence of coarse pumice on top of the fine in the region northeast of Crater Lake.

Except near the crest of the Cascade Range, the pumice fall mantles the surface with amazing fidelity. On steep, rocky slopes much pumice has been washed to lower levels, and elsewhere it has drifted into depressions to a depth of 50 feet, but in general erosion has done little to modify the thickness of the sheet. The meager rains which fall east of the Cascade crest percolate readily through the pumice and there is little transport of material save on the banks of the principal streams. So faithfully does the pumice sheet reflect the underlying surface that test pits excavated at the tops and bottoms of large hills show the thickness to vary only within a few inches over areas of several square miles. It is as if the ejecta had been "ducoed" onto the preexisting slopes. Where the thickness of the mantle exceeds 2 feet, hardly a single exposure of "bedrock" may be seen for miles.

The simplified isopachytes shown on the map (figure 16) are based on measurements made along highway and railroad cuts and on a large number of test pits.³ No attempt was made to plot the outermost limit of the pumice, for where the deposit diminishes to less than 6 inches in thickness, it is generally so mixed with soil that accurate estimates are no longer possible. Nor was it feasible to plot the isopachytes within the park boundaries, for close to the source, where erosion has been more marked and where the onrush of the later pumice flows obliterated much of the pumice fall, the thickness is extremely variable. Near the northern margin of the sheet, the ejecta are buried beneath younger pumice deposits from the Newberry volcano and beneath basaltic cinders erupted from scores of cones in that vicinity. Beyond the northern limits of these younger ejecta, the pumice from Crater Lake occurs only as a light sprinkle in the soil.

A few of the outstanding features of the pumice distribution deserve to be emphasized. The margins of the deposit do not diverge from the source as might be expected, nor does the pumice thin uniformly in radial directions. On the contrary, the isopachytes run more or less parallel, and along the south and southwest edges of the deposit the, pumice thins much more rapidly than elsewhere. Generally, the front of an ash or pumice fall widens with increasing distance from the vent; that it did not in this case suggests that strong winds were confined to a narrow belt. A similar lack of spreading was observed in the pumice blown across Argentina during the 1932 eruptions of the Chilean volcano Quizapu.⁴

Even 70 miles northeast of Crater Lake the pumice sheet is still a foot thick, and it therefore goes without saying that some of the finer ejecta must have spread for hundreds of miles. Probably the finest dust, like that from the eruptions of Krakatau in 1883, encircled the globe.

Perhaps no feature of the pumice fall is more striking than the scarcity of ejecta on the west and southwest slopes of Mount Mazama, even close to the rim of Crater Lake. For example, the veneer of pumice west of the Watchman and Hillman Peak is extremely and patchy; on Munson Ridge it is hardly more a light sprinkle, whereas on Vidae Ridge, a short to the east, the thickness locally exceeds 8 feet. Along the southwest rim of the caldera, the ice may be absent altogether. Red and Desert cones, on the northwest slopes of Mazama, though 2 and 4 miles, respectively, from the rim of the era, are so lightly veneered with pumice that it requires a careful search to find even a few fragments. Several explanations suggest themselves for the paucity of pumice so close to the source. The thinly covered slopes may have been blanketed with snow at the time of the eruptions; some of the slopes were later swept clean by the onrush of pumice flows; some may have been partly stripped by subsequent erosion. But the principal reason for the scarcity of pumice fall on the west and southwest sides of the volcano was simply that the winds were blowing in the opposite direction.

An observer would indeed have been safe in viewing the eruptions from the slopes of Union Peak, not more than 5 miles from the rim of Crater Lake. Occasional lumps of pumice up to 2 inches across fell on Union Peak, but by far the majority were less than 1/2 inch across, and even these were rare. The thick accumulation of pumice on Pumice Flat, near the east base of Union Peak, cannot be a product of these final explosions from Mount Mazama, for many of the lumps measure as much as 2 feet in diameter, and their pale-gray color contrasts distinctly with the pale-bluff tints of the final pumice fall. These coarser ejecta  belong to an earlier explosive phase of the volcano, and are probably products of the eruptions which preceded the dacite flows from the Northern Arc of Vents. The same is true of the coarse pumice in the flats on the northwest side of Union Peak and in the depression northwest of Bald Crater.

Very little pumice fell west of Annie Creek, and scarcely a trace is to be found on the west side of the Klamath graben. On the opposite side, however, the pumice thickens with astonishing rapidity. Near Klamath Agency, for instance, the thickness increases from a few inches to more than 6 feet in a distance of only 3 miles.

Less than 10 per cent of the total volume of pumice fell west of the summit of the Cascade Range, and much of this was swept away by the headwaters of the North and South Umpqua rivers. Thick banks of washed pumice occur far down the North Umpqua, and though most of it was derived from erosion of the later pumice flows, much of it represents products of the pumice fall washed from the neighboring hills. Certainly all the washed pumice found along the walls of the South Umpqua must be reworked pumice fall, for none of the later pumice flows entered the drainage basin of that river.

In the low country east of the Cascades, the thickness of the pumice varies gradually and regularly over long distances. Near the crest of the range, on the other hand, the variations are both rapid and irregular, and the isopachytes shown on the map are greatly generalized. To some extent, the variability in this high country is the result of erosion, but chiefly it seems to have been caused by swirling cross winds at the time of eruption. The influence of such disturbances on the thickness of ejecta was exemplified clearly during the eruption of Quizapu, already mentioned. Most of the pumice blown out during that eruption drifted eastward across Argentina, but instead of thinning regularly away from the source, the Quizapu pumice suddenly thickened in central Argentina and then continued to thin farther east. This local thickening was the result of complex eddies in the lee of the Sierra de Cordoba. We may reasonably imagine that turbulence of the winds among the peaks of the Cascades, by diminishing the transporting capacity of the air, likewise caused irregular dumping of ejecta.

Nevertheless, if these minor variations be ignored, the thickness of the pumice fall erupted from Mount Mazama varies with such regularity as to suggest that all of it was laid down within a short time. Had the eruptions continued for years, there would probably be far more irregular variation both in thickness and in distribution. Nowhere have any signs of erosion been noted within the pumice deposits themselves, nor are there any traces of interbedded soils. The probability is, therefore, that the whole eruptive episode was short-lived and that the explosions, if not actually continuous, followed each other at short intervals.