Distribution of Pumice and Scoria Flows

The outstanding feature of the distribution of the pumice flows is the manner in which they follow the principal canyons around Crater Lake (see map, figure 16). Clearly the explosions were not "directed" from the summit of the volcano as downward blasts, but their course was determined largely by the topography.

Fig. 16. Map showing distribution and thickness of Crater Lake pumice. (Drawn by A. W. Severy.)

On the southwest and west sides of the volcano, the flows did not begin to drop their heavy load until they had traveled about 5 or 6 miles from the source, for the slopes between the caldera rim and the headwaters of Castle, Copeland, and Bybee creeks are covered with only a thin and patchy veneer of pumice. Once the flows entered the canyons of those creeks, however, the pumice accumulated to depths of as much as 200 feet. This singular scarcity of pumice on the southwest and west flanks of the cone is only in minor part the result of erosion. The main causes were probably two: first, these bare slopes were then blanketed with snow, and, second, the flows rushed across them too swiftly to permit deposition. When their speed was checked on entering the U-shaped glacial canyons, the flows began at once to drop their detritus, and in that manner the valleys were choked and converted into the wide plains of pumice we see today.

Within the canyons, the margins of the pumice flows are usually sharp enough to be located within a few tens of yards, but where the flows were not confined by steep valley walls they spread to a feather edge, and it is difficult to see where they end against the underlying pumice fall.

The avalanches that followed the canyons of Annie and Sun creeks probably came to rest in the vicinity of Fort Klamath, having moved a distance of about 15 miles. Possibly they reached the edge of Upper Klamath Lake, which was then slightly higher than now and extended farther north. The flows that passed through Kerr Notch into the valleys of Sand and Wheeler creeks filled them to a depth of 250 feet in places. So rapidly did they move that even 13 miles from the source, after they had debouched onto the plateau at the foot of Mount Mazama, they rushed up the onset slope of Boundary Butte to a height of 200 feet.

On the east side of Mount Mazama there were no large glacial troughs; nevertheless the flows had not passed far beyond the present rim of the caldera before they were deflected into the narrow valleys of Scott and Bear creeks. These they followed down to the plateau, where they reunited and continued eastward for 10 miles or more to the edge of the Klamath Marsh. The more voluminous flows that raced down the northeast slope of the volcano along Desert Creek, combining with those which crossed the Pumice Desert and turned eastward through the depression between Timber Crater and Mount Thielsen, spread as far as Chemult, a distance of 25 miles, though the last 15 miles of their path lay across ground that was almost flat.

Part of the flow which crossed the Pumice Desert continued northward into Diamond Lake. How it crossed the lake is not: clear, but that it did so is apparent, for thick deposits of unstratified pumice may be found in the upper stretches of both Clearwater and North Umpqua rivers. Another branch of the same flow swerved westward into the headwaters of the Rogue River to augment those that swept down the northwest slope of the volcano. The latter did not begin to deposit much of their load until they reached the head of National Creek, between 8 and 9 miles from the vent.

Much of the pumice which poured down National, Bybee, Copeland, Castle, and Union creeks united to form a single flow in the valley of the Rogue, and the composite mass rushed another 20 miles almost to the site of the present village of McLeod.

Fig. 24. Size distribution of samples of pumice flow (lump pumice). Fraction less than 1/8 mm. omitted. (After B. N. Moore, Journal of Geology, vol. 42, p. 365, 1934. Numbers in parentheses refer to Moore's samples.)

As compared with the deposits left by glowing avalanches around other volcanoes, those of Mount Mazama are unusually large. Even in a straight line, the snout of the Rogue River flow is 35 miles distant from the source; the actual distance traveled must have been more than 40 miles. By contrast, the great "sand flow" in the Valley of Ten Thousand Smokes was 11.5 miles long, and few of the pumice flows and glowing avalanches of Pelée and Komagatake spread more than 4 miles from their source. The Crater Lake deposits compare, rather, with the widespread tuff sheets which encircle the caldera of Aso, Japan, or with those that form the extensive plateaus around the collapse depressions of Lakes Ranau and Toba in Sumatra and Lake Taupo in New Zealand.