We may now follow the margin of the lava northward from the vent along the caldera wall. Upward from the conduit, along the edge of the V-shaped part of the flow, several patches of coarse explosion breccia may be seen firmly adhering to the dacite within. The flow banding stands vertically, dips inward, or, in a few places, dips steeply lakeward, which, of course, it could not do if the lava had poured down a river valley in the opposite direction. Beyond the V-shaped part of the flow, the thickness of the lava diminishes to about 300 feet. Here, the banding at the bottom of the lava stands vertically and strikes parallel to the face of the cliff. In fact for a short distance the older, gently dipping andesites abut against a vertical wall of dacite. Clearly, therefore, either the dacite piled against an outward-facing cliff of andesite or some of it rose through a vertical fissure extending northward from the principal vent. Higher up, the banding of the dacite gradually flattens and dips away from the lake.
Close to the top of Skell Head, the margin of the dacite leaves the caldera rim and swings downhill for more than 1/2 mile. Throughout most of this stretch, it forms a line of cliffs, up to 200 feet high. Where they are highest, the cliffs are remarkably smooth and traversed by broad, almost horizontal flutings, and it comes as a surprise to find that these features are only in part the result of glacial erosion and are mainly original structures of the flow itself. Along this edge of the dacite, the lava is rich in spherulites and lithophysae lined with plates of tridymite.
Near the snout of the flow, there are several large caves, either at the base or a short distance above. One of these measures 35 feet across the mouth and 10 feet in height, and can be traced under the lava for a distance of 100 feet. It was not the escape of fluid lava from beneath a solid crust, nor was it explosion that produced these openings; they seem rather to have been caused by the rucking up of the viscous base of the lava and by shearing within the flow as layers moved differentially.
Along the southeast margin of the dacite, the banding, even at the very edge, is generally vertical and in certain places bulges outward. On the hummocky top of the flow, the planes retain steep dips and are arranged more or less concentrically. In other words, we are faced with the strange phenomenon of a thick lava flow, approximately a mile in length, in which the marginal banding rarely departs much from the vertical and the banding of the interior parts is also highly inclined over most of its surface. In view of the recentness of the eruption, the top of the flow can have been only slightly modified by erosion, and the steepness of the sides can have been increased only toward the base. If the original margins were restored, they would probably resemble those of the Watchman flow, and we should see that the banding in the lower parts lay horizontally or dipped inward at low angles, whereas in the higher parts the inward dip rapidly increased to verticality. Under the central part, the flow planes must, of course, conform with the slope of the floor, but toward the top they must curve so as to emerge at right angles to the surface. This peculiar structure has never been observed among any of the andesites of Mount Mazama except in the Watchman flow, but is typical of all the dacites. Presumably it results from shearing of layers of extremely viscous, almost solid lava over one another. In this sense the flow planes may be compared to the rising thrust planes so commonly seen in glaciers.
Although the structure of the Redcloud flow is fairly clear, that of the Cloudcap dacite is obscure. The topographic form suggests that Cloudcap itself may be a dome-shaped protrusion above a vent immediately beneath the summit. On this assumption the long flows of dacite forming Scott Bluffs may be considered as earlier products from the same source. The presumption is that the thick sheet of dacite on the caldera wall immediately below the southern extension of the Redcloud dacite is also a product of the Cloudcap vent (panorama, plate 23). At the base of this zoo-foot sheet, the lava merges into 20 feet of intensely welded and banded dacite tuff rich in lithic blocks of andesite, identical with the welded tuff of the Wineglass. Beneath this lie the blocky breccias, coarse lump pumice, and tuffs that form the Cottage Rock. Clearly, the growth of the Cloudcap dome was preceded by long and violent explosions.
***previous*** — ***next***
