Each point of sandstone lies against the foot of a mountain spur, as though it had been built for a _retaining wall_ to resist the out-thrust of the spur. Borrowing a name from the analogy, I shall call these elementary ridges _revet-crags_, and speak of the spurs which bear them as being _revetted_. The accompanying sketch is designed to illustrate the structure, but is not drawn from nature. In the view of Mount Hillers (Figure 27) the revetments may be seen, and in the bird’s-eye view of the Henry Mountains (Plate V), as well as in the Frontispiece, the revet-crags of Mount Ellsworth also are portrayed. The diagram of the north spur of Mount Ellsworth (Figure 14) shows the revet-crag of that spur at _V_. [Illustration: FIG. 15.—Revet-Crags. ] The revet-crags of Vermilion sandstone follow, in a general way, the line of maximum dip about the base of Mount Ellsworth; but a few of them rise higher, and one—that which joins the northwest spur—climbs until it is but little lower than the summit of the mountain. Outside the circle of Vermilion Cliff sandstone lies the Gray Cliff sandstone, and in a few places it takes the form of a revetment. Inside the same circle there are many revetments, constituted by trachyte sheets bedded in Shinarump shales (Figure 14). Conforming perfectly with the strata, the sheets yield by erosion forms which are identical with those afforded by hard sedimentary beds, and to the distant eye the impression of the arching structure of Mount Ellsworth is conveyed less by what can be seen of the strata than by ascending revetments of trachyte sheets, which simulate and interpret the strata. [Illustration: FIG. 16.—Mount Holmes, from the north. ] The laccolite of Mount Ellsworth is not exposed to view, but I am nevertheless confident of its existence—that the visible arching strata envelop it, that the visible forest of dikes join it, and that the visible faulted blocks of the upper mountain achieved their displacement while floated by the still liquid lava. The proof, however, is not in the mountain itself, but depends on the association of the phenomena of curvature and dike and sheet with laccolites, in other mountains of the same group. In the sequel these will be described, but it chances that the mountain next to be considered is even less developed by erosion than Mount Ellsworth. MOUNT HOLMES.