The lesser arch betrays no dikes nor sheets. The Vermilion Cliff sandstone covers it to the top. The greater is crowned by a few grand dikes which govern its topography. From the center a long dike runs to the south, a short one to the north, two to the east, and one to the west. The course of each is a mountain spur, and between them are amphitheaters and gorges. Clinging to the dikes are bodies of altered sandstone, but the great sandstone masses of the summit were unaltered and from them have been excavated the gorges. Along the dike-filled fissures there has been some faulting, but there is no reason to believe that the displacement is great in amount. Toward the flanks of the mountain there are a few sheets, the outermost of which is far within the line of maximum flexure. Their escarpments instead of facing upward like the revetting sheets of Mount Ellsworth, face downward; their buried and unknown edges are the edges toward the mountain. Their thinning toward the periphery of the arch is conspicuous to the eye in many instances, as is also the thinning of the dikes. Another peculiarity of dike form, one which has since been noted in a number of localities, was first detected in Mount Holmes. It consists in a definite upper limit. The dike so marked is often as even upon its upper surface as an artificial stone wall. The upper surface may be level or may incline toward one end of the dike, but in either case it is sure to be found parallel to the bedding of the strata which inclose the dike. This fact led to the suspicion, afterward confirmed by more direct evidence, that the flat top of the dike was molded by an unbroken stratum of rock bridging across the fissure which the lava filled (Figure 20). The converse phenomenon can be observed in the ridge which joins Mounts Ellsworth and Holmes. A great dike there forms the crest of the ridge for half a mile, its base being buried in sandstone; but at the end of the ridge the strata are seen to be continuous beneath it (Figure 21). [Illustration: FIG. 17.—Stereogram of the Holmes Arches; an ideal restoration of the form of the overarching strata. ] [Illustration: FIG. 18.—Ideal cross-section of the Laccolites of Mount Holmes. ] [Illustration: FIG. 19.—A flat-topped dike. ] [Illustration: FIG. 20.—Ideal Cross-section of Flat-topped Dikes; _a_, before denudation; _b_, after denudation. ] [Illustration: FIG. 21.—Ideal Cross-section of a Flat-bottomed Dike. ] [Illustration: FIG. 22.—Diagram to illustrate a hypothetical explanation of Flat-edged Dikes. ]