The observing traveler who is privileged to make the journey by steamer, threading its course in and out among the many islands and skerries of the Norwegian coast, will hardly fail to be struck by the remarkable profiles of most of the lower islands (Fig. 245). These profiles are generally convexly scalloped with a noteworthy regularity, and not in one unit only, but in at least two with one a multiple of the other (Fig. 246). As the steamer passes near to the islands, it is discovered that the smaller recognizable units in the island profiles are separated by widely gaping joints which do not, however, belong to the unit series, but to a larger composite group (Fig. 246 _b_). Frostwork, which depends for its action upon open spaces within the rocks, has here been the cause of the excessive weathering above the more widely gaping joints. ┌──────────────────────────────────────────────────────────────────────┐ │ PLATE 10. │ │ │ │ [Illustration: _A._ View in Spitzbergen to illustrate the │ │ disintegration of rock under the control of joints. │ │ (_Photograph by O. Haldin._)] │ │ │ │ [Illustration: _B._ Composite pattern of the joint structures within │ │ recent alluvial deposits. │ │ (_Photograph by Ellsworth Huntington._)] │ └──────────────────────────────────────────────────────────────────────┘ [Illustration: FIG. 245.—Island groups of the Lofoten archipelago off the northwest coast of Norway, which reveal repeating patterns of the relief in two orders of magnitude (after a photograph by Knudsen).] [Illustration: FIG. 246.—Diagrams to illustrate the composite profiles of the islands on the Norwegian coast. _a_, distant view; _b_, near view, showing the individual joints and the more widely gaping fractures beneath each sag in the profile.] High northern latitudes are thus especially favorable for revealing all the details in the architectural pattern of the lithosphere shell, and we need not be surprised that when the modern maps of the Norwegian coast are examined, still larger repeating patterns than any that may be seen in the field are to be made out. The Norwegian coast was long ago shown to be a complexly faulted region, and these larger divisions of the relief pattern, instead of being explained as a consequence solely of selective weathering, must be regarded as due largely to fault displacements of the type represented in our model (plate 4 C). Yet whether due to displacements or to the more numerous joints, all belong to the same composite system of fractures expressed in the relief. READING REFERENCES FOR CHAPTER XVII