Indeed, the reader must perceive that it is only because the actual facts of observation are thus arranged that the existence of the faults is inferred. Most of the faults are of moderate displacement; but just north of Meriden there is one whose movement amounted to two thousand feet; it cuts off the northern end of the main lava-sheet in Lamentation and the southern end of the same in the Hanging Hills group of lava-ridges. In following along the line between these two dislocated portions of the sheet, every ridge formed by the more resistant sandstones or conglomerates is cut off in a most systematic manner, precisely according to the pattern shown in the beveled surface of the model. The railroad crosses this great fault about a mile above Meriden, but the traveler will see nothing there to indicate the dislocation; its constructional effects have all been worn out.

But the region is not now a plain. It is a rolling lowland with occasional ridges formed on the resistant edges of the lava-sheets. Fig. 12. The cause of this is found in a moderate uplift of the whole country since it was reduced to a peneplain, introducing the second chapter in the history of its erosion. After this uplift a new cycle of erosive work was undertaken, and we now find ourselves at a moderate advance in this division of the valley's history. The softer beds have wasted away into lowlands, the harder ones still stand up as ridges. In the adjoining crystalline areas on the east and west, where most of the rocks are hard, the erosion of this cycle has made comparatively little progress; there the valleys are narrow and the interstream spaces are rolling uplands. In the Triassic belt, where most of the rocks are soft, the erosion of the same cycle has made much greater progress and reduced the area nearly to a second peneplain, except where the edges of the hard lava-sheets still hold up their crest lines to give some indication of the elevation that the whole surface once had. Here the valleys are broad and the interstream highlands are reduced to narrow ridges. This stage is indicated for our ten-mile-square area in Fig. 12, produced by removing from the previous form of the model certain little slips by which