All of these examples are truly only special cases of the one already described in which the Juniata left its original syncline for others to the south. The general case may be stated in a few words. A stream flowing along a syncline of hard beds (Carboniferous sandstones) develops side streams which breach the adjacent anticlines and open lowlands in the underlying softer beds (Devonian and Silurian). On these lowlands, the headwaters of side streams from other synclines are encountered and a contest ensues as to possession of the drainage territory. The divides are pushed away from those headwaters whose lower course leads them over the fewest hard barriers; this conquest goes on until the upper course of the initial main stream is diverted to a new and easier path than the one it chose in its youth in obedience to the first deformation of the region. Thus the Juniata now avoids the center and once deepest part of the old Broad Top lake, because in the general progress of erosion, lowlands on soft Devonian beds were opened all around the edge of the great mass of sandstones that held the lake; the original drainage across the lake, from its western slopes to its outlet just south of the Jack's mountain anticline, has now taken an easier path along the Devonian beds to the west of the old lake basin, and is seen in the Little Juniata, flowing along the outer side of Terrace mountain and rounding the northern synclinal point where Terrace mountain joins Sideling hill. It then crosses Jack's mountain at a point where the hard Medina sandstones of the mountain were still buried at the time of the choice of this channel. In the same way, the drainage of the subordinate basin, through which the main lake discharged eastward, is now not along the axis of the Juniata-Catawissa syncline, but on the softer beds along one side of it; and along the southern side because the easier escape that was provided for it lay on that side, namely, via the Tuscarora and Wiconisco synclines, as already described. The much broader change from the Anthracite to the Susquehanna was only another form of the same process. Taking a transverse view of the whole system of central folds, it is perceived that their axes descend into the Anthracite district from the east and rise westward therefrom; it is as if the whole region had received a slight transverse folding, and the transverse axis of depression thus formed defined the initial course of the first master stream. But this master stream deserted its original course on the transverse axis of depression because a lateral course across lowlands on softer beds was opened by its side streams; and in the contest on these lowlands with an external stream, the Susquehanna, the upper portion of the Anthracite was diverted from the hard rocks that had appeared on the transverse axis. The distance of diversion from the axial to the lateral course in this case was great because of