small depressions, which are valley systems in miniature, in which the tiny valleys converge into a main trunk leading into the nearest chasm. There are also tiny caves and hollows, that are sometimes mistaken for borings made by pholas. In the chasms the vegetation is most luxuriant, and the dark green fronds of harts-tongue, the delicate Lady-fern, and the graceful _Asplenium nigrum_, grow with a rare luxuriance. In these pavements every feature of limestone scenery is represented on a minute scale. There are the valley systems on the surface, determined by the direction of the drainage; the long chasms represent the open valleys and ravines, and the caves and hollows, for the most part, run in the line of the joints. The carbonic acid has left precisely the same kind of proof of its work within the caves as we find above-ground; and it would necessarily follow, that to it, as well as to the mechanical power of the waters flowing through them, their formation and enlargement must be due, as Professor Phillips has pointed out in his “Rivers, Mountains, and Sea Coast of Yorkshire,” pp. 30-1. From the preceding pages it will be seen that caves in calcareous rocks are merely passages hollowed out by water, which has sought out the lines of weakness, or the joints formed by the shrinkage of the strata during their consolidation. The work of the carbonic acid is proved, not merely by the acid-worn surfaces of the interior of the caves, but also by the large quantity of carbonate of lime which is carried away by the water in solution. That, on the other hand, of the mechanical friction of the stones and sand against the sides and bottom of the water-courses, is sufficiently demonstrated by their grooved, scratched, and polished surfaces, and by the sand, silt, and gravel carried along by the currents. The generally received hypothesis, that they have been the result of a subterranean convulsion, is disproved by the floor and roof being formed, in very nearly every case, of solid rock; for it would be unreasonable to hold that any subterranean force could act from below, in such a manner as to hollow out the complicated and branching passages, at different levels, without affecting the whole mass of the rock. Nor is there cause for holding the view put forth by M. Desnoyers[32] or M. Dupont,[33] that they are the result of the passage of hydrothermal waters. The causes at present at work, operating through long periods of time, offer a reasonable explanation of their existence in every limestone district; and those which are no longer watercourses can generally be proved to have been formerly traversed by running water, by the silt, sand, and rounded pebbles which they contain. In their case, either the drainage of the district has been changed by the upheaval or depression of the rock, or the streams have searched out for themselves a passage at a lower level.