Page images

rendered impermeable by the accumulation in them of loam and clay, and these may then be occupied by permanent lakes.

The influence of subterranean water is not, of course, confined to regions of soluble rocks. Whereever water circulates in the crust rocks are attacked, and their constituents become liable to chemical change. In this manner immense quantities of mineral matter are brought up from below, some of it to be thrown down at the surface, where in time it may form massive accumulations. The mechanical action of subterranean water is also recognised almost everywhere, and more particularly in places where the geological structure is weak, where rocks are in a state of unstable equilibrium. But the effect of underground water in bringing about rock-falls and landslips in such regions has already been sufficiently discussed.

Although caverns naturally occur most numerously and attain the largest size in the more readily soluble rocks, they are also met with in many other kinds. They appear, for example, not infrequently in lava. Some of the smaller of these are merely large blisters or bubbles, formed by the segregation of the absorbed water-vapour while the lava was in a semi-fluid condition. The more extensive lava-caves have a different origin. While lava is flowing it necessarily cools rapidly at the surface, and in this way becomes crusted over. If the crust thus formed be of sufficient thickness and strength, it remains steadfast, forming a kind of tunnel, out of which the still liquid lava issues. Such lava-caves are of common occurrence in Hawaii, ^Mexico, California, the Canary Islands, Iceland, etc. Some are only a few feet in height and breadth, others may be 20 to 30 feet broad, 6 to 10 feet in height, and many yards in length. In certain volcanic regions lava-caves obtained much larger dimensions, but there is reason to believe that these have been modified by subsequent erosion. One in Hawaii has a width at the entrance of 130 feet, a height of 20 feet, and a length of 260 feet. Another (the Raniaka Cave) is 1200 feet long. Water flowing in cavities under the lava-coulees of Auvergne (as in the neighbourhood of Clermont) has cut out courses in the subjacent granite, and issues at the lower ends of the lavastreams through natural arcades. And many similar examples of subterranean tunnels and caves might be cited from other regions, where the erosion has been effected chiefly by the mechanical action of water upon relatively insoluble rocks.

Mention may also be made of the great cavities which occasionally occur in faults. The spaces between the two walls of a fault or dislocation are usually filled up either with rock-dddrts, or subsequently infiltrated mineral matter, or with both. Now and again, however, the filling-up is only partial, and chambers of some size remain. These are often lined with finely crystallised minerals, and form what are known in Switzerland as "crystal-cellars."

Of caves solely due to erosion it is not necessary to

r s

say much. Shallow caves (rock-shelters) are frequently met with in river-valleys, where one can see that they owe their origin to the under-cutting action of the water. More extensive are the caves often excavated by the sea. These necessarily vary in appearance with the character of the rocks in which they are excavated. The presence of a cave indicates some weak structure—some rock or rock-arrangement which has offered less resistance to the attack of waves and breakers. Vertical dikes of basalt, for example, are often so abundantly jointed, that they are broken up and removed more readily than the rocks they traverse, although the latter may consist of "softer" material, such as sandstone. The highly jointed basalt, notwithstanding its superior hardness, is easily shattered. The mere force of the waves combined with hydraulic pressure in some joints, and the compression and expansion of air in others, suffices to rupture and burst the weak structure, and with each drop of the wave large and small fragments may sometimes be seen falling from the roof and sides of the cave. The cave thus increases in height as the sea works its way inland, until not infrequently it communicates with the surface by a "blow-hole," through which in storms not only spray but spouts of water, and even gravel and larger stones, are ejected. Similar caves are frequently formed in well jointed sandstones and in many other kinds of rock. They are very common, for instance, in Orkney and Shetland, and they are well known also in Cornwall and the West of Ireland. In time the whole roof of such caves may give way, and the latter then appear as narrow ravine-like or gorge-like inlets. This can happen only when the land-surface does not rise to any great height above the sea. When the rocks above a sea-cave are too strongly built or too thick to permit of a downfall of the roof, the cave may attain very considerable dimensions. But as all rocks are traversed by lines of weakness, a limit must be reached beyond which caves cannot be widened. By and by the rocks will cease to be self-supporting, and collapse must take place.

Caves of marine origin are seldom met with far removed from existing coast-lines. They are naturally confined to the latter, and to those lines of old sea-level known generally as " raised beaches." Their position at the base of old sea-cliffs renders them liable to early obliteration, for they tend to be obscured by, and eventually to be concealed underneath, a talus of ddbris. They are not singular, however, in that respect, for many of the most interesting and important of the limestone caverns of Western Europe have been hidden in the same way, their discovery having been the result either of mere accident or of patient scientific research.






ALL the varied topographical features of the land owe their origin either to subterranean or to superficial agents, or to both. This is true of elevations and depressions alike. It would seem possible, therefore, to classify hollows according to the mode of their formation. Not a few, however, are of complex origin, having resulted partly from hypogene and partly from epigene action. Indeed, we might group all basins roughly in two divisions, according as they owe their origin more or less directly to crustal deformation and fracture, or to the action of surface-agents. Epigene action, however, is so manifold and diverse—the agents of erosion, of transport, and accumulation act in so many different ways—that a more detailed grouping is desirable. Any classification adopted must be more or less arbitrary and in

« PreviousContinue »