Where the beds are level, with occasional partings of a slightly different composition, the country will develop terraces, like those of the Burren in northern Clare. Where they are folded, as in the Juras, scarps and dip-slopes follow one another picturesquely, the weathered edge of the bed, the true escarpment, being sometimes at an angle as steep as that of the dip. Hence a false effect of sharp peaks is produced, when these "edges" are seen end on at a distance.
The terrace-structure may be seen in miniature forms upon a rocky shore, where the blocks loosened from the escarpments of the successive beds are carried away by the waves. Frost-action is powerful in larger instances, and sends down huge blocks upon the lower terraces. A combination of shale bands and massive limestones, especially with a dip outward from the highland, leads to destructive landslips, since the sloping surface of shale is lubricated by water that passes through the limestone (see Fig. 9). Outward slips of the coast are thus common in Antrim, and have been extensive near Axmouth, two regions where chalk rests upon Liassic clays.
Broken ground, then, occurs freely under limestone scarps, and the falling blocks often prevent the growth of trees. The freshness of the rock-face above and of the talus below calls attention to spots where denudation is most active. Differences in the constitution of the beds are indicated by differences of the slope formed by denudation on the rocky walls. The huge cañons of Arizona afford effective illustrations.
These cañons owe much of their character to the presence of vertically jointed limestone. The small rainfall of the region has allowed the rivers to deepen their channels ahead of the wearing back of the walls. Yet even where valleys are widened by rain and other atmospheric agents, those formed in limestone will maintain the character of ravines. In the valley-sides of Derbyshire, or of the Franconian plateau, or of the Arve near Sallanches, where the crags rise a mile or more above the stream, we see how cañon-cutting is assisted by the joints in limestone. The ravine of the Dourbie, east of Millau in Aveyron, in the romantic region of the Causses, is a winding gorge two thousand feet in depth (Fig. 2). That of the Tarn, a little to the north, has only recently been penetrated by a road, cut out for the most part in a vertical rock-wall.
When we observe, especially from the stream itself, the details of these sheer valley-sides excavated in limestone, we again and again detect evidences of solution. High above the present water-level, the rocks are rounded, and are often undercut, so that they overhang (Fig. 3). In Millersdale in Derbyshire, above grass-grown taluses, the surface is still smooth to the hand, and we can picture the water swirling against it, and washing it away, as it does now in the bottom of the grim ravines of Carniola. It has been suggested, indeed, that some limestone cañons represent underground waterways, the roofs of which have fallen in. This may be true of the fine gorge of Cheddar, and in some cases is proved by the existence of rock-arches bridging across the hollow of the stream.
The characters of an unmitigated limestone region are best seen when we travel east of the Adriatic. Here what have been styled the karst landscapes become prominent, and may be followed through the Greek isles to the Levant. Something of the kind is realised in the terraced lands between the Rhône and the upper reaches of the Durance; lavender bushes form dull-green spots on almost barren hills, and the grey walls of old stone-built towns are barely distinguishable against equally grey hillsides. But towards Trieste the limestone lands are barer still. The small amount of insoluble matter yielded by the rock may accumulate in swallow-holes, which are here called "dolinas," a Slavonic word really meaning valleys. This residue appears in the dolinas as a red clayey earth, the "terra rossa" of the Italian-speaking Dalmatian coast. But on the surface of the plateaus it is washed or blown away as soon as it is extracted from the limestone. A. Grund[28] has suggested that the frequency of frost-action in more northern areas allows surfaces of limestone to be cumbered with loose blocks among which soil-patches may gather; hence we do not find karst-features on the plateaus of central Bavaria, Champagne, or the Cotteswold Hills. Something approaching to a karst appears in the wind-swept levels of southern Galway and of Clare, and exposure to strong winds has probably a good deal to do with the origin of the Causses and the Illyrian karstlands. At the same time, the amount of impurity in the limestone must strongly influence the resulting landscape. The noble woods in the limestone hollows of southern Ireland are rendered possible by the clay soils derived from the limestone, as much as by the sheltered nature of the ground.
In typical karstlands, water sinks in, and emerges again on low ground, where the surface-forms cut the level of the subterranean water-table. Streams that manage to hold their own for a time on the uplands often disappear into the clefts. Marshes may occur in hollows, but may have no outlet, except in vertical directions, upwards by evaporation and downwards through the dolinas. The dolinas correspond, as the Slavonic shepherds so aptly perceived, to the river-valleys of more normal areas. The landscape of flowing streams has to be sought for in a mysterious underworld, of which we can gain only a few glimpses. What we know is largely due to explorers of singular enterprise and resource, notably E. A. Martel and the "spelæologists" whom he has inspired.
A view over the plateau of Hercegovina shows us how deep gorges, rather than ordinary river-valleys, are prevalent where important streams run across a karstland (Fig. 4). The roads are carried, where possible, along the ravines, and the country possesses a double life, that of the broad uplands, where tanks have to be made to preserve the water, and that along the commercial highways, four or five thousand feet below. Even beside the rivers there is a sense of desolation in the barren whiteness of the rocks. The sunlight strikes on the wall of some theatre of the limestone, carved out in old times by a side-swirl of the stream, and the hollow glares like a white furnace in the hills. The river in summer shrinks among broad stony reaches, to which thin-flanked sheep are driven for a scanty pasture. Its clear green water gives no promise of alluvium for its banks. Limestone, even in temperate Europe, may create the features of a desert land.
The most extraordinary rock-scenery in Europe is due to limestone in the dolomitic state. It is not clear if the crags and pinnacles of Tyrol are caused by the change from calcium carbonate into dolomite, whereby a granular mass has arisen, weathering freely along its vertical joints. It may well be that these compact limestones have developed an exceptionally jointed structure under earth-stresses, and that faulting has intensified their tendency to break up into fort-like blocks. Stratified masses of more normal Rhætic limestones often provide a terraced structure near the mountain-crests; but in thousands of feet of underlying dolomite vertical clefts prevail entirely over planes of bedding. If, as is extremely probable, these dolomite-rocks arose from the composite masses that we style coral-reefs, stratification was none the less a marked feature as their limestone grew in thickness. This structure is still plainly visible; but the joints have been widened, and the mass is cut up into stupendous pinnacles and dominating towers. The Drei Zinnen near Landro, the deeply notched wall of the Langkofel and the Plattkofel, rising four thousand feet above a grassy upland of normal Lower Triassic strata, and the overhanging crests of the Sett Sass above Buchenstein, are types of a country where dolomite is pre-eminent, and where the zone of steep rock-weathering is marked by the most fantastic forms.
ON MARBLES
Any limestone the markings or colour of which render it suitable for ornamental purposes passes as a Marble. "Fossil marbles" are often mere grey limestones, in which the stems of crinoids, or the curved sections of shells, or the radiating patterns due to corals, please the eye with their variety on a polished surface. The Purbeck Marble that was so much used as a grey foil to the massive white columns of cathedrals throughout England is simply a freshwater limestone, of no great merit as a building stone, crowded with the shells of Paludina. The black marbles are limestones coloured by one or two per cent. of carbon, derived from the decay of organisms, and white shells may stand out in them conspicuously, in contrast with the ground. The red marbles of Plymouth and of Cork have become iron-stained, and at the same time secondary crystallisation has destroyed many of their original features. In Little Island, near Cork city, earth-movements have crushed the mass, which in consequence shows signs of solid flow. The breaking of a crystalline limestone under such stresses furnishes us with many handsome marble Breccias. The abrupt juxtaposition of angular masses of various colours, torn from beds originally distinct, renders some of these rocks almost too startling for the decoration of rooms of moderate size.
There seems no such thing in nature as amorphous carbonate of lime, and all limestones are therefore formed of crystalline particles; but the further crystallisation of this material produces a true marble, in which all traces of fossils may be lost. Heat and pressure underground probably facilitate this change, since even soft chalk is converted by igneous dykes into granular marble. But where the pressure is accompanied by the possibility of movement, the shearing action breaks down the grains, and a more delicate structure results.
We have already seen (p. 35) how dolomite may undergo striking mineral changes through advanced metamorphic action. Lime-garnets, wollastonite, diopside, and other silicates similarly develop in ordinary limestones exposed to the intrusion of an igneous magma. The extreme changes in such rocks will be described when amphibolites are dealt with.