cter for long stretches. All such inequalities and lifferences are due to the nature of the rocks and the mode of their arrangement. Bedded rocks, for ex

ample, owing to the regularity of their joints, tend to form cliffs with even faces. If the strata be horizontal, it is obvious that the cliffs must be vertical, or nearly so, since the rocks naturally yield along their approximately vertical division-planes. When a slice has been detached from the cliff, the new surface exposed is an even wall of rock. But as the beds entering into the formation of such a cliff are likely to yield unequally to weathering, the smooth wall of rock sooner or later becomes etched and furrowed. (Fig. 86.) Now and again, however, owing to the

sea

J

J

J

FIG. 86. SEA-CLIFF CUT IN HORIZONTAL STRATA.

jj, joints.

nature of the rocks, or to the rapid retreat of the cliffs, weathering has not sufficient time to effect any marked modification of the surface. When the strata, instead of being horizontal, are inclined, and the dip is inland, or away from the coast, the joint-planes necessarily have an inclination towards the sea, and the cliffs naturally slope in the same direction. (Fig. 87, p. 320.) On the other hand, should the strata

dip seaward, cliffs hewn out of them have a tendency to overhang, because the division-planes along which

FIG. 87. SEA-CLIFF CUT IN STRATA DIPPING INLAND.

jj, joints.

the rocks yield are now inclined away from the shore. (Fig. 88.) Cliffs having this structure are in a state of unstable equilibrium-the truncated beds being apt to slide forward-so that actually overhanging

FIG. 88. SEA-CLIFF CUT IN STRATA DIPPING SEAWARD.

jj, joints.

cliffs of this kind are not often met with.

Not infre

quently, indeed, when strata dip seaward at a relatively low angle they form natural breakwaters, and the waves do not succeed in cutting out a cliff.

In all cases, when the strike of the strata coincides

approximately with the trend of a coast-line-the dip being either seaward or landward-the forms assumed by cliffs are largely determined by the position of the strike-joints. The regularity of a line of cliffs is likewise greatly controlled by the position of the dipjoints, which, it will be remembered, cut the strikejoints at approximately right angles. If the former be somewhat wide apart, and not strongly pronounced or discontinuous, the sea-wall may run continuously for miles without any marked interruptions. On the other hand, should the dip-joints be in places more numerous and closely set, they will form lines of weakness, and thus allow the waves to sap and notch the cliff, so that all such cliffs tend to assume rectangular outlines, the faces of the sea-wall and the indentations that break its continuity being determined by the double set of joints. And the same holds true in the case of horizontal strata.

It goes without saying that the cliffs of a regular coast are evidence of marine erosion. The sea acts like a great horizontal saw, forming rock-shelves and terraces that increase in width as the cliffs are undermined and cut back. So effectually has the work been done in many cases that at high tide these terraces of erosion are completely covered. Frequently, however, islets, stacks, and low reefs and skerries appear-fragments of land which owe their preservation to the superior hardness of the rocks at the sea-level, or to some peculiarity of structure, such as the paucity or absence of joints. Lofty stacks are

perhaps most commonly met with in the case of horizontal or approximately horizontal strata, or of gently inclined beds, when the strike coincides with the general trend of a sea-wall. But smaller stacks, reefs, and skerries are usually most abundant when the coast-line cuts across the strike, and the truncated rocks differ much as regards durability. Such a coast-line is usually very ragged or frayed out. The cliffs are often approximately vertical, but usually show many narrow and broader indentations, while long parallel ranges of reefs, skerries, stacks, and islets diversify the surface of the terrace of erosion.

In

Of the various forms presented by the projecting bastions and towers of a line of cliffs, and by the islets and stacks of the sea-shelf, it is not necessary to say more than that these necessarily vary with the nature of the rocks and the geological structure. the case of horizontal strata they all have a tendency to assume pyramidal or conical shapes, and similar forms are usually seen in the cliffs of massive structureless accumulations like boulder-clay. Stacks built up of inclined strata are usually less regular in form. With a low dip the truncated beds are necessarily unstable, and the tendency to collapse is greater than it is in the case of a horizontal arrangement. But with a high dip the structure becomes more resisting, especially if the beds be thick and massive. When the strata are folded we not infrequently find that projecting headlands, islets, and stacks coincide with synclinal arrangements.

In short, it may be said

generally that the geological structures which best withstand the action of the eroding agents in mountainous and inland regions are just those which offer the most resistance to the assaults of waves and breakers. Finally, it must be borne in mind that the action of the sea in the reduction of a steep coastline is always more or less aided and modified by other epigene agents. Were it not for the action of springs and frost coast-cliffs would often be steeper and more abrupt than they generally are, the tendency being for cliffs of all kinds of structure to become benched backwards. Overhanging and absolutely vertical rock-walls are by no means so common as one might suppose; however steep a cliff may be, it usually has an inclination seawards. The accompany

FIG. 89.

SEA-Cliff Cut in Beds Dipping Seaward.

a a, cliff-face determined by master-joint; cliff may yield along several joints in succession, as at b-b.

ing diagram, representing strata dipping seawards, shows how a cliff may be overhanging or not according as the beds yield in a wholesale fashion along one joint-plane, or bed by bed along different joint-planes. The cliff-face a-a coincides with a master-joint. It is obvious, however, that yielding may take place ir