lently contorted, the beds are often crushed and confusedly shattered or jointed, while at the same time the rocks themselves may become metamorphosed, and eventually pass into the condition of schists. Rectangular outlines are thus gradually replaced by the jagged, rough, and abrupt configuration which is so characteristic of slaty and schistose or foliated rocks.

Amongst the crystalline schists rectangular outlines are not common. Now and again, however, when different kinds of schists rapidly alternate in successive sheets or beds, some will almost certainly weather more rapidly than others. The outcrops of the less yielding rocks will thus tend to project; but as jointing is usually irregular and confused, such outcrops seldom show rectangular outlines. Exceptionally, well-marked escarpments may be met with, but the general high dip and contorted character of the rocks forbid such formations. When steep wall-like outcrops of schists occur, they have very often been determined by the presence of normal faults or of thrust-planes. In short, while the foliation and pseudo-bedding of schistose rocks now and again give rise to surface-features which are more characteristic of truly bedded strata, yet such features are apt to be strongly modified by the vagaries of the jointing.

In amorphous crystalline masses, which show neither bedding nor foliation, the character of the joints usually varies with the nature of the rock.

In

granite, for example, there are usually three sets of joints, one of which traverses the rock in an approximately horizontal direction, or may have a dip now in one direction, now in another. The vertical joints often cut each other at right angles, but not infrequently they meet at more or less acute angles. In addition to these main joints, however, there are often others. Sometimes the joints are wide apart, and they then enclose large rectangular or rhomboidal blocks. At other times they are set so closely together that the rock when exposed breaks up into a mass of angular débris. As the character of the jointing varies in this way within narrow limits, the rock tends to assume broken interrupted contours. On the other hand, when the disposition of the jointplanes is more regular and better defined, the hori zontal joints maintaining their direction for some distance, granite not infrequently breaks up as if it were a bedded igneous rock. A mountain-wall so constructed rises in a series of gigantic steps, like tiers of cyclopean masonry, interrupted by entering and re-entering angles. Where the "horizontal" joints are much inclined a corresponding change in the direction of the main rock-ridges and reefs may be observed. Not infrequently, however, the horizontal jointing is obscure and ill-defined or even wanting, and the chief contours of the surface are then determined by the vertical joints alone. Under such conditions the mountain-slope shows irregular vertical or steeply inclined walls, ridges, and but

tresses, which often run into each other as they are followed upwards, and may eventually taper off to a point. (See Plate I.)

The influence of joints, however, is apt to be greatly obscured by the manner in which rocks themselves disintegrate and crumble down. The sharply angular rock-faces defined by joints are slowly or more rapidly eaten into by epigene action, and the rock exfoliates or crumbles down irregularly according to its character. Indeed, this rotting action has often proceeded very far before the joint-faces are laid bare. When a mass of rock, losing its support, falls away, the new surface exposed has already become to a larger or smaller extent disintegrated and decomposed, so that frost and rain are enabled rapidly to reduce and modify it. Hence the sharp irregular outlines which joints naturally tend to produce are, in the case of such rocks as granite, generally rounded off. Basalt-rocks in like manner often weather readily and become decomposed and disintegrated along planes of jointing, and thus give rise to a somewhat rounded and lumpy configuration. But there is often much diversity of surface displayed by one and the same rock-mass, the basalt in some places weathering rapidly into rounded forms, while in other places, especially where the rock is fine-grained and compact, the sharp angles of the jointing are better preserved. (See Plate II.)

The usually finer-grained rhyolites, trachytes, andesites, and phonolites are not as a rule so readily dis