the consequent expansion and contraction of its crust, have produced wrinkles and folds upon the surface, while constant oscillations, changes of level which are even now going on, have modified its conformation, and molded its general outline through successive ages.

In thinking of the formation of the globe, we must at once free ourselves from the erroneous impression that the crust of the earth is a solid, steadfast foundation. So far from being immovable, it has been constantly heaving and falling; and if we are not impressed by its oscillations, it is because they are not so regular or so evident to our senses as the rise and fall of the sea. The disturbances of the ocean, and the periodical advance and retreat of its tides, are known to our daily experience; we have seen it tossed into great billows by storms, or placid as a lake when undisturbed. But the crust of the earth also has had its storms, to which the tempests of the sea are as nothing, — which have thrown up mountain waves twenty thousand feet high, and fixed them where they stand, perpetual memorials of the convulsions that upheaved them. Conceive an ocean wave that should roll up for twenty thousand feet, and be petrified at its greatest height: the mountains are but the gigantic waves raised on the surface of the land by the geological tempests of past times. Besides these sudden storms of the earth's surface, there have been its gradual upheavals and depressions, going on now as steadily as ever, and which may be compared to the regular action of the tides. These, also, have had their share in determining the outlines of the continents, the height of the lands, and the depth of the seas.

Leaving aside the more general phenomena, let us look now at the formation of mountains especially. I have stated in a previous article that the relative position of the stratified and unstratified rocks gives us the key to their comparative age. To explain this I must enter into some details respecting the arrangement of stratified deposits on mountain-slopes and in mountain-chains, taking merely theoretical cases, however, to illustrate phenomena which we shall meet with repeatedly in actual facts, when studying special geological formations.

We have, for instance, a central granite mountain, with a succession of stratified beds sloping against its sides, while at its

base are deposited a number of horizontal beds which have evidently never been disturbed from the position in which they were originally accumulated. The reader will at once perceive the method by which the geologist decides upon the age of such a mountain. He finds the strata upon its slopes in regular superposition, the uppermost belonging, we will suppose, to the Triassic period; at its base he finds undisturbed horizontal deposits, also in regular superposition, belonging to the Jurassic and Cretaceous periods. Therefore, he argues, this mountain must have been uplifted after the Triassic and all preceding deposits were formed, since it has broken its way through them, and forced them out of their natural position; and it must have been previous to the Jurassic and Cretaceous deposits, since they have been accumulated peacefully at its base, and have undergone no such perturbations.

The task of the geologist would be an easy one, if all the problems he has to deal with were as simple as the case I have presented here; but the most cursory glance at the intricacies of mountain structure will show us how difficult it is to trace the connection between the phenomena. We must not form an idea of ancient mountain upheavals from existing active volcanoes, although the causes which produced them were, in a somewhat modified sense, the same. Our present volcanic mountains are only chimneys, or narrow tunnels, as it were, pierced in the thickness of the earth's surface, through which the molten lava pours out, flowing over the edges and down the sides and hardening upon the slopes, so as to form conical elevations. The mountain ranges upheaved by ancient eruptions, on the contrary, are folds of the earth's surface, produced by the cooling of a comparatively thin crust upon a hot mass. The first effect of this cooling process would be to cause contractions; the next, to produce corresponding protrusions, for, wherever such a shrinking and subsidence of the crust occurred, the consequent pressure upon the melted materials beneath must displace them. and force them upward. While the crust continued so thin that these results could go on without very violent dislocations, the materials within easily finding an outlet, if displaced, or merely lifting the surface without breaking through it, the effect would be moderate elevations divided by corresponding

depressions. We have seen this kind of action, during the earlier geological epochs, in the upheaval of the low hills in the United States, leading to the formation of the coal basins.

On our return to the study of the American continent, we shall find in the Alleghany chain, occurring at a later period, between the Carboniferous and Triassic epochs, a good illustration of the same kind of phenomena, though the action of the Plutonic agents was then much more powerful, owing to the greater thickness of the crust and the consequent increase of resistance. The folds forced upward in this chain by the subsidence of the surface are higher than any preceding elevations; but they are nevertheless a succession of parallel folds divided by corresponding depressions, nor does it seem that the displacement of the materials within the crust was so violent as to fracture it extensively.

Even so late as the formation of the Jura mountains, between the Jurassic and Cretaceous periods, the character of the upheaval is the same, though there are more cracks at right angles with the general trend of the chain, and here and there the masses below have broken through. But the chain, as a whole, consists of a succession of parallel folds, forming long domes or arches, divided by longitudinal valleys. The valleys represent the subsidences of the crust; the domes are the corresponding protrusions resulting from these subsidences. The lines of gentle undulation in this chain, so striking in contrast to the rugged and abrupt character of the Alps immediately opposite, are the result of this mode of formation.

After the crust of the earth had grown so thick, as it was, for instance, in the later Tertiary periods, when the Alps were uplifted, such an eruption could take place only through the agency of an immense force, and the extent of the fracture would be in proportion to the resistance opposed. It is hardly to be doubted, from the geological evidence already collected, that the whole mountain range from Western Europe through the continent of Asia, including the Alps, the Caucasus, and the Himalayas, was raised at the same time. A convulsion that thus made a gigantic rent across two continents, giving egress to three such mountain ranges, must have been accompanied by a thousand fractures and breaks in contrary directions. Such a pressure along so

extensive a tract could not be equal everywhere; the various thicknesses of the crust, the greater or less flexibility of the deposits, the direction of the pressure, would give rise to an infinite variety in the results; accordingly, instead of the long, even arches, such as characterize the earlier upheavals of the Alleghanies and the Jura, there are violent dislocations of the surface, cracks, rents, and fissures in all directions, transverse to the general trend of the upheaval, as well as parallel with it.

Leaving aside for the moment the more baffling and intricate problems of the later mountain formations, I will first endeavor to explain the simpler phenomena of the earlier upheavals.

Suppose that the melted materials within the earth are forced up against a mass of stratified deposits, the direction of the pressure being perfectly vertical. Such a pressure, if not too violent, would simply lift the strata out of their horizontal position into an arch or dome, and if continued or repeated in immediate sequence, it would produce a number of such domes, like long billows following each other, such as we have in the Jura. But though this is the prevailing character of the range, there are many instances even here where an unequal pressure has caused a rent at right angles with the general direction of the upheaval; and one may trace the action of this unequal pressure, from the unbroken arch, where it has simply lifted the surface into a dome, to the granite crest, where the melted rock has forced its way out and crystallized between the broken beds that rest against its slopes.

In other instances, the upper beds alone may have been cracked, while the continuity of the lower ones remains unbroken. In this case we have a longitudinal valley on the top of a mountain range, lying between the two sides of the broken arch. Suppose, now, that there are also transverse cracks across such a longitudinal split, we have then a longitudinal valley with transverse valleys opening into it. There are many instances of this in the Alleghanies and in the Jura. Sometimes such transverse valleys are cut straight across, so that their openings face each other; but often the cracks have taken place at different points on the opposite sides, so that, in traveling through such a transverse valley, you turn to the right or left, as the case may be, where it enters the longitudinal valley, and follow that till

you come to another transverse valley opening into it from the opposite side, through which you make your way out, thus crossing the chain in a zigzag course. Such valleys are often much narrower at some points than at others. There are even places in the Jura where a rent in the chain begins with a mere crack,

a slit but just wide enough to admit the blade of a knife; follow it for a while, and you may find it spreading gradually into a wider chasm, and finally expanding into a valley perhaps half a mile wide, or even wider.

By means of such cracks, rivers often pass through lofty mountain-chains, and when we come to the investigation of the glacial phenomena connected with the course of the Rhone, we shall find that river following the longitudinal valley which separates the northern and southern parts of the chain of the Alps till it comes to Martigny, where it takes a sharp turn to the right through a transverse crack, flowing northward between walls fourteen thousand feet high, till it enters the Lake of Geneva, through which it passes, issuing at the other end, where it takes a southern direction. For a long time mountains were supposed to be the limitations of rivers, and old maps represent them always as flowing along the valleys without ever passing through the mountain-chains that divide them; but geology is fast correcting the errors of geography, and a map which represents merely the external features of a country, without reference to their structural relations, is no longer of any scientific value.

It is not, however, by rents in mountain-chains alone, or by depressions between them, that valleys are produced; they are often due to the unequal hardness of the beds raised, and to their greater or less liability to be worn away and disintegrated by the action of the rains. This inequality in the hardness of the rocks forming a mountain range, not only adds very much to the picturesqueness of outline, but also renders the landscape more varied through the greater or less fertility of the soil. On the hard rocks, where little soil can gather, there are only pines, or a low, dwarfed growth; but on the rocks of softer materials, more easily acted upon by the rain, a richer soil gathers, and there, in the midst of mountain scenery, may be found the most fertile growth, the richest pasturage, the brightest flowers. Where such a patch of arable soil has a southern exposure on a