the wrinkles on the skin of a dried apple; they would not be nearly so long or so lofty as they are. Hypothesis of Columnar Expansion.-If compres sion has not been caused by contraction, it can only be due to some kind of expansion or extension of the earth's crust, and Mr. Fisher suggests that this is to be found in the production of fissures opening upwards from the liquid substratum, and filled with matter proceeding from it. Considering the amount of water which is always given off as steam during volcanic eruptions, he believes the liquid magma to contain a large amount of water-substance (dissolved in the state of gas), and thinks that all fissures originating on the under surface of the crust would be immediately filled with this gas, under a pressure of about 10,066 tons to the square foot; the gas would consequently exert this pressure in widening the fissure, that is to say, in compressing the rocks on either side of it. "When the rent reached the surface, the vapour would rush forth, and be followed by the magma itself, now appearing as lava, and thus a volcano would be formed; but this would be an exceptional occurrence. It would be only here and there that the vapour would escape at the surface, because its doing so at one point would relieve the internal pressure for a long distance." 1 The fissures are kept open and the lateral compression maintained by the injection of the molten material, which solidifies and adds to the mass of the crust, the upper ends of such injected fissures appearing on exposure as dykes of igneous rock. Mr. Fisher also considers the consequences which would ensue from the compression of a crust resting upon a fluid substratum, and shows that for every subaerial elevation above the mean surface there must be a corresponding protuberance dipping downwards into the fluid below, and, owing to the lower parts being softened by heat, a greater thickness will be sheared downwards than upwards. Assuming that under lateral compression two-fifths of the thickened part would go up and three-fifths would go down, and that the ratio between the specific gravities of crust and substratum is the same as that between those of 1 "Physics of the Earth's Crust," pp. 186, 284. H H granite and basalt, which is nearly as that of ice to water, he shows that the thickened tract would sag still further downwards. "Hence depressions would arise on both sides of the ridge, and the ocean, which covers the general surface, would be deeper than elsewhere along two channels parallel to and at some little distance from the ridge. But should the ridge be steeper on one side than on the other, as seems inevitable, the ocean would be deeper on the steeper side." This relation between ocean depths and mountain chains seems actually to exist, as already mentioned. Lastly, he considers the consequences of the transference Fig. 150. Diagram to show the Roots of a Mountain Chain. a, Crust. b, Substratum. c, Mountain chain. P, Downward protuberance. o, o, Ocean level. m, m, Lower mean level of the earth's crust. of sediment from the mountain chains to the oceans, and shows that on hydrostatic principles the areas from which matter is removed will rise, while the areas to which sediment is transported, being loaded, will sink. Hence he concludes that the vertical movements affecting the earth's surface are of two kinds; one resulting from compression, the other from the laws of hydrostatic equilibrium simply. As the result of the latter degradation and elevation are necessarily correlatives of each other; the degradation of the tract causes more rock to rise up, to be in its turn degraded and removed.2 IN CHAPTER II. EARTH-SCULPTURE, OR THE EVOLUTION OF SURFACE- N Part I., Chapters VI., VII., VIII., we described the action of the various agencies which are continually engaged in the work of erosion and denudation; but their operations were treated rather from a detritive point of view, as resulting in the collection of materials for the formation of new rocks, than as leading to the exposure of new surfaces, or the production of new forms on previously existing surfaces. It is clear, however, that agencies which sweep away so much rock from the land must in time cause very great changes in its physical features. An elevated table-land, for instance, by the unequal disintegration of its surface, and by the erosion of valleys out of its mass, may gradually be converted into a series of hill ranges more or less isolated from one another. Such a process is fitly termed Landsculpture, as analogous to the work of a sculptor, who carves effigies out of solid blocks of stone. Many persons express surprise that the formation of such prominent features as lofty hills and profound valleys should be attributed to such comparatively feeble agencies as those above mentioned; but it must be borne in mind that we are naturally apt to underrate the amount of work done by these erosive agencies, because we see that in any period of time during which we can observe their action the results produced are very small. On the other hand, when we look at the magnitude of the results which have been produced in past time, we are liable to suppose that the agencies which have operated in former times were much more powerful and destructive than those which are now in action around us. "When, however, we come to reason on the matter, we find it very difficult to imagine what these agencies could have been if they were altogether different from existing causes;' and equally difficult to suppose that existing agencies have ever acted with much greater intensity than at present, unless we assume the general physical laws of the world to have been different from what they are now" (Jukes). We shall therefore take it for granted, in accordance with the tenets of the Lyellian philosophy, that all the geological phenomena observable among stratified rocks are due to the same causes as those now acting in some part of the world, or to some modification and combination of these causes, such as we may reasonably suppose to have occurred in the course of the earth's history. To such seemingly insignificant and slowly acting causes, operating continually through long periods of time, must be attributed all the erosion of rock which gives to elevated land its cliffs and precipices, its hills and valleys, and all the varied features of the earth's surface. Share taken by different Agencies.-The detritive and erosive agencies already described may be grouped under two heads, according to differences in their mode of operation, and in the results of their action. 1. Marine agencies, which act along the margin of the land, and tend to produce an approximate level surface or plain. 2. Subaerial agencies, which act over the whole surface of the land, and tend to produce a system of valleys and watersheds, hollows and relative eminences. No part of the sea bottom could be raised into dry land without passing through the destructive plane of the sea level, neither could any part of the land be depressed without passing through the same plane, and being exposed for a shorter or longer period to the erosive action of the waves. These agents would wear down the inequalities of its surface and would produce an inclined plane, the extent and inclination of which would depend on the slope of the sinking or rising land, and the rate at which the movement took place. It must be remembered, too, that a tract of land which has been submerged and then upraised must have been twice subjected to this levelling process, and on its second emergence would present a wide area with a nearly level or slightly undulating surface. Such an area has been termed a plain of marine denudation, but would perhaps be more aptly called a surface of planation. When once any such tract had been brought almost up to the sea-level, only a very slight further uplift would be required to raise a large area of it through that level, and cause it to become dry land, so that there would be no time or opportunity for the formation of cliffs. Slight irregularities there doubtless would be, and indeed it must be understood that the words plain and planation are used in a general sense to convey the idea of a surface, where the vertical height of any inequality is small, as compared with the horizontal extent of the area. As soon as this surface produced by marine erosion is elevated into dry land, it is subjected to the detritive action of the subaerial agencies already described, and is ultimately carved out into new forms of hill and valley. In the present chapter we propose to describe some of the more striking forms and features which are developed in this process of rock-sculpture. The arrangement of the hills, ridges, and valley, in any given case, will depend partly on the original inclination of the surface of planation, and partly on the nature and structure of the rocks composing it. Conditions which determine the Position of Hills and Valleys.-The initial courses of the streams, and the general contour of the ground, will depend on the following circumstances: 1. The original slope of the surface. 2. The disposition and lie of the stratified beds. 3. The relative resistant power of the rocks, depending partly on their hardness, and partly on their chemical composition. 4. The presence or absence of igneous rocks. 5. The direction of any strong divisional planes, such as those of bedding, jointing, cleavage, or faults. Of these conditions, the second is perhaps the most important, so much depending upon whether the beds are horizontal, inclined, or curved, that it will be instructive to consider each of these three cases separately. A. Erosion of Horizontal Strata.—The simplest case which |