254 convincing geological evidence in favour of the secular con- traction of the globe, that during the process masses of sedimentary strata, many thousands of feet in thickness,have been crumpled and crushed, and that the ernmpling has often been accompanied by such an amount of heat and evolution of chemical activity as to produce a movement and rearrangement of the elements of the roeks,—this change sometimes advancing to the point of actual fusion. There is reason to believe that some at least of these periods of intense terrestrial disturbance have been followed by periods of prolonged volcanic action in the disturbed areas. Mr Mallet‘s theory is thus, to some extent, supported by inde- pendent geological testimony. The existence, however, of large reservoirs of fused rock, at a comparatively small depth beneath the surface, may be conceived as probable, apart altogether from the effects of the subsidence of the outer shell upon the inner nucleus. The connexion of volcanoes with lines of elevation, and consequent weakness in the earth's crust, is precisely what might have been anticipated on the view that the nucleus, though practically solid, is at such a temperature and pressure that any diminution of the pressure, by corrugation of the crust or otherwise, will cause the subjacent portion of the nucleus to melt. It is along the lines of elevation that the pressure must be relieved, and it is there that the consequent melting will take place. On these lines of tension and weakness, therefore, the conditions for volcanic excitement must be best developed. Water is there able soonest to reach the intensely-heated materials underneath the crust, and to give rise to the volcanic ex- plosions. The periodicity of eruptions will thus depend upon the length of time required for the storing up of sufficient steam, and on the amount of resistance in the crust to be overcome. In some cases the intervals of activity, like those of the geysers, return with considerable regularity. In other cases, the shattering of the crust, or the upwelling of vast masses of lava, or the closing of sub- terranean passages for the descending water, or other causes may vary the conditions so much, from time to time, that the eruptions may follow each other at very unequal periods, and with very discrepant energy. Each great outburst exhausts for a while the vigour of the volcano, and an interval is needed for the renewed accumulation of vapour. Section II.—Earthquakes. The phenomena of earthquake-motion having been dis- cussed in the article EARTHQUAKES, we shall notice here only those which have a marked geological importance from the way in which they affect the crust or surface of the earth, briefly describing the effects of earthquakes upon the surface of the land, and upon terrestrial and oceanic waters, the permanent changes of level occasioned by them, their distribution and geological relations, and their causes. 1. Efiects upon the Soil and General Smface of a Cozmtrg/.—The earth-wave or wave of shock under- neath a country may traverse a wide region and affect it violently at the time without leaving any trace of its passage. Loose objects, however, are apt to be displaced. Thus blocks of rock, already disengaged from their parent masses, may be rolled down into the valleys below. Large landslips are thus produced, and these may give rise to very considerable subsequent changes in the drainage of the localities where they take place. It has often been observed that the soil is rent by the passage of the earthquake. Fissures appear, varying in size from more cracks, like those due to desiccation, up to deep and wide chasms. “'here these cracks are numerous, and where, consequently, the ground has been much disturbed, permanent modifications of the landscape may be produced. Trees are thrown down and buried, wholly or in part, in the rents. These superficial GEOLOGY [u1. DYNAMICAL. effects may, in a few years, be effaced by the gradual levelling power of the atmosphere. Where, however, the chasms are wide and deep enough to intercept any rivulets, or to serve as channels for heavy rain-torrents, they are sometimes further excavated, so as to become gradually enlarged _into ravines and valleys. As a rule, each rent is only a few yards long. Sometimes it may extend for half a mile or even more. In the earthquake which shook the South Island of New Zcaland in 1848, a fissure was formed averaging 18 inches in width and traceable for a distance of 60 miles parallel to the adjacent axis of the mountain-chain. The subsequent earthquake of 1855, in the same region, gave rise to a fracture which could be traced along the base of a line of cliff for a distance of about 90 miles. Remarkable circular cavities are sometimes formed in the ground during the passage of the earth-wave. In many cases these holes serve as funnels of escape for an abundant discharge of water, so that when the disturbance ceases they appear as pools. They are believed to be caused by the sudden collapse of subterranean water-channels and the consequent forcible ejection of the water to the surface. 2. Ejfects upon, Terrcstrir1l Wuters.——Springs are tem- porarily affected by earthquake movements, becoming greater or smaller in volume, sometimes muddy or dis- coloured, and sometimes increasing in temperature. Brooks and rivers have been observed to flow with an interrupted course, increasing or diminishing in size, stopping in their flow so as to leave their channels dry, and rolling forward with increased rapidity. Lakes are still more sensitive to the tremors of the ground beneath. Their waters occasionally rise and fall for several hours, even at a distance of many hundred miles from the centre of disturb- ance. Thus, on the day of the great Lisbon earthquake, many of the lakes of central and north-western Europe were so affected as to maintain a succession of waves rising to a height of 2 or 3 feet above their usual level. Cases, how- ever, have been observed where, owing to excessive sub- terranean movement, lakes have been emptied of their contents and their beds left permanently dry. On the other hand, areas of dry ground have been depressed, and have become the sites of new lakes. Some of the most important changes in the fresh water of a region, however, are produced by the fall of masses of rock and earth. Landslips, by dannning up a stream, may so arrest its water as to form a lake. The barrier, if of sufficient strength, may be permanent, and the lake will then remain. Owing, however, to the usually loose, inco- herent character of its materials, the dam thrown across the pathway of a stream runs a great risk of being undermined by the percolating water. When this is the case, a sudden giving way of the barrier will allow the confined water to rush with great violence down the valley, and produce perhaps tenfold more havoc there than may have been caused by the original earthquake. When the landslip is of sufficient dimensions to divert the stream from its previous course, the new channel thus taken may become permanent, and a valley may be cut out or widened. 3. Effects -upon the Sec(.—The great sea—wave propagated outward from the centre of a sub-oceanic earthquake, and reaching the land after the earth—wave has arrived there, gives rise to much destruction along the maritime parts of the disturbed region. As it approaches the shore, the littoral waters retreat seawards, sucked up, as it were, by the advancing wall of water, which, reaching a height of some- times 60 feet, rushes over the bare beach and sweeps inland, carrying with it everything which it can dislodge and bear away. Loose blocks of rock are thus lifted to a consider- able distance from their former position, and left at a higher
level. Deposits of sand, gravel, and other superficial