1. It must be determined by evidence whether the accumulation to be explained happened on the land 'at its present level, or on the bed of the sea.
2. It must be determined by evidence what was the probable character of the climate in the countries where diluvial accumulations excite attention.
3. In proposing a general cause, such for example as the movement of glaciers, it must be shown to be adequate to satisfy all the minutes details of the phenomena, and not inconsistent with general limiting conditions established by extensive induction from facts observed in the earth itself, or admitted as parts of general cosmical theory. [REFRIGERATION OF THE GLOBE; and, in NAT. HIST. DIV., GEOLOGY and SURMARINE FORESTS.] The determination of the cause of the diluvial accumulations is of the highest importance in geological theory. It is impossible to doubt that to the same cause must be ascribed many considerable modifica tions of the pre-e.ristent surface of the land. If, abstracting our attention from the accumulated deposits which conceal the stratified and other rocky masses in the crust of the earth, we look at the actual firm of its surface, there appears little that is even difficult of explana tion by the application of known and real causes. The relative areas of sea and land; the peculiarities of outline of continents and islands ; the directions of mountain-ranges, and remarkable vales and plains ; the individual features of hills and valleys ; the degree in which the land is wasted in some quarters and augmented in others; and the rate of change which may take place in these respects;—all this may be satisfactorily referred to subterranean and submarine disturbances of different periods, to the effects of the sea upon the land when the land was not at its present level above the water, and to the operation of the atmosphere, rains, rivers, and inundations.
From this large field of research we shall select for brief illustration the outlines of land and sea, the directions of high and low ground, and the individual features of hills and valleys. The few examples needed will be drawn from the British Isles, but the explanations are of general application. To render the article, as reproduced in this
work, more accordant with the actual state of geology, illustrations of other parts of the subject are now added, some from foreign countries, but of general application equally with the former.
Much of the irregularity of outline, on a large scale, of the British Islands depends on the form in which the ancient bed of the sea was elevated into dry Land. Thus the line of the Hebrides, the prominent parts of Caithness, of Aberdeenshire, and of Argyleshire, are on axes of upward movement of the primary strata ; while the Great Caledonian Valley, from Fort William to Inverness, and the great basin of the Forth and Clyde, are in axes of depression of the strata. This latter hollow is margined on the south by the great axis of elevation of the Galloway and Lammermuir hills, reappearing beyond the Irish Channel south of Belfast, as the Argyleshire chain is resumed in Donegal. The Isle of Man, the promontory of Llcyn in Caernarvonshire, the east and west ridges and hollows of the strata which reach the sea in South \Vales and North Devon, give to Pembrokeshire, Glamorganshire, and North Devon remarkable and detailed alternations of promontory and bay. The Isle of Wight is formed on an axis of elevation from the Needles to Culver Cliff; while north of it are the axis of depression of the Solent, the axis of elevation of the wealda of Sussex, and the axis of depression of the estuary of the Thames.
Inland, the same ridges and hollows, and others of as great import ance, produce continuous chains of hills—the North-Western Highlands, the Grampians, the Lammermuir range of hills, the Wicklow monn tain.s, the Snowdon mountains, the Berwyn, and Malvern hills, and mul titudes of other narrow tracts of elevated land. Great faults, elevating or depresaiug one portion of a natural district, leave marks of inequality on the surface. Thus the great Pennine faults, ranging from New castle to Brampton, and thence to Kirby Lonadale and Settle, occasion differences of level in the ground of 1000 and 2000 feet for a length of 100 miles.