42               GEOLOGY OF NEW YORK CITY
	made up of their topmost films, would, in any case, be inevi­table. But that the deep deposits of sand and clay might rep­resent shore sands and clay beds formed in the weathering of the gneiss ridges of Manhattan Island ages before the Ice Age or its succeeding quaternary epochs, is a violent assump­tion. The results of the decomposition of the rocks of Man­hattan Island, formerly far more elevated than they are to-day, were doubtless removed to the south, and may, indeed, have formed cretaceous and tertiary beds beneath the surface of Long Island. The geological conditions are readily understood, as indi­cated by these deposits. This lower section of the island is underlaid by the same kind of rock as forms the hills, promi­nences and ridges northward, and over this floor a burden of loose strata has been accumulating, rising in a succession of beds, each individually homogeneous, or relatively so, as clay, sand and graveL These beds originated by water action upon drift material heaped up possibly north and south of them, at any rate near their present position, an action which washed out the clay particles and permitted their settling in clay beds. This action, also, parted the sand and gravel, and under the impulse of torrents or slow shore-washing completed the separation of each. During the long periods required for this gradual re-assortment, the island underwent changes of ele­vation, alternating, probably, over one hundred feet, since on the one hand gravel beds formed in swift and shallow streams are found fifty feet below water level, and must have been much nearer the surface when made, and, on the other hand, sand hills, also indicating water action, are recorded eighty feet above tide. In many places such modified beds were capped by mingled aggregates brought hither by ice which, in rolling hills of stones, pebbles, and boulders, formed the orig­inal surface of the island over much of its extent, the higher hills and more remarkable boulders seeming to have been con­centrated eastward about Corlears Hook. As present condi­tions supervened, river mud and other debris accumulated and