(3) The erosion of river valleys to great depths since the Glacial period fails to establish the great antiquity of the caves left on their sides or the high level gravels of their banks. Throughout the northern hemisphere, the river valleys are of old date, and were merely filled with loose detritus in the Glacial age. The sweeping out of this _débris_ would be a rapid process, more especially when changes of level were occurring, and when the rainfall was greater than at present. Besides, as Croll has well remarked, the actual configuration of our continents, the amount of drift still remaining, and the imperfect manner in which the river valleys have been cleared out, all testify to the comparative recency of the Glacial period.[91] These considerations would, indeed, materially reduce the antiquity which he claims on astronomical grounds for the ice age. (4) The growth of peat and the deposition of silt are very deceptive as indications of great antiquity. For instance, accurate observations made by a French engineer in the construction of docks at St. Nazaire,[92] show that in 1,600 years the Loire had deposited over Gallo-Roman remains six metres of mud. Relics of the Bronze age occur below these at a depth indicating 500 years previously as their date; and the beginning of the modern deposit of the Loire would, on the same evidence, be only 6,000 years ago. Hilgard’s observations on the delta of the Mississippi in like manner tend greatly to reduce our estimates of the time occupied in the deposit of the modern silt of that river. The peat deposit at Abbeville, at the mouth of the Somme, has been supposed to have required 30,000 years for its formation. But this estimate was based upon the present rate of growth; and, as Andrews has shown, the fact admitted by Boucher de Perthes, that birch stems three feet high stand in this peat, implies a much more rapid rate, which is also proved by the depth at which Roman remains have been found. In like manner the Scandinavian peats, to which a fabulous antiquity has been ascribed, have been proved to be comparatively modern by the depths at which metallic works of art are found in them. (5) The paucity of remains of Palæocosmic men in Europe, with their wide distribution, indicate that their sojourn was not long, or that the population was very small and much scattered. Even in a few thousands of years, an active and vigorous people, living in a country well supplied with food, must have multiplied greatly, and must have left considerable remains. On the theory that these men inhabited Europe even for 2,000 years, we have to suppose that the greater part of their remains have been swept away, or remain under the waters, or buried out of sight in diluvial sediments.