been the thickening of the talus which protected the loose cinders and bones from waste. We behold in many a valley of Auvergne, within fifty feet of the present river channel, a volcanic cone of loose ashes, with a crater at its summit, from which powerful currents of basaltic lava have poured, usurping the ancient bed of the torrent. By the action of the stream, in the course of ages, vast masses of the hard columnar basalt have been removed, pillar after pillar, and much vesicular lava, as in the case, for example, of the Puy Rouge, near Chalucet, and of the Puy de Tartaret, near Nechers.[1] The rivers have even in some cases, as the Sioule, near Chalucet, cut through not only the basalt which dispossessed them of their ancient channels, but have actually eaten fifty feet into the subjacent gneiss; yet the cone, an incoherent heap of scoriæ and spongy ejectamenta, stands unmolested. Had the waters once risen, even for a day, so high as to reach the level of the base of one of these cones—had there been a single flood fifty or sixty feet in height since the last eruption occurred, a great part of these volcanoes must inevitably have been swept away as readily as all traces of the layer of cinders; and the accompanying bones would have been obliterated by the Rodes near Aurignac, had it risen, since the days of the mammoth, rhinoceros, and cave-bear, fifty feet above its present level.
The Aurignac cave adds no new species to the list of extinct quadrupeds, which we have elsewhere, and by independent evidence, ascertained to have once flourished contemporaneously with man. But if the fossil memorials have been correctly interpreted—if we have here before us at the northern base of the Pyrenees a sepulchral vault with skeletons of human beings, consigned by friends and relatives to their last resting-place—if we have also at the
