Excavutiun of' Valleys. 43 



ing out oi- sweeping away of the subjacent pumiceous and other 

 beds of lighter ingredients. Further, there would have been 

 more confusion, and more unconformability, in the deposites, if 

 they had subsided from turbulent waters rushing through this 

 narrow valley, while it remained an open passage for the drain- 

 age of the higher country. But the waters of a temporary lake, 

 on the contrary, would arrest the progress of the matter intro- 

 duced into it, and allow it to spread over the bottom with some 

 degree of regularity, the violent erosive force of the current 

 being deadened, and each deposite being protected from destruc- 

 tion by the mass of water, while others were superimposed. 



A lake, therefore, being regarded as alone capable of explain- 

 ing in a satisfactory manner the phenomena of Perrier, it be- 

 comes an interesting matter of inquiry to decide at what epoch 

 the valley was dammed up, and at what period, in that great 

 succession of volcanic eruptions of which Auvergne has been 

 the theatre, the animals, whose remains were buried in the lake 

 lived and perished in these regions. 



Now, if the recent lava-stream, which has its course down 

 the same valley, had flowed on as far as the village of Perrier, 

 it would evidently have assisted us greatly in determining this 

 relative date ; for we should then have known how far the lacus- 

 trine formation had been eaten into when the " cheire " de- 

 scended, and, consequently, whether the lake had long ceased 

 to exist, before the more modern class of volcanoes in Auvergne 

 had commenced their activity. 



We have before stated, that this modern lava-current has ac- 

 tually flowed down the valley to a certain distfuice, viz. to the 

 spot where the village of Ourseyre, below Besse, stands ; but it 

 stopped short at Sauriers, some miles before reaching the locality 

 where the alluviums and breccias attest the former existence of 

 the lake. This deficiency, however, is in some measure supplied 

 by the existence, in the next adjoining valley of Nechers, of the 

 current of Tartaret, one of the most recent in Auvergne, and 

 which the geologist may look down upon from the summit of 

 Mont Perrier ; so that it requires no great effort of the imagina- 

 tion to suppose the lava transferred from the oine valley to the 

 other, and thereby to judge of the relations which a " cheire" 

 of this age would have borne to the ancient alluviums, had it 



