91 



composed of tertiary marls, capped with basalt $ but the basalt does 

 not here overlie the alluvions, as has been asserted. 



Phenomena perfectly analogous to those of Perrier are exhibited 

 on the Allier at St. Maurice, and in the hill of Monton, not far di- 

 stant: and these three sections, as well as that above mentioned at 

 the new passage of the Sioule, all concur in proving that many val- 

 leys in Auvergne, anciently excavated through gneiss and lacustrine 

 marls capped with old basalt, have at some remote periods been 

 filled up with transported matter, and afterwards been excavated a 

 second time, — generally to a depth below their original bottom. 



6. The authors conceive, with the writers already mentioned, 

 that a satisfactory explanation of these phenomena may be derived 

 from the effects of the latest volcanic eruptions of Central France. 

 For the more recent lavas appear to have dammed up the channels 

 of several rivers, and converted ancient valleys into lakes ; wherein, 

 as at Aidat and Chambon, alluvial matter is continually accumulating 

 at present. The modern lava of Montpezat, in the Vivarrais, has 

 thus obstructed the course of the Fontaulier, and given origin to a 

 lake, since filled with river alluvion and volcanic ashes; and these 

 deposits themselves, together with a part of the volcanic barrier, 

 have been subsequently cut through, by the action of the river and 

 the waters of the lake. The early and more copious lava-currents 

 of Auvergne must have occasioned larger lakes than those of recent 

 formation ; and these, as has been stated by other authors, seem to 

 have been gradually filled up, with materials introduced by rivers, 

 and occasionally by floods from the sides or craters of volcanoes, 

 probably during their moments of eruption ; through which accu- 

 mulations new valleys were excavated by the continued action of 

 the rivers : — as at Mont-Perrier, to the depth of about 100 feet ; and 

 at Maurice on the Allier, to 400 feet, below their original bottoms. 

 The high antiquity of these alluvial depositions is inferred from the 

 fact, that their lowest remnants occupy as elevated a position on the 

 sides of the valleys, as the lava-currents of intermediate age in Au- 

 vergne ; and from the compactness and enormous mass of the tra- 

 chytic breccias, which overlie and alternate with the alluvions. 



7. Lastly, since the sand and gravel containing the fossil bones, 

 found on two different sides of the mountain of Perrier, are overlaid 

 by a vast mass of trachytie breccia, it is concluded, that the ele- 

 phant, rhinoceros, hippopotamus, hysena, tiger, wild cat and other 

 quadrupeds, whose remains have been recently disinterred, must 

 have been inhabitants of this district, before the most recent cones 

 and lavas of Auvergne had appeared, or the valleys had been exca- 

 vated to their present depth ; and even before the fires of Mont 

 Dor were extinguished. 



Jan. 2, 1829. — A letter was read, addressed to R. I. Murchison, 

 Esq. For. Sec. G.S. &c. by G. W. Featherstonhaugh, Esq. F.G.S. 

 " On the Series of Rocks in the United States." 



Mr. Eaton has published, in Silliman's Journal of Science, (vol. xiv.) 

 a Synopsis of the rocks of North America. In the commencement of 



