ftO Messrs I-yell and Murchison on the 



La Begude, a few miles above Aubenas, after contributions 

 have been received from all the lava-currents of the Vivarais. 



At the above locality, the bed of the Ardeche is about 400 

 yards wide ; it is laid open in parts to the depth of fourteen 

 feet perpendicular, through sand and boulders of unknown 

 thickness. The upper surface at least of this great mass moves 

 on annually ; and some islands formerly covered with trees and 

 buildings have been overwhelmed. Here the prisms of hard 

 basalt have, by trituration, in a few miles, lost all traces of their 

 original form, — a fact which shews how much the quantity of 

 matter carried into the sea in the shape of sand, must exceed 

 that which descends in the form of pebbles. 



On regarding the void spaces once occupied by basalt in the 

 valleys of the Vivarais, the doctrine formerly advocated by Play- 

 fair is much illustrated and confirmed ; and we feel disposed 

 with M. Montlosier, Scrope, and other authors who have writ- 

 ten on the excavations of valleys in central France, to ascribe al- 

 most unlimited power to ordinary rivers, when a sufficient lapse 

 of time is assumed. Nature rarely affords us, as here, an accu- 

 rate measure of amount of destruction occasioned during pe- 

 riods of definite extent, or whose limits we can fix with reference 

 to other natural events. 



But the waste of the gneiss during the same lapse of time has 

 been incomparably greater, and the matter which has disappeared 

 from the innumerable valleys falling into the Ardeche would ex- 

 cite much more admiration, if it could be raised again from the 

 bed of the Rhone or the Mediterranean, and submitted in one 

 mass to our inspection. Yet the mountains and valleys have 

 preserved the same mutual relations, and the same general form, 

 since the epoch of the eruptions. The lavas still share with the 

 rivers the lowest levels, and if the basalt of Ay sac could be re- 

 melted, and again made to flow down from its crater, it would 

 descend in the same direction, and, after encircling in the same 

 manner the promontory of gneiss on which Antraigues stands, 

 would fill up again, to a certain height, the three valleys which 

 converge and meet below the town ; from whence the principal 

 lava stream would then wind down as now towards the Ardeche. 

 In these and similar cases, the lower parts of the valleys have re- 



