Excavation of Valleys. ST 



rier, the Fontaulier, having first pushed forward a delta across 

 the lake, brought down its coarse alluvion. It then opened it- 

 self a passage through the lacustrine alluvium, and is now gra- 

 dually eating out a ravine through the subjacent basalt. The 

 enormous magnitude of some boulders resting on the basalt, do 

 not imply a greater power than that of the present river, for we 

 saw masses of gneiss of as many tons in weight, which had been 

 recently brought by much smaller torrents into the Fontaulier. 



The lava current, after crossing the river at the point above 

 mentioned, spread itself over the small plain, where the town of 

 Montpezat stands, and filled the channel of the Pourseille, a 

 stream which unites with the Fontaulier a short distance below, 

 at the Castle of Pourchirol *. A few yards above the point of 

 confluence, the Pourseille falls in a cascade over a cliff, the 

 upper part of which, to the depth of thirty feet, consists of sco- 

 riaceous lava worn through by the stream ; below which is a 

 bed of prismatic basalt, thirty feet in thickness, vertical and 

 overhanging, for it has been undermined by a copious stream of 

 water which runs beneath, through a bed of sand and pebbles, 

 resting on gneiss. This ancient alluvion varies from three to 

 nine feet in thickness, and is composed of ferruginous brown 

 sand, and rounded boulders of gneiss, but none of basalt. The 

 gneiss underneath is worn down in an inclined plane to the 

 depth of perhaps thirty feet. In the natural grotto thus formed 

 under the basalt, the original pebble-bed may be studied in 

 nearly as satisfactory a manner as in the mine of Chaluzet be- 

 fore described j*. 



Below the Castle, the lava expands itself into a sheet about a 

 mile in breadth, and about two miles long, forming the rich and 

 level plateau called Champagne. The Fontaulier here flows on 

 the north of the plain, in a ravine between the gneiss and the 

 basalt, the latter presenting towards the river broken, and occa- 

 sionally vertical, cliffs of about 100 feet in height. One of the 



" See Scrope's Geology of Central France, Plate XV. 



-f- Mr Scrope, with whose general views on the excavation of valleys in 

 Auvergne we fully concur, has considerably over-estimated the thickness of 

 primary rock worn through imder basalt at the Castle of Pourchirol. Sec 

 Scrope's Geology of Central France, Plate XV. j), 181. 



