26 Messrs Lyull and Murchison o« the 



cut to the depth of fifty feet into the compact iava ; and still far- 

 ther down, in a narrow defile in the granite below Verrieres, it 

 has not only pierced, but entirely swept away, at many points, 

 the whole lava-current, — a break the more remarkable, because 

 the " cheire"" is again resumed in great force, where the valley 

 widens below, spreading near its termination at Nechers to the 

 breadth of half a mile. 



Although the Couze of Chambon produces so little eflfect at 

 its exit from the lake, it has formed at the upper extremity an 

 alluvial tract, several miles in length, and, by the continual in- 

 flux of matter from the heights of Mont Dor, it will fill, at 

 some remote epoch, the whole lake. 



In like manner, the upper end of the Lac d'Aidat is shallow, 

 and encroached upon by a marshy delta. The stream where it 

 issues has only cut down to the depth of a few feet, passing be- 

 tween the granite and the " cheire," although at St Saturnin, 

 about twelve miles below, it has not onlv cut through the 

 " cheire" (above 100 feet thick), but through about twenty 

 feet of subjacent fresh- water strata. 



In the case of Montpezat, there appears to us clear evidence 

 of the lava having occasioned a lake, and of the lake having been 

 subsequently filled, and this lacustrine deposite, with part of its 

 barrier, having been finally cut through. The south pier of 

 the bridge over the Fontaulier at Montpezat, rests on basalt, 

 constituting that part of the lava-current nearest to the crater. 

 Overlying this, on the right bank, are seen alluvial deposites, 

 about thirty feet in thickness, the lower half composed of layers 

 of volcanic sand, with a few small pebbles of gneiss, the upper 

 of coarse gravel, and rounded blocks of gneiss, several tons in 

 weight, the whole often cemented into a puddingstone by vol- 

 canic ashes. On the opposite bank, a terrace of this coarse al- 

 luvion is also seen resting on basalt. About thirty yards higher 

 up, on the same side of the river, the finer volcanic part of this 

 deposite is seen reaching down to the level of the river, in thin 

 sedimentary layers. We may suppose, therefore, that a lake 

 was formed, and that the lower extremity was principally filled 

 with volcanic matter, washed down, and perhaps in part ejected 

 from the adjoining cone. Over successive layers of this finer 

 matter, which at length overtopped the sloping sides of the bar- 



