34 Messrs Lyell and Murcliison on the 



tendency, it is said, formerly, in the lavas to spread themselves 

 out in sheets, instead of flowing, as afterwards, in long narrow 

 stripes. But the discovery of liver alluvions under so many 

 plateaus, helps greatly to explain the phenomenon. We have 

 remarked in this memoir, that the lava of Montpezat spreads 

 itself out into a wide sheet, where the breadth of the valley in- 

 creases greatly, and forms in these situations a columnar basaltic 

 plateau, the flat surface of which is at a considerable height 

 above the present river, and the foundation of which is also 

 seen to be an ancient alluvion. We have also stated, in other 

 cases, that the narrowest parts of several currents have been 

 completely annihilated by the erosion of a river, whereas the 

 same are again resumed in grea^jforce, where the valley widens, 

 as in the case of the current of Tartaret. 



It is obvious that, if these operations were continued for a 

 great period of time, the wide plateaus of lava would alone re- 

 main, the connecting portions having disappeared ; and as the 

 process of excavation, in every country, diminishes the slope of 

 the bottoms of valleys (where subterranean forces do not cause 

 fresh derangement), the inclined planes of the old plateaus 

 would be steeper than the present river beds, as seems so often 

 the case in Auvergne. There is no difficulty in regard to the 

 appearances now under consideration, so long as it cannot be 

 pretended that the old plateaus are wider than the flat alluvial 

 plains occasionally bordering the modern rivers in the same 

 district. But, so far is this from being the case, that many 

 rivers (as the Allier, in particular, near Vichy) exhibit large flat 

 plains, now strewed over with gravel and sand, of greater width 

 than any of the old plateaus. It is scarcely necessary to recall 

 the remark of Mr Scrope and others, that the oldest lavas, 

 however inexplicable their present position, must, when they 

 were in a liquid state, have attained, soon after issuing from 

 below, the lowest levels of the then surface, and consequently 

 have occupied the course of a torrent or river. 



Mont Perrier, near Issoire. {Montague de Boulade of MM. 



Deveze and Bouillet.) 

 As two works have recently been published on the relations 

 of the alluvial strata of Mont Perrier, we feel called upon to 



