1849.] LYELL ON THE STRUCTURE OF VOLCANOS, 229 



out in relief from the cliffs at the head of the great valley, have not 

 owed their preservation to their superior hardness, and consequent 

 power of resisting aqueous action. There are in the tertiary lime- 

 stones of the Val di Noto, in Sicily, circular valleys where the steep 

 boundary cliffs have been shaped out into a great succession of ledges, 

 separated by small cliffs, often producing an effect which I have 

 compared (see ^Principles,' 1st edit. 1833, p. 110, vol. iii.) to the seats 

 of a Roman amphitheatre. The precipitous rocks of white limestone 

 thus carved out are sometimes 500 feet high. The period of this 

 extensive denudation was very modern, geologically speaking, and we 

 may infer that when the sea had power to shape out such cavities in 

 rocks of uniform solidity and compactness, it may have exerted a far 

 greater denuding energy on such alternations of stony and incoherent 

 materials, as those now constituting the boundaries of the Val del 

 Bove. 



The dimensions of Etna are on a sufficient scale to have produced 

 a large crater of denudation, had a cavity been excavated in the sum- 

 mit or centre, instead of on the flanks of the cone. Suppose the vol- 

 canic mass not to have been cut away to a greater distance from the 

 axis of the mountain than the middle of the Val del Bove ; there might 

 have been a cavity formed three or four miles in diameter, encircled by 

 escarpments from 3000 to 4000 feet in height. The dikes in that 

 case would have been most numerous in the vicinity of the original 

 and principal centre of eruption. At the nearest point to this centre 

 now accessible is a rock already alluded to, called Giannicola, agreeing 

 in mineral composition with the lavas of Etna, but highly crystalline, 

 and massive, which Hoffmann describes as almost resembling gra- 

 nite in structure, and between 150 and 200 feet wide. The deeper, 

 therefore, we are enabled to see into the composition of the volcano 

 near its central axis of eruption, the more massive and crystalline are 

 the contents of upfilled fissures. 



I have offered in the 'Principles of Geology' an explanation of the 

 fact on which M. de Beaumont has dwelt with much emphasis, that 

 the more ancient parts of Etna have in the course of the last 2000 or 

 3000 years scarcely received any superficial accessions of lava and 

 scoriae (see 'Principles,' 7th edit. p. 398, 1847). 



The dome-shaped or conical mass was probably several thousand 

 feet less elevated when it was originally formed. After its bodily up- 

 heaval the eruptions would become more and more lateral and basal, 

 or in other words, the exogenous growth of the cone would shift its 

 chief place of development. Yet the focus of eruption continued in 

 the loftiest part of the cone where the lava still rises to a great height, 

 and often overflows, whenever lateral eruptions occur. 



Mr. Hopkins has suggested, that if the denudation of the Wealden 

 was anterior in great part to its elevation, the removal of an incumbent 

 weight of matter from the central area might have enabled the expan- 

 sive force to act with greater intensity on that space where so much 

 less pressure remained to be overcome. In accordance with this view, 

 we might expect that the Val del Bove, after the abstraction of volcanic 

 masses varying in thickness from. 500 to 4000 feet, would have become 



