214 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [DeC. 19, 



itself, like Stromboli, and been exposed to the power of tlie waves. 

 The inclination of the beds corresponds to that of the external slope 

 of the island, being greatest towards the central hollow axis, and less 

 steep near the sea. The intersecting dikes and veins are more and 

 more abundant as we approach the crater, and therefore are most 

 numerous where the slope of the beds is greatest. This is seen by- 

 aid of a transverse section of the entire succession of beds obtained 

 in the cliifs bounding the one deep baranco which extends from the 

 sea-coast to the crater. The origin of this deep ravine, a phsenome- 

 non, says Von Buch, common to all " craters of elevation," and which 

 recurs in the Great Canary, the island of Amsterdam, Barren Island, 

 and, as we shall presently see, in Santorin, is left wholly unexplained 

 by the hypothesis of sudden upheaval, unless we are prepared to 

 assume that the same engulfment which swallowed up the central 

 mass once filling what is now the hollow axis of the cone, has ex- 

 tended to one side and one side only of the encircling zone of rock 

 (see figs. 1 & 2, p. 208). Had there been several such gorges inter- 

 rupting the circular and solid girdle which encloses the caldera, it 

 might have been argued with some plausibility that such openings 

 were due to the fracture of a non-elastic mass, which, however slowly 

 upraised, could not expand and stretch, because even the less com- 

 pact beds were fortified by ribs of the unyielding stony substances 

 constituting the dikes. 



According to Von Buch, the mass upheaved in Palma fell back into 

 the middle of the crater *, but sections seem wanting to show that 

 the nature and structure of the bottom of the great hollow, where the 

 ground rises very considerably in the centre of the caldera, are such 

 as to lend countenance to this conjecture. The theory of denudation 

 briefly stated at the commencement of this paper may explain not 

 only the excavation of the caldera, but may account for its enormous 

 size ; and what is still more satisfactory, it absolutely requires the 

 existence of a great baranco through which the abstracted rocky ma- 

 terials or the missing portions of the cone have been swept out in the 

 form of mud, sand and gravel. To refer the evacuation of the Caldera 

 of Palma to explosion is inadmissible, for the same reason that M. de 

 Beaumont has very properly rejected a similar hypothesis in regard 

 to the Val del Bove on Etna, viz. because if so vast a volume of solid 

 matter had been blown out into the air, it must when it fell down 

 again have formed a dense bed of fine dust and angular fragments of 

 stone, such as does not envelope the surface or exterior slope of the 

 island. Sections of such an envelope would have been seen in the 

 ravines or barancos, some of them 500 feet deep, which radiate to- 

 wards all points of the compass, from the rim of the caldera to the 

 sea, without however interrupting hj their upper or shallower extre- 

 mities the continuity of that rim. As to the origin of these numerous 

 barancos, M. von Buch is of opinion, that the torrents now flowing 

 in some of them, even when the snow melts in the higher parts of 

 the truncated cone, are too inconsiderable to cause them. He sup- 



* Description des iles Canaries, p. 285. French edition, 1836. 



