1849.] LYELL ON THE STRUCTURE OF VOLCANOS. 215 



poses them to have been produced when the island was suddenly up- 

 raised, an hypothesis which I regard as inadmissible, because they 

 never intersect the rim of the escarpment. If on other grounds we 

 conclude that the elevation of Palma from the sea was gradual, we 

 are bound to reflect whether the waves may not have contributed as 

 well as torrents acting on rocks of unequal hardness to produce such 

 ravines. 



Santorin. 



After I had indulged in the above speculations in respect to the 

 origin of the Caldera of Palma, it occurred to me that the circular 

 gulf or crater of Santorin offered a serious objection to the theory of 

 denudation, because the boundary cliffs of the Gulf plunge suddenly 

 into very deep water. It is clear that while the land and sea stand 

 at their present relative levels, the bottom of a crater 1000 feet deep 

 could never have been hollowed out by the denuding force of waves 

 and currents : but learning from my friend Capt. W. H. Smyth, that 

 a new survey of Santorin had been recently executed, under the di- 

 rection of Capt. Graves, I obtained, through the kindness of Capt. 

 Becher, of the Hydrographical department of the Admiralty, an un- 

 published chart, in which the soundings around and between this 

 group of islands are laid down with great minuteness. Capt. Smyth 

 also allowed me to consult a paper recently communicated to the Geo- 

 graphical Society by Lieut. Leycester, who has been actively engaged, 

 together with Lieut. Mansell, in the late survey. From these sources 

 I have derived data by which it will appear that the case of Santo- 

 rin, so far from militating against, is, on the contrary, strongly con- 

 firmatory of the denudation theory, besides throwing no small light 

 on the mode in which new volcanic mountains are gradually formed 

 in the centres of many craters of denudation. The largest of the 

 three islands surrounding the circular gulf of Santorin is called by 

 Lieut. Leycester, Thera (see Map, fig. 4). It is of a horse-shoe form, 

 and has an external coast-line of eighteen miles. It is three miles 

 wide from east to west, and, as MM. Boblaye and Virlet ascertained, 

 consists of volcanic matter, with the exception of its southern part, 

 where Mount St. Ehas (fig. 4, D), 1887 feet high, occurs, being com- 

 posed of limestone and argillaceous schist. The volcanic mass is quite 

 independent of these older formations, and abuts against them. It is 

 made up of alternating beds of trachytic lava, tuff, and conglomerate, 

 which, as M. Virlet has shown, have a gentle dip outwards from the 

 centre of the Gulf, towards which they terminate abruptly in a steep 

 and often perpendicular cliff. That these beds, and similar ones 

 occurring in the other two eastern islands, Therasia and Aspronisi, 

 are the lower portions of a great cone or flattened dome, the centre of 

 which has disappeared, was the opinion arrived at by MM. Virlet and 

 Boblaye, in the French 'Expedition of the Morea.' In the cliffs, says 

 M. Virlet, the separate masses of trachyte and obsidian are seen to 

 mould themselves into the inequalities of previously existing surfaces 

 formed by fragmentary and conglomerate beds. Neither the solid nor 

 the incoherent masses constitute wide-spreading sheets, but are dis- 



