1859.] SCEOPE COIS^ES AND CEATEKS. 541 



or elliptical cavity is formed by the subsidence of the fluid, sur- 

 rounded by perpendicular cliffe, in which perhaps several shelves 

 may appear, to attest the different stages of this process. This again, 

 however, is evidently a process wholly distinct from, and even the 

 reverse of, " upheaval*." 



Concentric Craters. — One of the most remarkable and interesting 

 features of volcanic craters is their frequent occurrence one within 

 the other, in more or less concentric circles. In my former paper 

 (of April 1856) I referred to the recorded history of the eruptions of 

 Yesuvius during the last hundred years, and showed that within that 

 short period the cone had been no less than five times gutted by ex- 

 plosions of paroxysmal violence, and the craters so formed been subse- 

 quently as often filled again by the welling up of fresh lava and the 

 ejection of scoriae from the bottom in the intervals of these paroxysms. 

 And I called the attention of geologists (an attention I again sohcit, 

 since it is a point of special importance) to this alternate filling, 

 emptying, re-filling, and re-emptying of the central crater of a vol- 

 cano, as a normal process or general law of volcanic action, to which is 

 owing the so frequent appearance in volcanic districts of one or more 

 cones and craters within encircling annular clifi^-ranges, or portions of 

 such ranges — the "basal wrecks" (as Mr. Darwin justly calls them) of 

 still larger cones formerly existing there, and blown up by some early 

 eruptive paroxysms. As some of the most remarkable and best-known 

 examples may be mentioned, in addition to Yesuvius and Somma 

 (fig. 18, p. 542), those of Santorini in the Grecian Archipelago, of 

 Barren Island in the Bay of Bengal, of Bromo in Java, as described 

 by Professor Jukes, St. Helena, St. Jago (well described by Mr. Dar- 

 win), the Pic de Fogo in the Cape de Yerde Islands, the Cirque of 

 Teneriffe (fig. 4) (within which rise the cones of the Peak and 

 Chahorra), the Curral of Madeira, the Cliff-range around the volcano 

 of Bourbon (fig. 19), Antuco in Chili, Irasu in Costa Rica, the Campo 

 Bianco, and the Isle of Yolcano in the Lipari Isles (figs. 20 and 21), 

 and many others. The outer rings in all these instances, as well as 

 the interior cones and craters in nearly all, are of course considered 

 by the upheavalists as "elevation-"^ or " upheaval- craters;" but 

 only because, as I think I have shown, they have no correct con- 

 ception of the character of the eruptions which give rise to such, or 

 indeed to any craters. 



It is, in truth, singular how completely this theory has blinded 

 its advocates to the best- attested facts respecting even that trite and 

 most frequented of European volcanos, Yesuvius, which is within 

 view of the luxurious residences of Naples, and which consequently 

 has been visited and explored more than any other. They appear 



* Professor Dana thinks that there is no limit to the size of craters formed in 

 this mode, and considers that those of the moon may be pits of the kind. But 

 the moon-craters are not encircled by abrupt precipices ; they rather resemble, 

 in aspect and configuration, the tuff-craters of the Phlegrsean Fields, which owe 

 their origin to explosive sulsaqueous eruptions in shallow water. The remark, 

 however, is suggestive of the important bearing of the theory of the formation of 

 cones and craters upon astronomical no less than on geological questions. 



VOL. XV. PAET I. 2 Q 



