[footnote] *See the interesting little map of the island of Nisyros, in Roise's 'Reisen auf den Griechischen Inseln', bd. ii., 1843, s. 69. Sometimes half of the annular ledge has been destroyed, and in the bay formed by the encroachment of the sea corallines have built their cellular habitations. Even on continents craters of elevation are often filled with water, and embellish in a peculiar manner the character of the landscape. Their origin is not connected with any determined species of rock: they break out in basalt, trachyte, leucitic porphyry (somma), or in doleritic mixtures of augite and labradorite; and hence arise the different nature and external conformation of these inclosures of craters. No phenomena of eruption are manifested in such craters, as they open no permanent channel of communication with the interior, and it is but seldom that we meet with traces of volcanic activity either in the neighborhood or in the interior of these craters. The force which was able to produce so important an action must have been long accumulating in the interior before it could overpower the resistance of the mass pressing upon it; it sometimes, for instance, on the origin of new islands, will raise granular rocks and conglomerated masses (strata of tufa filled with marine plants) above the surface of the sea. The compressed vapors escape through the crater of elevation, but a large mass soon falls back and closes the opening, which had been only formed by these manifestations of force. No volcano can, therefore, p be produced.* [footnote] *Leopold von Buch, 'Phys. Beschreibung der Canarischen Inseln', s. 326; and his Memoir 'uber Erhebungscratere und Vulcane', in Poggend., 'Annal.', bd. xxxvii., s. 169. In his remarks on the separation of Sicily from Calabria, Strbo gives an excellend description of the two modes in which islands are formed: "Some islands," he observes (lib. vi., p. 258, ed. Casaub.), "are fragments of the continent, others have arisen from the sea, as even at the present time is known to happen; for the islands of the great ocean, lying far from the main land, have probably been raised from its depths, while, on the other hand, those near promontories appear (according to reason) to have been separated from the continent." A volcano, properly so called, exists only where a permanent connection is established between the interior of the earth and the atmosphere, and the reaction of the interior on the surface then continues during long periods of time. It may be interrupted for centuries, as in the case of Vesuvius Fisove,* and then manifest itself with renewed activity.