In part of the region above alluded to, as, for example, near Lentini, a conglomerate occurs in which I observed many pebbles of volcanic rocks covered by full grown _serpulae_. We may explain the origin of these by supposing that there were some small volcanic islands which may have been destroyed from time to time by the waves, as Graham Island has been swept away since 1831. The rounded blocks and pebbles of solid volcanic matter, after being rolled for a time on the beach of such temporary islands, were carried at length into some tranquil part of the sea, where they lay for years, while the marine _serpulae_ adhered to them, their shells growing and covering their surface, as they are seen adhering to the shell figured in p. 22. Finally, the bed of pebbles was itself covered with strata of shelly limestone. At Vizzini, a town not many miles distant to the S.W., I remarked another striking proof of the gradual manner in which these modern rocks were formed, and the long intervals of time which elapsed between the pouring out of distinct sheets of lava. A bed of oysters no less than 20 feet in thickness rests upon a current of basaltic lava. The oysters are perfectly identifiable with our common eatable species. Upon the oyster bed, again, is superimposed a second mass of lava, together with tuff or peperino. In the midst of the same alternating igneous and aqueous formations is seen near Galieri, not far from Vizzini, a horizontal bed, about a foot and a half in thickness, composed entirely of a common Mediterranean coral (_Caryophyllia caespitosa_, Lam.). These corals stand erect as they grew; and, after being traced for hundreds of yards, are again found at a corresponding height on the opposite side of the valley. [Illustration: Fig. 123. _Caryophyllia caespitosa_, Lam. (_Cladocora caespitosa_, Ehr.) _a._ Stem with young stem growing from its side. _a*._ Young stem of same twice magnified. _b._ Portion of branch, twice magnified, with the base of a lateral branch; the exterior ridges of the main branch appearing through the lamellae of the lateral one. _c._ Transverse section of same, proving, by the integrity of the main branch, that the lateral one did not originate in a subdivision of the animal. _d._ A branch, having at its base another laterally united to it, and two young corals at its upper part. _e._ A main branch, with a full grown lateral one. _f._ A perfect terminal star.] The corals are usually branched, but not by the division of the animals as some have supposed, but by the attachment of young individuals to the sides of the older ones; and we must understand this mode of increase, in order to appreciate the time which was required for the building up of the whole bed of coral during the growth of many successive generations.[152-A]