According to the most recent chemical analyses of the younger Silliman, the genus Porites, like many other cellular coral-trunks (Madrepores, Astræas, and Mæandrinas of Ceylon and the Bermudas), contains besides from 92 to 95 per cent. of carbonate of lime and magnesia, a portion of fluorine and phosphoric acid.[LL] The presence of fluorine in the hard skeleton of the polyps reminds us of the fluoride of calcium found in fish bones according to Morechini’s and Gay-Lussac’s experiments at Rome. Silex is mixed only in very small quantities, with the fluoride of calcium and phosphate of lime found in the coral-trunks; but one coral animal allied to the Horn corals (Gray’s _Hyalonema_, Glass thread) has an axis of fibres of pure silex, resembling a hanging tuft of hair. Professor Forchhammer, who has recently been engaged in a thorough analysis of sea-water in the most opposite parts of the earth’s surface, finds the quantity of lime in the Caribbean Sea remarkably small, it being only ²⁴⁷⁄₁₀₀₀₀, whilst in the Cattegat it amounts to ³⁷¹⁄₁₀₀₀₀. He is disposed to ascribe this difference to the numerous coral-banks near the West India Islands, which appropriate the lime to themselves, and thus exhaust the sea-water.[LM] Charles Darwin has with great ingenuity developed the genetic connection between shore-reefs, island-encircling reefs, and lagoon islands, _i. e._, narrow, annular coral banks which surround inner lagoons. According to his views, these three kinds of structure depend upon the oscillating condition of the bottom of the sea, or on periodical elevations and subsidences. The often-advanced hypothesis, according to which the lagoon-islands, or atolls, mark by their circularly enclosed coral-reefs, the outline of a submarine crater, raised on a volcanic crater-margin, is opposed by the great extent of their diameters, which are in some instances upwards of 30, 40, or even 60 miles. Our fire-emitting mountains have no such craters, and if we would compare the lagoon, with its submerged mural surface and narrow encircling reef, with one of the annular lunar mountains, we must not forget that these annular mountains are not volcanoes, but tracts of land enclosed by walls. According to Darwin, the following is the process of formation. An island mountain closely encircled by a coral reef subsides, while the _fringing_ reef that had sunk with it, is constantly recovering its level owing to the tendency of the coral animals to regain the surface by renewed perpendicular structures; these constitute first a reef encircling the island at a distance, and subsequently, when the inclosed island has wholly subsided, an _atoll_. According to this view, which regards islands as the most prominent parts, or the culminating points of the submarine land, the relative position of the coral islands would disclose to us what we