128 PACIFIC OCEAN (Ranai), Molokai (Morotoi), Oahu (Woahoo), Kauai (Atooi), and Niihau (Oneehoow), and four small barren islets, the entire area being 6100 square miles. The islands of this group are mountainous, and abound in active volcanoes; the great lake of fire, Kilauea, on the east side of the Mountain of Mauna Loa (13,760 feet) in Hawaii is probably the largest active crater in the world, while one of the largest known extinct craters is that of Mauna Haleakala (" The House of the Sun") in Maui, at a height of 10,200 feet above the sea; it is 12 miles in circumference. The Hawaiian Islands being within the zone of coral formation are surrounded by fringing reefs, and there is abundant evidence that gradual upheaval has taken place over the whole area which they occupy. There are beds of coral limestone in Molokai at a height of 400 feet, and in Kauai coral sand is found at an elevation of 4000 feet above the sea; in many other islands coral and lava are found interstratified. The three groups of the Bonin Islands known as the Parry, Beechy, and Coffin groups are composed of high rocky islets of a bold and fantastic outline, and are situated between 26 and 27 N. lat. The Ladrones or Mariana Islands (see vol. xiv. p. 199) have a total area of 395 square miles; they stretch for nearly 450 miles between 13 and 20 N. lat. and 144 37 and 145 55 E. long. These islands are all of volcanic origin, and their mountains contain several active volcanoes. The Caroline Archipelago (see vol. v. p. 1 25) lies about 170 miles to the south of the Ladrones, and, together with the Pelew Islands, has an area of 877 square miles. The Carolines embrace forty distinct island groups, five of which are basaltic and mountainous, though surrounded by coral reefs ; the remaining thirty-five groups are entirely of coral formation, and do not rise much above the sea-level. The Pelew Islands resemble the Carolines in their physical characters ; they present peculiarities in the arrangement of atolls which will be alluded to below. The Marshall Islands (see MICRONESIA, vol. xvi. p. 256) consist of two chains running parallel to each other, and composed of fourteen and seventeen small groups respec tively. They lie to the eastward of the Carolines, and are entirely of organic formation. The Gilbert Archipelago (see vol. xvi. p. 256) is cut by the equator. It contains sixteen groups of small coral islands, low and barren, but densely populated. In the South Pacific oceanic islands are scattered with the greatest profusion over a region between 5 and 25 S. lat. and 180 to 120 W. long. The northern part of the shallow water surrounding Australia, New Zealand, and the Malay Archipelago is occupied by the Solomon Islands, the New Hebrides, the bold rocky and mountain ous islands of Fiji with fine barrier reefs, the Friendly Islands, and Samoa or the Navigators Islands. Farther to the south there are the Society Islands, including Tahiti; they are lofty, of volcanic origin, and surrounded by very perfect barrier reefs. The Marquesas or Mendana Archi pelago, farther to the north, also consists of volcanic islands, but they are not fringed by reefs. The volcanic group of the Galapagos Archipelago is situated under the equator at a distance of 500 or 600 miles from the went coast of South America; it has been minutely described by Darwin. The extensive Low or Paumotu Archipelago lies to the south-east of the Society Islands, and runs parallel to them. It consists of about eighty atolls, some of them of large size, and all typical examples of this form of coral island. The total area of the islands of the Pacific is exceedingly small, especially when the vast number of groups that stud the ocean is taken into consideration. THEORY OF CORAL ISLANDS. The origin of coral islands was specially studied by Darwin during the voyage of the "Beagle" in 1831-36, and he shortly afterwards published a theory on the subject which has been full} detailed in the article CORAL "(vol. vi. p. 377). This theory was so simple, and it appeared so complete, that it acquired universal acceptance ; and the continuous action of subsidence in promoting the development of fringing reefs into barriers, and of barriers into atolls, was long unquestioned. In 1851 L. Agassi/: 1 expressed the opinion that the theory of subsidence was insufficient to explain the formation of the coral reefs and keys of Florida. In 1863 Carl Semper stated that an attentive study of the Pelew Islands showed the complete inadequacy of this theory, and in 1868 he reiterated his convictions. 2 In 1880 Mr John Murray published an abstract of his "Chal lenger" observations, 3 and gave a theory of coral island formation which claims to account for all the phenomena without calling in the aid of subsidence. It is pointed out that, with hardly an exception, the oceanic islands are of volcanic origin, and it is assumed that the various peaks which deep-sea soundings have shown to be scattered over the bed of the ocean, and rising to within various distances of the surface, are also, primarily, of volcanic origin. There is no evidence of any extensive submerged continent or mass of land such as Darwin s theory requires. Whether built up sufficiently high to rise above the surface of the sea and thus form islands, or brought up only to varying heights below the sea-level, these volcanic eminences tend to become platforms on which coral reefs may be formed. The erosive action of waves and tides tends to reduce all volcanic summits down to the lower limit of breaker action, thus producing platforms on which barrier reefs and atolls may spring up. Again, submarine eminences may be brought up to the zone of the reef builders by the deposit of volcanic and organic detritus falling from the surface, as well as through the agency of organisms secreting lime and silica, which live in profusion at great depths, especially on the tops of submarine peaks and banks. The great profusion of life in the tropical surface waters is insisted upon, and it is pointed out that this pelagic life supplies the reef-building corals with food, and that, when these surface creatures die and their shells fall to the bottom, they carry down with them sufficient organic matter to furnish food to the animals living on the floor of the ocean. As the result of tow-net experiments in the tropics Mr Murray estimated that, in the surface waters of the ocean, there were in a mass 1 mile square by 100 fathoms, 16 tons of carbonate of lime existing in the form of shells of pelagic Foraminifera and Molluscs. In this way it is urged that submarine banks are continually being brought within the zone of reef-building corals. Darwin admitted that reefs not to be distinguished from atolls might be formed on such submarine banks, but the improbability of so many submerged banks existing caused him to dismiss this explanation without further consideration. He was not, however, aware of the great number of submerged cones which recent soundings have made known, nor of the enormous abundance of minute calcareous organisms such as calcareous alg?e, Foraminifera, and Molluscs in the surface waters and of the comparatively rapid rate at which their remains might accumulate on the sea bottom. Xor had lit any idea of the comparatively great abundance of animals living at considerable depths. Coral-reef builders starting on a bank, whether formed by elevation or subsidence, by erosion or the upward growth of deep- sea deposits composed largely of organic remains, tend ultimately to assume the atoll or barrier form. When the coral reef or colony approaches the surface, the central portions are gradually placed at a disadvantage as compared with the peripheral parts of the mass, in being farther removed from the food supply which is brought by the oceanic currents, and consequently dwindle and die. In pro portion as the reef approaches the surface, the centre becomes cut off from the food supply and the conditions become increasingly uncongenial. At last an outer ring of vigorously growing reef corals encloses a central lagoon. The windward side of the reef grows most vigorously, not because of a larger supply of oxygen and greater aeration of the water, but because that is the direction in which the oceanic currents bring the food to the reef. As the atoll extends seawards from vigorous growth the lagoon becomes larger, chiefly from the removal of lime in solution by the action of the carbonic acid in sea water which flows in and out at e aeh tide. This solvent actio:i of sea water on dead calcareous organ isms was shown by the "Challenger s" observations to be uni versal. Mr Murray reverses the order of growth as given by Darwin for the groups in the Indian Ocean. He regards the Laccadive, Caro line, and Chagos archipelagos as various stages in the growth of coral reefs towards the surface, and he explains the various appear - 1 Bull. Mus. Comp. Zool., vol. i. 2 Verhandl. Physik. Med. Gesellsc/t. Wiirxburg, Feb. 1, 1868. 3 Proc. Roy, Soc. Edin., vol. x. p. 505.
