CORAL REEFS are formed of the whitish calcareous frame work of various marine organisms, of which corals (q.v.) and nullipores are the two most important. These reef-builders flourish only at moderate depths—the nullipores usually not below so fathoms and the corals not below 25—in the warmer oceans. The minute larval forms of corals and the still more minute spores of nullipores given off in myriads by the adult forms float pas sively in the ocean waters ; if they happen to drift with currents of fitting temperature to a shore of firm rock or to a shallow bank where pebbles or shells lie undisturbed by surface waves, they May attach themselves there and in time form a new reef.
Coral reefs are of three kinds : fringing reefs, barrier reefs and atolls. Fringing reefs are sea-level flats, up to a quarter or half mile in width built out upon salient parts of continental or insular shores, with fronts that fall off seaward to moderate depths. They are composed largely of dead reef rock, and are occupied by living reef-builders chiefly on their outer edge and slope. Reefs of this kind are common in the East Indies. Barrier reefs have, like fringes, a narrow or broad sea-level flat and an outer growing face ; but they are separated from the coast which they front by a salt-water lagoon, from half a mile to a score of miles or more in width and from 20 tO 4o or more fathoms in depth. The lagoon floor is usually rather smoothly strewn over with fine calcareous detritus. The inner shore is commonly occupied by a fringing reef, the growth of which is less vigorous than that of non-enclosed fringes. The outer or growing face of a barrier is continued in a gentle slope to 4o or so fathoms, followed by a steep pitch to great depths. Barrier reefs are frequently interrupted by passages or "passes," through which ocean-going vessels may enter the pro tected lagoon. When barriers encircle islands, the leeward arc of their circuit is, as a rule, interrupted by wide breaches, largely because of the leeward drift of lagoon sediments. The Great Bar rier Reef of north-eastern Australia, goo nautical miles in length, is the largest of its kind; the lagoon is from 20 tO 7om. in width. Many small and large islands rise from the lagoon within this great natural breakwater, and they as well as the embayed main land coast are little cliffed or not at all, indicating long-continued protection from ocean waves by the reef ; but the cooler and reef less coast farther south, also embayed, is island-free and strongly cliffed, indicating its long exposure to ocean waves while an inor ganic continental shelf was accumulating off-shore.
Atolls resemble barrier reefs, except that no island rises from their lagoon. The largest atolls are about 4om. in diameter; the smallest, less than a mile with no lagoon. They are rarely circular, usually irregular in pattern. The reef flats of barriers and atolls are often heaped with low islands of reef sand, which come to be covered with vegetation. The reef-islands of atolls are frequently inhabited, although exposed to the occasional danger of being over whelmed by earthquake waves. The origin and history of barrier reefs may be inferred less from the facts observed in the reefs than from those seen in the coasts that they adjoin or front ; and in the absence of neighbouring lands, sea-level atolls are inscru table structures. The three classes of reefs grade into each other. A fringe separated from its coast by a shallow water belt resembles a close-set barrier; the central islands of certain barriers are so small that the reefs may be called almost atolls. All classes of reefs are found either elevated above or depressed below sea-level. Vatu Vara is a small atoll in central Fiji, i,o3oft. in altitude. The large island of Timor in the East Indies has many elevated reefs at various altitudes on its slopes, the loftiest being almost atolls or atolls on the island crest, about 4,000ft. above sea-level. Chagos atoll, from 65 to 95m. in diameter, in the Indian ocean, is slightly submerged nearly all around its circuit. Reef-building corals have been dredged in the Ceram sea, Dutch East Indies, from a depth of over 7oo fathoms, 3om. distant from the nearest shore.
Three contrasted theories of coral reefs are as follows: Darwin (1838-42) explained barrier reefs as the upgrowth of fringing reefs on slowly subsiding coasts or islands, and atolls in the main as the continued upgrowth of barrier reefs until the central island sinks below sea-level. Murray (188o) rejected subsidence and ex plained barrier reefs by the outgrowth of fringing reefs from sta tionary coasts on their own talus. He believed the lagoon was exca vated by solution of the dead reef rock behind the growing reef front, thus repeating an idea of Semper's 0863-80. He sug gested also that atolls might be developed from barrier reefs by the gradual degradation of the central island, but his preferred view was that they are coral crowns on banks that were organically built up over still-standing foundations, usually volcanic cones, of whatever depth, thus repeating an idea of Rein's (187o). Guppy (1887) thought that coral reefs were formed on rising founda tions; atolls would thus crown shoaling but not emerging banks; barrier and fringing reefs would lie on emerged coastal slopes. He explained the lagoons of barrier reefs as covering platforms of abrasion cut in coastal slopes during a pause in their emergence, thus repeating in essence an idea of Tyerman and Bennet's (1832).
The inventors of these theories adopted them without making a suffiCiently thorough deduction of their consequences. Thus Guppy overlooked three significant points: (I) If fringing reefs are formed on rising coasts, they should lie on non-eroded slopes; but most fringes lie on eroded slopes, showing that subsidence had preceded reef growth, even if in some cases the subsidence were followed by upheaval. (2) If barrier reefs rise from abraded plat forms on rising coasts, the shore back of them should be cliffed and not embayed, but such shores are in nearly all cases embayed and not cliffed. (3) If atolls are based on non-emerged shoals, their limestones should lie on the non-eroded surface of the shoals, but several elevated atolls are known to rest unconformably on subaerially eroded foundations. which must therefore have sunk before the atolls were built upon them.
Murray overlooked five significant points: (I) Narrow fringing reefs would ordinarily be smothered by detritus outwashed from valleys in a stationary coast before they could grow out as barriers. (2) If a fringe on a stationary coast succeeded in growing out as a barrier, the shore behind it would not be embayed, but such shores are always embayed. (3) If lagoons are excavated by solution their floors should be covered by insoluble residue, instead of by accumulating calcareous detritus, as is usually the case. (4) If barrier reefs are transformed into atolls by the degradation of their stationary central island, then the islets of almost atolls should be low and flat ; but almost atoll islets are of steep-sloping, mountain-top form. (5) If most atolls are crowns on organically aggraded banks, then elevated atolls should show pelagic deposits between a non-eroded volcanic base and the coral crown; but only two elevated atolls—Roti in the East Indies and Barbadoes in the Lesser Antilles—are known to be underlaid by pelagic deposits, and in both cases these deposits rest unconformably on a subaeri ally eroded, non-volcanic rock, thus showing that island subsidence, at a rate too rapid to be compensated by reef upgrowth, had pre ceded a later and slower upheaval, with reef growth and emer gence.
Darwin failed to recognize three significant consequences of his theory : (I) Subsidence provides by far the best means of dis posing of the great volume of detritus that has been eroded from the coasts fronted by barrier reefs. (2) If barrier reefs have grown up from slowly subsiding foundations, the coasts from which they are offset should be embayed by the partial submergence of the coastal valleys. (3) If the formation of barrier reefs and atolls is associated with the subsidence of their foundations, their lagoon limestones—but not necessarily their external talus—should lie un conformably on the eroded surface of the foundation rocks. But while the unnoticed consequences of Murray's, Guppy's and certain other theories are contradicted by the facts of observation, thus invalidating those theories, all the unnoticed consequences of Dar win's theory are confirmed by the facts. (I) The volume of detritus that has been eroded from reef-fronted coasts would in nearly all cases have filled—often much more than filled—the lagoons and overwhelmed the reefs if the coasts had remained sta tionary. (2) The central islands of barrier reefs are embayed. as Darwin knew, but it was Dana who first showed (1849) that the embayments result from the entrance of arms of the sea into the valleys of subsiding coasts. (3) In the elevated atoll of Tu vutha in eastern Fiji, Foye has reported (1918) an uncon f ormable contact of the atoll limestones on a subaerially eroded volcanic foundation, and the same observer found in the neigh bouring Exploring isles a similar relation between the limestones of several elevated reefs and their volcanic bases. The elevated bar rier reef of Mangaia in the Cook group has recently been shown by Marshall (1927) to rest unconformably upon the slopes of a well dissected volcanic island. In all these cases the association of sub sidence with reef growth seems unquestionable, and the same asso ciation is therefore probable in the case of sea-level barriers and atolls; for it is not to be believed that deep-seated telluric forces selected reefs of exceptional origin to be elevated. Hence, inas much as the consequences of Darwin's theory, though they were not perceived by its inventor, have now been confirmed by previ ously unknown facts, the theory is strongly supported.
The novel glacial-control theory, recently put forth by Daly (1910-15), is based upon the similar depth of many reef-enclosed lagoons, which he believes cannot be explained by Darwin's sub sidence theory. He assumes that, as a rule, reef foundations have long been stationary; that many of the older volcanic islands of the Pacific had been degraded in pre-glacial times to low relief with deep-weathered soils : that with the coming of the Glacial period the ocean was lowered some 3o or 4o fathoms by with drawal of water to form continental glaciers and ice sheets; that even in the torrid zone the chill of the lowered ocean killed the reef-builders, the ocean waves abrading the reefs and worn-down islands to low-level platforms ; and that as the ocean rose and warmed in post-glacial time, barrier and atoll reefs grew up with it on the platform margins. But this theory is largely invalidated by the evidence given above of island instability, also by the pre vailing absence of cliffed shores within close-set barrier reefs in the coral seas; yet as a good number of plunging-cliff islands sur mounting imperfectly reefed banks occur in the marginal belts of the Pacific and Atlantic coral seas, it seems probable that Daly's factor of low-level abrasion—but not his postulate of insular sta bility—has there had application. In this and a few other respects Darwin's theory may be subordinately modified to advantage, as follows : Young volcanic islands are as a rule unfavourable sites for fringing reefs because of the abundance of down-washed detritus which soon forms a reef-smothering beach around the shore ; such islands are therefore attacked by the waves and cut back in cliffs; witness Reunion in the Indian ocean. Not until the subsidence of such an island has disposed of a great volume of detritus eroded from it—the isostatic subsidence of the island being likely be cause of its imense weight, as Molengraaff has suggested (1916) —will the eventual submergence of its rising, cliff-base beach permit reef growth to begin, either on the faces of the plunging cliffs or somewhat off-shore. The Marquesas islands, strongly cliffed and well embayed, seem to offer examples of incipient cliff-face reefs so conditioned. Thus subsidence favours the initia tion of fringing reefs as well as their further development and transformation into barriers. But not unless subsidence continues slowly and maintains embayments in the coastal valley mouths, where down-washed detritus will be deposited in bay-head deltas, are up-growing reefs likely to persist. Even then they may be drowned by rapid subsidence, as seems to have happened with a young barrier now submerged around Tutuila, Samoa. But if sub sidence continues slowly an on-shore or a near-shore reef may grow up as an off-shore barrier before the cliffs previously cut are wholly submerged; this appears to be the case at Tahiti, in the Society group, where the island spurs are cut off in cliffs that seem to plunge below sea-level. Yet here an extended pause has per mitted the filling of the drowned valley embayments with del tas which have now advanced into the lagoon, smothering many cliff-face fringes, and even the off-shore barrier appears to be en dangered by out-flowing floods of muddy fresh water. A renewal of subsidence would widen the narrowed lagoon and re-embay the valleys, and the barrier reef would then be rescued from the danger of muddy floods. Thus in time the early-cut cliffs would disappear and the inter-bay spurs of the diminishing island would slope gradually into the widening lagoon. This stage appears to have been reached in the other barrier-reef members of the Society group, for as one proceeds north-westward, the islands are found to be more and more dissected and degraded, and of smaller and smaller size as if increasingly submerged. Borabora, Darwin's type barrier-reef island, is the next to the last of the series. Sev eral atolls follow, the volcanic foundations of which have entirely disappeared. It would be difficult to imagine a better confirmation of Darwin's theory than these islands and their reefs provide. In Fiji the distribution of various kinds of reefs, both at sea-level and elevated above it, especially in the eastern part of the group, at first appears so confused that Darwin's theory has been held to be inapplicable there; but a closer examination shows that this old theory is really the only one which can reasonably account for the Fiji reefs. In the East Indies movements of elevation and sub sidence have been so active that typical sea-level barrier and atoll reefs are not often found there, but sea-level fringes and elevated reefs of various kinds abound.
The several groups of atolls in the central Pacific can be to-day explained only by analogy with elevated reefs elsewhere, as above intimated, except that on the atoll of Funafuti in the Ellice group a boring in the reef, made under the direction of the Royal Society of London, I,184f t. deep—a small measure compared to the pre sumable reef thickness—has shown that shallow-water organisms prevail to that depth in the reef rock, while deep-water organisms are found at similar depths on the exterior slope of the reef. This supports Darwin's theory. A number of reefs in Florida have been shown by Vaughan to have been formed at times of subsidence, but they are of small thickness. In view of what is at present known concerning the coral reef problem, it may be concluded that although Darwin's theory was abandoned by many students of the subject during the past 5o years, it may—slightly modified as above—regain in the next 5o years the general acceptance that it enjoyed through the middle of the last century.