130 



L WER IN VER TEBRA TES. 



peculiar cliaracter formed on the coral shallows and flats. This kind of coral reef is. 

 very common in sheltered localities in the Florida waters. 



A coral island, however formed, is continually changing its contour, on account of 

 the method of its formation and the liability to erosion of the soft rock of which it is 

 composed. A study of the causes of the different shapes which coral islands assume 

 is hio-hly interesting and instructive. In some cases their outline is due to the changes 

 in level in the foundations upon which they rest, elevation, and subsidence of the sea- 

 floor; in others to the direction of oceanic currents of the waters out of which 

 they rise. 



Coral islands have every variety of form, although elongated, circulai-, ring-shaped, 

 and crescentic forms predominate. A coral formation skirting the shore, called a 

 fringing reef, follows the contour of the coast except when the continuity is disturbed 

 by local causes. The same may be said of reefs separated from the coast line by a 

 lagoon, and known as barrier reefs. 



Circular or crescentic reefs are the most striking in shape, and their mode of for- 

 mation has been a cause of considerable speculation. The circular reefs are known as 



Fig. 12^'.— Kiiig-slKipeil coral isLaml nr .itoU. 



atolls, and are abundant in the Pacific and Indian oceans. Several atolls also occur 

 in the Atlantic and the Gulf of Mexico, of which the Marquesas Islands in the Florida 

 reefs is a well-known example. The Bermudas are, I think, erroneously ranked as 

 atolls by many authors. From circular reefs, or atolls, to the simple elongated or 

 crescentic coral island we find every form of ring-shaped islands. There is nothing to 

 show that all circular coral islands are formed in the same way, or owe their peculiar 

 outlines to identical causes. In some cases thcv result from a sinking of the sea floor, 

 and in others from the direction of ocean curi-ents in their immediate vicinity. 



Many coral atolls have been shown by Darwin and Dana to have been caused l)y a 

 slow sinking and corresponding growth of corals on a submarine base. Let us sujipose 

 an island favorably situated for coral growth to have a narrow, fringing reef around 

 its coast. Suppose also that in the geologic changes of the sea floor this mountain is 

 slowly sinking below the level of the sea. As the mountain settles the animals on the 

 fringing reef cause it to rise, jyari })assu, with the depression. The intensity of coral 

 growth is always at the periphery of the fringing reef on the side turned away from 

 the island, and exposed to the pure sea water. There the coral formation, by the 

 growth of the animals, is kept to the sea level, notwithstanding the sinking of the 



