THE FLORIDA REEFS. 83 



stated already by Darwin, Dana, and others, that no coral reef can 

 grow above the level o£ the lowest tides, and that all subsequent 

 additions of material must be due to accumulation of sediment 

 transported by the action of the tides and prevailing winds. 



Next come the clusters of coral heads, huge masses of astrseans 

 and of mseandrina, very limited in their distribution at the Tor- 

 tugas, as well as the more or less extensive patches of Madrepora 

 palmata (Fig. 51), and finally what is known as broken ground, 

 namely, the outer edge of the reef occupied mainly by clusters 

 of gorgonise (Fig. 52), which also reach upward into the shal- 

 lower region. Occasional patches may be seen of astrseans, ma- 

 drepores, and other reef -builders, which have extended below the 

 depths at which they generally flourish, where they are soon 

 killed or choked by the accumulation of fine coral sand and cor- 

 alHne sand or ooze of the deeper waters. This sediment fills 

 the broad and narrow flat channels dividing the three great 

 banks which compose the Tortugas, or separate the inner shoals, 

 banks, and islands. Finally come the lines of broken coral heads 

 and branches, mixed with dead corallines, shells of mollusks, old 

 serpulse tubes,^ gorgonise stalks, and the like. These form a 

 low dike, as it were, to be little by little pounded up by the 

 breakers into smaller fragments, and carried, either by the winds, 

 or waves, or currents, into the interior of the reefs, there to form 

 sand flats of more or less coarse materials, until on the western 

 faces of the banks the finest detritus is deposited in very steep 

 slopes, constantly shifting like those of sand dunes, and, like 

 them, running; forward and backward at the will of the winds 

 and waves. This continues until the particles have become ce- 

 mented together by the action of the carbonic acid contained in 

 excess in the salt water surrounding the reefs, and by the gluing 

 of the slight amount of animal matter which holds these parti- 

 cles together. Some of the slopes (according to General Wright, 

 of the Engineers) are as great as thirty-three degrees. 



All this fine material, composed of fragments of every animal 



^ Serpulse often form incrusting masses and dead corals from being too rapidly 



of considerable extent, acting-, as has been broken to pieces by the action of the 



noticed by Darwin, ranch as the patches waves. 

 of nuUipores do in protecting decayed 



