THEORIES OF BARRIER AND CIRCULAR REEFS. 151 



than the upward building of the coral-ground ; that every reef com- 

 mences as a fringing reef, but, in the progress of subsidence, was con- 

 verted first into a barrier and finally into an atoll. For, as the vol- 

 canic island went down, the corals would build upward on the same 

 spot ; and as the island would become smaller and smaller, and the 

 corals would grow faster on the outer side of the reef, where they are 

 exposed to the breakers, it is evident that the reef would become sepa- 

 rated from the island by a ship-channel, and thus become a barrier. 

 Finally, when the island disappears entirely, the reef, still building 

 upward, would become an atoll. These changes are represented in 

 the accompanying section (Fig. 125). As the changes are relative, 

 they may be represented either by the land sinking or the sea-level ris- 

 ing ; for the sake of convenience we use the latter. In the figure, I" I" 

 represents the sea-level when the reef was & fringe, I' I' when it was a 

 barrier, and I I the present sea-level, when it has become an atoll. The 

 ship-channel and the lagoon, though always lower, rise pari passu with 

 the reef proper. This is the- result partly of the growth of placid- water 

 species of corals, and partly of the drifting of coral debris from the 

 reef, and detritus from the volcanic island. It is seen that the corals 

 do not build a vertical wall, and therefore that the atoll is always 

 smaller than the coast-line of the original island. Consequently, if the 

 subsidence continues, a typical atoll is changed into a small closed 

 lagoon, and, finally, into a lagoonless island. These, therefore, indicate 

 the deepest subsidence. 



Fig. 125. 



Evidences. — 1. , This theory accounts for all the more obvious phe- 

 nomena of atolls, such as their irregular circular form, their size, the 

 steepness of their outer slopes, etc. 2. Every stage of gradation between 

 the fringing reef on the one hand, and the atoll on the other, has been 

 traced by Dana, strongly suggesting that they are all different stages of 

 development of the same thing. We have in the Pacific some high 

 islands, which are surrounded by a pure fringing reef ; others in which 

 the reef is a fringe on one side and a barrier on the other ; others in 

 which the barrier is one mile, two miles, five miles, ten miles, twenty, or 

 thirty miles distant ; others which are called atolls, but the point of the 

 original volcanic island is still visible in the middle of the lagoon ; others 



