REEFS OF FLORIDA. 159 



The seeds of the mangrove have also the faculty of shooting out 

 long roots and stems, even while still attached to the parent tree. 

 These sprouted seeds, falling into the water, float away, and if their 

 roots touch bottom immediately fix themselves, grow into mangrove- 

 trees, and commence multiplying in the manner described. Thus in 

 the shoal water (e') are found mangrove islands in which there is no 

 land, but only a mangrove forest, standing above water by means of 

 their interlaced roots. By these, however, sediments are detained, and 

 a true island is speedily formed. It is in this way that the small man- 

 grove islands in the shoal water on the south and west of Florida are 

 formed. They are entirely different from the wave-formed coral isl- 

 ands or keys. The hummocks in the Everglades have probably a 

 similar origin, although some of them may possibly be of coral 

 origin. 



Florida Reefs compared with other Reefs. — In comparing the reefs 

 just described with other reefs, it will be seen that the former are 

 unique in two respects. 



1. No other reefs continuously make land. In fringing reefs there 

 is a small accretion about the shore-line of the previously-existing land, 

 but this process is quickly limited. In barriers and atolls, according 

 to Darwin, there is always loss of land, only a small fraction of which 

 is recovered by coral and wave agency. But under these agencies 

 Florida has steadily advanced southward more than 100 miles, and the 

 area thus added to the continent is at least 10,000 square miles. It 

 seems utterly impossible to account for this, except by supposing some 

 other agency at work preparing the ground for the growth of success- 

 ive reefs. 



Probable Agency of the Gulf Stream. — Since corals can not grow in 

 water more than sixty to one hundred feet deep, it is evident that, un- 

 less subsidence goes on pari passu with the growth of the corals, a coral 

 formation can not be more than one hundred feet thick. But there is no 

 evidence of subsidence on the coast or keys of Florida. On the contrary, 

 the height of these parts is precisely the usual height of toave- formed 

 islands, although no longer exposed to their action. It follows, there- 

 fore, that the corals must have built upon an extensive submarine bank, 

 produced by some other agency. Furthermore, since the reefs were 

 formed successively one beyond another, it is evident that there must 

 have been a progressive formation of this bank from the north toward 

 the south. The dotted lines (Fig. 128) show successive positions of the 

 tank of the reefs. Such a progressive extension of a bank can only 

 be formed by sedimentary deposit. It is almost certain that in some 

 way the Gulf Stream is connected with this sedimentary accumula- 

 tion. It is to this agency, therefore, that we attribute the formation 

 and extension of the bank upon which the corals grow. 



