6 FLORIDA REEFS. 
even an extensive interruption in the succession of the keys before we reach 
the Tortugas. These last, however, as well as the bank west of these keys, 
belong none the less to the main range of keys, from which they are only 
separated bv a more extensive and deeper depression. West of Sand Key 
the reef itself becomes gradually less elevated, luitil it is finally lost where 
the ship channel, south of the Marquesas, expands into the broad depression, 
separating that group of keys and shoals from the Tortugas. 
In order to understand fidly not only the topography, l)ut also the mode 
of formation of all these keys and reefs, it must be remembered that the 
risino- reefs, which form more or less continuous walls, reaching- at unequal 
heio-hts nearly to the surface, or above the level of the waters, are only 
a particular modification of those formations growing upon coral grounds 
under special circumstances. It has been ascertained, whenever similar 
investigations have been made, that living corals do not occur in depths 
exceeding twenty fathoms, tiiat the reef-building species prosper from 
a depth of about twelve fathoms nearly to the surface, and that diflereut 
species follow each other at successive heights. Now, if we keep in mind 
these facts, we shall see that all the coral-bound islands of the West Indies, 
as well as of the main-land of Central America, constitute an extensive coral 
field, divided by broad, deep channels, over which the coral reefs extend, 
with different features, according to the depths in which they occur and the 
changes which their own growth has gradually introduced upon the locali- 
ties where they are found, influenced .and modified to some extent also by 
the direction of the prevailing currents and the action of the tides. 
The formation of the main range of keys in their primitive condition as 
a reef, — for, as we shall see hereafter, they have been a submarine reef 
before they rose as islands above the level of tlie ocean, — the formation 
of this range, we repeat, at gradually greater distances from the main-land, 
as we follow their course from east to west, has been simjily owing to the 
depth of the bottom from which the reef has risen. It has followed the line 
of ten or twelve fathoms' depth ; and if there is so wide an interruption 
between the Marquesas and the Tortugas, it is because the ground is deeper 
over that space. Again, if the Pine Islands have a northwesterly direction, 
while the main range runs more from east to west, it is no doubt because 
the bod}^ of water emptying from the northern part of the gulf, along the 
western shores of the peninsula, has, for a time, run chiefly over that field, 
while the tract of mud flats between the keys and the main-land was filling 
