Described hy Mr Darwin. 39 



depth of water on the outside, and small shallow lagoon chan- 

 nels between them and the land. They are generally found 

 in seas that shelve gradually. The distinction between the 

 last two classes of reefs has reference chiefly to theoretical 

 considerations, as will be shewn by and by. 



Theory of Atolls — Land that has subsided or is subsiding. — 

 It must be kept in mind, as already stated, that reef-building 

 corals do not live at a greater depth than 20 or 30 fathoms, 

 or, to take the exti*eme in round numbers, say 200 feet. This 

 fact is of fundamental importance in reference to every theory 

 of coral reefs. 



1. The earliest opinion was, that these i-eefs were built up 

 in the ocean from unfathomable depths. But this is at once 

 disposed of by the fact just stated. 



2. At a more recent period some naturalists, struck by the 

 generally circular form of the reefs, and the steepness of their 

 exterior sides in many instances, supposed that they were 

 based on the craters of submarine volcanoes. To this idea 

 there is the conclusive objection, that it does not apply to long 

 narrow reefs like Bow Atoll, 30 miles by 6, or Menchikoff 

 Atoll 60 miles in length, or the larger rings, composed of 

 smaller rings, of the Maldives. That submarine craters, if 

 they reached the proper height, would afford fit foundations 

 for atolls, is probable, and such may exist ; but that all the 

 numerous atolls scattered over the ocean rest on such a basis 

 is inadmissible. 



3. It has been supposed that the atolls rest on the sum- 

 mits of the submarine mountains. But this fails in explain- 

 ing the existence of those which appear in groups. The low 

 Archipelago, for instance, contains 80 atolls, scattered over a 

 space of 840 geographical miles by 420, and not a single 

 island of ordinary rock. How can we believe that a chain or 

 group of mountains extending over such a vast area had 80 

 summits, all reaching within less than 200 feet of the surface, 

 and not one rising above it ? And this is not a solitary case ; 

 for the objection applies equally to the Gilbert group, 300 miles 

 in length ; the Mai>hall group, 520 miles by 240 ; and the 

 Maldive and Lacadive group, 1000 miles in length by 100 in 

 breadth — none of which contain a single island of any other 



