20 J!' J. SOLLAS [january 



level of low tides. Sandbanks resting on this form the dry land. 

 Since, he observes, every particle of the atoll which lies within the 

 reach of observation consists of coral, it is only just to conclude that 

 the whole structure, including the table-mountain, is formed of the 

 same material. Not perhaps a strictly logical conclusion, yet, as 

 events have proved, in the main correct. 



Chamisso's opinion was not destined to remain long unchallenged, 

 for two famous French naturalists — Quoy and Gaimard — asserted, as 

 the result of their observations, that the coral-rock of an atoll is only 

 skin-deep, i.e. it forms, according to them, a mere superficial crust, not 

 more than about 25 feet in thickness; the rest — Chamisso's "table- 

 mountain " — being, on this view, of volcanic, or at all events of inor- 

 ganic, origin. 



lew of the arguments by which it was attempted to sustain this 

 erroneous conclusion strike one as being very satisfactory, but they include 

 one highly important observation, viz. that reef-building corals do not 

 live at greater depths than 2 5 feet below the level of low tides. Sub- 

 sequent inquiry, while fully confirming the existence of a limit, has at 

 the same time extended it down to a depth of as many as 25, or 

 perhaps even 40, fathoms. Yet, even with this modification, the un- 

 expected discovery of Quoy and Gaimard seems to stand in flagrant 

 contradiction to the views of Chamisso. If corals cannot grow below 

 a depth of 25 fathoms, how could they possibly have built up islands 

 of over 100 fathoms in thickness ? 



The answer to this question, as is well known, was given by 

 Charles Darwin. If we admit the truth of both the apparently con- 

 flicting statements, it is obvious that the corals at the base of a reef 

 100 fathoms in thickness must have been situated within the limit of 

 25 fathoms at the time they were alive. But in order to bring them 

 within this limit it is only necessary to suppose that the foundation 

 on which they grew originally stood 75 fathoms nearer the sea-level 

 than it does now ; or, in other words, that since the lower layers of the 

 reef were alive and flourishing, the ground which supported them has 

 sunk 7 5 fathoms deeper in the sea. No fact is better established than 

 the rise and fall of islands situated in mid-ocean, and thus there is 

 nothing antecedently improbable in this supposition. But once grant 

 it and Darwin's explanation of atolls naturally follows. Thus let a 

 be an island with its summit rising 100 fathoms above the sea ; let 

 its shores become peopled with corals, which extend seawards down 

 to the limit of 25 fathoms, beyond which, as we admit, they cannot 

 proceed : a reef is thus started, which will continue to grow, rising 

 upwards till it reaches the level of low tides : when this is attained 

 upward growth will cease, and the reef will begin to pass into decay, 

 from the shore-edge outwards. So long as the island remains stationary, 

 neither rising nor falling with respect to the sea-level, this is practically 

 all that will happen, and the final result is a reef not much exceeding 



