— 12 — 



ayilraulic pressure, includes the assumption, and throws into 

 the opposite scale, the heavy improbability that their tops would 

 in gencral , remain standing at nearly an equal height and that 

 cnough for the habitudes of the coral builders , and besides these 

 considerations , we have to conclude, that these volcanic walls 

 even of the dyke form, but certainly those of the circular or 

 nearly so, would long retain heat, thereby enabling the coral- 

 lines to live and work at depths, where the temperature of 

 the water was elsewhere too low for them, par exaraple the 

 crater in the isle of Amsterdam (South Indian ocean) has been 

 open to the sea some two hunderd years and yet, the tempe- 

 rature of the water rising through it's walls, is still raised to 

 boiling heat. Whereas on the idea of subsidence the case is 

 utterly dark, no light being shed by it on the onochis operan- 

 di, whereby the upper platforms of these isles were by tlie 

 coral builders first raised above the sea level, and afterwards 

 retained at that elevation, wJdlst the foundations were under- 

 going the process of subsidence, with reference to the oceanic 

 // extents of more than a thousand miles by several hundreds, 

 // scattered over with islands, none of whicli rise to a greater height 

 // than that to wliich the sea can throw fragments , or the wind 

 // heap up sand. " We shall defer the notice , till we come to 

 tlie more special statement of the situation etc. which we sec 

 looming a hcad. 



IX «llaving endeavoured on geneval groiinds, to shew tlint tlie belisf 

 »of a general subsidence is almost necessary, to account for the exis- 

 »tence of a vast numbcr of recfs on one level, we will uow seehow far 

 » tlie samo idea will apply to the peculiar configuration in the severai 

 » classes. Let us imagine an island, nierely fringcd by recfs, extending 

 » to a short distam-e frora the shore , in which case, as we have rcmarked, 

 »there is no difficulty in undeistanding their structure. Now let this is- 

 » land siibsidc by a series of movcmeuts of extreme slowness, the coral 

 » at each interval growing up to the surface (level of low water sjiiing 

 »tidcs), a little reflection will shcw , that a reef encirding the shore at a 

 sgrcaicr or less distance , according to the amonnt of subsidence, would 

 » be jiioiluced. If we suppose the sinking to continue, the encircled is- 

 » huid must by the submcrgcncc of the central land, but upwaid growih 

 » of ilic ring of coiiil, bc convcrtcd into a l'igo'in island-" 



