154 Couthouy on Coral Formations 



ged as to render walking, or to speak correctly, clambering 

 over them a difficult and fatiguing task. At the foot of the 

 cliff, back of the lowland, are frequent caverns, from whose 

 roofs depend numerous stalactites from the size of a pipe-stem 

 to that of a man's body, the little drops of water at their ex- 

 tremity sparkling like so many diamonds wherever the light 

 from a crevice falls upon them. The floors are also covered 

 with stalagmitic incrustations in every degree of hardness, 

 and assuming a great variety of forms. Those into which I 

 entered, descended for a few feet at an angle of about 30°, 

 like an arched vault, and then expanded into an irregular cir- 

 cular grotto, with a level floor, whose ceiling was from four to 

 fifteen feet in height. Some of these caves are capable of 

 holding at least three hundred persons. 



From one of the plains on the north side, where there is a 

 village with some two hundred and fifty inhabitants, a steep 

 ascent leads to the summit, which presents a broad table land, 

 declining a few feet toward the centre, where we may sup- 

 pose the lagoon to have been situated. Near the eastern ex- 

 tremity, a few yards from the bank, are two knolls gently 

 rising to a height of perhaps forty feet, which I presume to be 

 remains of the ancient fragmentary ridge, formed when this 

 table land was the surface reef, the main portion having been 

 undermined and worn away by the action of the surf on the 

 south-east or windward side during tlie period of elevation. 

 To this cause, I imagine, is also to be assigned the sloping 

 form of the island in that direction, while the sheltered lee- 

 ward side has preserved its original sheer descent. The 

 dense growth of forest, and tangled luxuriance of under 

 growth, prevented any minute observation during my hurried 

 visit, but I recollect that the whole surface of the table land, 

 and the ascent of the q\\S. for eighty or a hundred feet below 

 it, was covered with fragments of coral conglomerate, the 

 species imbedded in which were the same with those found 

 on the reef below. Indeed, the entire mass of the island is a 

 reef-rock in various stages of consolidation, the lower portion 

 approximating to a solid limestone, the cellular coralline struc- 

 ture being in some fragments hardly perceptible, and the 



