760 NATURAl. HISTORY OF ALDABRA ISLAND — AHBOTT. 



of the lagoon. There are hundreds of other smaller islets, varying 

 in size from a few acres to a few square yards. 



The atoll is entirely of coral formation. Darwin, in his " Coral Reefs," 

 relying iirincipally on the reports of Oapt. Moresby, did not regard it 

 as a true atoll. The rocks of which it is composed were said to be " vitri- 

 fied.'" The rock certainly resembles lava in its outward appearance; 

 but it is easily broken, and the fracture displays a white interior and 

 numerous fossil corals that are in a scarcely altered state. The rock 

 gives a peculiar ringing sound wlieii struck. 



The principal difference biitween Aldabra and other coral islands is, 

 that it seems to be of very ancient ibrmation and has undergone an 

 elevation of 15 to 20 feet. The island is flat, composed almost entirely 

 of naked coral rock, rough and Jagged, completely honeycombed in 

 every direction with pits and iissures. Scarcely any soil exists, except 

 ing where a small quantity of rich nu)ld, formed by decomposing coral, 

 has accumuhited in hollows of the rock. The sea face is an overhang- 

 ing clilf of rock, but iu a few places, especially on lie Picard and ou 

 the west coast, are sandy beaches and h)w sand-hills. Upon the south 

 coast are Dune Jean Louis and Dune du Meche, sand hills, which reach 

 ().") feet abo\e sea level — the highest i)oints in Aldabra. 



Nearly the whole surface is covered Avith a dense, almost imi)ene- 

 trable scrub of tangled bushes. No large trees now exist excei)t the 

 mangroves, which attain a height of (10 feet and a diameter of a loot 

 or more. Formerly some large trees existed, as shown by the decay- 

 ing stumps and fallen logs, occasionally 3 feet in diameter, still to be 

 found ui)0u lie Picard. In some places the larger mangroves are dead 

 over areas of several acres. The disappearance of the larger trees can 

 only be attributed to a diminution in the rainfall. 



The supi)ly of fresh water is very scanty, only obtainable in hollows 

 in the rock, except at one place near the southeast corner of Grande 

 Terre. Here there is a sort of spring, filling a hollow in the rock 6 

 by 2 feet, and 5 feet deep. This seems to drain a considerable area, as 

 the level of the water can not be appreciably lowered !»>■ baling out. 

 The water is of poor (piality. 



The rainfall is scanty and very irregular. Sometimes many months 

 elapse during which not a drop of rain falls, and, on the other hand, C 

 inches have been registered in a single night. 



The islets in the lagoon aie of very i)ecnliar form, generallj' more or 

 less mushroom-shaped. The level of their flat summits is a few feet 



level of present reef. 

 Fig. 1. — Section of au islet in tlie lagoon. 



above that of spring tides. They are» evidently the remains of the 

 ai!.cient floor of the lagoon. All the other parts having been cut away 

 by the action of the water, the sides of all are undermined, and the 

 smaller frequently present a very perfect mushroom-shape, as shown 



