THE ALDABRA ISLES 21 i 
detailed maps were got up later by Dr. Abott, an Ameri¬ 
can, and Dr. A. Vbltzkow, a German, the latter of 
whom sailed from North Madagascar to Aldabra for the 
purpose of investigating its Natural History. 
Aldabra, according to the description of the last- 
named very trustworthy investigator, is an oval atoll 
with a major diameter of 19 miles and a maxini'um 
breadth of yi miles. It has three small entrances, the 
largest of which, that on the north side, is only half a 
mile wide. The lagoon is shallow and is partly dry at 
low tide. The belt of land projects from 3 to 6 ft. 
above the limit of high tide and is steep towards the 
sea; its breadth varies from half a mile to 3 miles. 
The geological structure everywhere shews us 
only recent coral. Whether the underlying rocks are 
volcanic or composed of primitive rock, cannot be deter¬ 
mined owing to the fact that these rocks do not 
come to the surface. Possibly we have to do with 
a pinnacle approaching the surface with an atoll resting 
upon it. The island seems to have been originally 
a flat reef, gradually worn and washed away in the 
middle during a slow upheaval. Voltzkow remarks 
that numerous islets of mushroom shape occur 
in the interior of the atoll. I look upon these as 
secondary formations, for the mushroom shape, which 
I also have observed in the erythraea-like reefs, agrees 
with the on the part of the coral-polype in its need 
for light. 
o 
There are two good-sized islets in the interior of the 
lagoon: the long Euphrates Island, which is 3 miles dis¬ 
tant from the inner edge of the belt, and Cocoa-Nut 
Island in the east of the lagoon, which, as its name indi¬ 
cates, is covered with cocoa-nut palms. 
There are no brooks or rivers, so that it is necessary 
for the settlers to collect during the rainy season sufficient 
water for their needs and to store it in reservoirs. On 
