FRYER— FORMATION OF ALDABRA, ETC. 405 



a narrow belt of Pemphis and then the swamp. The west of Main Island belongs almost 

 entirely to the " champignon " type of country, though at Couroupa a depression in the 

 rock (2 or 3 acres) has been filled with sand and supports a thick jungle. 



Picard Island has a narrow sandy belt along the south-west shore, on which the 

 settlement is situated, and behind is a varied jungle and certain peculiar rocks which will 

 be described later : the north of the island is largely Pemphis scrub, though opposite 

 Grande Passe is a small sandy area covered with old Casuarina trees, marked on the 

 chart as the "Admiralty Observation Spot." 



I have now indicated the general nature of Aldabra and in the next subdivision 

 I propose to deal in greater detail with its component rocks, from which even without 

 expert knowledge certain deductions of great importance can be made. 



Mr Hughes, of Bangor University, who has very kindly undertaken their examination, 

 writes (19th June, 1910), "All my work up to the present seems to support your theories 

 completely ; I cannot think of any other possible explanation," so his paper, which will 

 appear subsequently, may be expected to confirm the general principles of my deductions, 

 though no doubt he will modify many of my details. 



C. Structure*. 



The rocks of Aldabra fall naturally into two divisions : 



(a) those which have a high percentage of calcium in the form of calcium carbonate 

 and are obviously of coral reef origin ; 



(b) certain exceptional rocks in which the acid radical in calcium carbonate has been 

 replaced by phosphoric acid derived from guano. 



(a) Coralline Limestones. There are three classes of coralline limestone, of which 

 the first forms a zone encircling the atoll and constituting the outer portions of the 

 land-rim. It is characterised by being composed very largely of reef corals, which have 

 retained their structure very perfectly, and are in most cases in the position of growth, 

 a phrase which means that the different coral colonies are still in the same positions 

 as when alive and stand on the same bases. This rock therefore has been formed by 

 a reef flourishing beneath the sea. From it we can deduce with certainty that the land 

 has been formed by the elevation of a reef in relation to the surface of the sea, and 

 not by the piling of broken fragments during a long period of subsidence, as is demanded 

 by an explanation based on Darwin's theory. From the facies of the corals we can deduce 

 that before elevation the reef was existing in quite shallow water. (Plates 23, 24.) 



The outer portions of the whole atoll were searched diligently in relation to this type 

 of rock and it was never found to be missing; indeed the most striking impression was 

 given by these elevated corals, which on the face of the cliffs from top to bottom stood 

 out in a way which can be realised only from the illustrations, while on the top of the 

 cliffs so perfect were they, that it was difficult to believe they were long dead. 



* [Mr J. O. Hughes, who has been examining the rocks chemically, has kindly read through this section. 

 J. S. G.] 



SECOND SERIES— ZOOLOGY, VOL. XIV. 52 



