FRYER— FORMATION OF ALDABRA, ETC. 427 



however it is either bare or (near the lagoon) is covered by a thick layer of dark 

 brown guano. The surface of the rock is almost as rough as the "champignon" of 

 Aldabra but, unlike it, was obviously composed of coral, which differed from anything 

 I saw elsewhere in the region by forming fields of unbroken colonies all in position of 

 growth, even the short spike-like form of Madrepora projecting from the surface as if 

 still on a growing reef. At Aldabra corals in position of growth were abundant on the 

 cliffs and cliff faces but never were they found seated on the surface of the ground as 

 if elevated but yesterday. There is no doubt therefore that Astove is an elevated atoll, 

 and in addition I deduced that no great rain-water denudation had occurred ; but 

 Professor Gardiner considers that the corals might still project undamaged on the 

 surface even after the island had been much reduced in level by denudation, a supposition 

 which is upheld by the fact that Aldabra, Assumption and Cosmoledo must all have 

 suffered greatly from this cause, and it is almost inconceivable that an island distant less 

 than a hundred miles should have escaped. 



Metamorphosed rock is scarce, but a little was discovered near the lagoon to the 

 south and as usual contains phosphatic inclusions. 



The eastern part of the land-rim was not visited but it was obviously narrow and 

 much heaped with sand. 



The lagoon is extremely shallow and is hardly more than a foot deep at low tide ; 

 the bottom is covered with a very fine white sand or mud, which on windy days is so 

 churned up that the whole lagoon looks white and the shores are covered with white 

 foam. As the tide descends all suspended material is swept to sea and the loss is thus 

 very large. Judging by this fact and by the form of the lagoon cliffs (where existing) 

 it is evident that the lagoon is increasing both in size and depth. Considering its 

 extreme shallowness and the absence of small islands and rocks, I am inclined to think 

 that the Astove lagoon is a primary lagoon the greater part of which existed at the time 

 of elevation. 



The single pass is narrow (150 yards), and even shallower than the lagoon, being 

 almost dry at low tide : the bottom is of coral rock, apparently metamorphosed, and 

 practically devoid of any living corals or debris, though the east shore of the pass is 

 somewhat piled with rounded lumps of coral rock. This pass is obviously of recent 

 formation and there is no doubt that a little time ago the lagoon was absolutely enclosed 

 except for subterranean connections, of which I remarked one in the north of the lagoon. 

 In future, more passes will undoubtedly form, the land-rim in places on the north and 

 east being very narrow. The reef outside the pass is obviously not a reef which has 

 grown up in situ but a ledge of elevated coral rock cut down to low water-mark ; owing 

 to the rush of water from the pass, it has not been covered either with coral or debris, 

 though it is somewhat encrusted with lithothamnia. There is a marked tendency to form 

 a series of descending terraces, down which the water from the lagoon rushes on its way 

 to the sea. The remainder of the atoll fringing reef is doubtless of the same character, 

 though it has been piled with coral debris and sand much in the same way as the 

 Aldabra reef. 



The time I spent on Astove was insufficient to properly investigate its structure 



