ON INVESTIGATIONS IN THE INDIAN OCEAN. 207 
containing much phosphoric acid and apparently apatite. Bones, the 
teeth of elasmobranchs, and remains of other organisms, at present 
unidentified, were found both in the calcareous rock and in the brown 
‘conglomerate.’ A small area near the basin was covered with 
pinnacles of the ‘ shell rock’ as at Esprit. 
Specimens of these rocks were sent home in advance, and some 
were sectioned and reported on as ‘volcanic glasses’ in the issue of 
‘ Nature ’ of May 13, 1909. 
There is little doubt but that this identification is entirely erroneous, 
for since my arrival I have myself examined them briefly, and find that 
almost all of these strange rocks, with the exception of the ‘ shell 
rock,’ are largely composed of calcium phosphate, the colouring being 
apparently due to a considerable percentage of iron. 
Arrangements are being made for the proper examination of these 
rocks, the formation of which is at present shrouded in mystery, but 
which, it is hoped, will eventually throw considerable light on the early 
history of Aldabra. 
The four passes into the lagoon are interesting, and perhaps give a 
ciue to its formation. They have usually deep central channels, with 
reefs on either side. Small rock islets are present on these reefs, and 
it appears certain from their existence that the passes are steadily 
increasing in size, and that their reefs are really the remains of the 
land kept up to low-tide level by growing coral. Live coral extends for 
some distance into the lagoon, there being in all cases a luxuriant bed 
just inside the pass. At the mouth of the pass all corals are largely 
encrusted with Lithothamnia, and further seawards many are com- 
pletely killed by these alge. 
Besides the existing passes, it should be noted that there seems a 
likelihood of at least three more being formed—at Camp Frigate the 
Mangrove swamp extends right through the island to the sea, and no 
doubt a certain amount of water already traverses the land at that 
point. In Polymnie Island, at one place the swamp is within 100 yards 
of the shore, and a pass will probably be formed in time; at Dune Jean 
Louis there is only a quarter of a mile between the sea and the swamp, 
and if the lagoon erosion continues, no doubt Main Island will be 
divided at this point. It is worthy of note that fresh passes seem always 
to be formed by lagoon erosion, and not from the seaward side. 
The lagoon itself is very shallow, and the bottom sandy in the 
middle, changing into fine mud as it approaches the mangrove swamp. 
Eyerywhere one is forcibly struck by the extent of the erosion in the 
lagoon. Judging by its maze of small islands and mushroom-shaped 
rocks, at least one-third, or even more, of the lagoon can be shown to 
have been land at one time. At spring tides the amount of fine mud 
carried out to sea in suspension is very large, and it is obvious that the 
lagoon is still growing in size. There is some difficulty in accounting 
for the rapid transformation of the rock into mud, as boring animals are 
not common. 
As regards the vegetation, it is impossible to say much until the 
specimens collected have been worked out. The mangrove swamps 
Swat 
