December. 2 1,- 1-905] 



NA TURE 



]S = 



considerable variety of other alga', but the edge and outer 

 slope were, as elsewhere, covered bv corals and nullipores. 

 The reef, however, to the west, where there is a flat 

 extending out for some miles with about 16 fathoms of 

 water, closely resembled similarly situated reefs in the 

 Chagos, but the greater variety of its organisms was 

 equally marked, though individual species were 

 common. 



Leaving Coetivy on September 25, w : e proceeded to a 



point about midway between Madagascar and Farquhar 



Atoll, both to ascertain the depth and the comp: 



ation. The latter was almost the same as a" Mauritius, 



situated 9 degrees to the south, while the depth, 1856 



fathoms, precludes the idea of anv close connection between 



the 1 wo localities. Farquhar, which 



we then visited, was (as, indeed, were 



all the reefs we subsequently saw) re- 

 markable for its almost completely 



covered " varech " reefs, both rim and 



lagoon. Its land attains a height of 



more than 70 feet, and is clearly oi 



the same formation as that of Coetivy ; 



it shows iim tin,- of elevation, and it 



has not been formed, as has been 



stated, by submarine deposits. The 



section of the reef also showed the 



outer slope to be quite similar to that 



of other atolls. 



From Farquhar we proceeded to 



sound between the chain of islands that 



extends between Madagascar and the 



Seychelles, and which would appear to 



indicate a line of former connection. 



Between Farquhar and Providence, 



32 miles, we found 890 fathoms, and 



between the latter and Alphonse- 



Francois, 155 miles, 2170 fathoms. 



while there were already soundings of 



052 fathoms between Alphonse and the 



Amirante Group, 46 miles, and of 1150 



fathoms between the latter and the 



Seychelles, 32 miles. As the depth on 



either side is only about 2300 fathoms, 



any connecting ridge is comparatively 



low and of doubtful importance. 



Providence was particularly interest- 

 ing, being simply a great reef, 28 



miles long by 7 miles broad to the 



100-fathom lines. Off it we took 



twelve dredgings, obtaining a rich 



fauna down to about 100 fathoms, 



below which the bottom was exceed- 

 ingly barren of life. From one dredg- 

 ing at 744 fathoms, 3 miles to the 

 ■ if the reef, we obtained about 



5 cwt. of stones, the largest about 



2 liit in diameter. "We have here no 



means of properly ascertaining their 



nature, but similar rock has not, so 



far as we are aware, been hitherto 



described off anv coral reef. It is 



almost entirely insoluble in acids, and 



is largely formed of different crystals, 



organic deposits practically not enter- 

 tto its composition. Some masses 

 looked like solidified ash or clay, while others appeared 

 rather like vulcanic bombs. All were more or less coated 

 with manganese, but we do not know its thickness, prefer- 

 ring to keep the specimens intact for proper examination on 

 our return to England. However, it is clear that the 

 of this rock in such a position will havi to be 

 carefully considered in connection both with the formation 

 of Providence Reef and with the existence of any former 

 land connection between the Seychelles plal 

 Madaga 



Pierre Island, 17 miles to the west of Providence Reef, 

 and with a depth of 1088 fathoms between, is peculiar in 

 having no fringing reef. It is simply an elevated coral 

 island, reaching to .1 height at present of about 30 feet, 

 surrounded by overhanging cliffs, so that landing is ex- 

 tremely difficult. Its rock is entirely coral. 



Alphonse and Francois are sandbanks on the rims of 

 two reefs, scarcely 2 miles apart. Both reefs are of atoll 

 formation, the lagoon of Alphonse (not shown in any 

 chart) being 3 to S fathoms deep and of considerable 

 size. 



The Amirante Islands are likewise sandbanks, no parts 



les~ of any being" more than 10 feet above the high-tide level. 



The hills represented in the separate enlarged plans of 



L> 'Arros, St. Joseph, and Desroches do not exist, and 

 probably owe their presence thereon to the imagination 

 of the draughtsman. 1 Desroches is really an atoll by 

 itself, lying 10 miles to the east, and being separated bv 

 a channel 874 fathoms deep. The rest of the islands and 

 reefs lie on a bank about 50 miles long by 20 miles 



SEYCHELLES 



-Ojrr„ . -c JO"! 1 " 



jC 



\\ 



Chart of the Indian Ocean between Madaga 



nd the Seychelle 



broad, with an average depth of about 30 fathoms. Eleven 

 separate reefs reach the surface, of which St. Joseph alone 

 has a lagoon, being really a small atoll with about 

 4 fathoms of water in the centre. With the exception of 

 Eagle, D'Arros, and Bertant, all tin- reefs lie on the edge 

 of the bank, but its edge is in most places covered by at 

 least 8 to 10 fathoms of water. Its slope is steeper than 

 is customary off coral reefs, no possible dredging ground 

 existing between 60 and 500 fathoms. 



All the islands of the Amiranti Group, with the exception 



'! Louise and Eagle, are now planted for cocoanut 



oil, but the indigenous vegetation still remains in places. 



The land plants and animals are almost the same as at 



Coetivy and in the Chagos, lie- additions due to the 



NO. 1886, VOL. 73] 



1 If we had had any idea of thi- 

 Cosmoledo and perhaps Aldabra. 



sho.ild have probably visited 



