204 



NA TURE 



[December 17, 1908 



1HE PERCY SLADEN TRUST EXPEDITION TO 

 THE INDIAN OCEAN. FURTHER 

 EXPLORATIONS.' 

 '"pO complete the work of the above expedition I left in 

 ■'• June last for the Seychelles, accompanied by Mr. H. 

 Scott and Mr. J. C. F. Fryer as naturalists. We arrived 

 there on July lo, but were unfortunately detained on Long 

 Island, the quarantine station, for ten days. The time, 

 however, was by no means wasted, as the island was fairly 

 representative of low-country coco nut cultivation, antl 

 contained a fairly rich insect fauna, mainly of introduced 

 species. We had boats also, and were thus enabled to 

 examine the reefs in its vicinity, collecting particularly the 

 sedentary organisms. 



On_ July 23 I sent Mr. Fryer to examine Bird and 

 Dennis, two islands seventy miles to the north of Mah^, on 

 the edge of the Seychelles 'Bank. He remained a fortnight 

 on each, examining them in all their aspects. The former 

 was barren, with a vast number of sea birds breeding 

 upon it, while the latter was planted in coco nuts. Both 

 proved to be formed entirely of coral material, rock and 

 sand, and both were found to be gradually washing away 

 into the sea. The surface reefs around them are insignifi'- 

 cant, and they gave the impression of being the remains 

 of reefs once widely extending along the north edge of the 

 whole bank. On his return Mr. Fryer at once started 

 down to visit the islands of Aldabra, Assumption, Cosmo- 

 ledo, and Astove, among which he is to work for 

 four months. These islands lie to the north-west of 

 Madagascar, and due west of Farquhar .'\toll, visited by 

 the Scalark in 1905. Assumption and Aldabra form the 

 most western group, the former being a crescent-shaped 

 bank, 3J miles long, and the latter a ring-shaped atoll, 

 19 miles long by 7 miles across, almo.st completely sur- 

 rounded by land, with a shallow lagoon. 



Astove and Cosmoledo form a second group, the former 

 a ring-shaped atoll, 2 miles long, completely surrounded 

 by land save for one passage to the south, and the latter 

 an atoll, oJ miles by 7 miles, lagoon 4 fathoms, ring much 

 broken up, with eight main islands. .'Vldabra Is well 

 known as the home of gigantic land tortoises. Some rock 

 from it in my possession contains abundant silica, a fact 

 which makes the accompanving abstracts from Mr. Frver's 

 preliminary report on its' a'djacent islands of peculiar 

 interest. 



On July 24, accompanied by Mr. Scott, I went to 

 Silhouette, the second highest island in the Seychelles, 

 2467 feet. We settled at 1600 feet, within the indigenous 

 jungle, of which a square mile still exists. Here Mr. Scott 

 remained until October i, obtaining a rich collection of 

 Its inse<:t fauna, which, allowing for the difference in the 

 size and position of the islands, appeared In be to a large 

 degree comparable in its nature to that of the Sandwich 

 Islands. The island itself is about 12 square miles in 

 extent, with rugged granite hills, and two bavs with fiats 

 covered with coco nuts. These owe their 'existence to 

 former fringing reefs, the level of the island having 

 changed to the extent of at least 30 feet within compara- 

 tively recent times. The coco nu't is grown up to 1200 

 feet, but the trees, as is also the case in Mah(5, are infested 

 with a fungoid disease, and do not bear well. 



.\fter a month in Silhouette I returned to Mah^, every 

 part of which I visited in the ensuing seven weeks. For 

 the most part I w^as occupied in examining its geographical 

 teatures, rocks, reefs, and jungles, and' in collecting its 

 plants, 'of which I obtained upwards of 2000 sheets. I 

 have little of general interest to add to mv report in 

 Nature, January 25, 1906. Mah(5 showed the same change 

 of level _ as Silhouette, and my former impression, that 

 there might have been a more' ancient elevation of about 

 2on feet, proved to be erroneous. Only about 2A square 

 mijes of the indigenous jungle are now left, and that is 

 being graduallv affected by the reckless destruction of its 

 larger trees. Mr. Scott is to collect its insects during 

 the next four months, but it can scarcely be supposed that 

 more than a small percentage of its indigenous fauna 

 still manages to survive. This iungle area lies in the 

 centre of the island, the north and south thirds of which. 



reports see Nature, April 13, .August 

 ecember 2t, 1905, and January 25, 1906. 



October 5, 



NO. 2042, VOL. 79] 



being almost completely deforested, have become physio- 

 logically dry, to the almost complete destruction of their 

 indigenous flora and fauna. Much of the land, too, has 

 been destroyed by cassava planting, which is followed on 

 the steep hill-sides by the washing away of the soil, con- 

 verting them into bare glacis. 



J. St.ANLEV G.ARDINER. 



.Astove, where we arrived first, is an atoll about 2 miles 

 long by li miles broad; so far as I could tell it is entirely 

 composed of elevated coral as a basis, with sand distributed 

 in various places. The seaward beach in most parts is 

 formed of sand, but in places coral rock forms small cliffs, 

 showing very evident washing away. There is only one 

 pass. This is narrow, and from the present rate of wash- 

 ing away must be of fairly recent date. On the westward 

 side of the pass arc " coral rock " cliffs, while on the 

 east a good deal of piling up of big blocks has occurred. 

 These blocks are all of coral rock, not dead corals. They 

 appeared to have come partly from the present land 

 (washed out) and partly from the reef, which, so far as I 

 could see, was composed of coral rock only, and was not 

 ordinary dead reef such as I saw at Bird Island, to the 

 north of the Seychelles. The coral rock interested mc 

 very much. In places one could see regular fields of coral 

 with all the corals in their natural positions, while here 

 and there are small holes, 5 feet to 6 feet deep, with sides 

 all encrusted with corals exactly as they grew. There was 

 absolutely no question of piling up. The whole place is 

 evidently exactly as it was underneath the sea. In parts 

 of the island the corals have been more " metamorphosed " 

 into rock (not retaining their original structure), but 1 

 could trace no correlation between the occurrence of this 

 rock and its position on the island. In the north of the 

 island there ■■iri' some dunes about 50 feet high, purely of 

 wind formation. I dug a hole on the landward side of 

 one, and found guano underneath with a sort of shingle 

 below. I think the north-west monsoon must have been 

 much stronger once, or cyclones more frequent, to have 

 driven this shingle inland. 



The lagoon is very shallow: bottom of fine coral (?l 

 mud, which makes the whole lagoon white, and forms a 

 froth all round the shore. There are one or two small 

 islands near the pass. There will probably be two mor. 

 " passes " formed soon, one to the N.N.E. and another to 

 the S.E. The reef to the west is sandy, with little living 

 coral. It falls directly to " no bottom " without any 

 slope, so that a ship cannot anchor, but is moored by 

 lines to the reef. I searched the reef opposite the pass, 

 but found no living corals, only a piece or two of recently 

 killed coral thrown up. As I have said before, it is rock, 

 and not reef. In one place it drops lo the sea. in terraces, 

 forming small waterfalls as the tide goes out. 



The next island we went to, Cosmoledo, represents a 

 further stage in the .Astove condition. There are only a 

 few islands left. The settlement is on Menai, with a 

 fishing station on Wizard. The rock is the same as at 

 .Astove, but naturally there is very much more sand. -All 

 round the reef the remains of islands stick up like rock 

 mushrooms. On Menai Island there is a mangrove swamj) 

 on the lagoon side, trees mostly small, as the larger have 

 been cut for their bark, used for tanning. There is a 

 sand-dune, with some very old Casuarina. Everything 

 was very dead and burnt up, and not nearly so attractive 

 as at -Astove. I went also to Wizard, Goelette, and North- 

 East Islands. Wizard is very sandy, but has the distinc- 

 tion of having three good wells. I got a positive measur.- 

 as to the rate of erosion, 15 yards having gone in th.- 

 last fifteen years (measured by Spur's old house, which is 

 now in the sea). Goelette is of no interest except as con- 

 taining some guano : North-East Island I only stayed on 

 for a short time ; all the guano has been dug, and is now 

 being taken to the Cape. I caught a large lizard there, 

 which occurs on none of the other islands, but it seems 

 to me to be onlv a variety of the universal small one. 

 We had wretched weather, very squally, and I seem to 

 have spent most of my time in a whale-boat trying to 

 get to somewhere. 



The next island. Assumption, is the most interesting 

 of the three; it is not an atoll, and the settlement is a 

 new one, so that everything is untouched. The rock is 



