1891.] B.Train— The Vegetation of the Coco Group. 287 



Little Coco consists of several ridges the highest having an elevation 

 of 200 feet. The ridge jungle is much as in the other islands, but the 

 level land is more largely composed of a basis of coral-shingle than is 

 the case in. the other two islands and the undergrowth is not quite so 

 dense as in the level land on Great Coco. The coco-nut fringe is quite 

 as uniform as in the Great Coco, but there is only one point, — at the head 

 of a shallow bay in the middle of the west coast, — where the belt is as 

 much as thirty yards wide. Daring his visit to this island the writer 

 was able to cut his way from west to east across the highest ridge ; to 

 cross in another part along more level and frequently swampy ground ; 

 to work through a lagoon that occupies the south-western part of the 

 island, and to skirt the whole coast on two different occasions. 



The islands have all the physical features of the Andaman islands 

 of the main chain as opposed to those of the Archipelago lying to the 

 north-east of Port Blair ; the rocks indeed recall at once those of Ross 

 Island and of the shores of Port , Blair in South Andaman. They are 

 also equally like those forming Diamond Island, off the Arracan coast 

 at the mouth of the Bassein river and, as in these localities, are best 

 seen at points where the inland ridges end in abrupt headlands or are con- 

 tinued as long reefs exposed wholly, or in part, at low-tide.* Such reefs 

 not infrequently rise into outlying islets. These islets are some distance 

 from the main island, and are bare and rocky, or jungle-clad, according 

 to size and exposure, those off the west coast being all very bare. The 

 bays between the headlands are mostly wide and shallow, and are filled 

 up, except opposite the mouths of creeks, with an accumulation of coral 

 debris that becomes at times banked up, causeway-like, between the shore 

 and an outlying island; these causeways are in some instances becoming 

 stocked with the mangrove-vegetation of the neighbouring creeks. 



The floor of these shallow bays is remarkably flat and uniform and 

 is, at the sea-edge of the bay where the reef ends, generally rather 

 shallower than it is within, so that at low-tide each bay consists of a 

 long shallow pool, one to two feet deep, separated from the sea itself by a 

 long low bank of exposed coral. The bottom of such a pool is usually 

 covered by a close meadow of Gymodocea cillata, but though this species 

 is so common there seems to be no other marine phanerogam present. 

 AlgcB, too, are remarkably inconspicuous, being of small size and very 



* For further notices of the physiography of the islands the reader ia referred 

 to Alcock ; Nat. Sist. Reports in Hoskyn, Administration Reports of the Marine Survey 

 of Indii 1889-90, pp. 14, 15; 1890-91, pp. 11, 12; where also notices of the fauna, 

 particularly marine, will be found. In Hume ; The Islands of the Bay of Bengal in 

 Stray Feathers, vol. ii, pp. HI — 119, an account of these islands will also be' found ; 

 there the ornithology of the group is exhaustively discussed. 



