264 N. ANN AND ALE. 



There is no tide at Barkuda and We have no evidence of the existence of any- 

 constant diurnal movements of the water. Distinct occasional changes in level are, 

 however, often due to the wind. A steady breeze from the south-west drives the 

 water of the Rambha Bay past the island and often causes the level to sink by two 

 or three inches. As a breeze of the kind as a rule sinks in the evening the water 

 returns at night and is higher on a calm morning than it is after a breezy day. On 

 the other hand I have known the level to rise at least five inches in the course of a 

 day owing to a storm north-east of the island which banked the water up round it. 



Major R. B. Seymour Sewell, I. M.S., Surgeon Naturalist, Indian Marine Survey, 

 has described the hydrography of Rambha Bay in a paper to be published in Vol. 

 V of the Memoirs of the Indian Mtiseun. This paper will throw much light both 

 on these phenomena and also on the piling up of dead weed on the shore of the 

 island and the consequent formation of free sulphur as a result of the action of a 

 bacterium on the decaying matter.' The chain of events is of faunistic importance, 

 because it has prevented the existence of a true marginal fauna of amphibious 

 molluscs and Crustacea such as is frequently found in similar situations. 



II. PALEONTOLOGY. 



The rock of Barkuda is non-fossiliferous but the soil of its interior and the sand 

 of its shore contain large numbers of moUuscan shells that may be classed as sub- 

 fossil and utilized to throw some light on the age of the island. Before describing 

 the shells I must refer to the store of specimens now being laid up for future palaeonto- 

 logists by two pairs of sea-eagles {Haliaetus leucogaster) that have made the island 

 their home. These birds, as is a common habit of the species, nest in large Banyan 

 trees and do not desert their nesting-place when the breeding-season is over but 

 remain constant to it throughout the year. They catch their food in the lake and 

 bring it home to devour. Their prey consists mainly of three items, the sluggish 

 water-snake, Chersydrus gvanulatus, the swimming-crabs, Scylla serrata and Neptunus 

 pelagians, and the fish, Triacanthus hrevirostris, perhaps the most abundant in the lake. 

 Fish of several genera of Siluridae are also captured. The snakes the birds swallow 

 whole as a rule, but the ground under the tree in which they live is covered with the 

 remains of fish and the broken claws of crabs, and scattered bones and claws that 

 they have dropped can be found on almost any part of the island. Triacanthus is a 

 laterally compressed fish with a very solid skeleton and with three long and stout 

 bony spines that have given the genus its name. One of these is situated in the dorsal 

 fin, while one on each side of the lower part of the body represents the pectoral fin of 

 that side. To give support to the lateral spines, which can be * locked ' in such a 

 way that it is impossible to press them together against the sides without breaking 

 them, the pectoral girdle is greatly strengthened, while the skull is also remarkably 

 hard and bony and has in lateral view a curious superficial resemblance to that of a 

 land-tortoise. The three spines, the skull and the pectoral girdle are connected in 



I Annandale, Proc. As. Soc. Bengal, XIV (n.s.) (tqi8), pp clxix-clxx. 



