366 D. Prain — TTie Vegetation of the Coco Group. [No. 4, 



over intervening seas and in another way along continuous land. As an 

 example may be mentioned Oyrooarjpus Jaequinii, whose progress from 

 island to island is clearly a sea-assisted process, yet whose dispersal inland 

 when it is once established is greatly aided by wind because of its curious 

 dipteroid fruit. It might even be suggested that the wings of this fruit 

 may be sufl&cient to account for its transmission across intervening seas ; 

 but no one who has carefully observed the fall of its fruits is likely to 

 consider this possible. Another very pertinent instance is Terminalia 

 Oatappa, a species distributed by ocean currents over all the coasts of 

 the Andaman Sea, but which nevertheless occurs far inland as well as on 

 the beaches. The explanation of its inland dispersal is extremely simple, 

 for rats and frugivorous bats are extremely fond of the fleshy part of its 

 fruits while they leave uninjured the stone and kernel. Both these 

 animals are apt when disturbed while eating to carry off in their mouths 

 the fruit they may be devouring, ultimately dropping it some distance 

 from the place where the parent tree grew. But though bats occur in far 

 off lonely islands like Batti Malv and Barren Island, and though their 

 presence there indicates the possibility that animals of the kind may, like 

 fruit-eating birds, carry undigested seeds from one island to another, it 

 is clear, since they do not swallow the stones of Terminalia Gatappa that 

 they are not to be held respossible for the passage of that species across 

 intervening seas. The further spread of these species within new locali- 

 ties by agencies quite distinct from that necessary to account for their 

 initial appearance is, it will be admitted, amply demonstrated.* Other 

 examples are Pisonia aculeata and excelsa which are perhaps introduced 

 by the sea along these coasts. If they are, however, it is quite certain 

 that their presence inland may be amply accounted for owing to their 

 sticky fruits having become attached to birds or animals that have come 

 in contact with them.f 



* Residents in India are familiar with the treatment of " conntry-almonds " by 

 the large " flying-foxes ; " frnits carried off by them, and with a portion bitten out of 

 the fleshy side, may be constantly found dropped at considerable distances from the 

 trees on which the almonds grew. In Barren Island there is no doubt that the 

 fruffivorous bats which exist there are partly responsible for the same thing, and the 

 writer had an opportunity of witnessing the rats, which abound on that island, engaged 

 in the same act, these creatures having come down to the shore for the fruits that 

 are common there and when disturbed scampering off up gullies with frnits in their 

 mouths. 



f A striking instance of the possibility of their becoming attached to the bodies 

 of passing animals was witnessed by the writer on a path between Rangachang and 

 Ali Musjid in South Andaman in April 1891. Though some miles from the sea a 

 considerable number of Pisonia excelsa trees occurred at the place, and the path 

 was strewn with their fruits. A tree-snake was seen which had become entangled 

 in a fallen panicle of these so that all escape was impossible, its every movement in- 



