402 D. Prain — The Vegetation of the Coco Group. [N'o. 4, 



Of tlie above, Gyclea peltata, Ab7-us precatorius, Acacia pennata, Al- 

 hizzia Lehbeh, Albizzia procera, Ipomcea Turpethum, and Pollia zorzogon- 

 ensis occur in I^arcondam. With few exceptions they are widely distri- 

 buted species ; five are cosmopolitan, one extends throughout the Eastern 

 Hemisphere and thence to Australia and Polynesia, two extend from 

 Africa and Asia to Australia, one occurs in Asia and Africa, one in 

 Asia and Australia ; only eight species, or less than half of the group, 

 are confined to South- Eastern Asia. Of these latter, three are confined 

 to the countries east of the Bay of Bengal ; two of them, Panicum 

 javanicum, and the particular variety of Pollia zorzogonensis that occurs, 

 are moreover distinctively Malayan, as opposed to Indo-Chinese, plants. 

 On the other hand one species, Dendrocalamus strictus, is as distinctively 

 an Indian or Indo-Chinese plant. 



We have now in conclusion to consider the "remanent " species, a 

 list of which is given in the subjoined table ; in a few cases where in- 

 troduction is remotely possible the agency that may have been respon- 

 sible is indicated. 



" where they happened to alight, to make way for the keeper or his men." {Hume ; 

 Stray Feathers, vol. ii, p. 113-4.) 



In this passage we have all the evidence that is required to show that not merely 

 the usual visitants but even unusual ones may at times be driven to, or seek shelter 

 on these islands when in an exhausted or injured condition, and to show that this is 

 as likely to happen to grain-, pulse-, and seed-eating species as to fruit-, or fish-, 

 crustacean- or mollusc-eating ones. It is of little moment that the fish- or crastacean- 

 eatino- species must always be the more numerous, if we know that species of the 

 other kind are at any time driven to the islands in this state. All of these exhausted 

 and injured creatures certainly do not recover or escape destruction. As regards 

 those that simply die, when the insects that abound have eaten all but their bones, 

 the seeds that may have been contained in tbeir crops must fall aside and may 

 terminate and grow. And as regards those that are killed and devoured it would be 

 remarkable if a few of the seeds in their crops did not thus fall aside and obtain an 

 opportunity of germinating. 



Even if no other creature that exists in these islands were capable of, or likely 

 to catch and eat such exhausted birds, the presence of a large lizard— Ht/drosaurus 

 salvator— which is very common, is sufficient to account for the destruction of 

 many of them. During our visit to Little Coco one of the officers of the " Investi- 

 gator " shot two Garpophaga bicolor — right and left— by the side of the lagoon near 

 the south end of the island. Before his attendant could reach the birds, which had 

 fallen among the Fandanus bushes of the sea-fence, one of these lizards had 

 already eaten all but the wings and head of one pigeon and had torn open the breast 

 of the second ere it could be interrupted in its feast. On a previous occasion 

 a, Hydrosaurus was killed as it was apparently about to seize a Teal that had just 

 been shot and had been laid down on a rock in the same lagoon. On opening up 

 this creature its stomach was found to contain already a large land crab, two fishes, 

 and a quantity of grassy roots (apparently those of Scirpus subtdatiis). 



