ON THE DISPERSAL OF PLANTS. 281 
‘pumice, pebbles, a little mould, and a very considerable 
number of the seeds of Towrnefortia argentea, with a few other 
small seeds. There were no trees in the vicinity that could 
have supplied the Tournefortia seeds, and it was evident 
that this log had been washed away from some other part of 
the coast where it had been lying for a long time near a 
Tournefortia tree. During unusually high tides and gales, 
sand, seeds, and pumice had been washed into the empty 
burrows of the Teredo; and after a time the log was swept 
away and stranded in its present position.’ I examined some 
other honeycombed logs, just beached, in which the Teredo 
was still fresh, though dead. No sand, pumice, or. seeds 
occurred in their cavities. 
Mr. Ross informed me that not infrequently, later on in the 
year than the time of my visit, he has observed the seeds of 
Pemphis acidula, Seevola Kenigit, and Triumfetta procumbens, 
sprouting from pumice on the beach, The pumice is gener- 
ally old pumice drift which has been washed up during gales 
under the littoral trees, where it becomes partially covered 
with sandy soil and leaves, and seeds often drop into the 
crevices. When such old pumice is carried off by an un- 
usually heavy sea and stranded on another part of the coast, 
the seeds often germinate, but the crabs soon bite off the 
shoots. I have never found seeds in recently-arrived pumice, 
though I have observed grass growing in the cells of pumice 
that has lain a long time on the ground. It is, however, 
easy to perceive from the foregomg remarks how logs and 
pumice that have been lying for some time on the shores of 
an island under the shadow of littoral trees, may often be 
swept away into the open sea and carry seeds in their crevices 
over a wide extent of ocean. 
The numerous Pandanus seeds that occur in the vegetable 
drift on these coasts im certain times of the year, at once 
explain how this tree originally reached these islands. The 
origin of some of the littoral trees does not admit of so easy 
an explanation. Take, for instance, Suriana maritima, which, 
in spite of the circumstance that it is found on nearly all 
tropical shores, both insular and continental, has only 
established itself here within the last 20 years. It is not 
included in the lists of Darwin and Forbes, and the Cocos 
Islanders take such an interest in the plants of their atoll 
that there is no reason to doubt the recent appearance of the 
shrub on the weather side of Gooseberry Island. Its seeds 
are small, rather soft, and are ill suited for drifting on the sea, 
since they sink after floating between two and six days. 
