ON THE DISPERSAL OF PLANTS. 299 
little that does not push my argument home concerning the 
stocking of this cora! atoll with its plants. I have described 
its original vegetation, before man’s disturbing influence 
began; I have demonstrated by observation and experiment 
the facilities for dispersal possessed by the plants, and have 
pointed out the reason of the absence of certain familiar 
species; I have dwelt in detail on the seeds and fruits brought 
by the currents to these islands, and have ascertained the 
direction in which they drifted; I have followed these ocean 
waifs to their principal home on the coasts of the Indian 
Archipelago, and have observed the parent plants growing 
on those shores; lastly, I have referred to an instance of an 
island in these seas, absolutely bare of vegetation, where the 
process of plant-stocking has been carefully observed, and 
where amongst the first of the flowering plants to grow from 
the seeds and seed-vessels stranded on its shores were those 
of familiar Keeling Island species. This method of itself 
would be conclusive, as long as the facts are trustworthy, 
and for this I can safely vouch. 
CONCLUSIONS. 
The principal points of this paper may thus be summed 
up :— 
(1.) The evidence goes to show that the cocoa-nut palms 
established themselves on these islands before their occupa- 
tion by man. 
(2.) Several coral island plants, not recorded by Darwin 
in 1836, but which there are good reasons for believing 
originally existed in these islands, occur in my collections 
(sce page 272). 
(3.) Crabs, by eating the seeds stranded on the beaches, 
are important agents in preventing certain common littoral 
plants from establishing themselves on these islands 
(page 276). Thus it has happened that not more than 
one-fourth of the numerous seeds and seed-vessels brought 
by the currents have found a home on this atoll. 
(4.) It is well known that the familiar fruits of Barring- 
tonia speciosa, Calophyllum inophyllum, Terminaha Catappa, §¢., 
will float for a long time in sea-water unharmed ; and the 
experiments recorded on page 287 show that several other 
coral island plants will germinate after floating from six to 
seven weeks in sea-water. 
(5.) The flora has been mostly derived through the agency 
of the currents from the, adjacent coasts of the Indian Archi- 
VOL. XXIV. Y 
