Fortunate Islands 
Suppose one of these navigator seeds has stranded 
on the coast of a desert island, then it has still various 
difficulties to meet with, If it is a mangrove seedling, 
it will only be able to take root and grow if it has 
happened on a muddy foreshore. It will perhaps ger- 
minate, but is sure to die if the coast is sandy or 
coralline. The usual associates of the mangroves can 
only establish themselves if there is at least the be- 
ginning of amangrove swamp. There are, for instance, 
none of the mangrove association in Tahiti, because 
apparently there is no place suitable for them. 
But if the newly arrived navigator plant is an ordinary 
coral strand plant, it will at once take root and begin to 
grow vigorously. Sometimes they are destroyed by 
crabs, which nibble off the young shoots, but this seems 
not to be a common casualty. 
One can see this happening with such a common 
navigator as Entada which, once established on the 
shore of an island, soon spreads itself all along the 
coast, and may even pass inland and be found straggling 
over the trees and bushes at a long distance from the 
shore, 
It has apparently happened that the descendants of 
some of these navigators have not only taken to an 
‘inland life, but have become so modified and changed 
as to constitute entirely new and peculiar species only 
found in one of these island groups. This seems to 
explain at least the various island species of Vigna, 
Premna, Canavalia, Erythrina, Sophora, and Ochrosia, 
for each of these genera has a navigating species. 
But one cannot entirely account for all the plants in 
South Sea islands by these explanations. Many are 
neither navigators nor descendants of them, nor are 
their seeds, so far as we know, carried either by wind 
or by birds. 
281 
