400 
eastern line. The species that are common to these islands and to 
Ceylon are more difficult to explain. If we felt certain that they are 
species of distinctively Ceylonese type and that they occur, out of 
Ceylon, only in these islands, we might suppose that Ceylon birds are 
occasionally driven by storms as far as the Coco Group and consider 
the dispersal of the seeds of such species as one of the indirect sequels 
of cyclones of unusual severity. The birds even need not be different, 
as regards species, from those commonly found in the Andamans ; they 
need only be individuals that have followed the western instead of the 
eastern line of migration southward, and that under exceptional cir¬ 
cumstances have passed directly from one line of migration to the other, 
carrying in their crops seeds or fruits that are characteristic of the line 
of migration from which they have been driven. If the species are 
not of Ceylonese type, their occurrence both in Ceylon and the Cocos 
may, as has been said already, only indicate that they have been brought 
directly from Malaya or Australia by southern birds that migrate to Cey¬ 
lon as well as to the Coco Group but do not go as far north as peninsular 
India. 
The remaining sub-group consists of species with seeds or fruits 
that are eaten by birds of different kinds, not for the sake of any pulpy 
portion, but on account of the nutritious properties of the whole fruit 
or seed. We have to realize that the dispersal in this case is not, as in 
the case of pulpy fruits the seeds of which are afterwards voided, an 
oi’dinary circumstance, inasmuch as the seeds are eaten for their own 
sake and are of necessity digested by the birds that eat them. But 
though it is not perhaps a common occurrence—the numbers of migrat¬ 
ing grain- or seed-eating individuals considered—for newly-arrived 
birds to be killed, there is no doubt that a certain proportion, tired out 
by their long flight, must fall victims to raptatorial birds immediately 
on their arrival, the grains or seeds that their crops may contain 
falling aside and possibly germinating. Besides this means of introduc¬ 
ing such species, and, even if the results be slight, it must nevertheless 
be in constant operation, there is the further possibility of similar species 
being introduced during severe cyclones, owing to birds that have been 
driven to land being captured and devoured, while exhausted by the 
buffeting of the tempest, by birds or beasts of prey. In this way not 
only the grain- or seed-eating species that ordinarily visit the islands, 
but species both of this and of the fruit-eating class that do not usually 
reach the group, may conceivably arrive and as conceivably bring with 
them the seeds of plants that birds which are normal visitants have 
no opportunity of meeting with or may not care to eat. It has to 
be admitted, however, that species for which this mode of introduction 
210 
