52 
Tansley and Fritsch. 
cases the possession of a peripheral floating tissue) is not in the 
first instance to be regarded as a direct adaptation to sea-transport, 
since allied species of the same genus which live inland usually 
possess the same character developed to a less extent. The coast- 
species therefore presumably owe their wide distribution to the fact 
that the fruits or seeds of their ancestors were able to float for 
a considerable period because they possessed a character which 
had some quite distinct origin and significance. 
This character helped them to colonise localities inaccessible 
to their accidentally less favoured congeners, and, under the stress 
of natural selection, has been increased and specialised in relation 
to the peculiar habitat thus adopted. Schimper gives an interesting 
account of his direct observations on the drift at Tjilatjap in Java, 
and we propose to add here some observations on the drift seen at 
one or two spots of the Ceylon coast. 
By far the best example of drift on the actual sea front was 
seen at Kalutara (Oct. 28, 1900.) The river here, the vegetation 
of whose banks has been fully described above, flows into the 
sea very obliquely, a long sandy spit separating the estuary from 
the open sea. This spit is less than a hundred yards broad, 
and its centre is occupied by the regular Pes-caprae formation. 
Fig. 16. Fruit of Heritiera littoralis. natural size. 
On each side of this vegetation, between it and the water, the sand 
was largely covered with drift. Starting from the sea, the sand 
sloped pretty steeply up to a ridge which evidently corresponded 
to the high-water mark of normal tides. Along this ridge, or more 
often, along a second ridge, probably reached only by the highest 
tides, there was a thick line of drift, forming in places a dark 
blackish brown humus of disintegrated vegetable debris, several 
inches deep, among which were numerous fruits and seeds. The 
drilt was composed of parts of plants, mainly twigs, leaves, and bits 
