24 
Ridley .— The Distribution of Plants. 
belonging to the order Bignoniaceae, contains nine species, of which three are 
endemic in North Australia, one in Portuguese East Africa, three in the forests 
of southern India, one in the Irawaddy district, Burma, and one in Lower Siam. 
All these appear to be quite local plants in distribution. They are middle- 
sized trees with long-tubed, fragrant white flowers opening in the dusk and 
falling in early dawn, with long pods of winged seeds, the wings of the seed 
being thin and hyaline and longer than the body of the seed, so that they 
are easily dispersed by wind, as is the case in most plants of the order. 
But D. Rheedii is an inhabitant of tidal-river mud. Closely resembling the 
other Indian species in habit, foliage, and flowers, and most closely allied 
to the Burmese D. semdata ; it differs most remarkably in its seeds. 
Instead of having thin hyaline wings longer than the body of the seed at 
each end of it, the seed has at either end a short, oblong, corky prolongation, 
quite unsuited for wind dispersal and quite unlike any other Bignoniaceous 
seed. When dropped from a height it falls straight to the ground, while 
a winged seed of D. semdata flutters, rotating as it goes, to a considerable 
distance. By the shortening and thickening of the wing of the seed it has 
been adapted for sea-dispersal. 
As has been mentioned, the other species of the genus are confined to 
limited areas, but this species occurs in mangrove swamps and tidal rivers 
all round the Bay of Bengal as far as Ceylon, and to Travancore and 
Malabar on the west coast of the peninsula along the coasts of Burma, the 
Andamans and Nicobars, the Malay Peninsula and the Malay Islands 
from Sumatra to New Guinea, and the Philippines to New Caledonia and 
the Solomon Islands. 
The distribution of this plant shows clearly the superiority of sea- 
dispersal, both in time and distance, over dispersal of winged seeds by wind, 
and, further, as all the other plants in the order have the peculiar thin¬ 
winged seeds possessed by the other species of Dolichandrone , we may 
certainly assume that D. Rheedii is derived from one of the thin-seeded 
species, probably D. semdata , Seem., of the banks of the Irawaddy, as it so 
closely resembles it that specimens of the two plants have often been taken 
for each other. The seed-wings in D. Rheedii persist, but have been con¬ 
verted into shorter, thick corky floats. This species is therefore younger in 
time than* the other winged-seed species, yet its distribution is far wider 
than that of any other species: it is in fact another case which militates 
strongly against the age and area hypothesis. 
Most of the large orders of plants contain one or more species whose 
fruit is adapted especially for sea-dispersal; usually one species only occurs 
in a large order, the rest being inland plants having no such adaptation. 
In these cases it will be almost invariably found that the distribution of the 
sea-dispersed plant is very much wider than that of the inland species. 
Thus one may instance Calophyllum (Guttiferae), of which we have twenty- 
