Ridley . — On the Dispersal of Seeds by Wind. 363 
Coral atolls, like Cocos Island, are also deficient, but with other islands it 
is different. 
In Krakatau, after the eruption in 1886, Dr. Treub found that out of 
2 6 plants 11 belonged to this group ; Penzig, revisiting it in 1896, found 62 
plants, of which 16 are powder-seed plants. In Christmas Islands out of 
170 I found 33. South Trinidad out of 11 has 4; the Admiralty Islands 
out of 1 18 have 41 ; Kerguelen out of 27 has 6 ferns and lycopodiums, but 
it is also accredited with 154 mosses and 26 hepatics ; St. Paul and 
Amsterdam Islands have 19 powder- seed plants (exclusive of cellular 
cryptogams) out of 33 plants. 
These facts are, I think, sufficient to settle to a considerable extent the 
part played by the wind in seed-dispersal ; of the three classes of fruit and 
seed modified for wind-dispersal, that of winged seed and fruit is the 
slowest. The species migrate very slowly and are, usually at least, unable 
to cross any large tract of sea by this means alone. Plumed seeds and 
fruits, though easily and probably quickly disseminated over open country, 
for which they are most suited, are liable to be stopped in their migrations 
by dense forests. They can, at least occasionally, cross successfully large 
areas of sea. 
Powder-seed , on the other hand, has the most rapid transit probably 
of any form of seed, and is most widely diffused. 
HENRY N. RIDLEY. 
c c 2 
