Fortunate Islands 
seeds must reach it, but they must also be able to grow 
upon its special soil. 
When Dr. Schimper visited a lava-flow (eruption of 
1843) in Western Java, he found that there were no trees 
upon it) The surface had been no doubt inhabited first 
by the blue-green algz (p. 20), and then by ferns which 
were still numerous enough. But though the rich 
tropical forest was in full vigorous development beyond 
the lavas, the only plants which had managed to settle 
upon its surface were orchids and other Epiphytes such 
as occurred elsewhere on the branches of the forest-trees, 
as well as a peculiar rhododendron, an insect-devouring 
Nepenthes, and others which belonged to dry or rocky 
habitats, 
All of them were adapted to obtain their nitrates and 
other food not so much from the soil as from the atmos- 
phere or from peaty material. 
Mr. Ridley has shown that when a bare tree branch 
in a tropical jungle is colonised by Epiphytes, the process 
is not unlike that which happened on Krakatoa. First 
blue algze establish themselves, then it is overgrown 
by mosses and ferns, and then the orchids and other 
flowering plants plant themselves upon it. 
The seeds of these orchids are exceedingly minute, 
and remind one of the dust-like spores of algze, mosses, 
and ferns; they are easily blown by the wind or 
carried by birds from branch to branch, or to desolate 
lava-flows, and even to lonely islands far out to sea. The 
stalks of these orchids generally hang down, or droop 
over the branches, so that the flowers are quite con- 
spicuous, but when in fruit the stalk is erect, so that the 
tiny seeds are shaken out of the capsules by any violent 
wind. 
On these lonely islands of the Pacific, then, one would 
expect a large percentage of the plants to possess small 
277 
