a oe that ean thrive in diffe 
115 
of ferns and still lower plants. Very remarkable was the disco- 
very that the bare surface of the eruptive matter belched forth 
by the vulcano was carpeted with a slimy layer; it proved to 
consist of blue-algae, which are coated with a viscous layer 
that easily retains water, and so already contributes to the 
disintegration, the breaking up and weathering of the volcanic 
matter. At the same time this moist coating afforded a wel- 
come opportunity for the development of fern-spores which 
were not slow to avail themselves of the chance. Both the 
blue-algae and the ferns were blown across as spores by the 
wind. Of course the pioneer-ferns were only such species as 
were used in their natural homes also to open sunlight and to 
a non-waterretaining soil, such as the silverfern (Ceropieris 
calomelanos Und.) and other kinds. We must therefore represent 
ourselves how in 1886, three years after the eruption, the 
coast of Krakatau was furnished with littoral plants, and more 
in the interior with algae and ferns. A few grasses and bushes 
were also to be met with already here and there. Investigations 
have however disclosed that there are also other plants, whose 
seeds and fruits can float in the seawater for some length of 
time without losing their germinative power; they also will 
therefore be washed ashore and can then develop as part of 
the new strandflora; but gradually they are killed, chocked or 
ousted by the real beach-plants. 
Now for several years research stagnated, till Prnzic came 
with some other naturalists among whom was also Treus, eleven 
years later, in 1897. They found the flora much enriched, and 
what was most remarkable was that plant communities had 
already been formed. By plant-communities are meant groupings 
often belonging to different families of the vegetable kingdom 
all adapted in combination and living under special conditions — 
in which those plants | are wont to occur. Now on Krakatau 
the building-up of similar formations had already commenced, 
_ @ process witht ie ll Sp a3 
| Ann. Sard. Bot, Bates. Vol. XXX. i as 
poe ogress. Of course the different for- a 
mations are not always to be rigidly kept apart; on the border- = 
lines they will overlap and in — — and there are ‘Plaats aa 
™~ 
