288 H. B. GUPPY, M.B., 
It might have been thought that I had proved that Keel- 
ing Atoll has been mainly stocked with plants by the waves, 
assisted by birds. It is, however, quite open to some to 
suggest that the arguments can only be clinched by 
taking an instance of an island that is absolutely bare of 
vegetation, and then following the process. This has been 
accomplished by Dr. Treub* in the case of the volcanic 
island of Krakatoa, which, as he conclusively proves, was 
entirely deprived of vegetation by the thick covering of 
fiery ashes and pumice that invested its slopes, from the sea 
margin to the summit, at the time of its great eruption in 
1883, forming a soil proved by chemical analysis to be com- 
pletely sterile. When he visited the island nearly three 
years after the eruption, he found stranded on its shores 
seven seeds and fruits, all of which are very familiar amongst 
the seeds and fruits stranded on the Keeling Islands: they 
were those of Heritiera littoralis, Terminalia Catappa, Cocos 
nucifera, Pandanus (two species), Barringtonia speciosa, and 
Calophyllum inophyllum. Excluding the first-named, and 
remembering that the species of Pandanus were not identified, 
it may be said that nearly all these trees have established 
themselves on Keeling Atoll. What, however, is more im- 
portant, is the fact that on the shore Dr. Treub observed 
growing nine young plants, of which at least four, Calophyllum 
imophyllum, Hernandia peltata (syn. sonora), Ipomea pes capre, 
and Scwvola Kenigii, belong to species well known in the flora 
of Keeling Atoll. Seven species of flowering plants had 
already begun to ascend the slopes of the volcano, two of 
them being Tournefortia argentea and Scevola Kenigii, so cha- 
racteristic of the weather shores of the Keeling Islands. I 
cannot doubt but that the waves were mainly instrumental 
in presenting the island of Krakatoa with these familiar coral 
island plants, which were evidently derived from the shores 
of the Sunda Strait, and from the coasts in its vicinity. Dr. 
Treub, however, ascertained that ferns formed the prevailing 
vegetation (eleven species haying been collected), and that a 
thin coating of alge, which covered the pumice and ashes on 
the slopes, prepared the soil for the growth of the fern- 
spores, the ferns in their turn performing the same service 
for the flowering-plants. It is evidently to the agency of 
the winds that we must attribute the presence of the ferns 
* Annales du Jardin Botanique de Buitenzorg, vol. vii, p. 213. See also 
Nature, August 9th, 1888. . ; 
