68 



in Krakatao may be easily explained. No. 9, 10, 1 1, 12, 14, 15 and 17 of 

 T r e u b's list, all of them typical kremnophytes, may quite well have 

 been growing before the eruption in the locality investigated in 1886, 

 an old basaltic slope, covered by the eruption with ashes and 

 pumice; they may (as was already argued on pp. 31, 32) have 

 survived the catastrophe in the form of rhizomes or of spores. Very 

 soon after the eruption the covering layer was here and there washed 

 away. In such spots the surviving rhizomes or spores may have 

 developed into plants, the offspring of which may have spread over 

 the permanently covered parts. I think it much more probable that 

 the rich fern-vegetation which T r e u b found within three years after 

 the eruption in the locality investigated has arisen in this manner, 

 than that the spores of all these species were carried over by wind 

 over a distance of at least 20 km. to suitable habitats in such num- 

 bers as to allow the vegetation in question to develop so rapidly 

 quite anew. Alas nothing is known of the time the spores of these 

 ferns need to develop into adult plants under the conditions of growth 

 prevailing at Krakatao. On this point also the investigations after 

 the re-establishment of the vegetation have been insufficient. 



No- 16 and 18 (Ptcris tripartite S\v. and Pteridium aquilinum Kuhn) 

 may quite well be remnants of the old vegetation of the higher 

 parts of Krakatao. In many primeval forests in the mountains of Java 

 sunny or slightly shadowed spots occur, caused by large trees having 

 fallen to the ground. Not rarely also, especially in swampy localities ') 

 or on steep ravine-walls or dry ridges, the soil is locally unfit for a 

 dense vegetation of trees. In such sunny spots it often bears a grass-, 

 fern- or bush-jungle. In the former woods of Krakatao, which were 

 never investigated and of which \vc Ao\v absolute!} not/ung, many 

 such open or slightly shadowed spots may have been present and it 

 is quite possible that they bore one of these 2 ferns (which never 

 are found in each other's company: Pteris tripartite) needing much 

 more water than Pteridium aquilinum). Both ferns may have survived 

 the eruption by means of their rhizomes or their spores; both of 

 them may, after the eruption, have found fit habitats in the higher 

 not saline parts of the island; the spores of both may have been 



] j Sucli localities, by constant marshiness unfit for a dense vegetation of trees 

 are found in Java i.a. on Mount Cede, Mount Patuha, at tlie base of the twin-moun- 

 tains Wajang-Windu and at the southern slope of Mount Hjang. It is remarkable that 

 whilst sweet water marsh-forest has developed in many swampy parts of the plains 

 (i.a. near Djatiroto in Java and in vast localities in Sumatra) such has not happened, 

 to my knowledge at least, anywhere in the mountains of Java. 



