41 



somewhat remote one, may be considered as practically not hap- 

 pening at all. During the latest 20 years I have travelled on foot 

 over a distance of several hundreds of kilometers along the shores 

 of Java and the smaller islands in the vicinity and never found there 

 a single coco-nut which after having been washed ashore, had ger- 

 minated and rooted. Schimper when investigating the beach near 

 Tjilatjap (on the southern coast of )ava) made the same experience 

 as myself; he expressly states ty ,,|unge Cocospalmen findet man gar 

 nicht" and I mentioned already that the coco-nuts, which Ernst-] 

 and his companions found washed ashore on Krakatao, were pierced 

 like a sieve and quite empty. Ernst has figured ;! ) a rather large 

 coco-palm growing on the south-eastern side of Krakatao in a spot 

 where the beach was decreasing. According to the legend this tree 

 was growing on the beach behind the high-tide mark but the figure 

 itself shows that it was already attacked by the sea and therefore 

 every moment liable to be destroyed. This tree might quite well have 

 been planted at some distance from the sea which may have reached 

 it by washing away the beach: a shore so much inclined as the one 

 figured in the left half of the plate is only found in localities where 

 the beach decreases. 



As a proof that Cocos nucifera can be spread by sea-currents 

 Docters van Leeuwen mentions '*) the fact, that he found 

 on the (inhabited) island Sebesi (north of Krakatao) ,,some big 

 ,,coco-nut plants growing on the mostly inaccessible cliff-walls a few 

 ,,yards above the sea, probably cast up there by the high waves. Close 

 ,,to the water and higher up not a single coco-palm was to be seen". 

 But cliff-walls inaccessible to Europeans navigating on the sea below 

 need not at all be so to nimble-footed natives coming from the inte- 

 rior, and therefore the argument advanced by Docters van 

 Leeuwen is not conclusive. His assertion that the coast is (and 

 was?) never visited, should he confirmed, if possible, by a minute 

 investigation in loco before it may be taken for correct. I shall revert 

 to his paper on Sebesi in Chapter XII. 



How dangerous it is to eliminate man when accounting for the 

 presence of coco-palms in uninhabited islands appears from the 

 following communication of W. Botting Hemsley in Nature, 

 April 10 1890, p. 537: ,,The question whether the coco-nut palm is 



!) Indo-Malayische Strandflora (1891), pp. 161, 162. 



2 ) Neue Flora Vulkaninsel Krakatau (1907), p. 28. 



:! ) Neue Flora Vulkaninsel Krakatau (1907), tab. V, fig. 8. 



'<) Ann. lard. Bot. Buitenzorg XXXII (1923), p. 142. 



