47 



mouths of rivers. On the beach this tree grows only where sweet 

 or but slightly brackish ground-water rises near the surface ') On 

 Krakatao it has repeatedly been found, i.a. in 1897 by P e n z i g, in 

 1906 by Ernst, in 1908 by myself, in 1919 by members of the 

 Congress of Natural Science. The specimens found in 1908 were all 

 very young; this proves once more that young seedlings on the 

 beach often do not grow up into trees. The reproduction is effected 

 exclusively by fruits -) which are exceedingly fit to be carried by water 

 and are very frequently found washed ashore The specimens of this 

 tree which are found in the interior of |ava are without exception 

 either cultivated or the offspring of the latter, growing quite near. 



4. Hernandia sonora L. 



This name has been given to different species. The one found 

 by T r e u b was probably Hernandia peltata Meissn., a tree which in 

 the western half of the Dutch Indian Archipelago is rather common 

 on sandy beaches ;! ). The beaches of Krakatao are a fit habitat for 

 this species which was found there at several excursions. The 

 reproduction is effected exclusively by the fruits which possess 

 floating power '') and are often found washed ashore '). 



5 and 6. Two Cyperaceae. 



This vague indication is of no value for the study of the 

 renewal of the flora. The Cyperaceae which until now were found 

 on the north-western side of Krakatao on or near the beach were, 

 as far as I know, Cyperus pennatus Lmk,, Fimbristy/is spathacea 

 Roth and Remirea maritima Aubl, This last species, occurs only 

 on young beaches consisting of very loose sands and possesses 

 a long creeping rhizome which, at short distances, emits young 

 individuals, whilst the other two species grow in tussocks and occur 

 not only on the beach but also rather far behind it; they prefer a 

 less permeable substratum. The fruits of these 3 species can be 

 spread by water. It is needless to discuss the Cyperaceae found by 



') Where this is not the case (as on the coral-island Pulu Lang in the Java-sea, 

 investigated by S c h i m p e r) either the seeds do not germinate or the seedlings die 

 early [Sc him per. Indo-Malayische Strandflora (1801), p. 190|. 



a ) Described by S c h i m p e r, Indo-Malayische Strandflora (1891 ), p. 168 tab. VII 

 fig. 22 under the name of Cerbera odollam Caertn. 



3 ) The only other species of Hernandia occurring in the East Indian Archipelago, 

 H. ovigera L, is no littoral tree | Cf. Reorders en Valeton, Bijdrage VII, p. 110 

 and R u m p h i u s. Herb. Amboinense II, 257. tab- 85 (Arbor regis). 



4 ) Described by S c h i m p e r, Indo-Malayische Strandflora (1891), pp. 176. 177. 

 tab. VII. 



8 ) G u p p y found them in the drift-zone on the Keeling-islands. 



