OCEANIC DISPKKSAL OF PLANTS. 297 



PALM^E. 



Orania aruensis, Beccari. (Plate LXV., I.) 

 Oranialaruensis, Beccari, Malesia, i. p. 76. 



Sea-beach, Arrou. 



Only the one germinating seed represented in our plate was collected. Dr Beccari 

 furnished the name from a drawing we forwarded to him. This palm was described from 

 specimens collected in the island of Wokan, one of the smaller islands of the Arrou group, 

 where it inhabits very dense, often inundated forests. In a paragraph, loc. cit., p. 9, on 

 the distribution of palms, Dr Beccari calls attention to the fact that the fruits or seeds 

 of very few palms are adapted for wide dispersion. Some, he says, like Cocos nucifera 

 and Nipa fruticans, are able to resist the action of salt water; and he had also observed 

 that the fruits of Orania are often thrown up uninjured from the sea. "Ho osservato 

 che anche i frutti di Orania sono spesso rigettati intatti dal mare." But what is more 

 extraordinary, he states that the cassowaries often swallow the fruits of palms which 

 are in no wise fleshy, and carry them considerable distances. In the Arrou Islands, for 

 instance, he saw heaps of the seeds of Orania aruensis in their excrement, yet every one 

 was from 55 to 60 millimetres in diameter — that is, from about 2\ to 2^ iuches ; the one 

 represented in Plate LXV. is barely 2^ inches in diameter. Of course, although this be 

 the case, the cassowaries could only assist in the local dispersion of the species. It is 

 certainly remarkable how very local the majority of palms are. 



GNETACE/E. 



Gnetum rumphianum, Beccari. (Plate LXIV., D.) 

 Gnetum rumphianum, Beccari, Malesia, i. p. 182. 

 Funis gnemoniformis, Rumph., Herb. Amboin., v. p. 11, t. 7. 



Sea-beach, Arrou. 



We are also indebted to Dr Beccari for the determination of this seed, of which there 

 are several in the collection. In all cases, the outer fleshy envelope of the seed has dis- 

 appeared, leaving bare the second envelope, which is crustaceous and traversed longi- 

 tudinally by about fifteen prominent ribs, and is free from the body of the seed except 

 at the base. As shown in the plate, the seeds have begun to germinate by breaking 

 through the apex of this second envelope, which, being free from the body of the seed 

 within and splitting longitudinally, might easily be mistaken for a pericarp. None 

 of the seeds is in a sufficiently advanced stage of germination to enable us to fuliy 

 explain the mode, which is certainly very singular, and apparently essentially the same as 

 in Gnetum gnemon, as described by Bower (Quarterly Journal of Microscopical Science, 

 X. S. xxii. p. 278, t. 25). In the latter the seed is about a cmarter the size of ours. The 



(BOT. CHAIX. EXP. — PART III. — 1885.) U °° 



