OCEANIC DISPERSAL OF PLANTS. 307 



tion. Of course, the waves of the sea may sometimes bury the cocoa-nut as well as cast 

 it ashore. Moreover, under certain conditions, the cocoa-nut may germinate freely without 

 being buried or partially buried, or even without being fixed to the soil. See Tonning's 

 reference to the cocoa-nuts being cast ashore in Norway, ante, p. 277. 



Sagus sp. 



At Malanipa Island, off the coast of Mindanao, Philippines, Mr Moseley found a young 

 sago-palm, which was just beginning to form a stem, washed up just above the ordinary 

 beach-line, and firmly rooted, though in an inclined position, and growing vigorously 

 (Notes by a Naturalist, p. 368). 



FRUITS AND SEEDS TAKEN FROM THE CROPS OF FRUIT PIGEONS, 

 SHOT IN THE ADMIRALTY ISLANDS. 



In Mr Moseley's Notes, reproduced at p. 229, he records some observations on the 

 part played by carpophagous birds in the dissemination of plants. The collection of fruits 

 he mentions having taken from the crops of a number of a Fruit Pigeon (Ca rpophayu, 

 rhodinolcema 1 ) contains about a dozen species, of which we have identified : — 



Elaeocarpus sp. 



This fruit is the most numerous of all, and perhaps the least nutritious. Within the 

 thin fleshy coat is an exceedingly hard five-celled indehiscent endocarp. The species of 

 Elaocarpus are numerous in Tropical Asia, and a few inhabit Polynesia, Australia, and 

 New Zealand. 



Soulamea amara, Lam. 



Soulamea amara, Lam., ante, p. 235 ; Ann. "Wiener. Mus., i. t. 1G. 



A single empty fruit. This plant is in Dr Guppy's collection from the Solomon Islands, 

 lately received at Kew. It is the Rex amaroris of Rumphius (Herb. Amboin., ii. 

 p. 129, t. 41). 



Rourea ? 



Only one small undeveloped worm-eaten fruit. One species of Rourea has been found 

 in the Fijis, and one in North Eastern Australia. 



Rubiacea ? 



The mericarps, as well as the entire fruits of what we regard as a member of the 

 Rubiaceae, are, next to the Elaocarpus, the most numerous in the collection. We have 

 not been able to determine the genus. 



1 Narr. Cliall. Exp., vol. i. p. 566, et passim, 1885. 



